ARAM, 13-14 (2001-2002), 51-72 H.H. CURVERS 51

THE LOWER TOWN OF (1200-300 BC) A PRELIMINARY SYNTHESIS

HANS H. CURVERS

This preliminary synthesis is an effort to clarify the Iron-Age developments in Beirut. After a long period of archeological inaction the recent excavations1 in Beirut provide a new impetus to the research of the Iron Age and beyond. As this contribution is but a first step towards a periodization of the Iron Age developments throughout we do not yet try to define a specific chronological frame. The first half of the 1st millennium BC is the start of a new cycle of rise and fall in the civilization process in the Levant. The previous centuries are often considered a Dark Age in the history of the Eastern Mediterranean and Near East. The period of decline started at the end of the Late Bronze Age in Egypt, the Hittite Empire and the Mycenean Empire (ca. 1200 BC). These cultures had produced a material culture that found its way towards the outlying vassals that lived in the Levant. It is through the presence of art objects and textual sources that we can start to appreciate the history of the area. In the past the first part of the Iron Age (c. 1200-) was viewed as a period of decline compared to the Late Bronze Age. The excavations in Israel, Jordan, Lebanon and Syria, however, seem to suggest that after a relatively short period of transition a new cycle of rise and decline of civilization in human history started. The influences of and ex- change with distant regions are reflected by monuments, graves, architecture and techniques of fortification throughout the Near East and the Mediterranean world. Within this perspective we have summarized the recent findings in one of the maritime centers on the Lebanese coast and we will synthesize the first re- sults and interpretations published in periodicals that are not widely distrib- uted2. These publications focus on the results of the excavations of the tell of 1 We are grateful to the efforts of the representatives of the General Directorate of Antiquities in Lebanon, SOLIDERE (Societˆé Libanais pour le Development et la Reconstruction of the Bei- rut Central District) and UNESCO who have cooperated and allowed for the archeological exca- vations in the BCD. 2 L. Badre, “Bey 003 Preliminary Report, Excavations of the American University of Beirut Museum 1993-1996”, Bulletin d’Archéologie et d’Architecture Libanaises 2 (Beirut, 1998), pp. 6-94; N. Karam, “Bey 013 rapport préliminaire”, BAAL 2 (Beirut, 1998), pp. 95-113; U. Finkbeiner and H. Sader, “Bey 020 Preliminary Report on the Excavations 1995”, BAAL 2 (Bei- rut, 1998), pp. 114-166; H. H. Curvers and B. Stuart, “The BCD Infrastructure Archaeology Project, 1995”, BAAL 2 (Beirut, 1998), pp. 167-205. 52 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS

Beirut3. Subsequently we will add more detailed observations on the excavated sites and present a general layout of the lower town. We will also include the discoveries of a representative quarter of the 5th-4th century BC in BEY 0104. Finally, the discoveries of the cemetery associated with these remains will be dealt with in another communication (Stuart, this volume), but they will con- tribute to this attempt of synthesis. All these new observations complement the approach and synthesis of Mouterde and Davie5. To obtain a background for the preliminary results of the excavations in Beirut we used the available written documents: internal sources, such as the epigraphic documents and the external sources in the Egyptian, Babylonian and Assyrian archives. The corpus of written sources in Beirut and beyond is still small and the few epigraphic sources that we know of have limited historical value6. No royal inscriptions of the type found in the monuments of Egypt and Assyria have yet been found. A royal stele such as the stele of Mesha, king of Moab, is absent. As very few monumental buildings of this period have been discovered in the coastal sites of Lebanon, monumental building inscriptions are also lacking. Although petty kings have resided in the coastal ports of Lebanon we did not find their royal archives. Helen Sader has summarized the textual evidence on Beirut and concludes, “Beirut was in the eighth century BC a fortified city under the political hegemony of Sidon.”7 The inscriptions of Asarhaddon, King of Assyria, list Beirut in a series of geographical names, which can be associated with modern communities on the Lebanese coast: Gi’ (Jiyyeh), Inimme (Naameh), Hildua (Khaldeh), Qartimme (Kafrshima? according to Sader8), Bi’ru (Beirut), Kilme, Bitirume (Batroun), Sagu (Chekka), Ampa (Enphe). Badre continues, “It is only during the Persian period of the 4th c. BC that Pseudo Scylax briefly mentions in his Periplus ‘the city and the north- ern harbor’ of Beirut”.9 Linda Hall cites Mouterde’s reference to Pseudo

3 See also L. Badre, Les découvertes archéologiques du centre-ville de Beyrouth, CRAIBL 1996, (Paris, 1996) pp. 87-97; L. Badre, Beyrouth: découverte d’une cité fortifiée, Liban, l’autre rive. Exposition présentée à l’Institut du monde arabe du 27 octobre 1998 au 2 mai 1999 (Paris, 1998), pp. 76-78; L. Badre, Un entrepot de commerce phénicien, Liban, l’autre rive, (Paris, 1998), p. 114; U. Finkbeiner, Beyrouth: la cité phénicienne révelée, Liban, l’autre rive, (Paris, 1998), pp. 112-113; H. Sayegh, Urbanisme phénicien à l’époque perse, Liban, l’autre rive, (Paris, 1998), p. 114-115. 4 H. Sayegh, “Bey 010 Les souks, secteur nord/est”, BAAL 1 (Beirut, 1995), pp. 235-269; H. Sayegh, “Urbanisme phénicien”, Liban, l’autre rive, (Paris, 1998), p. 114-115. 5 R. Mouterde, “Regards sur Beyrouth, phénicienne, hellénistique, et romaine”, MUSJ 40 (Beirut, 1964), pp. 164-189; M.F. Davie, “Maps and the Historical Topography of Beirut”, Berytus 35 (Beirut, 1987), pp. 141-163. 6 W.A. Ward, “Ancient Beirut” in Beirut – Crossroads of Culture (Beirut, 1970), pp. 27-37. 7 H. Sader, “The localisation and setting of Ancient Beirut”, in U. Finkbeiner and H. Sadr, “Bey 020” (1998), p. 117. 8 H. Sader, “Localisation”, p. 116. 9 L. Badre, “Bey 003”, p. 12. H.H. CURVERS 53

