Bronze Age Syria PDF

Dialogues between Ancient near Eastern Texts and the Archaeological Record: Test Cases from
Bronze Age Syria

Author(s): Marie-Henriette Gates
Reviewed work(s):
Source: Bulletin of the American Schools of Oriental Research, No. 270, Ancient Syria (May,
1988), pp. 63-91
Published by: The American Schools of Oriental Research
Stable URL: http://www.jstor.org/stable/1357005 .
Accessed: 15/11/2012 11:34
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
The American Schools of Oriental Research is collaborating with JSTOR to digitize, preserve and extend
access to Bulletin of the American Schools of Oriental Research.
http://www.jstor.org
This content downloaded by the authorized user from 192.168.82.204 on Thu, 15 Nov 2012 11:34:41 AM
All use subject to JSTOR Terms and Conditions
Dialogues Between Ancient Near Eastern
Texts and the ArchaeologicalRecord:
Test Cases from Bronze Age Syria
MARIE-HENRIETTE GATES
Department of Classics
University of North Carolina
Chapel Hill, NC 27514
Scholarship on the historical ancient Near East pursues two often independent
approaches, epigraphic and archaeological. While the two approaches may meet
when one requires confirmation from the other, especially when putting historical
events into archaeological contexts, the two disciplines rarely are brought together
on an equal footing to reconstruct those ancient cultures. Since the ancient Near
East provides much written evidence to supplement the cultural material that
represents its mute legacy, it is advantageous to utilize both sources of information,
particularly to determine the function of common categories of artifacts that invoke
no obvious interpretation and to identify functional contexts more securely. I will
argue the virtues of a joint approach by considering thefunctions of several types of
finds in light of textual references. The illustrations come from Bronze Age Syria,
as follows: I) ceramic commercial containers; a brewery at LB I Tell Hadidi; 2)
nonceramic finds, including animal bones from Mari and Selenkahiye and metal
annaku [tin?]; 3) architectural units, namely the function of Zimri-Lim palace at
Mari. In all these exercises, considering texts and finds together provides a likely
reconstruction, but separately they would produce little conclusive result.
R econstruction of the literate cultures of of those records. (One need only think of the
the ancient Near East has drawn on two imbalance between the archival information from
classes of contemporary records: written Old Babylonian Period Mari and contemporary
documents in the form of inscriptions on clay Aleppo: Mari's richness overcompensates any de-
tablets, stone, or other objects on which trans- sire to know more from elsewhere.) But despite
mission of a text remained the primary objective, difficulties with textual interpretation, the written
or inscriptions that served as labels on items documents by themselves speak at great length
whose major function was not literary; and cul- and eloquently.
tural documentation from anepigraphic material Interpreting the archaeological record, on the
remains uncovered in excavations at ancient sites. other hand, requires steps beyond the initial trans-
There is a signal discrepancy in the intensity of lation. In conjunction with a strictly descriptive
information those two sources transmit. The writ- primary publication, such as an excavation report
ten texts and archives preserve historical moments (the archaeological equivalent of a transliteration),
(Liverani 1973), personalities, cultural behaviors, the material must be submitted to an analysis that
and literary and religious movements, which can represents a modern construct of architectural/ art
be pieced together into a testimonial of varying, historical elements that may not in fact reflect
yet satisfactory, complexity for a spectrum of culturally significant features; a cultural/ethnic
phases. Especially with economic archives, ana- study (identifying, for instance, pottery cultures)
lysts can reconstruct detailed social and economic along general historical lines; an impersonal,
patterns, although in doing so, they risk disre-
garding the sporadic and indeed enclosed nature To MachteldJ. Mellinkon her seventiethbirthday.
63
64 MARIE-HENRIETTE GATES BASOR 270
"normative" charting of human response to the viewer, although not necessarily as the artist in-
challenge of existence at a specific time; or a tended. I shall apply comparable methods to less
combination of those approaches. The archaeo- expressive archaeological objects, in particular to
logical context too often provides only a sparse pottery but to other types as well, to show that
backdrop for the textual records, and interpreta- Goetze's maxim is appropriate to all aspects of
tion plays on very general themes. The archaeo- research in this field: "only when monuments and
logical record, moreover, is hampered by such documents agree with each other, can we hope to
factors as uneven transmission of remains, gaps in be on safe ground" (Goetze 1957: 24).2 Goetze was
sequence, flukes in discovery, the excavation or referring to the elucidation of historical facts. This
nonexcavation of key sites (that factor fluctuates article will extend his precept to more pedestrian
with trends in scholarship; it also depends on cultural remains through illustrations from Bronze
what one hopes to extract from the sites), top- Age Syria, and will demonstrate that archaeo-
onymic uncertainties in large part dependent on logical material can be coaxed into speaking more
whether a site is epigraphic, and the excavator's articulately.
scholarly abilities and interests. Perhaps because
of those difficulties a schism has developed (if it CERAMICS
has not always existed) between the epigraphic
and the archaeological approaches to the ancient Pottery is the most commonly found category of
Near East. The epigrapher and the archaeologist artifact in historical ancient Near Eastern sites. It is
work with tangent material, but may interrelate therefore often the primary evidence used to assess
their investigations only in passing (Gibson 1972; the cultural and chronological parameters of exca-
one exception is the Mallowan-Gadd partnership: vations; as such it appears in excavation reports as
Gadd [1940]). The schism is by no means restricted essential documentation. A few generalizations on
to ancient Near Eastern studies. Even Binford sees the type of ceramic analysis characteristic of those
the two disciplines as separate; but he is concerned reports need to be stated: First, the attention pot-
with "the dynamics of past ways of life and... tery receives in publications is often inversely pro-
the conditions in the past that brought them into portional to the architectural preservation at the
being" (Binford 1983: 20), whereas written testi- site. Ceramic finds from stratigraphic soundings
monials are the preserve of the historian. tend to be presented in far more detail than mate-
However, efforts to make use of the written/ rial recovered in well-preserved buildings whose
epigraphic documents and the material/archae- architecture warrants lengthy discussion; compare,
ological remains in tandem can yield spectacular for example, the ceramic reports from Tyre (Bikai
results. Winter's studies on the Iron Age Karatepe 1978) and the Middle Bronze Mari palace (Parrot
reliefs (1979) and on Neo-Assyrian visual propa- 1959: 114-45). Tomb contexts are exceptions,
ganda (1982) brilliantly illustrate the fruitfulness, because tombs tend to preserve artifacts (compare
and essential need, to recreate "a fabric woven of Du Mesnil du Buisson's Qatna publications of
whole cloth-a mesh of interlocking threads that tombs and settlements (1927; 1930). Second, the
must combine historical, linguistic, topographic, publications emphasize imported wares, which
archaeological, and artistic skeins before the man- attest to contacts with other areas-i.e., trade or
tle is complete" (Winter 1979: 51). One can go ethnic movements-and thus provide chronologi-
beyond conjecture, however rationally it may have cal or cultural indexes of interest to the historian.
