site located in northeastern Iraq, 13 km (8 mi.) southwest of Kirkuk (35°21′47′′ N, 44°15′30′′E). Kirkuk (ancient Arrapḫa-āl-ilāni) was the source of tablets that had been sold on the antiquities market. Nothing was known about the origin or historical context of these tablets, which were marked by a peculiar dialect of Akkadian and contained Hurrian onomastica that had previously been encountered in letters of the Amarna archive in Egypt and in Kassite-period texts from Nippur. The search for an uninhabited site more suited for excavation led to Nuzi, where a cache of similar tablets had been discovered near the main mound.

Nuzi comprises a main mound and several smaller tells. Oriented northwest-southeast, the mound measures about 200 × 200 m, with an average height of 5 m above the surrounding plain. Only three of the smaller tells were investigated: one prehistoric site (Qudish Ṣaghīr), about 5 km (3 mi.) to the south, and two small mid-second millennium BCE sites, about 300 m to the north. Occupation of the main mound extended from the prehistoric to the Mitannian/Middle Assyrian period, followed by meager remnants of habitation in the Neo-Assyrian, Parthian, and early Sasanian periods (the “late period”). Identified as Gasur by recurrent late third-millennium BCE textual references and as Nuzu/Nuzi in the mid-second millennium BCE, the site is known primarily for the latter occupation. Its stratified Mitannian/Middle Assyrian period archives of more than five thousand sealed tablets span five or six generations of a mainly Hurrian population in the kingdom of Arrapḫa, the easternmost province of Mitanni. Published extensively over the years, the archives sparked intense study of the social, economic, religious, and legal institutions of the Nuzians/Hurrians, whose presumed identity with the biblical Horites (ḫōrī[m]) prompted several comparisons with the patriarchal customs of the Hebrew Bible. Nuzi replaced Kirkuk as the generic term for analogous tablets and seal impressions from neighboring mounds, including Kirkuk/Arrapḫa and Tell al-Faḫḫar/Kurruḫanni. The associated material culture, notably painted ceramics, glazed wares, and glass and faience objects became diagnostic of the so-called Nuzi period.

Excavations and Division of Finds.

At the request of Gertrude Bell, director of antiquities in Iraq, Edward Chiera led the first campaign to Nuzi In 1925–1926, under the joint auspices of the Iraq Museum and the American Schools of Oriental Research. Work began on the small northwestern tell, in the house of Šurki-tilla (two buildings) and the house of Teḫip-tilla. The excavated material was divided between the two supporting institutions.

The subsequent four campaigns, sponsored jointly by the Fogg Art Museum and the Harvard Semitic Museum in Boston and the American Schools of Oriental Research, were directed by Chiera (1927–1928), Robert H. Pfeiffer (1928–1929), and Richard F. S. Starr (1929–1930, 1930–1931). The finds were divided between the Iraq Museum and Harvard University. From the northwestern tell, Chiera moved to the northeastern tell, where he excavated the house of Zike and the house of Šilwa-teššup before turning to the main mound. Pfeiffer introduced a grid system. In addition to completing Chiera's excavation of the palace and adjacent domestic structures to the southwest and northeast, Pfeiffer initiated two test pits in rooms N120 and L4 to establish the sequence of occupation. Starr concentrated on the northwestern ridge and the temple complex. Once the upper levels associated with the widespread Nuzi tablets had been cleared, work focused on an investigation of the lower strata in the temple area (temples A–G), along the north-western ridge (strata I–IV), and on the southeastern edge in the region of the city wall (strata V–VIII). Pit L4 was extended down to virgin soil, providing a continuous sequence of fifteen occupation levels (pavements I–XII).

Stratigraphy and Chronology.