Scylax and translates “the city and its port which is turned ‘a little to the north’”.10 Furthermore, most of the written sources from the area such as the scrolls of parchment and sheets of papyrus, did not survive the humid climate of the Lebanese coast. However, in addition to these written sources a growing cor- pus of ostraca (potsherds with inscriptions) has come to light. The date of these ostraca, although written in Phoenician script, is closely related to im- ports of “Attic ceramics”.11 Sader continues her historical introduction: “the site of pre-classical Beirut has been protected over the centuries because it was not resettled after the center of the Roman city had moved westward. The exclusion of this area from the Medieval and Ottoman settlements left it unoccupied. Its partial use as a Muslim cemetery prevented the growth of the city in this direction. Ironically, most of the damage caused to the archaeological remains goes back to the last seventy years”.12 If we also include the evidence retrieved from Bey 00313 as indicated on the plan of the tell area prepared for the General Directorate of Antiquities14, we can add more “culprits” to the list of destroyers of a major part of the tell: the builders of the Crusader Castle. They dug a ditch of 14 m in width along the south and east walls of the castle tower15. In BEY 032 we observed that the Crusaders built a retaining wall along the south edge of the ditch to prevent the ancient remains from falling into the ditch.16 We could describe this construction as an act of archeological resource management avant la lettre. The external sources provide the historical milestones of the first half of the 1st millennium BC. The campaign of Shishak, king of Egypt, resulted in the destruction of cities such as Megiddo (925 BC). One of the Assyrian period reliefs carved into the rocks of the Nahr el-Kelb, c. 10 km north of Beirut, is dated to 842 BC. In this year Shalmaneser III campaigned against Hazael of Damascus. In his palace at Imgur-Bel (Balawat) in northern Iraq he had Tyre, the fortified island city, depicted on one of the bronze strips decorating the gate.17 In the annals of this campaign the king boasts that he descended to the 10 L. Jones Hall, this volume; R. Mouterde, “Regards”, p. 156. 11 H. Sader, “Phoenician Inscriptions from Beirut”, in L. H. Lesko, (ed.), Ancient Egyptian and Mediterranean Studies, In Memory of William A. Ward, (Providence, Rhode Island, 1998), pp. 203-213. 12 H. Sader, “Localisation”, p. 121-122. 13 L. Badre, “Bey 003”, p. 168. The plan has been wrongly included in the Curvers-Stuart contribution on page 168 with the caption “Fig. 1 BCD masterplan and location of the BCD In- frastructure Project (IPP-UVA) sites”. Badre refers to this plan as Figure 1b on pp. 26, 28 and 32. 14 U. Finkbeiner and H. Sader, “Bey 020”, p. 167. 15 See n. 13. The ditch is located in the squares 110-120/270, 110-120/260, 110-120/250, 100- 80-240. See n. 14 in the area with the text “south tower” and “BEY 003”. 16 See n. 14. Wall indicated in green in Bey 032. 17 C.L.W. King, Bronze Reliefs from the Gates of Shalmaneser, (London, 1915), pls. 13-14. 54 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS mountains of Ba‘alrasi, situated down at the coast. He also collected tribute from Sidon, Tyre and king Jehu, descendant of Omri18. Sennacherib launched his campaigns against the area in 701 BC. He is held responsible for the de- struction of Lachish stratum III. The flight of Luli, king of Tyre and Sidon is depicted on his reliefs.19 Asarhaddon left his stele in the rocks at the Nahr el- Kelb. The king left his image covered with a text boasting the victory over Sidon and Tyre after his campaign in 676-671 BC20. In Zinjirli in southeastern Turkey he expressed his dominance over the conquered kings on a stele.21 Nebuchadnezzar left his marks in Lachish stratum II in 586 BC. The latter date is traditionally used for the end of the Iron Age in Israel and beyond. We can imagine, however, that the material culture did not change significantly or im- mediately after this date. Even the fall of the Babylonian Empire in 539 BC may not be visible in the material culture beyond Babylon itself. The town of Beirut seems to have depended on Sidon till the early third century.22 For reasons of lumping rather than splitting we prefer to discuss the devel- opments in a large chronological frame; therefore, we use the end date of 300 BC in the title of this contribution. In our preliminary reports on the period between 1200 and 300 BC we have used a local designation “Beirut Period VI”.23 This wide range allows for and challenges more refinement and recon- struction of the developments in the realm of the material culture. We are pre- paring a more detailed presentation of the data discussed here for publication in a special edition of Talanta.

THE STRONGHOLD (Fig. 1-2)

The stronghold of the Iron Age settlement of Beirut is located on the tell, which is situated on the cliffs of the northern half of the semi-peninsula. The features described below have been indicated by numerals on the simplified plan in Figure 2. The earliest remains (see Fig. 2:1) are found within layers and associated with floors that yielded EB II-III ceramics.24 Leila Badre exca- vated three architectural complexes I-III (see Figs. 2:1-3), of which two be- long to the six fortification systems (see Figs. 2: 2, 3, 4, 9, 10, 12). Uwe Finkbeiner and Helen Sader exposed a fragment of the Bronze Age enclosure wall (see Fig. 1:4), a major part of the glacis and the stepped entrance to the

