Semitist and Islamist, called “one of the greatest Orientalists of our times” (Moscati, 1968, p. i). Levi Della Vida’s main works include the publication of Syriac and Arabic manuscripts, the edition of important Western Semitic inscriptions (e. g., from Karatepe in Turkey and all the Punic inscriptions from Tripolitania, North Africa) and historical syntheses of tire ancient Semitic and Islamic worlds. His analytical works display profound erudition, subtle intuition, and an exemplary philological metlrod; in his work as a historian he combined positivistic methodology with tire idealistic tendencies of contemporary philosophy and historiography.
Levi Della Vida had been a pupil of Ignazio Guidi at the University of Rome. His collaborator, as early as 19ii, in his translation of the Arabic sources for Annali dell’ Islam was Leone Caetani. In 1913 Levi Della Vida became professor of Arabic at the Istituto Universitario Orientale in Naples and in 1916 was a professor of Semitic languages at the University of Turin; in 1920 he succeeded Guidi in teaching Hebrew and Semitics at Rome. In 1931 he refused, with only eleven otlrer Italian professors, to swear allegiance to the fascist government and was consequently removed from Iris office. From 1932 to 1939 he was employed in the Vatican Library, charged with studying Arabic and Syriac manuscripts, but as a result of political and religious persecution, he was forced to leave Italy. He went to the United States, where he taught Semitics at tlie University of Pennsylvania until 1945 (and again in 1946-1947). At the end of World
War II he remrned to Rome, where he taught tlie history of Islamic institutions until 1956, when he retired. He was a member of the most prominent academies in Europe and America; he received a doctorate honoris causa at the universities of Paris (Sorbonne), Algiers, and Jerusalem, and he received Lidzbarsld’s prize for Semitic epigraphy. His name and image appear on a medal awarded by the University of California to smdents of Islam.
BIBLIOGRAPHY
Garbini, Giovanni, ed. Giorgio Levi Della Vida nel cenunario della na-scita, i886-ig6j. Studi Semitici, n. s. 4. Rome, 1988. Includes a bibliography of works by Levi Della Vida.
Gordon, Cyrus H. The Pennsylvania Tradition of Semitics. Atlanta, 1986. See especially pages 62-64.
Michclini Toed, Franco. “La storiografia orientalistica in Italia fra otto e novecento.” In La cuitura slorica italiana ira olio e novecento, edited by M. Martirano and E. Massimilla, vol. 2, pp. 113-144. Naples, 1991.
Moscati, Sabatino. Ricordo di Giorgio Levi Della Vida. Orientis Antiqui Collectio, 7. Rome, 1968. With a bibliography of Levi Della Vida’s works.
Tessitore, Fulvio. “Giorgio Levi Della Vida nella storiografia italiana tra otto e novecento.” Introduction to Giorgio Levi Della Vida,/lraii ed Ebrei nella sloria, edited by Francesco Gabrieli and Fulvio Tessitore, pp. 11-56. Biblioteca di Saggistica, 17. Naples, 1984.
Maria Giulia Amadasi Guzzo
LIBRARIES AND ARCHIVES. The emergence of writing in the ancient Near East, shortly before 3000 bce, resulted in two kinds of texts, as the Mesopotamian evidence shows: groups of records drawn up for administrative purposes and texts used in training scribes. The former are usually designated archival texts; the latter, commonly called school texts, in time formed the contents of libraries.
Archival texts were drawn up by scribes/managers as an aid to memory and proof of transactions, to take stock of goods and personnel, to register transfers, and to meet tlie demands of accountability. Eventually, contracts, letters, and a variety of reports and notes were included, written to control economic and political processes and kept for consultation. Such collections, uneartlted during excavations, are called archives. School texts, reflecting tlie complexity of cuneiform script, comprise a variety of sign and word lists together with student exercises. [See Cuneiform.] In the middle of the third millennium school texts were joined by tire earliest literary texts (e. g., hymns, incantations, proverbs), written by and for the scribes responsible for the archives and usually found in tlie same locales. They attest the scholarly scribal activity that eventually developed libraries: collections of nonadministrative, scholarly, and traditional texts written and collected for educational purposes and professional needs. [5ee Scribes and Scribal Techniques.] They keep turning up in houses (schools and dwellings of master scribes), temples, and even palaces, always connected with tlie literati who collected and used tlrem and who usually lived or worked in the places the texts were found.
Archives. The definition of archives as collections of records accumulated during the period of a particular administrative task performed by a person or institution may be too broad. Some would add to it “performed ex officio and still present with those who made and used tlaem.” Others would even wish to reserve tlie term for “a collection or repository of records no longer in use but preserved for their historical value and stored under special conditions.” In the ancient Near East, such restrictions were usually ignored. Coherent bodies of administrative and legal records from large private households are in principle no different from those created by officials of a palace or temple—both of which also functioned as a large household. [See Palace.] Many archives have been found in situ, left behind when a disaster struck the building in which tliey were kept and used. A convincing example is the archive in room L 2769 of palace G at Ebla (c. 2400 bge) . [See Ebla.] Thousands of tablets had been stored on wooden shelving along walls that collapsed. The room was only accessible from adjoining room L 2875, a typical scriptorium, with low benches along the walls on which tablets could be laid out. [5ee Tablet.] Similar discoveries have been made elsewhere: in tlie houses of Assyrian traders in Kanes (Anatolia, nineteenth century bge), in the residence of Ur-Utu at Sippar/Amnanum (Tell ed-Der, destroyed c. 1630 bge), and in the house of Shil-watesshub at Nuzi (house A on the northeastern mound, destroyed c. 1330 bge). [5ee Kanes; Sippar; Der, Tell ed-; Nuzi.]
Many complete or partial archives have been found in secondary locations: old records, having lost their value, might be discarded or put to secondary use as fill for walls and mud-brick benches or for leveling floors (not surprising in a country without building stone). Reconstruction and terracing suggest that parts of older buildings were displaced together with their contents, including archives. The bulk of the archival texts of the Temple of Inanna at Nippur (Ur III period, twenty-first century bge) were discovered in distinct pockets in the fill of the foundation platform of a temple (?) erected there two tlrousand years later. [See Nippur.] Hundreds of records of the same period were discovered at Ur, stacked five to six layers deep under the pavement of a building dated to the Kassite period (c. 1300 bge). [See Ur; Kas-sites.] Archival texts from the period of Assyrian domination of Mari were used as fill for mud-brick benches built in the palace of the next king, Zimrilim (eighteenth century bge). [See Mari.] The so-called Persepolis fortification tablets (Achaemenid period) by tlieir very name perpetuate their secondary use as fill for walls. [See Persepolis.]