Three stratigraphic schemes were used simultaneously and only tentatively integrated into a single system. Their only contact is during the final Nuzi period between pavement I, stratum II, and temple A, which were also linked by a complicated drainage system and the Nuzi tablets. The date of the tablets hinged until recently on a letter bearing the seal impression of Sauštatar, son of Parsatatar, king of Mitanni, whose reign fluctuates within the fifteenth century BCE. It now appears that the seal was probably an heirloom and that a later date for stratum II conforms with the evidence from other sites where Nuzi-related material is dated by associated texts to the mid-fourteenth and thirteenth centuries BCE. The range of the Nuzi archives is currently posited between 1445/1425 and 1350/1330 BCE, following the low chronology. The variant dating of the latest Nuzi texts and the destruction of the site depends, then, on the Aššur-mutakkil synchronism in the first or second generation at Nuzi and on historical-political factors that best match the socioeconomic turmoil reflected in the latest Nuzi records.

Too little remains of the uppermost stratum I material to extend the duration of the Nuzi period much beyond the stratum II destruction date. The origin of this period is less certain. Although architectural and material remains of the Nuzi period are represented in the lower pavement II, strata III–IV, and temples B–E, Starr's correlation between stratum IV and temple E is contested by Ruth Opificius, who relates stratum IV to temple F, which she considers Old Babylonian by its pottery and figurines (Das altbabylonische Terrakottarelief, Untersuchungen zur Assyriologie und vorderasiatischen Archäologie, 2, Berlin, 1961, pp. 18–19). The latter would be compatible with Starr's own comparison between the pottery of strata IV and VII.

Most scholars agree on the synchronism between pavements IIA and IIB, strata V–VII, and temple F. One Old Babylonian tablet and five early Old Assyrian tablets from pavement IIA and one Ur III tablet found in or above pavement IIB–III (a rubbish deposit) provide a tentative chronological range (c. 2100–1650 BCE) for pavements IIA, IIB, and IIB–III, whose contents are considered transitional between those of the upper Nuzi pavements and the lower Gasur pavements. Opinions diverge over the association of the Gasur pavement III, dated by tablets to the Old Akkadian period, with temple G and stratum VIII by virtue of their relative position directly below the transitional phase. Structural as well as ceramic links between temples F and G and between strata VII and VIII support a later Ur III/Isin-Larsa date instead.

Pavements III to VIII were disturbed by numerous graves, so that any demarcation of the Old Akkadian occupation, linked variously with pavements III–V and III–IX, remains inconclusive. Henry W. Eliot (in Starr, 1939) has tentatively identified pavement VI as Early Dynastic, VII as Late Uruk, and VIII–IX as Uruk. The earliest deposits, pavements XA–XII, belong to the Ubaid period.

Prehistoric period.

Represented in pits L4 and G50 and on Qudish Ṣaghir, the lowest levels for the prehistoric period are characterized by packed clay (pisé) walls and incised, knobbed, and painted wares. The painted wares are superseded by unpainted wares in pavement X, which sees the introduction of mud-brick construction and infant burials. A cache of marble stamp seals from G50 is attributed to this period on typological grounds.

Gasur and transitional periods.

The Gasur occupation is encountered in pits L4 and G50 and, according to Starr, in the City Wall (stratum VIII) and temples G and F. It is identified and dated by 222 tablets attributed to the early Old Akkadian period on linguistic, orthographic, and paleographic grounds. Most notable among the mainly business documents, in which Semitic names predominate, is an inscribed clay “map” that accompanies a record of land and appears to locate an estate within the district of Gasur. The small finds include a terra-cotta mold for casting animal-shaped amulets and a copper statuette from temple G. The cultural pottery is typified by incised gray-ware vases, zoomorphic vessels, house-shaped stands, and relief decoration in the form of applied snakes, scorpions, and quadrupeds. The glyptic evidence comprises impressions of one cylinder seal and two stamp seals and five cylinder seals (including a shell seal mounted on a copper pin that accompanied a burial).

Nuzi period.

The Nuzi period was fully excavated over the entire mound and on both northern tells. The evidence of architecture, texts, glyptic, and luxury products provides the most complete picture of a provincial town on the eastern edge of the Mitannian kingdom toward the end of its hegemony. As such, the Nuzi material complements the evidence from Alalakh levels V and IV, which represents a provincial town on the western edge of the Mitannian kingdom during its formative and mature phase.