18 R. Mouterde, “Regards”, p.155. 19 D. Harden, The Phoenicians, (London, 1962), Pl. 50. 20 F.H. Weissbach, Die Inschriften Nebukadnezars II im Wadi Brisa und am Nahr el-Kelb. (Leipzig, 1906). F.H. Weissbach, Die Denkmäler und Inschriften an der Mündung des Nahr el- Kelb. (Berlin, 1922). 21 F. von Luschan, Ausgrabungen in Senschirli I, Berlin, 1839), Pp. 11-29, Pl. 1. 22 A.H.M. Jones, The Cities of the Eastern Roman Provinces, (Oxford, 137, reprint 1998), p. 446. 23 H.H. Curvers and B. Stuart, “BCD 1995”, p.76-77. 24 L. Badre, “Bey 003”, p. 12-22, figs. 2-7. H.H. CURVERS 55 stronghold (see Fig. 2:17). Naji Karam exposed the eastern limits of both the Bronze Age enclosure wall and the glacis (see Figs. 2:5, 8). Cut into these re- mains he excavated a sequence of semi-circular towers belonging to the Hel- lenistic settlement (see Fig. 2:18). Hans Curvers and Barbara Stuart excavated remains of the stronghold in several operations located at the outskirts of the Iron Age enclosure. In the operations at the western end of the tell several glacis and a ramp leading up from the lower town to the stronghold were ex- posed (see Figs. 2:11, 14). In an operation south of the stronghold, fragments of the glacis overlaying pre-glacis architecture and part of a Hellenistic fortifi- cation system were found (see Figs. 2:7, 8, 19). First we describe the majority of these remains in more detail. Subsequently we will try to synchronize their proposed dates and, on the basis of the proposed synchronization provide a more coherent view on the stronghold of Beirut. Complex I (see Fig. 2:2) consists of two walls that enclose a mud-beaten floor, and a third wall that was added to this construction in a later phase25. The First Fortification Wall is interpreted as a possible retaining wall parallel to the long wall of complex I and on top of a natural or artificial slope covered with clay.26 Complex II (see Fig. 2:3) consists of a wall provided with pilasters. In two other sites remnants of this wall seem to have been exposed. There, however, the pilasters were not observed. In BEY 020 two phases of an enclosure wall have been exposed. 27 “The Middle Bronze Age city-gate in Bey 003 lies in the line of the earlier phase of the Bronze Age wall (in BEY 020) and is con- structed in a similar way out of similar stones”.28 Pilasters in the BEY 020 walls (see Fig. 2:4) are missing; one can imagine that the pilasters were built in the gate area to stress the monumental character of the entrance. More to the east, yet another fragment of an enclosure wall was exposed in BEY 013 (see Fig. 2:5).29 “Quant à la hauteur elle varie entre quatre mètre (..) et plus de sept mètres vers l’extrémité (…). Ces gigantesques dimensions en font un monu- ment tout à fait exceptionel et l’on ne peut que l’identifier à l’un des remparts de la cité antique de Beyrouth.”30 A clear date is not provided; it is in the cap- tions of the publications on the ‘archeological crimes’ committed in Beirut that one can catch a glimpse of the date. Karam describes a picture similar to the ones published in his preliminary report: “Les restes du rempart du IIIe mill.: toujours menacés”.31 The pottery retrieved from a pocket of yellowish clay at

25 L. Badre, “Bey 003”, p. 22-24. The excavator offers various comparisons for the pottery, ranging in date between MB I and MB II. 26 L. Badre, “Bey 003”, p. 26-28. 27 P. Jablonka, “Stratigraphy and architecture”, in U. Finkbeiner and H. Sadr, “Bey 020”, pp. 124-135 28 P. Jablonka, “Stratigraphy”, p. 124-135 29 N. Karam, “Bey 013”. 30 N. Karam, “Bey 013”, p.107. 31 N. Karam, Beyrouth, l’histoire qu’on assassine. (Beirut, 1996), p. 63. 56 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS the bottom (or below?) the wall is compared with Early Bronze IV pottery from Tell Arqa. “Il est bien évident que cette suggestion restera provisoire en attendant l’analyse de tout le materiel céramique découvert au pied de cette structure <>”.32 The possibility that the pocket with the EB IV ceramics is related to an occupation on top of the bedrock suggests a stratigraphic posi- tion anterior to the construction of the high wall. The material deposited against the rempart blanche seems to belong to the substructure of the glacis33 (cf. BEY 020 and BEY 003), and therefore allows a date of the rampart be- tween the end of the third millennium and Iron II. We assume that topographic conditions asked for the construction of the wall against the cliffs and consider the rempart blanche to have been part of the (Late) Bronze Age enclosure. Badre suggests that the gate (see Fig. 2:3) of 2.10 m width in the Bronze Age perimeter was created by the construction of two piers on both sides of the opening in the wall. In a later phase an L- shaped wall was added to the entrance to create a limited access to the gate.34 Badre dates the “Monumental gate (…) with certainty to the Middle Bronze Age”.35 Complex III which includes a silo (or well), rock-cut chamber and a jar burial (see Fig. 2:6) was found within the perimeter of the Crusader Castle. The remains are dated to the Middle-to-Late Bronze Age.36 Glacis I, which represents the third fortification system (see Fig. 2:9), par- tially covers the monumental gateway. The pottery found in the levels over this glacis suggests that “the use of Glacis I ended in the LB II period”.37 The fourth fortification system is represented by Glacis II (see Fig. 2:10); “the date of this glacis is deduced from the periods of two destruction levels above and below the glacis. The first destruction level over which the glacis was built consists of four layers which covered Glacis I. These provide a ter- minus post quem toward the end of the Late Bronze II Period to Glacis II”.38 The excavator uses the material deposited in the layers over Glacis II to obtain a terminus ante quem for the use of Glacis II. The dates established on the ba- sis of a preliminary analysis of the pottery retrieved from the deposits below and above Glacis II suggest “that the glacis [II] was built in the Late Bronze II Period and remained in use until the end of the Iron I Period”.39