Remains of archives and coherent files have even been recognized among tlie oldest administrative records known, those from level IV at Uruk (before 3000 bge), where tliey had been deposited as part of the general debris in trash pits or used as architectural fill. [See Uruk-Warka.] Such hoards may be tteated as archives if they exhibit enough coherence, but as random parts of larger collections, they must be used with caution for historical reconstructions. The best example of an archive as a respository of records no longer in use was found in the archival building of the “tablet hill” at Girsu (Telloh). [See Girsu and Lagash.] The archive originally contained some thirty tliousand tablets dating to the end of the third millennium, they were carefully filed, in chronological order, in five or six layers on benches in two adjacent galleries. Such huge archives and repositories must have been restricted to large institutions, where, it may be assumed, administrative continuity and storage facilities were accompanied by qualified scribes who respected old records.
Even when violent destruction disturbed an original locale and scattered tablets, tire findspot and the contents of the records can help to identify the source. Records dealing mainly with land transfer, dating to the fourteenth cenmry BCE, were found in debris filling tire area of court IV in the southern part of the palace at Ugarit (destroyed c. 1180 bge) . Found in the higher levels of the debris, they proved to have fallen from an upper story and had to be distinguished from records dating to the second half of the thirteenth century BCE found in the same area near the floors (particularly in archive rooms nos. 30 and 31). Upper stories could of course also have held current archives, as was the case at Ma§at Hoyiik (Tapigga, with a Hittite palace), Mari, Nineveh, and Hattusa/Bogazkoy (in building E of tlie citadel, where they had fallen down the slope); [5ee Nineveh; Bogazkoy.]
Another source of confusion is archives having been transferred in antiquity. Tablet labels with Babylonian dates show that tlie diplomatic correspondence of Zimrilim of Mari was moved to a room (no. 115) adjoining the palace’s interior court, where it was inspected by the scribes of the conqueror Hammurabi. In tlie wake of building new capitals or palaces, tlie chancery archives of Assyrian Idngs were moved, as was deduced by Simo Parpola (1983) from the sometimes remarkable location and composition of certain archives. Such operations imply selection because tlie collections exhibit gaps.
All the examples mentioned thus far concern cuneiform archives. There is no reason to doubt that these observations are also valid for archives consisting of records on papyrus, leather, or wood. [See Papyrus.] Destruction (with conflagration) and climate have almost completely robbed history of such collections. They can only be inferred from the existence of tlie clay sealings or bullae originally attached to certain records that have survived, [See Seals.] This applies also to the hundreds of wax-coated wooden writing boards (codices) used in Mesopotamian chanceries and libraries, whose existence is known from colophons and library records. Hoards of bullae with royal seals have been discovered at Haltusa/Bogazkoy, in the so-called western building at Ni? antepe, in building D at Buyiikkale, and in the City of David, Jerusalem, once attached to official documents. Sealed bullae, discovered in Judah from about 600 bce are considered to be remnants of a burned archive (Avigad, 1986), and the numerous bullae found in an archival chamber with extensive shelving, room 61, of Sennacherib’s palace at Nineveh are the remains of an archive containing records on papyrus originating in Egypt and the Levant.
Only under special circumstances have parts of archives not written on clay survived. From ancient Soutli Arabia records in the shape of inscribed wooden sticks are loiown that survived in the dry climate (Ryckmans et al, 1994). The so-called Samaria papyri (fourtlt century bce) were discovered in a cave in Wadi ed-Daliyeh near the Jordan River, (Cross, 1969) and several hoards, including the Babatha archive (c. too Ce), carefully packed in a leather bag, have been discovered in caves west of the Dead Sea (Yadin, 1971). [5eeDead Sea Scrolls.] Such hoards had been carried to refuges in inhospitable and dry places. A few hoards of inscribed potsherds (ostraca) are also known that could be called archives: the Samaria ostraca (eighth century bce), file Lachish ostraca (c. 600 bce), and the Arad ostraca (i. e., the archive of Elyashib, c. 600 bce). [See Samaria Ostraca; Lachish Inscriptions; Arad Inscriptions.] The first consist of administrative bookings (later digested in a papyrus ledger?); the second and third comprise small groups of letters. The following observations apply almost exclusively to clay tablets because of the lack of substantial archives of noncuneiform texts found in situ.
Storage and filing. A small, private archive could be stored in a jar or basket in a residence. Large institutions had special archival rooms for the systematic storage of records, Large archives, with valuable and sensitive records, were kept in sealed rooms and admission was restricted (ex officio or by mandate). Special committees were appointed to enter fiie sealed rooms of absent or dead Assyrian traders in Kaneg (Veenhof, 1986, p. 12). The location of archival rooms was often dictated by practical considerations: archives for registering regular deliveries and expenditures near fire entrance of a palace (e. g., the eastern and western archives of the palace at Ugarit); the archive of “the Icing’s meal” in die kitchen area of the palace (e. g., at Mari); a chancery archive near the rooms where the Idng and his counselors met; tablets with international treaties in the archival room of a temple, where they had been deposited before the gods who had to guarantee them (e. g., at Hattusa), [&e Ugarit.] This implies that large institutions usually had several archival rooms for different spheres of domestic and international activity: agriculture, industry, husbandry, commerce, the military. At Archaic Urulc (Eanna, level IV, area of fiie Red Temple) “substantially coherent and discrete administrative. .. archives, that is, tablets from an individual accounting, , , unit” have been observed (England, 1994, p. I4ff.). Practical considerations may explain archival decentralization—for example, in the case of production centers (for textiles, leafiier, metals, and even agricultural production) outside temples and palaces, such as the Isin Craft Archive (c. 2000 bce, with nearly a thousand records). This may account for the absence of archives in Early Dynastic temples (however, fitat may also be because such patrimonial organizations knew only a limited separation of official and private spheres, and thus administrative archives are to be found in the houses of its high officials).
Three methods of storage are usually distinguished: a container system, an open-shelf system, and a pigeonhole system. Many written sources mention jars, baskets, boxes out of wood or reed, trays and leather bags. The first are found regularly in excavations. Some containers even have an inscription identifying the contents, as was the case with some of the ten jars holding file archive of the regular offerings of the Assur temple (c. i too bce, found in storeroom 3', under the responsibility of Ezbu-leshir). Traces of baskets and boxes are more elusive, but their use can be deduced from the remains of wood (Gasche, 1989, p. 3off.; Stein, 1993, p. 24ff) and metal (copper nails in the archival room of Shilwa-Tesshub at Nuzi). The use of shelves is well attested in Mesopotamia and elsewhere (at Ugarit and Hattusa; in buildings A and K at Buyiiklcale [Veenhof, 1986, p. 13]). Along tliree walls of the main archival room in tlie palace at Ebla were clear traces of where triple shelving had been fixed to file walls. Mud-brick benches for storing containers holding tablets may have been an alternative to shelves and were suitable for laying out tablets for inspection. Room 17 of Ur-Um’s residence (see above) was equipped with benches. The pigeonhole system was used in the library room of the Nabu temple at Kltorsabad, in Sennacherib’s palace at Nineveh (room 61, used for papyrus records), and in the NeoBabylonian Temple of Shamash at Sippar (room 355, wifit fifty-six niches built of mud brick, wood, and reeds, in three tiers along three walls). Less sophisticated were brick boxes used as filing cabinets (in fite scribal office of the northwest palace at Kalhu).