been achieved, with confirmation from an inde- Third-in contrast-local pottery is summarily
pendent source.1 There are certainly instances classified according to shape, ware, and decora-
where the effort to stretch interpretation between tion; it often occurs divorced in the publication
the verbal and material evidence may raise associa- from its pertinent architectural, as opposed to
tions containing enough uncertainty that we may stratigraphic, context. The ceramic repertoire thus
consider the results misleading, or at best specious. exists solely as a manifestation of a period or
It is nonetheless a direction that I wish to defend, culture, not as a collection of objects that had
and to justify the gamble as ultimately necessary. active functions within their original context.
The examples cited involve art-cylinder seals and Yet one must assume that more precision in
sculptural programs-for which the visual impact understanding the pottery finds from a given
remains an accessible language to any informed context can often reveal the nature of that context
1988 TEST CASES FROM BRONZE AGE SYRIA 65
since "the form of a pot is determined (although Habur Ware beakers did serve as standard mea-
not precisely so) by its intended function" (Renfrew sures at Chagar Bazar, it is highly unlikely that
1977: 3). The analysis must go beyond the standard they performed the same role elsewhere. Standards
shape typologies (bowl, pot, storage jar), and aim of measure varied considerably from one state to
instead for more specific labels: unguent pot, wine another despite a fairly uniform terminology for
jar, grain pithos. Efforts to do that in Near Eastern weights and volumes; they may have varied even
archaeology have been sporadic at best. Ellison's within a city, texts often specify which standard
pioneer studies in the identification of the ways they are using (Gelb 1982: 585-90). The broad
Mesopotamian household vessels could be used in distribution of Habur Ware beakers throughout
food preparation (especially Ellison 1984) repre- northern Syria and northwest Iran therefore can-
sent a major methodological innovation for the not imply, as Gerstenblith (1983: 76) and others
discipline. She is correct that context is not an suggest, that they served in a standardized and
exclusive indication of a vessel's function in food hence uniformly controlled commercial network.
preparation, which may be done both indoors and Certainly trade is involved, but it would be mis-
outside (Ellison 1984: 63). However, it is more leading to envisage it along such modern lines.
likely that the significance of a context may emerge The method behind both interpretations relies on
through the ceramic (or other) finds. Henrickson rationalizing the presence of such vessels in the
and McDonald (1983) illustrate the potential suc- general light of text references to extensive trade
cess of such an assumption and its deliberate appli- throughout Middle Bronze Age Syria, rather than
cation in their treatment of two early village sites in on integrating the evidence the vessels represent
central western Iran. By basing their ceramic on the one hand and the specifications the texts
analysis on ethnographic morphological studies, in represent on the other. Such theorizing may offer
combination with volumetric estimates for the the correct interpretation, but often it does not
vessels from both sites, they distinguish between present an argument that one can adopt except on
the superficially similar repertoires and conclude faith.
that the one was for seasonal occupation and the The intellectual leap from classifying a vessel by
other was appropriate to year-round settlement its basic shape to identifying it correctly by its
(Henrickson and McDonald 1983: 639-40). original intended function is difficult and perhaps
The prehistorian's goal, as this example illus- specious. Yet a variety of features may well re-
trates, is to understand "ancient society and eco- spond to analysis. Traces of vessel contents in all
nomy" during a phase for which there exists no but rare instances disappear (Matson 1965: 205-6),
epigraphic evidence. But one can and should apply usually removed in sherd-washing; and chemical
similar techniques to situations where abundant analysis is too time-consuming and costly to do
written documentation on the economy does sur- routinely. However, even outside Egypt-where
vive, and where the sources and volumes of con- exceptional preservation of liquids does occur (see
tainers specific to certain traded and consumed Lucas 1948: 380 [Cypriot vessel from Amarna];
goods are abundantly attested. Merrillees 1962 [Cypriot base-ring juglets])-some
In Near Eastern archaeology, most efforts to discoveries can be applied to a broader context
isolate the functions of specific vessels have been (below, containers from the mid-second millen-
based on logic rather than on methodical com- nium B.C. cemetery at Baghouz). Shipwrecks also
parisons of specifications described in texts with offer great potential in preserving sealed jars, as
the vessels themselves. The treatment of "Habur vividly demonstrated by recent findings off the
Ware" beakers is an example. Mallowan (1936) Turkish Mediterranean coast of the Late Bronze
presented them as possible grain measures corres- merchant boat from Ka? (Bass 1986). Finally,
ponding to a standard system, because the major- grain is regularly found in situ in bins and pottery
ity, he noted, occur in two sizes: single and double containers, a fact that the pottery catalogue should
(0.10 m and 0.20 m high). Mallowan did not cal- but rarely does include as adjunct to individual
culate their volumes, surely a determining factor entries. Surface pitting and wear patterns on
in his proposal, or their relation to the standard ceramic vessels (see Hally 1983) are only imprecise
measures used in the contemporary Chagar Bazar indicators of use; the abrasions may be secondary,
economic texts (the sutu kinate and the "official" especially on sherds, and those features are not
sutu of Samag/dUTU: Gadd 1940: 30).3 Even if the systematically recorded either in the field or in
reports. Smoke-blackening, in contrast, is easy to beakers, jars, storejars, pitchers, juglets, and bot-
recognize and is usually noted; but the nature of tles. Many of them can aspire to more descriptive
the blackened vessel ("coarse cooking pot") often terminology, especially those with special features
suggests the same conclusion. (In theory, smoke- that demand a functional explanation. The fol-
smudging is desirable for the insides of wine-jars lowing example, using vessels from a Late Bronze
[Matson 1965: 206], but is unattested in ancient context at Tell Hadidi in western Syria, illustrates
Egypt, or, by extension, in the ancient Near East the extent to which one can coordinate textual in-
[Lucas 1948: 28-29]). Surface finishes may consti- formation and ceramic typology with some
tute a more reliable index. Coating a container reward.