32 N. Karam, “Bey 013”, pp. 107-108, figs. 5-8. 33 cf. L. Badre, “BEY 003”, p. 48; cf. Jablonka, “Stratigraphy”, pp. 126-127, Fig. 6. 34 L. Badre, “Bey 003”, p. 28. 35 L. Badre, “Bey 003”, p. 30-31, Fig. 12. We would date the context on the basis of the latest remains which suggest MBII. A date in the LB period would also be appropriate for the very shallow bowl (Fig. 12:11) for just one southern and one northern comparison we suggest R. Amiran, Ancient Pottery of the Holy Land. (Jerusalem, 1969), Pl. 38:16 (LBII) and G. F. Swift Jr., The Pottery of the ‘Amuq Phases K to O and its Historical Relationships, Unpublished Ph.D. thesis, University of Chicago, (Chicago, 1958), p. 208, fig. 7 (Amuq M). 36 L. Badre, “Bey 003”, pp. 32-34. 37 L. Badre, “Bey 003”, p. 54. 38 L. Badre, “Bey 003”, pp. 60-64. 39 L. Badre, “Bey 003”, p. 68. H.H. CURVERS 57

Personal observations in the field indicate the presence of various sloping stretches of stones. These stretches could represent the top surfaces of various stone-paved glacis. The first glacis (Glacis I) enclosed the Bronze Age gate and its L-shaped chicane, the others would have enclosed their predecessors. Leila Badre has included the various stretches into two glacis. However, they could also reflect the dynamics of the fortification of ancient Beirut. The use, repair and construction must have been an ongoing routine in the maintenance of the fortifications. Repairs seem to have been necessary after attempts to conquer the stronghold and heavy rains.40 This interpretation fits the observa- tions in BEY 032 (see Fig. 2:11) where two to four glacis scales were ob- served41, and BEY 114 where two or three superimposed glacis pavements were exposed (see Fig. 2:7 and Fig. 3). The latest results in BEY 020 also seem to point towards the existence of several phases in the construction and use of the fortification, including steps, stairways and a guardhouse.42 As all observations made on the glacis fortification on the tell of Beirut sug- gest a sequence of a series of stone-paved glacis, the dating of these (re-)con- structions should be synchronized. Leila Badre suggests a Middle Bronze date for Glacis I and a Late Bronze to Iron Age I date for Glacis II. The analysis of the pottery retrieved from below the outer Glacis in Bey 032 and above the lower Glacis revealed that the assemblage between the Glacis can be best com- pared with Sarepta E/D contexts.43 Whereas the material over Glacis II shows the best match with Sarepta C.44 These comparisons indicate a date in the Iron II/III range. The outer glacis in BEY 020 is overlaid with “Iron II/III layers”.45 The sub-recent infrastructure systems of Azmi Bey Street seem to have ob- literated all evidence that could contribute to a clear-cut solution on the se- quence of the various glacis in BEY 003. The evidence from the surrounding sites may be all that is left and an Iron II-III date for the glacis the most reli- able preliminary date. This date seems to fit the general picture of Levantine architectural developments, in which great urban projects started in Iron II (c. 1000 BC). The fifth fortification system is described as the “Casemate Wall” or “Casemate-Wall Building”; a construction of interconnected elements each with two or more rooms.46 Some of the rooms were stacked with amphorae and suggest a use of this structure as a storage facility. Badre compares the

40 P. Jablonka, “Stratigraphy”, p. 127-128 41 O. Nieuwenhuyse, “Bey 032”, in H.H. Curvers and B. Stuart, “BCD 1995”, p. 178-180, fig. 5. 42 U. Finkbeiner, this volume. 43 A. Jamieson, “Bey 032”, in H.H. Curvers and B. Stuart, “BCD 1995”, p. 182-183; W.P. Anderson, Sarepta 1, Late Bronze Age Strata of Area II, Y. (Beirut, 1988), pp. 422-423, Pls. 35-37. 44 A. Jamieson, “Bey 032”, p. 183; W. P. Anderson, “Sarepta”, pp. 422-423, Pls. 33-34. 45 U. Finkbeiner and H. Sadr, “Bey 020”, p. 164, Pl. 4c. 46 L. Badre, “Bey 003”, pp. 76-88, Figs. 40 a-b. 58 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS amphorae and the two fragments of basket-handle amphorae to finds from Tell Keisan level 4-5; subsequently, she suggests a date for the “Casemate-Wall Building” in the 7th century BC.47 The discussion on the date of Keisan Levels 5-4 has been opened by Humbert who prefers a higher date for the tran- sition at 700 BC instead of the previously proposed 650 BC.48 Salles and Gilboa, however, prefer the previously proposed date for the beginning of Keisan Level 4.49 If we use the criteria established and applied at Tell Dor for the dating of the “Casemate-Wall Building”, a date in the Persian Period or 5th – 4th century BC seems appropriate. The “Casemate-Wall Building” re- vealed amphorae with a sack-shaped instead of a biconical body, and ampho- rae with a low neck instead of a relatively high neck.50 Yet a sixth fortification system was exposed south of the Casemate con- struction; a retaining wall apparently built in segments runs parallel to Glacis II. The remains of this sixth fortification system are also exposed in other areas of the tell (BEY 02051, BEY 114; see Fig. 3). Considering the constructional problems that the builders of the “Casemate-Wall Building” had to solve – construction of a building at the outskirts of the tell – one can imagine why the sixth fortification system was built. In addition to this retaining wall, an intri- cate system of more retaining walls may have provided the necessary stability for construction on the edge of the previous Iron II-III fortifications.52 It is too early to assess all the details of this long series of fortification sys- tems that date back to the second millennium BC. We gave this extended de- scription because it serves as an introduction to the developments to be ob- served in the lower town of Beirut. Leila Badre concluded her preliminary re- port on BEY 003 with, “Recent discoveries in various sites of the city must lead to a thorough research program of the extra muros extension of ancient Beirut. (…) It seems quite possible that at least from the Late Bronze period,