Architecture and architectural decoration.

Stratum II structures on the main mound include a palace and temple complex surrounded by well-constructed municipal buildings along the northwestern ridge and denser residential sectors on the southeast, southwest, and northeast. Five contemporary buildings, associated by their archives with wealthy landowners and high-ranking officials, were excavated outside the citadel walls on the northern tells.

The tripartite palace plan, incorporating old and new elements, is divided into a formal area consisting of rooms grouped conventionally around courtyards; an agglutinate administrative section south of the outer courtyard; and residential quarters beyond the throne room to the southwest. The main entrance, though not preserved, would have led into the outer courtyard (M94), which was bordered by benches. A connecting room with a hearth (M89) gave indirect access to the inner courtyard (M100), which was flanked by reception rooms and apartments. This separation of the public from the official domain, while reminiscent of the palace plans at Mari and Tell Asmar, becomes a characteristic of Neo-Assyrian palaces. The entrance to the unusually juxtaposed anteroom (L20) and audience hall (L4/11) is marked by two freestanding brick pillars that may have supported a portico or canopy. The supposed chapel directly north of the audience hall was devoid of cult fittings, and Starr's identification of a kitchen in the administrative sector has also been queried on practical grounds. Luxuriously furnished with a sophisticated drainage system, marble paving, and silver-coated copper door studs, the palace is best known for its well-preserved frescoes, also found outside the citadel in the house of Prince Šilwa-teššup, which resembles the palace on a smaller scale. The more elaborate examples show figured designs in panels of solid red and gray framed by geometric patterns arranged in an architectural scheme of bands and metopes. The Hathor head, bucranion, and palmette tree are recurrent motifs that also appear on ceramics, cylinder seals, and plaques. The designs reflect the current taste for foreign, in this case Egyptian and Aegean, imagery.

During the Isin-Larsa/Old Babylonian Period (temple F), the sanctuary was transformed from a single temple (G29) into a double temple (G29, G53) complex with forecourts with subsidiary rooms. Where originally G29 was based on a double-flanked main room (temple G), both temples are Herdhaus types—arranged on the standard Mesopotamian bent-axis scheme, with the hearth or altar and cult podium at the short, southeastern end of a rectangular room that was entered from its northern corner. Apart from structural revetments, neither temple was altered significantly before its destruction in the mid-fourteenth century BCE. The contrasting furnishings of the two cella suggested different patron deities. Temple G29, with its wall decoration of glazed terra-cotta nails, a sheep's head, a boar's head, zoomorphic jars, and glazed terra-cotta lions, was attributed to Ishtar, the Babylonian goddess of love and war, whose association with lions extends back to the Akkadian period. This cella also contained nude female figurines, female amulets, and a unique ivory statuette of a seminude female in Hittite attire, who may represent Ishtar's Hurrian counterpart, Šawuška. Temple G53, marked by its conspicious lack of contents, was attributed to the Hurrian storm god, Teššup, one of Ishtar-Šawuška's numerous partners, who headed the Mitannian pantheon and whose name is a common onomastic element at Nuzi. Evidence for the cult of Teššup first appears in the early second millennium BCE and supports the later dating of temple F to the Isin-Larsa or early Old Babylonian period.

Tablets and seal impressions.

The excavated tablets from Nuzi were stored as public and private archives on the main mound and on both northern tells. Although associated with the upper strata (I–IV), the vast majority come from stratum II. A relative chronological framework is derived from the five-generation scribal family of Apil-sîn and that of the real-estate magnate and crown official Teḫip-tilla. Through ration and personnel lists, contracts, legal records, and letters, the texts, written in a Middle Babylonian dialect of Akkadian mixed with Hurrian idiosyncracies, reflect rapidly declining socioeconomic conditions in the province of Arrapḫa: the grain surplus was decreasing and both debts and litigation were increasing, forcing the majority to cede their property rights to the privileged few. The last years are marked by intensive military activity on both the northern and southern frontiers, but Nuzi's destruction is generally attributed to the Assyrians.