Tablets in large archives could be classified by format, size, and style: there are links between the physical appearance and the contents of records (e. g., multi - or single columned, square, round or oblong, with or without a sealed envelope). In the Ur III period, for example, so-called round tablets are used in Lagash for recording the assessment of fields and (their) expected yield (Veenhof, 1986, p. 14). From fire seventh century bce onward, key words, summaries, dates, and names were written in Aramaic on the edges for quick identification. Containers and stacks of tablets could be identified by their labels, in use since the third millennium and particularly common during the fitird dynasty of Ur. The labels were usually attached to baskets, trays, and boxes with tablets called pisan-dub-ba, “tablet” (dub) basket (pisan) in Sumerian (Veenhof, 1986, p. i6ff.). The texts written on tliem identify the subject, transaction, occasion, date (they frequently caver an accounting period of one montli or one year), and place of the group of records. Unfortunately, almost all of them (their number approaches five hundred) lack an archaeological context, so they cannot be linked to particular files. The variety of forms bears witness to systems with different, at times ad hoc criteria of classification. Some labels, moreover, may be transport rather than filing or storage labels. Clay bullae, with seal impressions and short inscriptions are also used. Old Assyrian traders used them also. to identify records by provenience, format (sealed or not), type (letters, copies), and subject matter, frequently adding personal names.
Contents. Archives differ widely as to scope, size, and institutional setting. Archives of great institutions may cover various areas and levels of administrative activity, such as production (agriculture, husbandry, crafts), trade, foreign relations (gifts, tribute, booty), jurisdiction, personnel management, the army, cult. The huge archive at D rehem (Puz-rish-Dagan, Ur III period), however, only registered, painstakingly, the delivery and disbursement of livestock in the public sphere. The so-called customs archive from Mari consisted only of letters of clearance received by traders entering Mari’s territory upon payment of an import duty. The letters were submitted to the overseer of the traders upon tire latter’s arrival in the city. Private persons and families usually kept a single archive, covering many activities. This could result in large holdings, with at times more than a thousand records (Old Assyrian traders in Kanes [Veenhof, 1986, p. 9, n. 33], the Tehip-tilla family in Nuzi [Maidman, 1979], the Murashfi firm at Nippur [Stolper, 1985]). Large archives of high officials can be a mixture of records received or kept ex officio and those bearing on private affairs. That of Shamash-hazir, an official of King Hammurabi of Babylon in Larsa charged with the allotment and administration of crown land, contained more than a hundred letters of tire Idng’s, as well as his wife’s, correspondence. The business reflected in tire huge archive of the chief lamentation priest (Ur-Utu) of the goddess Annunitum at Sippar/Am-nanum (see above) also reflects his responsibilities as administrator of the temple. Archives of expert scribes and scholars, risen to the ranks of senior officials, may contain private archival records in combination with public records and a personal library of scholarly texts, as was tire case witlt that of Rap’anu at Ugarit.
Administrative records of institutions, which aim at controlling the movement of goods and persons, frequently exhibit standardized formats, styles, and layouts. They comprise both a small tablet for registering single, daily events and monthly and yearly summaries and balanced accounts—ledgers drawn up for the purpose of accountability. The transactions recorded are deliveries, transfers, expenditures, distributions, and accounting. There are always lists of personnel (receipt of rations, wages, fields), of material in stock (inventories), of special expenditures (offerings, gifts), as well as letter orders (as proof of disbursements). Such administrative archives usually cover a limited period—one or two generations at most—with a clear numerical preponderance of records of the last few years before their abandonment or destruction. Many records, especially of daily transactions, after having been consulted could be discarded, stored separately or “erased” (with red lines over tire tablets, as at Mari); legal documents (notes of debt) in due time could be “broken” or “Idlled” (in the language of the texts). Rules for discarding documents as such are not Imown and may have varied from one administration to another. Certain records, such as a comprehensive yearly account, cadastral texts, and some legal documents, were kept longer, for later reference. The size of a public arcltive is also conditioned by the measure of bureaucratic control, which of course was never complete and showed remarkable fluctuations from deliberate measures and historical developments. Increased standardization and central control are already attested during tire Old Akkadian period (Naram-Sin) and the Ur III period (from the middle of the reign of Shulgi, c. 2050 BCE).
A royal chancery archive, such as tlrat from the Neo-Assyrian period, had a wide scope: political correspondence, legal documents (treaties, grants, conveyances), records concerning the military (conscription, booty, equipment), official religion (expenditures for the cult, building of temples), and the court (royal family, harem, courtiers, officers, specialists, lists of disbursements of luxury items such as wine, oil, and garments). A variety of reports supplied information on military campaigns, conquered areas, production and works, and on ominous features relevant to tlie royal house (signs on earth and in the heavens, dreams, prophecies). Such palace archives could also contain private records of palace officials, either because tliey kept them in their offices, or because the king wished to exercise control over powerful and potentially dangerous individuals, such as officers of the guard, royal charioteers, and tliose controlling the harem (Parpola, 1983). Royal archives usually also contain selections of historical state documents, such as earlier royal correspondence. Thus, such archives could take on the character of a royal reference library.
Institutional and state archives are a mine of information for file historian interested in recovering a past administration’s main features, discovering how power was wielded, and the stattis, duties, and careers of tlie main officials. Dated records allow tlie reconstruction of historical events, and the statistical evaluation of large archives make quantitative approximations possible. The king’s obligations toward the cult and the temples make such archives a source for the study of religion and of the king’s role in it.
Private archives contain a variety of records reflecting the activities of their owners as traders, landowners, priests, and officials. Among these tltere are nearly always letters, botli those received in function (administrative duties, contractual links with the authorities) and those of a private nature. Letters were essential for overland traders (see above) and make up half of the records of tlie Old Assyrian traders. Because of their informative and virtual legal value (sealed message tablets contained orders and authorizations) archival copies were kept. In Babylonia, recipients of such letters were advised “to preserve them as proof of my order” and the same may be true of the Sumerian letter orders of tiie Ur III period. [5ee Sumerian; Sumerians.] Private archives always contain a number of long-term legal documents, such as marriage, adoption, and inheritance contracts and especially title deeds of real property. They were carefully preserved in their sealed envelopes and may go back several generations (some even more than two hundred years). They were tire core of the small archive of Silli-Ishtar from Kutalla, which he left wrapped in reed mats and covered by mud bricks when he had to flee from his house in about 1740 BCE, never to return. When Ur-Utu’s house in Sippar/ Amnanum was destroyed by fire, someone ttied to salvage a box of such records but lost them on the way out and they were found before the door to the court (Gasche, 1989, p. 32 with pi. 12.i).