47 L. Badre, “Bey 003”, p. 84. Badre favors apparently the higher date of the Keisan transi- tion between level 4 and 5, although comparisons with Tell Keisan Level 4 already suggest a date later than the middle of the 7th century BC. 48 J.B. Humbert, “Récents travaux à Tell Keisan (1979-1980), Revue Biblique 88 (1981), pp. 373-398. 49 J. F. Salles, “A propos du niveau 4 de Tell Keisan, Levant 17 (1985), pp. 203-204. A. Gilboa, “The Typology and Chronology of the Iron Age Pottery and the Chronology of Iron Age Assemblages, in E. Stern et al. Excavations at Dor, Final Report. Vol. 1B, Areas A and C: the Finds. (Jerusalem, 1995), p.2. 50 E. Stern, The Material Culture of the Land of the Bible in the Persian Period. (Warminster, 1982), Fig. 152; Type H-6 (biconical and high neck), Type H-8 sack-shaped or bag-shaped body and low neck.; E. Stern, “Local Pottery of the Persian Period” in E. Stern et al., “Dor”, pp. 62- 63. 51 U. Finkbeiner and H. Sadr, “Bey 020”, p. 128, Fig. 6. 52 L. Badre, “Bey 003”, General Site Plan, between pp. 6 and 7. The walls are situated in excavation unit 230/80 (NW corner): W 127 (6th Fortification system), W 300 and W 508 are proposed here to belong to the foundation system of the “Casemate-Wall Building”. See also U. Finkbeiner and H. Sadr, “Bey 020”, Topographical plan of the archaeological remains (cen- tral district NE) between pp. 166 and 167. H.H. CURVERS 59 the intra muros city may have turned into a fortified citadel while a lower town developed further to the south and towards the west.”53 To allow for a discussion on the dates of the features exposed within the perimeter of the ancient tell of Beirut, we have included a series of alternative dates for the main features described above. The synchronization is far from complete, but all discussions with Leila Badre, Uwe Finkbeiner, Helen Sadr, Andrew Jamieson, Olivier Nieuwenhuyse and Francesco Nunez Calvo have contributed to the development of the chronological table below. However, I take full responsibility for the suggested synchronization and I repeat that it should serve future discussion and refinement.

Badre Curvers Complex 1 and Fortification system 1 (see Figure 2:2) MB IB MBII

Rempart Blanc (see Figure 2:5) LB Bronze Age wall (see Figure 2:4) LB Complex 2, or Fortification system 2 (See Figure 2:3) MB IB LB

Complex 3 (see Figure 2:6) MB-LB –

Pre-Glacis architecture (see Figure 2:8) LB (or IAI?) Pre-Glacis entrance? (see Figure 2:17) LB?54

Glacis I or Fortification system (see Figure 2:9) LBII IA II-III Glacis with steps (see Figure 2:17) IA II-III Glacis I -II (see Figure 2:11) IA II-III Glacis I-III (see Figure 2:7) IA II-III Glacis II or Fortification system 4 (see Figure 2:10) LB-IA I IA III

Ramp (see Figure 2:14) IA III – Persian Casemate-Wall building or Fortification system 5 (see Figure 2:13) IA II IA III – Achaemenid Retaining wall(s) or Fortification system 6 (see Figure 2:12) IA III IA III – Achaemenid

Circular towers (see Figure 2:18) 3rd c. BC Casemate fortification (see Fig. 2:19-22) 3rd c. BC

53 L. Badre, “Bey 003”, pp. 91-92. Although I have suggested different dates for the indi- vidual architectural features, Leila Badre’s remark summarizes Beirut’s urban history perfectly. 54 U. Finkbeiner, this volume and personal communications. The balks of BEY 020 were re- moved to find the bottom of the Glacis. This information was necessary for an appropriate deci- sion-making process on the location of the retaining wall separating the Tell from the present Cadmus Street. The excavations revealed a platform that was enclosed by a glacis fragment. A relation to the nearby Bronze Age walls could not be excluded. We have to await more detailed reports and more thorough analysis of the pottery. With Helen Sader and Francisco Nunez we checked the pottery of BEY 105 and similarities with the earliest material of the ‘guarded stair- way’ were recognized. 60 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS

The table indicates that the sequence Bronze Age retaining wall/enclosure with gate, construction of the first glacis and the second glacis, and the con- struction of retaining walls/ “Casemate-Wall Building” stretches from the Late Bronze Age to the Achaemenid period. The site seems to have been with- out major construction projects in the Iron I period. The tradition of construct- ing retaining walls around the tell is continued in the Hellenistic period (3rd century BC). A casemate wall stretches all along the perimeter of the acropolis (see Fig. 2:19). A tower has been exposed at the eastern end of the oval mound (see Fig. 2:20).55 In 1995 remains of a substantial wall were retrieved from a large infrastructure trench SE of the tower. In 1999 a second tower and the adjoining eastern and southern city wall were exposed in the southeastern part of the BCD. These discoveries allow for the reconstruction of a fortified acropolis and the eastern limits of the fortified lower town of Hellenistic Bei- rut. If we assume that Beirut depended on Sidon till the early 3rd century BC56, we can imagine that the construction of the towers and city wall is of a slightly later date. The early 3rd century was characterized by the end of the kingdoms in Tyre, Sidon and Awad (ancient Aradus) and “the deposition of the Phoenician dynasties carried with it the dismemberment of their dominians; the dependent cities were detached and converted into separate re- publics”. For Beirut this may have been the chance to obtain the rights to for- tify the city in order to keep and seek independence from its previous masters in the period in which Ptolemaic and Seleucid rulers fought for control over the Levantine regions (e.g. southern Syria, and Judaea). The pres- ence of a wealthy trade colony in Delos and the city obtaining the right to mint its own coins in the 2nd century BC point to the success of this policy57. Future discussions on chronology should start with the following remarks that it is “unlikely that all defensive walls discovered represent stretches of chronologically successive circuits. At any one time some elements must have coexisted to give a multiple trace defensive system” 58 or that it is “an open question at this point whether the results outlined above will remain valid through the end of subsequent studies” 59or “La poursuite des fouilles au sud du glacis et des tours s’avère, donc, necessaire pour mieux comprendre l’importance de l’expansion urbaine de Beyrouth à la fin du premier millénaire avant notre ère”.60 To understand the history of the ancient tell the excavated “structures should evidently be understood in relation to similar structures exposed elsewhere on the slopes of the ancient mound”.61