with bitumen, burnishing its surface, or lining it
with gypsum will, at least in principle, reduce DUGNiG.DUjR.BR= namzitu =
porosity (Matson 1965: 205)4 or keep out the TELL HADIDI LATE BRONZE VATS
damp (e.g., a Middle Bronze grain storage jar
lined with gypsum from Rimah [Postgate 1978: Excavations at Tell Hadidi in inland western
71]). Such finishes point, in a general direction, to Syria were carried out as part of the Euphrates
the vessel's purpose as liquid or dry-goods con- Tabqa Dam salvage project in the mid-1970s
tainer. However, "sweating" can be a desirable (Dornemann 1979; see also Dornemann, above).
feature for liquid storage containers in hot cli- Because of the constraints of limited seasons,
mates, and ceramic oil storage containers do not investigations focused on stratigraphic soundings
require a surface finish since the oil itself will pro- to assess the chronological range of this large site,
duce the same result (Henrickson and McDonald which was occupied from the late third millennium
1983: 633). Decoration on pottery may serve a Early Bronze period through the Late Bronze
purpose other than aesthetic, and can thus pro- Age. The best preserved structure to be excavated
vide indirect clues to a vessel's intended use. Mark- was discovered in Area H on top of the mound,
ings on Middle and Late Bronze jars in Syria and called the "Tablet Building" because of its
(notably at Baghouz, Terqa, and Mari, below) most outstanding finds (Dornemann 1979: 144-
may refer to shippers or contents, although details 45). It consists of a square complex of seven or
are elusive. The raised bands applied to the bodies eight rooms arranged on three sides of a large
of large-capacity store jars, such as those at Mari rectangular court; it was destroyed by intense fire.
(Parrot 1958: 155), may be perceived as decora- The pottery lying in situ on the floors, and 17
tion but in fact must have allowed users to tie cuneiform tablets with sealings date the building
ropes around the vessels to aid in transporting comfortably to LB IB, in the early 15th century
them securely. Here again the decoration hardly B.C. (Dornemann 1981: 42, 59; carbon-14 dates
adds to an interpretation for the vessel based on between 1540/1510 and 1450/1400). The tablets
its context, size, and shape, but in other cases it conveniently provide the name for Late Bronze
will be a significant clue. Merrillees (1962) thus Hadidi as URUA-zu,the city Azu (Dornemann
has proposed that the shape and decoration of the 1979: 145-46), and Yaya son of Huziru and his
common base-ring juglet intentionally may have family as owners of the building itself. The pottery
imitated a slashed poppy pod as advertisement for from this building is in its own way equally vocal,
the product inside. for it indicates at least one of the sources for
Finally, the shape, size, and characteristic attri- Yaya's livelihood: he made beer on a large scale.6
butes of the vessel often best suggest the use for The admirable preliminary report on the Late
which it was designed; and it is at this point that Bronze pottery from Area H calculates that over
ethnographic comparanda and information in the 30 percent of the 127 vessels recovered from the
ancient texts may supply the details necessary for rooms of the Tablet Building were large storage
a correct identification. While one cannot label jars, broad-mouthed vats, and handled craters.
tablewares with any confidence, at least long- and Mugs, cups, jars, and pitchers-again many with
short-term containers, whether for dry or liquid handles-formed the second largest class of shapes,
goods, require universal specifications that can be while small bowls and platters accounted for the
recognized: size, ratio of diameter to height, wide smallest category, 9 percent (Dornemann 1981: 32-
or restricted mouth.5 In standard Near Eastern 33). The pottery is of industrial rather than fine
pottery classifications, those vessels are listed as quality, and none of the finds in the report suggest
c e
~I/ I Fa la ilb
f
i'S.~ ~i n
IIV J VII
vllw
Fig. 1. The Late Bronze I Tablet Building and selected
vessels from Tell Hadidi Area H. After Dornemann 1981:
I V. v VI L VI5 a. vat from Room III (fig. 3:2); b. vat from Room VI (fig.
7:4); c. soaking vat from Courtyard VIII (fig. 8:2); d.
tankard from Room III (fig. 4:9); e. pitcher from Court-
yard VIII(fig. 9:12); f. beer crater from Room II (fig. 10:7);
the Tablet Building.
tm
a purpose for the building other than utilitarian.7 where the absence of pottery suggested to the
Large quantities of carbonized grain (species not excavator that the grain had been stored in perish-
given) were found in Rooms III, IV, and VII: able containers such as sacks or baskets (Dorne-
around (but not in) two large storage vessels in mann 1981: 33). Architectural features include a
Room III in association with a number of grinding drain leading from Room Ib to the outside, a
stones, inside three of the Room IV storage vessels, possible stairwell in Rooms Ia/b, and an oven in
and in a thick layer on the floor of Room VII, the central courtyard.