Lawsuits over property rights were occasions to consult and produce tablets, and verdicts confirming tire rights they embodied became part of such archives. When lost, they could be “revived” and rewritten on tlie basis of oral testimony before the judges. When real property was sold, such earlier title deeds—which contained valid descriptions of the property—had to be handed over to the new owner. This explains the presence of seemingly out-of-place records found in private archives. They might also have contained contracts ceded, given as pledge, or deposited there for safekeeping by friends. Also discovered in private archives are small hoards of records belonging to junior members or employees of a household (who had no houses of their own). Witliout careful excavation and recording (many private archives have been dug illicitly), such data and links are lost. Excavated archives tlrat did not undergo conservation procedures (numbers lost), were poorly published (ignoringthe links between textual and archaeological data), or were divided among various museums (e. g., archives from Assur, Nippur, Nuzi, Ur) are responsible for the loss of essential information.
Well-analyzed private archives are a mine of information for social history. The large variety of legal contracts and judicial records are the main source of what is known of ancient law. Official archives as a rule contained few contracts; only in certain periods and cases (notably records dealing with tlie transfer of real estate, especially when encumbered with servicies and duties) were private contracts registered or abstracted by the authorities. Official courts of law occasionally kept copies of verdicts (e. g., the collection of tablets with final verdicts passed by city governors in the Ur III period). However, the general rule was that those verdicts and the files to which they belonged would be given to the winning party to store. The enormous variety of letters sheds light on many aspects of daily life, without which what is known of family life, trade, and personal religion would have been rather scant. Such letters, moreover, are a prime source for studying ancient languages, especially idioms.
Libraries. Collections of scholarly and traditional texts written and used by scribes during and after their training developed into libraries. The texts were also used in training pupils, which explains why such collections, especially in early periods, are frequently found in or near the archival rooms where the scribes worked (e. g., at Shuruppak/Fara, Abu Salabikh, Ebla). [5ec Fara; Abu Salabilch.] Expert scribes, witli specific duties and interests and perhaps more free time, eventually began writing down or composing additional texts to serve their personal needs, such as ritual, divinatory, liturgical, medical, and astrological texts and incantations not part of the school’s standard curriculum. Some worked as master scribes in a palace (chancery) or temple, others earned tlieir living by running a school in their house, where a mixture of school and professional texts and private archival records is usually discovered. Schools in a capital city may have had links with the chancery (where official inscriptions and royal hymns must have been composed). This may explain why some royal texts (hymns, letters, inscriptions—especially of the kings of Ur III and Isin) found their way into tlie curriculum. When Sumerian disappeared as a spoken and soon after as a written language (after 2000 bce), it remained indispensable for scholars as the basis of the writing system. [S’ee Writing and Writing Systems.] Schools and scholars devoted a great deal of attention to copying traditional Sumerian texts, which led to the growth of a land of standard corpus, eventually arranged in a normative sequence. This corpus, registered by the first lines of compositions in various catalogs, and to a varying degree actually discovered in schools (especially in Nippur), can be considered a Idnd of standard library. Texts outside this corpus, written and kept for specific professional needs (e. g., collections of incantations or omens) could be added.
Aldtadian compositions (epic and mytliical texts, prayers, wisdom texts, historiographic compositions) were composed, written down, and copied in these schools, but what is luiown of this Aldtadian corpus as part of school and scholarly libraries is more limited. [See Altkadian.] In the following centuries many such libraries are known, usually in the houses or work places of master scribes. Convincing examples have also come to light at Ugarit, discovered in private houses, such as tliose of Rap’anu (see above), of “tire high priest” (Mesopotamian lexical texts, Hurrian texts, mytlrical and epic texts in the alphabetic cuneiform script of Ugarit, written by the scribe Ili-malku), and of “the
Human priest” (Hurrian and Ugaritic texts in alphabetic script). [See Hurrianj Ugaritic.] Such libraries are also found in temples, where the scholars who wrote and used them worked (e. g., Temple Mj at Emar/Meskene on the Middle Euphrates, c. 1200 bce). [See Emar.] The many divinatory and incantation texts it contained reveal the specialization of its owner, who also had a small sample of literary texts (wisdom. Epic of Gilgamesh, a ballad). The same pattern is observable in Babylonia after the end of the second millennium BCE. The traditional corpus was by then more or less fixed (the term canonized has been used in this context). Many collections discovered in houses and rooms of temples contain a sampling of this standard corpus, together witlr more specialized texts (e. g., incantations, divinatory and medical texts). Good examples are the library of the incantation priest in Assur, the one found in a house in the Assyrian town of Huzirina (Sultantepe, four hundred different texts), and that of Anu-belshunu, a scholar attached to the Bit-Resh Temple of Seleucid Uruk. [5ee Assur.] The closest approximation of a temple library (without an admixture of archival texts) was discovered in 1986 in room 355 of the Shamash temple in Sippar (c. 500 bce), stored in more than fifty niches in the walls; however, its full size remains unknown (Al-Jadir, 1987).
The evidence for palace libraries is very limited. The postulated library of Tiglath-Pileser I in Assur (c. 1100 bce) has been shown to be the collection of an Assyrian scribal family from the first decades of tire twelftlr cenuiry bce. The palace chancery may have incorporated in part some important older state documents (palace edicts, perhaps also the Middle Assyrian laws), but that does not make it a library. That important documents were carefully preserved, perhaps as separate collections, is not surprising. A collection of treaty tablets was stored in the great temple at Hattusa, probably because tlrey had been deposited there before the gods who had to watch over tlrem. In the same city, in building A of the citadel house, an important collection of older judicial decisions and royal edicts was found registered in a catalog, perhaps in the royal reference library.
The only real palace library, and perhaps the only library tliat fully deserves the name, was created by King Ashur-banipal in about tire middle of the seventh century bce in Nineveh. It was discovered and excavated by Austen Elenry Layard and his successors between 1850 and 1932. [5ee the biography of Layard.] From royal letters and colophons on tablets it is known that the king wished to collect all scholarly and religious texts considered important. Ele may have built on an already existing royal collection, but his own contributions were very substantial. Scribes were sent out to collect and copy tablets and, after his victory over Babylonia, hundreds of texts (both on clay tablets and on wax-coated writing boards, including polyptychs) were confiscated from houses and temples. Tablets copied for the library were inscribed with special colophons (some in ink, on existing tablets) that state tire king’s intentions and identify tlrem as royal property. All of these texts are now in tire British Museum, as the K(uyunjik) Collection. The total number of different tablets (many compositions are represented by several—at times up to five or six—copies) has been estimated at around fifteen hundred, with perhaps two hundred thousand lines of text. The collection comprises the whole of the canonical corpus, the stream of tradition.