55 N. Karam, “Bey 013”, p. 109, Fig. 4, Pl. 1. 56 A.H.M. Jones, “Cities”, p. 446. 57 A.H.M. Jones, “Cities”, pp. 234-253. 58 L. Badre, “Bey 003”, p. 92. 59 U. Finkbeiner, “Pottery” in U. Finkbeiner and H. Sader, “Bey 020”, p. 136. 60 N. Karam, “Bey 013”, p. 112. 61 H.H. Curvers and B. Stuart, “BCD 1995”, p. 182. H.H. CURVERS 61

THE LOWER TOWN (Figures 1, 4 and 5)

The data retrieved from the occupation levels in the lower city, contempo- rary with the stronghold, suggest dates in the Late Bronze Age, Iron Age and Achaemenid period (Beirut Periods V-VI). In both the stronghold and lower town the remains of the Hellenistic period have been retrieved, but they will not be covered in this preliminary synthesis. The Beirut Periods V and VI re- mains have been retrieved from excavations in BEY 01062, 011, 019, 031, 04563, 046, 057, 064, 082, 084, 089, 090, 101, 105, 116, 118, 119, 124, 12564, 126, 127. The excavations in BEY 01065 and BEY 105 have furthermore re- vealed remains dating back to the Bronze Age. The stratigraphy in two sites helped us in defining the stratigraphic position of the earliest remains of the Bronze Age and Iron Age (BEY 105 and 114). Therefore, we will focus first on the architecture and stratigraphy of these sites. The other sites have yielded stratigraphic deposits of the first millennium BC that preceded the architecture dated to the Hellenistic, Roman and Byzan- tine periods (= Beirut Period VII). Two sites (BEY 118 and BEY 119) yielded two superimposed layers with architectural remains (Beirut Period VI) that re- vealed interesting features regarding the urban layout of Beirut. BEY 113, 126 and 127 have yielded features dated to the first millennium BC. These three groups of sites allow for a first interpretative hypothesis about the use of space in Iron Age Beirut. BEY 114 (see Figs. 1 and 2) is located at the edge of the tell. In 1995 Uwe Finkbeiner and Helen Sader started investigations in this area.66 In a small trench their team was able to trace some remnants of the glacis. In 1997 the construction of the new BCD infrastructure necessitated a second phase of in- vestigations in this zone. Again we were able to locate the glacis. Due to the construction of a stormwater culvert we had to descend below the level of the glacis. Taking advantage of the glacis being already damaged by the construc- tion of a pre-war building (No. 391), and thus combining the past and future destruction with archeological excavations, the archeological excavations re- sulted in stratigraphic information at the border of the stronghold and lower town. As the presence of a concrete foundation made our collected ceramics less reliable in this site we were limited to stratigraphic observations in the west and east sections. Both sections revealed stratigraphic deposits of layers earlier than the glacis. In the east section a red-brown layer was associated with the corner of a wall/