The Tablet Building pottery conforms well with (see Oppenheim 1950). The green-malt was sun-
material from Hama G and Alalakh IV (Dorne- dried or roasted (Oppenheim 1950: 15), an activity
mann 1981: 41). What appears unusual is its very for which the Tablet Building's large open court
limited repertoire of shapes, which can only be with an oven must have been designed. Either
explained as representing the equipment for a pounders or sieves (Room III had an especially
specialized industry. Moreover, the key to that large supply) were used in the next stage, treating
industry certainly lies in another peculiarity that the dried malt, which in turn was mixed in the
Dornemann underlined: a large number of the namzitu with a fermenting agent and stirred. The
vats and storage jars have perforated bases, a final product was poured into containers to settle,
feature that he proposes made them unsuitable for and ultimately decanted into storage and shipment
storing liquids, and by default suggests they were containers (Dornemann 1981: 16, 45, n. 43). The
used to store grain. But the evidence from Room tebibtu, or clarifying vat, was placed on a wooden
III, where grain was discovered around rather stand-which may explain the wood from Room
than inside the jars, and the larger percentage of III-rather than as shelving (contra Dornemann
jars found empty, would imply that the explana- 1981: 32). The Hadidi Tablet Building contains in
tion for the perforated bases must be sought its inventory all of the equipment, including grain
elsewhere. The grain on the floor of Room VII (barley was preferred, but emmer is also attested
could represent a shipment from outside, delivered [Oppenheim 1950: 13]) necessary to a brewery.
just before the building's destruction-i.e., before Besides the pierced vessels there are smallerjars for
the grain had been emptied into protective con- storage and shipping (e.g., Dornemann 1981:
tainers (one assumes pottery ones); alternatively, figs. 6:2-6), tankards to fill them (Dornemann
the storage jars could have been standing empty 1981: fig. 4: 7-10), and perhaps even kraters for
at the time of the fire, or even may have been on-the-spot consumption.'?
plundered. But none of these possibilities explains It would be a final flourish to recognize in the
the holes in the bottom of the jars. "crude" features of the basalt seated statue in the
If one accepts that the perforations indicate a courtyard (Dornemann 1981: 33) the patron god-
specific function for the vessels, the textual sources dess of the brewing profession presiding over the
provide the information necessary to solve the productivity of her devotees. Even without invok-
puzzle. The cuneiform lexical texts refer to a large ing her, a good case can be made for identifying
pottery vessel with perforated base as a DUGNiG.DUR the Area H Late Bronze Age "Tablet Building"
BUR = namzitu, a vat used in the brewing process with the breweries described in the Mesopotamian
(Salonen 1966: 183, 189-93; Ellis 1977: 33-34). written sources.
The perforation in the base allowed the ferment-
ing mash to drip into a collecting vessel set under- ON CONTAINERS FOR SPECIFIC
neath (Ellis 1977: 33, 34, n. 28, cites wide-mouthed COMMODITIES
Mesopotamian vessels as examples). The pierced
pottery vessels found in the Tablet Building at Pottery vessels are produced as household wares
Hadidi (and incidentally classed by Dornemann as intended for continual reuse, and as containers for
"vats") correspond to the requirements both of the the storage of goods or the shipment of specific
namzitu and of brewing on a large scale. The commodities. While those functions can overlap,
smallest example illustrated, a pierced vat from the characteristics of the second group must be
Room VI (Dornemann 1981: fig. 7:4) has a capa- determined initially by the vessel's original raison
city of roughly 23 1.; the vats from Room III d'tre. If Binford (1968: 22) is correct that "data
(Dornemann 1981: fig. 3:2) and the courtyard, VIII relevant to most, if not all, the components of
(Dornemann 1981: fig. 8:1, 3), are much larger, at past sociocultural systems are preserved in the
roughly 90 1. and 175 1.8A second group of pierced archaeological record," and if at least some of
vessels, the large jars of 300-350 1. capacity9found those commodities-Crawford's "invisibleexports"
again in Room III and the courtyard, VIII, could (1973)-can be deduced from their containers, the
not have been easily handled; their constricted textual documentation again will provide the cor-
mouths indicate that they must have been used in relations necessary to decode the vessels.
another stage of the brewing, perhaps the initial The avenues for determining a vessel's function
soaking of the grain to prepare the "green-malt" have already been presented; they can be supple-
which formed the base product to be fermented mented here by further indexes particular to pot-
tery that is associated with the storage or transport guents; Maniatis et al. 1984; Geva 1982: 69-71 on
of perishable goods. This pottery may be local, pickled fish in the classical world). In the Near
where it is difficult to assess, or imported from East in the second millennium B.C. again the
another pottery "culture." The intruder on a local Aegean and Cypriot wares stand out as containers
ceramic horizon is easiest to spot, of course; it is for goods imported into the Levant and Egypt
therefore with this simpler issue that I will begin (Merrillees [1962: 287-92] discusses Cypriot base-
the discussion. ring juglets and the opium trade; Gittlen [1981:
Pottery shipped as part of a commercial trans- 55] refutes his contentions. Stubbings [1951] and
action rather than as a token of ethnic movements Shelmerdine [1984] present the Mycenaean stirrup-
had two possible purposes: either it was traded for jar as a container for precious scented oils.)