Catalogs (also those listing all the individual tablets in a lengthy series, with tablet numbers and catch lines) and labels made the library accessible. The majority of tablets consist of magical, medical, divinatory, and ritual texts, indicating that it focused on what was “good for the kingship,” especially texts that would protect tire king and his family against evil. There is also a substantial group of literary texts, however. Notwithstanding all this information supplied by the lung’s scribes, it is still not easy to reconstruct the library: in the nineteenth century the rich harvest of texts from at least two palaces and a temple seems to have been mixed up: the southwest palace (rooms 40, 41), built by Sennacherib in about 700 bce, Ashurbanipal’s own northwest palace (the successor to the older Succession Palace), and the Temple of Nabu. In addition, there was a certain amount of mixing of archival records with library texts. It seems tlrat the king considered certain royal archival documents wortli preserving alongside scholarly and literary texts. This means that, although there is no doubt tlrat Nineveh contained a large, specifically royal library, the border between royal archive and royal library must have been somewhat fluid— how fluid can only be established if the originally separate royal text collections can be distinguished. A thorough analysis of the libraries of Nineveh remains a task for the future.
BIBLIOGRAPHY
Al-Jadir, Walid. “Une bibliotheque et ses tablettes.” Archeologia 224 (1987): 18-27.
Aniaud, Daniel. Recherches au pays d’Astata. Emar VII4: Textes de la bibliotheque. Paris, 1987. Full edition of all texts found in temple M,, as a sample of a library of a priest-specialist.
Avigad, Nahman. Hebrew Bullae from the Time of Jeremiah: Remnants of a Burnt Archive. Jerusalem, 1986.
Charpin, Dominique. Archives familiaks et propriite privee en Babylonie ancienne: Btude des documents de "Tell Sifr.” Geneva, 1980. Paradigmatic analysis of a small private archive of the Old Babylonian period discovered in ancient Kutalla.
Courtois, Jacques-Claude. “Ras Shamra (Ugarit). I. Archralogie.” In Suppliment au Dictionnaire de la Bible, fasc. 52-53, cols. 1222-1280. Paris, 1979. Illustrated overview of the various archives and libraries discovered at Ugarit.
Cross, Frank M. “Papyri of the fourth Century B. C. from Daliyeh.” In Neto Directions in Biblical Archaeology, edited by David Noel Freedman and Jonas C. Greenfield, pp. 41-62. New York, 1969. Englund, Robert K. Archaic Administrative Texts from Uruk. The Early Campaigns. Archaische Texte aus Uruk, Bd. 5. Berlin, 1944. Edition and analysis of the earliest cuneiform records, from c. 3200 bce. Foster, Benjamin R. “Archives and Record Keeping in Sargonic Mes-
Opotamia.” Zeitschrift fiir Assynologie 72 (1982): 1-27. Basic analysis of early archival practice, with good evidence on filing systems.
Gasche, Hermann. Im Babylonie au lye siicle avant notre ere: Approche archeologiqtie, probUmes el perspectives. Ghent, 1989. Chapter 2 describes and analyzes the large and carefully excavated archive of Ur-Um at Tell ed-Der.
Gibson, McGuire, and Robert D. Biggs, eds. The Organization of Power: Aspects of Bureaucracy in the Ancient Near East. Studies in Ancient Oriental Civilizations, no. 46. Chicago, 1987. Pays ample attention to administrative archives as instruments of bureacracy, especially during tlie Ur III period in Mesopotamia.
Hunger, Hermann. Babylonische und assyrische Kolophone. Alter Orient und Altes Testament, 2. Kevelaer, 1968. Presents and analyzes the information on the tradition and copying of literary cuneiform tablets incorporated in libraries.
Jones, Tom B. “Sumerian Administrative Documents: An Essay.” In Sumerological Studies in Honor of Thorkild Jacobsen on His Seventieth Birthday, June 7, 1974, edited by Steven J. Lieberman, pp. 41-62. Chicago, 1976.
Krecher, Joachim. “Kataloge, literarische.” In Reallexikon derAssyriol-ogie und vorderasiatischen Archdologie, vol. S, pp. 478-485. Berlin, 1976-
Lieberman, Steven J. “Canonical and Official Cuneiform Texts; Towards an Understanding of Assurbanipal’s Personal Tablet Collection.” In Lingering over Words: Studies in Ancient Near Eastern Literature in Honor of William L. Moran, edited by Tzvi Abusch et al., pp. 305-336. Atlanta, 1990.
Maidman, M. P. “A Nuzi Private Archive: Morphological Considerations.”I (1979): 179-186.
Nissen, Hans J., et al. Archaic Bookkeeping: Writing and Techniques of Economic Administration in the Ancient Near East. Chicago, 1993. Highly informative presentation, in word and image, of the earliest records for describing and interpreting the earliest administrative systems (third millennium bce) .
Oppenheim, A. Leo. Ancient Mesopotamia: Portrait of a Dead Civilization. Chicago, 1964. The introduction and chapter 5 deal witli the scribes, tlieir creative efforts, libraries, and recording.
Parpola, Simo. “Assyrian Library Records.” Journal of Near Eastern Studies 42 (1983): 1-30. Presents the records of acquisitions from.'; Ashurbanipal’s library in Nineveh from confiscations in Babylonia.
Posner, Ernst. Archives in the Ancient World. Cambridge, Mass., 1972. Describes and analyzes the data from tire ancient Near East in the wider framework of what is known from antiquity.
Ryckmans, Jacques, W. W. Muller, and Y. ‘Abdallah. Textes du Yemen Antique inscrits sur bois. Louvain-la-Neuve, 1994.
Shawe, J. “Der alte Vorderorient.” In Handbuch der Bibliolhekwissen-schaft, edited by F. Milkau and G. Leyh, vol. 3, pp. 1-50. Weisbaden, 1955-
Sigrist, Marcel. Drehem. Bethesda, 1992. Overview and analysis of the largest administrative archive (dealing with livestock) of the Ur III period.
Soldt, W. H. van. Studies in the Akkadian of Ugarit: Dating and Grammar. Alter Orient und Altes Testament, 40. Neukirchen-Vluyn, 1991. A meticulous reconstruction and analysis of tlie archives and libraries discovered at Ugarit (see pp. 47-231).
State Archives of Assyria, to vols. to date. Helsinki, 1987-. Full edition of file Assyrian state archives discovered at Nineveh, witlt introductions and translations, including royal correspondence, astrological reports, oracular queries, treaties and loyalty oaths, prophecies, legal transactions, administrative records, and court poetry.
Stein, Diana L. DasArchiv des Silva-Tessub, vol. 8, The Seal Impressions. Wiesbaden, 1993. Overview of the various archives discovered at Nuzi as introduction to a detailed analysis of one archive, with the use of texts and seal impressions.
Stolper, Matthew W. Entrepreneurs and Empire: The Murasu Archive, the Murasu Firm, and Persian Rule in Babylonia. Istanbul, 1985.