62 H. Sayegh, “Bey 010”, pp. 235-269. 63 R. Thorpe, personal communication. 64 Excavated by dr Leila Badre and Nadine Panayot. 65 H. Sayegh, personal communication. 66 U. Finkbeiner and H. Sader, “Bey 020”, Pl. 2b-c. 62 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS foundation of natural stones. The substructure of the glacis covered these re- mains. In the deep trench in the northern part of the site architectural remains were clearly noted below the glacis construction and the retaining wall (see Fig. 2: pre-glacis I, pre-glacis IA, and pre-glacis IB). This architecture seems related to a wall south of the bottom of the glacis in the west section (Fig. 2: pre- glacis I). At least two levels with architecture pre-date the glacis construction. We do have evidence of the glacis sequence in the deep sounding but the stratigraphic position of the exposed architecture, protruding from below the glacis and retaining wall, suggests that we are dealing with architecture previ- ous to the construction of the Iron Age fortification. The pottery associated with the architecture suggests a date similar to the earliest levels of BEY 105 and the published Late Bronze Age material of BEY 003. Therefore, a tenta- tive date in the Late Bronze Age is proposed. The walls exposed in BEY 114, if indeed belonging to the Late Bronze Age, belong to a settlement outside the Bronze Age wall excavated in BEY 003, 013 and 020. The remains in BEY 105 seem to support the existence of a Late Bronze Age lower town in Beirut. BEY 105 (see Figs. 4, 5) is situated in an area south of the Iron Age harbor. The remains appeared when new basements were added to the building (founded in the French Mandate period) to construct an underground car park. The Ottoman constructions preceding those of the French Mandate period had already heavily truncated the archeological remains. On top of a red-brown layer that was reached slightly below the future bottom slab of the basement, two parallel walls are assigned to the earliest level of the site. A preliminary analysis of the pottery suggests a date in the Late Bronze Age. In the subse- quent Level 2, small cubicles into which and against which platforms of circu- lar and rectangular shape were built, characterize the architecture. The remains of Levels 3 and 4 were heavily truncated. Level 5 was preserved in the eastern part of the building, apparently undisturbed by the Ottoman foundations. This level yielded ceramics clearly to be dated to the 5th and 6th centuries AD. For our synthesis of the lower town in Beirut Levels 1 and 2 are the most relevant and informative. The excavations in BEY 019, BEY 031, BEY 057 and BEY 081 have all revealed architectural vestiges belonging to the first millennium BC. All sites revealed walls and rooms belonging to multi-room houses that resemble the well-preserved houses in BEY 010. The sites are located in the area south of the Iron Age harbor. The location of the harbor in the zone northwest of the stronghold (see Fig. 1) is deduced from the presence of a stone-paved ramp in BEY 032. The problem of the ramp and its connection east and west is an interesting one, and can only be conjectured. As said before, we assume the ramp is leading to- H.H. CURVERS 63 wards the harbor in the west. In the east we assume that the ramp gives access to an area defined by the glacis of the Iron II and III periods. Observations in BEY 003 suggest that the glacis turned inward; we assume that the western glacis did the same. This would create a wide access to the stronghold. Goods brought up from the harbor found their way to the storage facilities within the enclosed center of the city. The “Casemate-Wall Building” was built either as a part of a new fortification structure filling the gap between the glacis or to expand storage space within the stronghold. A monumental access to the stronghold is furthermore supported by the out- line of the cliff, which we have only tentatively indicated in Figure 367. We assume the existence of Bronze Age and Iron Age architecture on this promon- tory, overlooking the harbor. Its location on the cliff provided a natural defense. Only the southern part of the site needed additional man-made fortifi- cations. Circulation patterns between the harbor, the lower town and the stronghold necessitated a wide but secured access at the western end of its for- tification system. BEY 119 and BEY 118 have revealed interesting information on the infra- structure of the 5th-4th century BC city of Beirut. In both sites canalizations were exposed. The canalizations consisted of two parallel walls of medium- sized natural cobbles. On top of the two parallel walls large rectilinear boul- ders covered the canal below. One canalization in BEY 119 clearly ran along an E-W street, another canalization seems to have originated within the build- ing. In BEY 118 a canalization of similar construction method as in BEY 119 was found in a narrow corridor. The limited exposure of the architecture does not allow for a clear interpretation of the corridor. It is either a space between two buildings or a corridor within one building. The canalizations crossed an outer wall and ran into the central canalization of again an E-W street. In comparison with the splendidly preserved architecture exposed in BEY 010 we can only conjecture about the reasons why the N-S streets in this so- phisticated urban quarter of ancient Beirut did not yield any canalizations. A fragment of a canalization was found in the E-W street at the south end of the sites. We have been able to trace its continuation inside a building to the east of BEY 010, but there the associated pottery and building elements suggest a later date. The excavations in nearby BEY 011, BEY 021 and BEY 090 did not reveal the presence of preserved canalizations. The exposed architecture in this northwestern corner of the lower town sug- gests that the site was located at a rather high elevation overlooking the harbor of ancient Beirut. BEY 010 has revealed complete plans of some multi-room houses, whereas we can only conjecture similar houses around the site. Be- sides its location and the method of construction, the infrastructure system of

67 M. F. Davie, “Maps”, Figs. 2-6. 64 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS this quarter also stands out among the architecture retrieved from excavations in the lower town. The infrastructure system collected the stormwater at least in two E-W streets located in the northern half of the quarter. In contrast to this E-W drainage sys- tem, the N-S streets were void of such a system. They probably drained the stormwater and domestically used water on their surfaces or in less substan- tially constructed canalizations or trenches. With the help of the natural slope of the zone the water was subsequently evacuated in the direction of the sea. As a similar pattern of canalization was observed in BEY 119, be it more fragmentary, we assume that the southern and western lower town had a drain- age system in its E-W streets that also collected the water from the N-S streets. BEY 105, BEY 113, BEY 126 and BEY 127 have yielded information that allows us to develop a general idea about the layout of the Iron Age- Achaemenid town. The remains in BEY 105 consist of small cubicles. Com- parison with the architecture in BEY 010 and BEY 019 underlines the limited dimensions of the cubicles. The architecture south of BEY 105 in BEY 031 is as irregular as the architecture exposed in BEY 105 but slightly larger. The architecture in BEY 119, due south of BEY 031, however, appears to be closer to the spacious and sophisticated architecture in BEY 010 and BEY 019. Associated with the cubicles in BEY 105 we also exposed a number of fea- tures that are missing from the architecture exposed in BEY 010, BEY 019 or BEY 119. In the cubicles or against the wall stubs of the cubicles, circular or rectilinear platforms of ca. 80 cm diameter and up to 40 cm high could have carried ovens. Furthermore a sequence of silos was found, probably in all phases surrounded by four walls. In the earliest phase of the sequence a deep silo was connected to a shaft which suggests that the shaft allowed access to the deeper (subterranean?) part of the silo (see Fig. 5). We assume that the si- lo’s upper structure (stone?) disappeared when the silo ran out of use and was replaced by the later silos. Another feature that supports the special nature of the site is only partially preserved. A rectangular bin with outlet in its southeastern corner was con- nected to a canalization provided with lime-plastered bottom and walls. The canalization widened out in a northern direction, where we were unable to find more related features. In BEY 126, located in buildings 200, 180 and 1155 – the eastern end of Lauffray’s Basilica Civica68 – we were able to retrieve vestiges belonging to the Hellenistic period and earlier. These building levels preceded the so-called Basilica Civica or, more likely, the colonnaded hall of the bath complex was constructed in the first century BC69.