its aesthetic appeal-as luxury ware-or because Leonard's study (1981) on Mycenaean export jars
of its contents-as a container. The evidence is in the eastern Mediterranean can serve as a model
now irrefutable that Cyprus at least exported pot- approach to correlating the features of foreign pot-
tery as a luxury ware to the eastern Mediterranean tery, vessel morphology, and textual evidence (in
and Egypt during much of the second millennium this case Linear B archives) to traded commodities;
B.c."MIt has recently been found stacked for safe moreover, it illustrates the returns such an ap-
transport in a large storage jar in the Late Bronze proach can anticipate.3
shipwreck near Ka? (Bass 1986: 274, 279-80; Pulak The identification of containers designed for
and Frey 1985: 24). Indeed, the visual appeal of specific commodities from second millennium sites
these wares is certainly indicated by the efforts of in the Syrian interior is far more complex. There
Levantine potters to imitate them; wasters of local the ceramic horizon is fairly uniform, a trait shared
copies have been excavated at Ras Shamra in a with Mesopotamia proper; political and, espe-
potter's workshop (Saade 1979: 151; for imitation cially, cultural units extend over far broader areas;
Mycenaean pottery, see Baramki 1973: 192-97, and the active trade networks favored transactions
mentioning wasters; and Monchambert 1983: 26; inside those related cultural spheres-the Middle
at Tarsus, local Cypriot-style Black-on-Red and Bronze Mari archives reveal that restricted market
White-Painted wares were discovered in situ in a clearly. With few exceptions (perhaps Nuzi Ware?),
kiln [Braidwood 1940: 193 n. 1]). Some imported no obvious substitutes exist for the outstanding
vessels were unfit to contain anything, and can be ceramic imports found along the coast. On the
readily added to the luxury list. The Mycenaean other hand, the written documentation regarding
rhyton, for instance, is a funnel that could hold Middle Bronze trade in inland Syria is, thanks to
nothing.12There is thus no doubt that pottery was Mari, unusually rich and detailed. It therefore
prized and traded for its own merits, at least in invites a more refined analysis of the archaeo-
the second millennium Levant. However, Aegean logical assemblage to which it must refer. The
and Cypriot trade was confined to the coast for following discussion approaches the issue from the
bulk goods such as pottery (see n. 11), which is combined perspectives of texts and artifacts.
attested no further inland than the Transjordan
(Dornemann 1983: 21), the BiqCaValley (Kamid- MB TOMB GIFTS; CONTAINERS FOR BEER
el-Loz) (Hachmann 1970), Alalakh (Gates 1981), AND WINE; MARI PALACE ROOM 162
and Hadidi. (Dornemann [1980: 230], a Mycenaean
pyxis for perfumed unguents). The Cypriot milk- The funerary equipment deposited in wealthy
bowl sherd from Rimah (Lloyd 1938: 137) is a Middle Bronze Age tombs in Syria generally in-
singleton. But trade in other luxury wares may cludes an assorted collection of pottery. While the
explain the presence of foreign vessels in the Syrian very large collection in the EB IV/MB I hypo-
interior. geum at Til Barsip consisted solely of empty-
The second explanation behind the occurrence even new-vessels (Thureau-Dangin and Dunand
of nonlocal wares at a site is that they simply were 1936: 97; pl. 20; Dornemann [1979] reports on
containers for imported goods. Mediterranean and similar tombs at Hadidi, and Matthiae [1985] on
European scholars have argued that trade took those at Ebla; the coastal Levant provides many
place in specific perishable goods such as milk, examples), there is no reason to suppose that this
oil, perfume, or pickled fish and that such items northwestern practice was the custom in other
were transported in appropriate vessels (Renfrew areas. In the mid-second millennium cist tombs
1977: 3; 1973: 4; Leonard 1981, on Mycenaean un- south of Mari at Baghouz, where the preservation
B._l w 2ns.
BAGHOUZ Z 203
SAGHOUZ Z 122 AGHOUZ 203
MARI M. 857
TEROA TPR 4:18
MARI M.1581
Fig. 2. Assorted Middle Bronze containers. From Bag-

houz, after Du Mesnildu Buisson 1948:beer jarfromTomb
Z 122 (pl. 69), wine jars from Tomb Z 203 (pl. 78). From
Terqa, TPR 4:18 (Kelly-Buccellati and Shelby 1977: 25,
fig. 11). From Mari,"jarreglobulaire a anse bifide"M. 857
(Parrot 1959: fig. 84); bitumen-coated container M. 1581
from Room 162 (after Parrot1959: pl. 36); pitcher M. 898
MARI M.898 from Court 51 (Parrot1959:fig. 83).
was exceptional, the remains of a meal intended 166), the origin of those vessels at Baghouz, Mari,
for the deceased were found in shallow dishes on and Terqa appears to be the Levant.
a wooden table, as for example in Tomb Z 122 The possible function of the pitcher/flasks, be-
(Du Mesnil du Buisson 1948: 75; pls. 50, 51). It yond their origin, is pertinent to this study. Since
follows that the large ovoid jar beside the table no other container accompanied the meat offering
was a container for a beverage, in this case cer- in Baghouz Tomb Z 203, one must assume that
tainly beer since nearby lay a rasp-like metal the two replaced the standard beer jars and that
strainer commonly used with a straw for sipping they contained a drink appropriate to the funerary
the unstrained drink.14 Beer need not be exclu- meal. Water is of course a possibility, although it
sively linked to those jars, of course; but given the would be surprising to find a pedestrian drink
popularity of that drink, the Baghouz vessels must stored in what seem to be exotic wares. It can be
represent the standard beer jar, otherwise well argued that the camel bones buried with the tomb
known only from representations in art. suggest that the deceased was a nomad of western
Tomb Z 203 at Baghouz was the only Middle stock, and the flasks his native equipment.'6 But
Bronze/Late Bronze Age tomb that had pottery the size of these vessels seems inappropriate for a
containers of a non-Mesopotamian type, which canteen-ca. 30 1. for the Terqa example, 10 1. for
was also remarkable because of the camel (and those from Mari'7--nor would that explanation
sheep) bones found in front of its entrance (Du suit the large quantities excavated in specialized
Mesnil du Buisson, 1948: 83). The only ceramic areas of the Mari palace.18 It is more likely that
finds in the tomb consisted of two globular pitchers the pitchers contained wine, which the Mari arc-
or flasks with tall, narrow necks and a single loop hives list as a regular import into the palace during
handle made of two strands of clay and attached the reign of Zimri-Lim.