Veenhof, Klaas R., ed. Cuneiform Archives and Libraries: Papers Read al the 3oe Rencontre Assyriologique Internationale, Leiden, 4-8 July 19S3. Leiden, 1986. Papers devoted to many archives in the ancient Near East by some of the best specialists in the field, with a general introduction and much bibliographic data.
Weitemeyer, Mogens. “Archive and Library Technique in Ancient Mesopotamia.” Libri 6 (1956): 217-238. Careful study by a professional Assyriologist and librarian.
Yadin, Yigael. Bar-Kokhba: The Rediscovery of the Legendary Hero of the Last Jewish Revolt against Imperial Rome. London, 1971. The discovery and contents of the remains of papyrus archives in caves west of the Dead Sea.
Zettler, Richard L. The Ur III Temple of Inarma al Nippur. Berliner Beitrage zum Vorderen Orient, 11. Berlin, 1992. Comprehensive analysis of an early temple on the basis of its archives.
Klaas R. Veenhof
LIBYA. Situated at tlie northeast corner of Africa, Libya consists of tliree regions: Cyrenaica, Tripolitania, and Fez-zan (ancient Phazania). Cyrenaica and Tripolitania are semiarid Mediterranean coastal zones on opposite sides of the Gulf of Syrte. Arid Fezzan, dotted with oases, separates them at the Sahara’s nortliern edge.
For the ancient Greeks, Libya was all of northern Africa between the Nile River and Gibraltar; to them, its light-sldnned inhabitants were all “Libyans,” though tliey distinguished subgroups such as Garamentes, Maures, and Nu-midians. Phoenicians traded with Libyans beyond the Pillars of Hercules (Herodotus 4.196). [5ee Phoenicians.] The Per-iplus of the Erythraean Sea of Hanno the Carthaginian, a partly fanciful, partly factual account of a sea voyage to “Libyan lands beyond the Pillars of Hercules” (presumably tlie Atlantic coast of Morocco), mentions Libyphoenicians, punicized native peoples. [5ee Phoenician-Punic.] The name also bears tenuous connection with biblical Lehabim (Table of Nations, Gn. 10:13) and Lubim (2 Chr. 12:3).
The Greeks may have learned the name from the Egyptians, for whom, from the Old Kingdom onward, Libya formed a hostile western desert frontier, In his mortuary complex at Abu Sir, Sahure (fiftli dynasty) boasted a victory over Libyans, The Story of Sinuhue (twelftli dynasty) mentions Tehune-Libyans vanquished by Egypt. Rameses II (nineteentli dynasty) built a chain of forts in die western desert against Libyan incursions. On reliefs at Medinet Habu, Rameses’ son Merneptah recorded his defeat of a coalition of Sea Peoples and Libyans (written RBW or LBW, vocalized Lebu or Libu). During tlie twenty-second-twenty-third dynasties (c. 945-725 bce), pharaohs with Libyan names like Sheshonq and Osorkon controlled Lower Egypt.
Cyrenaica took its name from Gyrene, tlie earliest Greek foundation in Africa (631 bce, following Herodotus 4-i50-151). Cyrenaica was also known as Pentapolis for its five major cities; Apollonia (Sosuza), Gyrene (Shahat), Ptole-mais (Tolmeta/Barce), Tauchira (Arsinoe/Tocra), and Euesperides (Berenice/Benghazi) (Stillwell, 1976). The Persians under Cambyses possessed tlie region, and Ptolemy I annexed it after Alexander’s death. In 96 bce Ptolemy Apion bequeathed Cyrenaica to Rome, after which it was united with nearby Crete as a Roman province. [5ee Gyrene; Pto-lemais.]
Tripolitania (“three cities”), colonized by Phoenicians in the seventh century bce, had tliree principle cities: Leptis Magna (Labdah), Oea (Tripoli/Trablus), and Sabratha (Stillwell, 1976). [5ee Leptis Magna.] Until 146 bce, Tripolitania formed the eastern province of a west Phoenician realm dominated by Carthage. [See Carthage.] After Carthage’s destruction, Rome ceded Tripolitania to its Numi-dian allies. Tripolitania later became a Roman province famous for olive oil production.
The monumental urban ruins at Cyrenaica and Tripolitania have atti'acted archaeologists, as have traces of Paleolithic and Neolithic peoples that once inhabited a less desiccated Fezzan. The archaeology of later native Nortli Africans, the Berbers, is still poorly understood, though their language and inscriptions are Icnown (Camps, 1986). One remedy for this neglect has been the UNESCO Libyan Valleys Survey, devoted to the archaeology of rural Tripolitania. The survey has discovered extensive Roman-period (tliough not necessarily Roman-designed) floodwater irrigation systems. [See Irrigation.] These supported a significant inland occupation based on mixed agriculture (grape, olive, cereals) integrated witli mobile sheep/goat pastoral-ism, the prevailing mode of subsistence before and after the Romans (Journal of Libyan Studies, 1989).
[S'ee also North Africa.]
BIBLIOGRAPHY
Camps, Gabriel. “Libyco-Berber Inscriptions.” In Encyclopaedia of Islam, new ed., vol. 5, pp. 754-757. Leiden, 1986.
Journal of Libyan Studies 20 (i 989). Special issue entitled “Libya; Studies in Archaeology, Environment, History, and Society, 1969-1989,” edited by D. J. Mattingly and J. A. Lloyd.
Stillwell, Richard, ed. Princeton Encyclopedia of Classical Sites. Princeton, 1976.
Joseph A. Greene
LIMES ARABICUS. The Latin term limes Arabicus means “Arabian frontier” and appears in literary sources in the fourth century ce. By tliis period the eastern frontier of the Roman Empire was divided into frontier districts (Lat., limites), each under the command of a dtix. South of the Euphrates River these districts extended along the edge of the North Arabian desert and formed the empire’s southeastern frontier. The limes Arabicus extended from the region just soutli of Damascus to Wadi el-Hasa, near tlie southern end of tlie Dead Sea. By the fourth century the frontier zone south of Wadi el-Hasa was the limes Palaesti-nae, which in turn extended to the Red Sea at Alla ('Aqaba). The limes was defended by Roman garrisons based in forts and towns along the frontier. Their primary mission was to protect the sedentary provincial population and the lucrative caravan ttaffic that passed through the region against the raids of nomadic Arab tribes (Saraceni, or Saracens) from the adjacent desert. [See Hasa, Wadi el-; ‘Aqaba.]
Only two literary sources explicitly refer to tlie limes Arabicus. The church historian Rufinus (Hist. Eccl. 2.6), describing the assault on the frontier by the Saracen queen Mavia during the reign of Valens (364-378 ce), refers to the “towns and cities of tlie limes Arabicus.” The historian Am-mianus Marcellinus (31.3.5), describing events leading up the battle of Adrianople in 378 ce, refers to a certain Mun-derichus, who was later dux limitis perArabiam, “duke of the limes through Arabia.”