68 J. Lauffray, “Forums et Monuments de Béryte”, BMB 7 (Beirut, 1944-1945), pp. 35-57. 69 For our new interpretation we owe much to Rueben Thorpe’s analysis of the BEY 045 ex- cavations, see also this volume. H.H. CURVERS 65

Besides a wall fragment belonging to the period under discussion, we ex- posed two circular pits both cut into the bedrock (see Fig. 6). Related to one of the pits we found a horizontally smoothed (working) platform cut out of the bedrock. The relation between the platform (slightly higher in elevation) and the pit is suggested by the gentle slope of the platform in the direction of the adjacent pit. The northern circular pit was provided with a central shallow de- pression and a ledge around its edge. The shallow depression resembles simi- lar features in the bell-shaped cisterns of Beirut. The ledge suggests that the pit was lined with stones and belonged to a higher level (possibly the level at which we found the horizontal platform). The presence of a post-hole due north of the pit also suggests that we are dealing with a feature constructed into the bedrock and in contemporary use with the earliest architecture in the area. In BEY 127, further to the southeast of BEY 126, the remains of yet another rock-cut circular pit were exposed below the Roman period northern section of the Cardo Maximus.70 In BEY 11371 fragmentary walls dated to the Iron Age do not allow for an interpretation of the site as purely domestic as BEY 010 and its surroundings or an interpretation as a working area as BEY 105. The remains of a rectangu- lar pit cut into the bedrock suggests the presence of a burial, although no bones were found and the feature therefore has not been included in the contribution on the cemeteries in Beirut.72 Near this rectangular rock-cut pit a circular pit was exposed. Around the edge of the circular rock-cut pit a circular wall of medium-sized cobbles suggests the presence of a semi-subterranean silo (see Fig. 6). Pottery retrieved from this site dates back as far as the 7th century BC. Its architectural stratigraphy suggests its use from that date till the Ottoman period. In BEY 066, located below Martyrs Square and its statue, rectilinear rock- cut pits were found with human skeletons still in situ.73 In a first effort to describe the general layout of the Iron Age city we have divided the city into the fortified oval stronghold, the harbor, the lower town and the cemetery (see Fig. 1).74 After a preliminary analysis of the remains excavated throughout the lower town of Iron Age Beirut (1995-1999), we can provide more details. The lower town was provided with a drainage system that both used the natural slope of the western part of the lower town to evacuate stormwater, and used domestic water.

70 J. Lauffray, “Forums”, Plan I, the unnumbered lot between building 1170 and building 400. 71 J. Lauffray, “Forums”, Plan I, BEY 113 is located in the area of the Souq Sursock, south of the forum oriental. 72 B. Stuart, this volume. 73 B. Stuart, this volume. 74 H.H. Curvers and B. Stuart, “BCD 1995”, p. 178, Fig. 4. 66 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS

The houses west of the harbor together with the more sophisticated rock-cut tombs to their west seem to represent the houses of the more well-to-do. Due south of the harbor storage facilities suggest the storage of grain. The platforms for ovens suggest a bakery, whereas the bin with plastered canaliza- tion suggests a possible second activity or an activity related to the bakery (rinsing of grain to float the hulls and other unwanted impurities or rinsing of fruits to be processed with the dough?). The circular rock-cut pits and the horizontal, smooth rock-cut platform sug- gest activities such as the cleaning of sponges or processing of murex shells. Clear evidence for both activities was not found, therefore these suggestions are mere hypotheses. In other excavation areas concentrations of crushed murex shells were found occasionally. In one case they were associated with a large reservoir or basin provided with a thick lime plaster coating. The date of the latter was Roman-Byzantine. The location of the rock-cut rectilinear pits with human bones suggests that in the south and southeastern zone around the city the less well-to-do inhabit- ants of the city were buried. In this zone also the city undertook the daily ac- tivities that resulted in the production of smoke and smell.

CONCLUSION75

The first analysis of the remains exposed during the 1993-1999 excavations in the have revealed a more detailed picture of the lay- out of the pre-classical town. Discoveries at the edge of the stronghold already suggested a closer link of the stronghold with the lower town (the ramp, see Fig. 2:14 and the stairway, see Fig. 2:17). We assume a temple, palace and other administrative institutions located within the stronghold. One access to this enclosed administrative center was by way of steps on the southeastern part of the oval stronghold. Another more monumental access is conjectured in a place where the eastern glacis and the western glacis turn inwards (see Fig. 2:15). The ramp connecting the administrative center with its central stor- age facilities and the western harbor is the only remnant of this gateway. The lower town of Iron Age Beirut knew a quarter of sophisticated houses at the higher slopes opposite the stronghold and overlooking the harbor. The harbor area seems to be characterized by the presence of both dwellings and working areas. In the west the outer zone of the lower town was predominantly in use as a cemetery, including rock-cut shafts and adjoining caves. The south- ern and eastern border zones of the lower town were in use as working areas

75 I thank Barbara Stuart for listening to all my proposals in an attempt to synchronize the discoveries on the Tell and lower town, and last but not least for editing the contents of this essay. H.H. CURVERS 67 and occasionally as burial grounds. One can imagine that the processing of murex shells, the cleaning of sponges, the storage of grain happened here. This zone was located favorably regarding the prevailing western winds in Beirut. Both the location of the various quarters and their characteristic functions and the use of the natural topography for the sewage system suggest that the inhab- itants of Iron Age Beirut were well aware of how to combine nature, climate and topography with the needs of Iron Age life. 68 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS

Fig. 1. The acropolis of Beirut (= Tell Ancien). H.H. CURVERS 69

Fig. 2. BEY 114: West section with Pre-glacis, Glacis and Post-Glacis features. 70 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS

Fig. 3. Northwestern quarter of the Lower Town of Beirut. H.H. CURVERS 71

Fig. 4. Southeastern quarter of the Lower Town of Beirut. 72 THE LOWER TOWN OF BEIRUT (1200-300 BC): A PRELIMINARY SYNTHESIS

Fig. 5. Features exposed in the working areas of the Lower Town.

Fig. 6. Settlement structure of Iron Age Beirut.