high on the shoulder (Baghouz type M; Du Mesnil Toward the end of the 18th century B.C., the
du Buisson 1948: 45, pls. 67, 78). The excavator commerce in wine delivered to the Mari palace
compared them to vessels from his soundings in involved several districts of inland Syria. The
the Qatna area, correctly noting that handles are Carchemish area was a major viticultural center
standard for the ceramic horizon of westernmost (Finet 1974/77: 122; see also ARMT VIII: 78,
Syria and Palestine and foreign to inland Syria where Karsum is now collated as Carchemish
and Mesopotamia (Du Mesnil du Buisson 1948: [ARMT XXI: pp. 110-12 (Durand) concerning
47).15 Identical pitchers ("jarresglobulaires a anse the merchant Yaqim-Addu's wine transactions]).
bifide") were discovered in large quantities at Mari Aleppo was also active in the trade as a distribu-
in the Middle Bronze palace, however, and more tion point for coastal wines and on its own ac-
recently a few have appeared from the Terqa ex- count (Finet 1974/77: 122).19A third region, the
cavations (from Mari, Parrot 1959: 116-17; from upper Habur, supplied wine, but on a more modest
Terqa, Kelly-Buccellati and Shelby 1977: 11; pl. scale, to judge from the Mari tablets. Finally, it
6:23). Their manufacture is peculiar, and described has been recently demonstrated that vineyards in
in detail in the Terqa report. The body was turned the Terqa area, upriver from Mari, were also com-
on the wheel as two deep bowls, which were then mercially exploited (ARMTIX p. 272 [Birot] with
attached rim to rim with a flat strip of clay. The collation in ARMT XXI p. 104 [Durand]). The
tubular neck/spout and the double-strand handle wine was shipped in jars that represented a stan-
were also set into the clay strip. Characteristic of dard of measure, for bills of sale refer to numbers
the potting technique are concentric wheelmarks of jars (DUGGESTIN= karpatu), never to liquid
on the body of the pitchers (carefully noted on the volumes (on karpatu, see ARMT VII: p. 351
drawings of the Baghouz and Mari examples) and, [Bottero]; Finet 1974/77; recently, ARMT XXI:
inside the vessel at the base of the neck, a projec- p. 192 [Durand]). The fixed capacity is judged to
tion that corresponds to the rims of the initial have been 10 qa, roughly equivalent to 10 liters.20
bowls (visible too in the Baghouz and Mari profile Once inside the Mari cellars, the wine was sealed
drawings). Since that technique was also applied and stored in its original jars (Finet 1974/77:
to the manufacture of Syro-Palestinian "pilgrim 126), or-depending on its quality-cut, perhaps
flasks," a hallmark of the coastal Late Bronze Age cooked, and rebottled (Durand in ARMT XXI,
but with possible MB II antecedents (Amiran 1970: p. 105-9). The fate of emptied jars is unknown;
one would think that some of them would have venience. Those jars have a capacity of roughly
been recovered in the palace inventory during 50 1. each (M. 1581: 52.47 1. + 3.67; the neck adds
excavation.21 a further 0.1 1.), which implies that they were used
The extent and diversity of the area that pro- for storage rather than transport. The bitumen
duced wine in Middle Bronze Age Syria preclude coating indicates they were used for liquids (but
that the wine containers shipped into Mari had not oil), but I cannot match them to any parallels,
anything in common beyond their standard vol- in part because no parallels have been attested for
ume, ca. 10 1. One would suppose that the Habur them. However, what might be described as "deco-
wines were readily distinguished by their contain- rative" markings painted on them are reminiscent
ers, for which the striped Habur Ware jars have of a practice illustrated at Terqa (Kelly-Buccellati
been proposed as likely candidates.22 A typical and Shelby 1977: 7) and especially at Baghouz (Du
example from Mari (M. 1584: Parrot 1959: 134, Mesnil du Buisson 1948: 48-51). Du Mesnil du
fig. 92 and pl. 36) is small at ca. 8 1., however, and Buisson read the markings as labeling the contents
its characteristic wide mouth would make stopper- of the jars. Or, might they be shippers' marks?
ing for long-distance transport a difficult task.23 Since similar markings occur on a selection of the
Better options are the ovoid jars with short, jarres a anse bifide (Parrot 1959: 117), the two
narrow collars such as M. 1585 (Parrot 1959: pl. vessel types must be related in some way.
36) at 15.68 1. + 1.23 1.; M. 1590bis (Parrot 1959: A fair number of the jars from Room 162-to
pl. 36), no dimensions given; and M. 1590 (Parrot judge by the site photo of the room and of the
1959: pl. 36: 11.83 1. + 0.93 1.). These Habur im- balks, which are piled high with sherds-belong
ports were stored with at least 30 other vessels in to yet another category of container. These are
Room 162 of the Zimri-Lim palace as part of a large, ovoid pitchers with a strap handle con-
deposit resembling a magazine (discussed below). necting shoulder to rim, and a characteristic tri-
Other jars in the Room 162 deposit include the lobate spout (cf. M. 898 from Court 51: Parrot
"jarres a anse bifide" (above), which on the cir- 1959: 114-15; fig. 83, pl. 35). Parrot does not
cumstantial evidence from Baghouz may be pro- include the Room 162 examples in his pottery
posed as wine jars from western Syro-Palestine.24 catalogue of the type, but he does describe them
Their presence at Terqa can be explained as tax in the architectural discussion of the room in
payment-the miksum, payable in silver or in kind sufficient detail that their identity with the pub-
(Finet 1974/77: 123)-on imports shipped down lished piece is certain (Parrot 1958: 32; fig. 33).