The Roman frontier in this region was created by Trajan, who annexed the Nabatean kingdom in 106 ce as the new province of Arabia. [See Nabateans.] Trajan then built a major road (Via Nova Trajana) that extended from southern Syria to the Red Sea. Most forts of the later limes Arabicus were built along or a short distance east of this road, which was completed between iii and 114. The original Roman provincial garrison consisted of Legio III Cyrenaica, based at the provincial capital of Bosra in the nortli, and perhaps ten auxiliary units based in towns and forts along the frontier. [See Bosra.] Some strengthening of the Arabian frontier occured under the Severan dynasty (193-235). Inscriptions document the construction of several forts around the northwest outlet of Wadi Sirhan, a natural migration route between southern Syria and the interior of the Arabian Peninsula, and by many milestones reflecting road repairs. The frontier appears to have faced an increased Saracen threat by tlie late tliird century, when the frontier defenses were significantly strengthened by Diocletian (284-305). He rebuilt tlie road system, constructed or rebuilt many new fortifications, and reinforced the provincial army with new units. He also redrew the provincial boundaries, transferring southern Transjordan, tlie Negev, and Sinai (formerly part of Arabia) to the province of Palestine. [See Negev; Sinai.] The military frontier advanced to tlie desert fringe and Roman forces campaigned deep in the desert, apparently as far as Jawf (Duma) at the southeastern end of Wadi Sirhan. The Notitia Dignitatum (Oriens 37), dated to about 400 CE, reveals that the dux Arabiae commanded two legions, eight elite cavalry vexillations, six cavalry units or alae, and five infantry cohorts. These forces were styled limitanei—that is, troops that manned the limes. The limes Arabicus appears to have remained well fortified until the late fifth century, when tliere is evidence for tlie abandonment of some forts and a decline in the quality of its garrisons. In about 530, Justinian effectively demobilized most of tlie remaining units by stopping their pay and turned over primary responsibility for the defense of the southeastern frontier to the Ghassan-ids, Christian Arab clients. By this date the limes Arabicus, in the sense of a military frontier, effectively ceased to exist.
Altliough descriptions of some individual sites were made by several nineteenth-century travelers to the region, the first systematic survey of most of the Arabian frontier was conducted by Rudolf-Ernst Briinnow and Alfred von Do-maszewski in the 1890s. Their focus was largely architectural and epigraphic. As a result of the subsequent destruction of many sites, tlieir work remains fundamental. Subsequent surveys identified many other smaller sites, such as watchtowers, that formed part of tlie frontier system. As late as the mid-1970s, however, no fort of the limes had been excavated and no comprehensive survey had been conducted. The Limes Arabicus Survey, directed by S. Thomas Parker in 1976 under the auspices of the American Schools of Oriental Research, collected pottery from most major sites along the entire length of tlie frontier in Transjordan. This evidence, combined with excavation results from several key sites (e. g., el-Lejjun/Betthorus and Umm el-Jimal), led to the first historical synthesis of the Roman frontier (Parker, 1986). [5ee Umm el-Jimal] Since the late 1970s several scholars have focused on more limited sectors or individual sites on the frontier, producing important new evidence. A lively debate has begun about the nature of the frontier, with some questioning the level of the Saracen threat while others defend the traditional notion about tlte purpose of the frontier system.
The fortifications of the limes include watchtowers, ca. v-tella (forts), fortresses, and fortified cities. The last category is beyond the scope of this discussion, but various sources suggest tliat Roman army units were based within some towns and cities. Hundreds of structures identified as towers exist along the frontier; they seem to date to many periods, but they were not all military structures. The Romans often reused existing Iron Age and Nabatean structures but also built some new towers at key locations. Few have been excavated, further complicating their dating. Analysis of surface pottery and architecture has, however, permitted some understanding of these structures: many were erected on elevations witlt extensive views, usually intervisible with
LIMES ARABICUS. Figure I. Plan of Dafaniya. Roman fort built around 300 CE along the Arabian frontier. (Courtesy S. T. Parker)
Other towers, and thus presumably served as the “eyes and ears” of the frontier system.
A number of castella have now been excavated, including forts at Umm el-Jimal, Qasr al-Hallabat, Kdrirbet es-Samra, Umm er-Rasas/ICastron Mefaa, Qasr Bshir, Khirbet el-Fi-tyan, Da'janiya (see figure i), and Humeima. [See Qasr al-Hallabat; Umm er-Rasas; Humeima.] The most common type of casiellum is the so-called quadriburgium, or rectangular fort with towers projecting from the four corners. This type may date as early as the early third century, as evidenced by Qasr al-Hallabat, dated epigraphically to 213 ce. Perhaps the best-preserved and most securely dated castel-lum is Qasr Bshir (see figure 2), on the Kerak plateau east of the Dead Sea. [5ee Kerak.] This quadriburgium is dated to 293-305 CE by an in situ Latin inscription. The interior plan consists of ranges of two-storied rooms surrounding a central courtyard; the ground-floor rooms served as stables and die second-story rooms housed a garrison of about 150 men. Most other excavated castella also date to the tetrarchy, including those at Umm el-Jimal, Khirbet es-Samra, Khirbet el-Fityan, and Da'janiya. Recent excavation by John Oleson of the large (approximately 3 ha, or 7 acres) fort at Humeima, on the Via No va Trajana between Petra and Aila, suggests a second-century date and is thus the earliest-Icnown Roman fort along tire frontier.
By the fourth century at least three and possibly four legions were based along the Arabian frontier. The castra of the Legio III Cyrenaica has been identified as an annex attached to the city wall of Bosra but is as yet unexcavated. Arabia’s second legion, IV Martia, was based in a fortress at el-Lejjun, east of the Dead Sea on the Kerak plateau. At 4.6 ha (i I acres), el-Lejjun is typical of Late Roman legionary fortresses with its massive projecting U-shaped interval towers and semicircular corner towers. Erected in about 300 CE, the fortress seems to have been designed for a garrison of about two thousand men, but this was later reduced to about one thousand, following a reconstruction in the late fourth cenmry. The fortress is the most extensively excavated Roman military site along tlie frontier, with portions of the fortifications, headquarters, barracks, bath, church, and other structures exposed. The legion remained in garrison until the Justinianic demobilization of the early sixth century. The fortress itself was destroyed in an earthquake in 551 CE.