The tablets bore the seals of people from all walks of life in Nuzi and in the surrounding towns. They thus provide a unique source of information on sealing practices and on the relationship between seal design, text type, user, rank, period, and origin. The style is characterized by the extensive, often unmasked use of point and tubular drills; the seals' material is sintered quartz (frit). The iconography is heterogeneous, drawing on a long tradition of assimilation and adaptation. Superimposed on native designs devoted to hunting and animal rituals centered on a tree, are court fashions that shift with the political tide. During the first generations, which coincided with the Mitannian confederacy, contact with the west brought Egyptian and Aegean iconography, such as the Hathor mask, the ankh sign, the sphinx, the griffin, and the winged disk. In time, the tree was replaced by the winged disk and confrontation scenes between human and divine figures, influenced by provincial Kassite-Babylonian traditions from Elam and the Hamrin, gave way to a revival of ancient Mesopotamian contest scenes between mythical beasts—which presumably had survived on the eastern periphery of the region. Among the Akkadian revivals are the storm god on his lion-dragon mount and mythological representations, including scenes from the myth of a Ḫedammu-type dragon and, possibly, the death of Huwawa from the Epic of Gilgamesh. Although associated with the Hurrians in later literary traditions, the origin of these myths is unknown.

Small finds.

The material culture associated with the stratum II tablets became diagnostic of the Nuzi period and, by extention, of the Hurrians, who migrated across the Fertile Crescent from the east. Apart from the difficulties of relating any category of archaeological material to an ethnic sector of a composite population, the redating of stratum II to the mid-fourteenth century BCE places the origin of much of the characteristic Nuzi material in the west.

Nuzi ware (formerly called Hurrian ware and Subartu ware) combines elements from a number of earlier ceramic traditions in Mesopotamia with foreign elements of design. Distinguished by a white pattern painted on a dark (red-brown to black) background, this ware is distributed from the Zab valley east of the Tigris River to the ῾Amuq plain and Orontes valley in the west and as far south as Babylonia. At Nuzi, geometric motifs predominate over natural ones. The characteristic shapes include the slender goblet (high cup) with a small foot or button base, and the shoulder cup. Both are thin, fine-grained drinking cups derived from Babylonian prototypes of the Isin-Larsa period. Found predominantly in palaces, temples, and “manor houses,” Nuzi ware is considered a luxury product, along with the great quantities of faience, multicolored glass, and glazed wares that are widely distributed during the Mitannian period. At Nuzi, these products include zoomorphic amulets, figurines, beads, vessels, cylinder seals, and plaques. The most common glazed objects are terra-cotta wall nails. They appear to be connected with the cult and may have been dedicated by individuals to decorate the walls following an ancient tradition reaching back to prehistoric times at Qudish Ṣaghīr and to Mesopotamia in general. Cultic paraphernalia cast in copper or bronze include small statuettes and a cylindrical post stand with excised sides and couchant lions on the upper edge. In a military context, bronze scales were used for the armor of men and horses. Most unusual is a bronze dagger, its hilt decorated with inlaid iron plaques secured by iron rivets. Terrestrial and meteoritic iron was a rare and valued metal used primarily for ornamentation until the late period.

Late period.

Structural remains of the late period were concentrated on the northwestern ridge and many low adjacent mounds. Three silver coins of the Parthian king Vologases III and a silver coin of the Sasanian king Shapur I provide a chronological range from the late second to the late third century CE for the settlement and cemetery.

[See also Hurrians; Mitanni.]


Primary Sources

Preliminary reports by the respective excavators published in Bulletin of the American Schools of Oriental Research, numbers 18, 20, 29, 30, 32, 34, 38, and 42 (1923–1931) were followed by a final report on the excavations by Starr (1938, 1939) and an ongoing series of text publications. Discrepancies between the preliminary and final reports in the designation of room numbers may account for some of the variation in descriptions of room contents, particularly tablets, which were recorded and published separately. For the excavations, see Richard F. S. Starr, Nuzi: Report on the Excavation at Yorgen Tepa Near Kirkuk, Iraq, 2 vols. (Cambridge, Mass., 1939). For a complete listing of primary text publications up to 1992, see Jeanette Fincke, Die Orts- und Gewässernamen der Nuzi-Texte, Répertoire Géographique des Textes Cunéiformes, vol. 10 (Wiesbaden, 1993), especially pages xvi–xxx.