the Euphrates by boat. But the vast majority were The site photographs show these pitchers to be in
recovered from the Mari palace, a sign that their the same size range (ca. 0.40 m high) as the jarres
contents were destined for a royal palate. The Mari a anse bifide, and thus as the published pitcher
records indicate clearly that the wine commerce M. 898; it, and by extension the Room 162 pitch-
involved direct trade between supply point and the ers, represent capacities of roughly 11 1. (11.77 1./
palace. Since the only stopping places were for the + 0.82). The affinities of those pitchers with coastal
payment of taxes, the absence of such jars at Syro-Palestine lie in their trilobate spouts and
intermediate sites is to be expected. handles,25 which are foreign even to such Middle
Finally, two other types of vessels from the Bronze sites as Hadidi. Their size makes them suit-
large repository in Room 162 need to be discussed able for shipping wine or oil, another staple the
(Parrot 1958: pi. 16.1). The first is represented by Mari documents indicate was delivered in jars that
two jars, M. 1581 and M. 1582, similar in size and served as transport units of measure (DUGkarpatu).26
shape, found lying together (Parrot 1958: pl. 16.2; It must be emphasized that, given the variety listed
1959: 132-33; fig. 91). They are large, ovoid, with in the Mari archives of products that must have
tall, straight necks and exterior surfaces painted been shipped in ceramic containers and with the
with bitumen applied in a broad band across the further illustration furnished by the Ka? shipwreck
base and the shoulder up to the rim. The plain (see n. 13), the options for reconciling jar with
area on the body of each jar is further "decorated" product are numerous. However, the identification
in bitumen with a Catherine's wheel. Parrot (1959: of the jarres a anse bifide as wine jars seems at this
133) refers to a third jar coated in the same stage reasonable. The intended functions of the
fashion, but with a "shield with five radiating other types of jars from Room 162 will be pursued
lines" substituted for the wheel; he gives no pro- (below) from another perspective, that of the role
of Room 162 and the northeastern wing of the millennium (Meijer 1980: 123). The context seems
Mari palace in the light of the Mari archives. a few centuries early for the horse in Mesopotamia,
but not necessarily for a site in northwestern Syria
NONCERAMIC ARCHAEOLOGICAL FINDS on a transit route to Anatolia. Middle Bronze Syr-
AND THE TEXTUAL RECORD ian economic texts refer to an active trade in horses
from Qatna, Carchemish, and Chagar Bazar; but
Because the written documents preserve refer- those cities probably acted as market centers for
ences to practices or situations in the cultural breeding grounds in Anatolia (Sasson 1966: 176).
sphere that could otherwise be neither posited nor The discovery of a red-burnished depas in a con-
even imagined, one can turn to them for explana- temporary Selenkahiye tomb also indicates Ana-
tions when the archaeological record contains tolian connections and reinforces the implications
puzzling features. The possibilities of this ap- of trade that the horse bones supplied. If there is
proach have already been demonstrated in the case anything at all surprising about the second mil-
of the mash-tuns from Tell Hadidi. Other non- lennium horse trade, it is that more horse bones are
ceramic categories of finds respond to the same not reported from archaeological contexts. Per-
method. haps the gaps have resulted from the special
requirements surrounding the disposal of dead
Animal Bones horses.
The occurrence of animal bones in unusual con- Metal

texts, and of unusual species in otherwise unre-
markable contexts, can sometimes be explained by Metal also falls into the category of "perishable
documentation preserved in the ancient texts. goods" or "invisible exports" in the archaeological
Parrot (1964: 11, fig. 5) discovered a complete record, since the raw materials were both alloyed
gazelle skeleton in a deposit of animal bones and worked after delivery, and often reused for
associated with the Early Dynastic Dagan temple centuries (see ARMT XXIV: [Talon] pp. 219-20;
at Mari, mid-third millennium B.C. While explana- Durand 1983). Reconciling the ancient text refer-
tions could be contrived for the other bones- ences to metals with the relatively scanty number
perhaps as sacrifices to the deity-the complete of actual finds has been clouded by the difficulties
skeleton and the species were both unexpected. in understanding the exact Sumerian and Akkadian
Parrot lamely suggested that the gazelle wandered technical vocabulary on one hand, and in tracing
into the temple enclosure and died there of exhaus- the various metals back to their sources, whether
tion, but he admitted that such an animal would through textual allusions or scientific analysis, on
seem out of place in a residential area, and even the other (most recently Moorey 1985). Scholars
more so in a sacred one. Yet references in the writing on the metals trade have accordingly split
Chagar Bazar archives several centuries later may into two camps, one comprising those for whom
indeed demonstrate why a gazelle would be appro- the third and second millennium archives represent
priate to a temple. Among the Chagar Bazar primary evidence to which the archaeological data
inventories are grain disbursements to an employee must conform (e.g., Leemans 1968; Dossin 1970;
responsible for the care of birds and gazelles, an Heltzer 1977); and the other, championed by
association that implied to Gadd that, like birds, Muhly (e.g., 1977), underlining the difficulties in
gazelles were used for divination (Gadd 1940: 32). understanding the impressions given by the texts
If that is true, then the Mari gazelle did belong in when those texts are contradicted by archaeologi-
the Dagan temple, and its presence in the archae- cal implications. A key topic in the scholarly
ological record reflects an ancient situation. And debate of the past several decades has been the
while the Chagar Bazar evidence need not be ancient Near East's source(s) of tin, a metal used in
considered conclusive, it at least offers an option increasing amounts after 2000 B.C. for manufac-
that relies on more than rationalization. turing bronze. The recent discovery of major tin
It is easier (although not necessarily more valid) deposits among the polymetallic ores in the Bol-
to recreate the circumstances leading to the deposit kardag range of the Cilician Taurus Mountains
of horse bones, and presumably a horse, in EB (southern Turkey; Yener and Ozbal 1987),27and
IV/MB I Selenkahiye, at the close of the third the concurrent projects to analyze Bronze Age
metal artifacts from the ancient Near East (Yener duce an arsenical bronze through natural alloying;
1983; Berthoud 1979) promise to clarify many it is even doubtful that arsenic was ever isolated
uncertainties affecting the texts and the Sumerian/ from this combination by ancient Near Eastern
Akkadian terms. In the context of this paper, I metalsmiths (Yener and Ozbal 1987; Moorey 1985:
intend simply to present one potential concordance 4). Thus a word for arsenic need not exist. The
as argument for the virtues of a combined front in polymetallic copper ore-perhaps URUDU in the
yet another category of materials. Ebla texts in contrast with A.GAR5.GAR5 (Waetzoldt
The complexities involving the translation of the and Bachmann 1984: 4)-was not refined of its
Akkadian term annaku will serve as example here. desirable other elements before being worked; so
Annaku generally appears to mean "tin," although the Sumerian language had no impulse to break
in certain contexts "lead" seems the more appro- down the technical word syntactically to reflect its
priate metal (see below regarding "arsenic"); the various component metals (Moorey 1985: 5, 15-
debate is well summarized by Moorey (1985: 124, 17). Behind the word annaku may lie the same sort
127). Annaku occurs, moreover, in combination of disregard for technical categorization. Annaku
with silver in the mid-third millennium B.C. Ebla represents a combination of ores as they exist in
texts, in lists grading inventories of gold and silver mines exploited by ancient Near Eastern metal-
according to their purity and type (Waetzoldt lurgists. Whether it means tin, lead, or arsenic is a
1981: 366). The association with the other grades distinction that only modern scholars demand.