Fartlrer south, on the limes Palaestinae, the Legio X Fre-tensis was transferred from Jerusalem to Aila ('Aqaba) by the end of the third century. Its fortress at Aila has not been identified. Just east of Petra is the great fortress at Udruh, which is virtually a twin of el-Lejjun in terms of its size (4.7 ha, or 12 acres) and the number and plan of its gates, interval towers, and angle towers. The original internal plan of Udruh was obscured by later Islamic occupation. Although the excavator, Alistair Killick, has proposed a Trajanic date for its construction, no specific evidence has been advanced
LIVtES ARABICUS. Figure 2. Qasr Bshir. Roman fort built between 293 and 305 CE east of the Dead Sea. (Courtesy S. T. Parker)
„........-/“a'- ¦' "Lca
To support this early date. Rather, the dose architectural parallels with el-Lejjun and other late third-early fourth-century fortifications strongly suggest that the extant fortifications at Udruh were erected in the Diocletianic era. It has plausibly been suggested tltat this fortress was originally erected for Palestine’s other legion, VI Ferrata, which may have been transferred from northern Palestine to Udruh in about 300 CE but was disbanded or destroyed during the fourth century. Neither the legion nor the site appear in the relevant chapter in the Notitia {Oriens 34).
[See also Fortifications, article on Fortifications of the Hellenistic, Roman, and Byzantine Periods; Roman Empire; and Transjordan, article on Transjordan in the Persian through Roman Periods.]
BIBLIOGRAPHY
Banning, E. B. “Peasants, Pastoralists, and Pax PomayKi.-Munialism in the Southern Highlands of Jordan.” Bulletin of the American Schools of Oriental Research, no. 261 (1986): 25-50.
Briinnow, Rudolf-Ernst, and Alfred von Domaszewski. Die Provincia Arabia. 3 vols. Strassburg, 1904-1909. Still indispensable survey of tlie architectural evidence for the limes Arabicus.
Graf, David F. “Rome and the Saracens: Reassessing the Nomadic Menace.” In L’Arabie preislamique el son environnement historique et culmrel: Actes du Colloque de Strasbourg, edited by Toufic Fahd, pp. 341-400. Leiden, 1989.
Isaac, Benjamin. The Limits of Empire: The Roman Army in the East, Rev. ed. Oxford, 1992. Revisionist study tliat downplays the level of external nomadic tlireat, emphasizing instead the problem of internal security.
Kennedy, David L. Archaeological Explorations on the Roman Frontier in North-East Jordan. British Archaeological Reports, International Series, no. 134. Oxford, 1982.
Kennedy, David L., and Derrick N. Riley. Rome’s Desert Frontier from the Air. Austin, 1990. Lavishly illustrated with aerial photos and plans of many key sites along tlie Arabian frontier.
Mayerson, Philip. “The Saracens and die Limes.” Bullelin of the American Schools of Oriental Research, no. 262 (1986): 35-47.
Parker, S. Thomas. “Peasants, Pastoralists, and Pax Romana: A Different View.” Bulletin of the American Schools of Oriental Research, no. 265 (1986): 35-51. Reply to Banning’s article.
Parker, S. Thomas. Romans and Saracens: A History of the Arabian Frontier. American Schools of Oriental Research, Dissertation Series, 6. Winona Lake, Ind., 1986. Historical and archaeological survey of the limes Arabicus.
Parker, S. Thomas, ed. The Roman Frontier in Central Jordan: Interim Report on the Limes Arabicus Project, tpSo-igSs. 2 vols. British Archaeological Reports, International Series, no. 340. Oxford, 1987. Report on excavations at el-Lejjun, Khirbet el-Fityan, Rujm Beni Yasser, and Qasr Bshir and survey of tlie Roman frontier east of the Dead Sea.
Parker, S. Thomas, ed. The Roman Frontier in Central Jordan: Final Report on the Limes Arabicus Project, igSo-igSp. z vols. Washington, D. C., forthcoming. Results on excavations at el-Lejjun and Da'janiya, survey of tlie frontier east of the Dead Sea, and a historical syntliesis of diis sector.
Parker, S. Thomas. “The Nature of Rome’s Arabian Frontier.” In Roman Frontier Studies igSg: Proceedings of the XVth International Congress of Roman Frontier Studies, edited by Valerie Maxfield and Michael Dobson, pp. 498-504. Exeter, 1991. Review of recent scholarly debate on the subject.
Sartre, Maurice. Trois etudes sur TArabie romaine et byzantine. Collection Latomus, vol. 178. Brussels, 1982. The first essay deals with the borders of the province; the second widi the region’s Arab tribes.
Speidel, Michael P. “The Roman Army in Arabia.” In Aufstieg und Niedergang der romischen Welt, vol. 2.8, edited by Hildegard Tem-porini and Wolfgang Haase, pp. 687-730. Berlin, 1977. Definitive study of the Roman army in this province.
Speidel, Michael P. “The Roman Road to Dumata Qawf in Saudi Arabia) and tlie Frontier Strategy of Praetensione Colligare.” Historia 36 (1987): 213-221. Important epigraphic evidence about the peneua-tion of Roman forces deep into the Arabian desert during Diocletian’s reign.
S. Thomas Parker
LIONS. Not found in the fossil record of the Levant until the Late Pleistocene, lions (Panthera leo) arrived as tropical elements entering an already diverse fauna. If artistic and literary evidence are any guide, lions were luiown all over the ancient Near East. Witli more than 150 citations in the Hebrew Bible, they apparently were familiar in biblical times. Lions were finally hunted to extinction in Palestine in tire nineteenth century CE. Elsewhere in the region tlrey were not exterminated until after World War I. Several subspecies of lion ranged throughout the greater Near East until modern times, but it is impossible to determine which were closest to ancient Levantine populations.
Lion imagery was ubiquitous in the cultures of the ancient Near East. In religious and royal symbolism the lion appears as aggressor, victim, and protector of gods and kings. Given its prominence in literature and iconography, it is surprising that the bones of lions are such very rare finds at archaeological sites. The most spectacular discovery has been the skull and mandibles of a lioness found on tite floor of a twelfth-century bce Pre-Philistine temple in Jaffa, Israel. Half a scarab seal found in close proximity led the excavators to suggest that a lion cult was practiced in the temple. The right mandible has two sets of deep cut marks that are consistent with opening the oral cavity from the basal surface of the jaws while leaving the head attached to tlie skin. Two lion bones associated with Iron 1 pottery in a noncultic Iron II building were recovered from Tel Miqne/Ekron. A third lion bone was of Iron II date. At Tel Dan, one lion foot bone was recovered from the mid-ninth-century bce deposit in the altar room complex in file sanctuary; the cut marks on this bone are consistent with skinning. A fragment of a mandible was found in a metalworldng area of Iron I date at the same time. These bones are in collections studied by Paula Wapnish and Brian Hesse (see Wapnish and Hesse, 1991).
At the site of Hesban, in Jordan, an ankle bone of Roman date and a shoulder from a lion cub (date uncertain) were recovered. Farther afield, from the site of Habuba-Kabira in nortliern Syria, four lion bones were noted, two from the Uruk period (fourtli millennium) and two from the Middle Bronze Age. From Lidar Hoyiik in southeastern Anatolia, a lower jaw and lower leg bone were discovered in Late Bronze deposits, and one first phalanx noted from the Hellenistic period.
Documentation of the lion in ancient Egypt comes almost exclusively from artistic representations, from the O
World History