Secondary Sources

  • Cecchini, Serena Maria. La ceramica di Nuzi. Studi Semitici, 15. Rome, 1965. Useful assemblage of Nuzi ware and summary of conventional evaluations, though now superseded by new finds and more critical investigations.
  • Dietrich, Manfried, Oswald Loretz, and Walter Mayer. Nuzi-Bibliographie. Alter Orient und Altes Testament, Sonderreihe, vol. 11. Kevelaer and Neukirchen-Vluyn, 1972. Complete listing of primary and secondary publications on Nuzi through 1971.
  • Eichler, Barry L. “Nuzi and the Bible: A Retrospective.” In DUMU-E2-DUB-BA-A: Studies in Honor of Åke W. Sjöberg, edited by Hermann Behrens et al., pp. 117–119. Philadelphia, 1989. Critical review of the impact of Nuzi scholarship on biblical studies from the early comparisons between socio-legal customs at Nuzi and those in the Bible to more recent skepticism regarding the Hurrian background of the customs at Nuzi and their relevance for the dating and interpretation of specific biblical narratives and institutions.
  • Hrouda, Barthel. Die bemalte Keramik des zweiten Jahrtausends in Nordmesopotamien und Nordsyrien. Istanbuler Forschungen, vol. 19. Berlin, 1957. Important, if outdated, examination of Nuzi ware in relation to other painted ceramics, which in one way or another were misconstrued as reflections of Hurrian culture.
  • Mellink, Machteld J. “A Hittite Figurine from Nuzi.” In Vorderasiatische Archäologie: Studien und Aufsätze Anton Moortgat zum fünfundsechzigsten Geburtstag gewidmet, edited by Kurt Bittel, pp. 155–164. Berlin, 1964. Unusually convincing description of one Hittite/Hurrian component in Nuzi iconography, citing art historical and archaeological as well as textual and historical arguments.
  • Porada, Edith. Seal Impressions from Nuzi. Annual of the American Schools of Oriental Research, 24. New Haven, 1947. Pioneering art historical study of the Nuzi glyptic based on a representative sample of seal impressions from the five-generation archive of Teḫip-tilla.
  • Stein, Diana L. “Khabur Ware and Nuzi Ware: Their Origin, Relationship, and Significance.” Assur 4.1 (1984): 1–65. Redefinition of these two painted wares in terms of their shape, decoration, distribution, and date. Many of the results need to be tested against new evidence from recent excavations in the Khabur triangle.
  • Stein, Diana L. “Mythologische Inhalte der Nuzi-Glyptik.” In Hurriter und Hurritisch, edited by Volkert Haas, pp. 173–209. Konstanzer Altorientalische Symposien, vol. 2. Constance, Germany, 1988. Interpretation of mythological imagery in the Nuzi iconography relating, in particular, to Teššup and Ishtar-Šawuška.
  • Stein, Diana L. “A Reappraisal of the ‘Sauštatar Letter’ from Nuzi.” Zeitschrift für Assyriologie und Vorderasiatische Archaeologie 79.1 (1989): 36–60. Casts doubt on the conventional dating of Nuzi and favors a later date based on sealing practice as well as historical and archaeological factors.
  • Stein, Diana L. Das Archiv des Šilwa-teššup. 2 vols. Wiesbaden, 1993. Comprehensive textual and art historical analysis of the sealings and sealing practice in the private archive of Prince Šilwa-teššup; based on a reedition of the texts and line drawings of the seal impressions. The use and designs of seals are investigated in specific legal, administrative, social, and regional contexts within the framework of one community, from about 1440 to 1330 BCE. The results compliment, amend, and extend Porada's study (see above).

Diana L. Stein