of metals in the Ebla lists implies that the silver- In contrast to the exercises demonstrated in
annaku alloy recorded there is a natural alloy, not the previous sections of this article, the contri-
an artificial one. The confusion is easily resolved bution of the archaeological data to the textual
(perhaps too easily) if polymetallic ores such as the record is the primary source for resolving diffi-
ones from the Bolkardag mines were the metals the culties in the translation of technical terms from
Ebla craftsmen and their colleagues elsewhere the Sumerian-Akkadian glossaries. Their ultimate
used. Silver from the Bolkardag range occurs in solution cannot be reached through abstract refine-
high concentrations in combination with tin, gold, ments proposed by the philologists, but rather
lead, zinc, and iron (Yener and Ozbal 1987). It was through the careful examination of correspon-
a major resource, since over 800 mines were dences between the termini technici and the per-
worked, some as early as the Neolithic period. The tinent materials in the context of their ancient
Ebla silver-annaku alloy would correspond readily technologies. Although many ancient materials
to such an ore, with annaku referring to both tin have disappeared without trace, metals, and the
and lead, which exist at Bolkardag in combination. increasingly refined scientific tests to which they
Furthermore, it may well be that the term annaku are being submitted, preserve trace elements of
referred to precisely this sort of combination, in great promise.
which the desired metal was one or the other, but CORRELATION OF TEXTS WITH
the same term could cover both. That would ARCHITECTURAL UNITS
certainly satisfy the situation of much third mil-
lennium metallurgy, where there is increasing evi- The preceding discussions associated textual
dence for the role of natural alloying through the references to single items in the archaeological
use of polymetallic ores in developing metal tech- record-brewing vats, beer and wine jars, faunal
nology (Yener and Ozbal 1987). remains, metals-which then admitted broader
The circumstances behind the absence in the implications about the contexts to which the finds
Sumerian-Akkadian vocabularies of a word for refer, the brewing vats serving to identify a brew-
"arsenic"-the significant alloy with copper for the ery, the wine jars acting as tokens of extended
production of arsenical bronzes, especially in the commercial activity. Those items acquired a char-
third millennium (see Waetzoldt and Bachmann acter far more vocal than their basic features
1984)-can be explained in precisely the same would suggest: instead of simply acknowledging an
fashion. Copper occurs in polymetallic combina- imported vessel, one can envisage trade in a
tions with a number of other metals, among them specific commodity. Just as the structuralist his-
arsenic; that is again the case in the Bolkardag torians (notably Liverani) caution us to see modes
range. Working a copper/arsenic ore would pro- of behavior rather than basic historical fact behind
ancient written documents, so we can understand ing to interpret the parts of an architectural unit as
from the archaeological remains further shades of intricate as the Mari palace.29 The written testi-
significance beyond the material manifestations of mony forms an inalienable corpus of material that
a given culture. must be integrated into the architectural entity.
In the case of economic archives recovered from Rather than present an exhaustive application of
an architectural complex, one can expect that the possible allusions in the Mari texts to specific areas
written documents can shed light not only on the of the palace, I will illustrate the potential of this
intellectual world of that complex, but on the method by investigating references to two depots
physical setting of that world as well. More simply, familiar to Zimri-Lim's bureaucracy, then by trac-
the written testimony describes not only the activi- ing the sources of those references to their appro-
ties of people during the period covered by the priate architectural contexts. The named depots
archives, but places those activities within specific were picked at random, not selected over others
physical surroundings. Indeed, the archaeologist because they ultimately could provide answers.
usually labels architectural features by function: One would expect comparable results with the
room or court, kitchen or throneroom, house, other named sectors of the palace.
palace, or temple. He assigns those functions on
the basis of structural characteristics, of interior or Zimri-Lim's Palace at Mari: Bit Raqqi and Bit
exterior furnishings, and of finds recovered from Kuprim [Table 1]
the architectural units themselves. Written records
from a complex may identify the entire building: Among the over 1,000 tablets excavated and
Zimri-Lim's palace at Mari, for instance; or sug- now published in toto from Rooms 160 and 215 of
gest its purpose, as with Delougaz' interpretation the Middle Bronze Age Mari palace (ARMTXXI
of the north palace at Tell Asmar as a factory [Durand] and ARMT XXIII/AAM I [especially
(Delougaz 1967; cf. Margueron 1979: 9-13). To Soubeyran]), 30 refer specifically to the profes-
take this approach one step further: when the sional activities of one Nur-ili, chief of the scented
archives from an architectural complex are exten- oil workshop during Zimri-Lim's fourth regnal
sive, and refer to or imply specific activities taking year (Z. L 4': "year of the throne of Samas"). In all
place in specific parts of the complex, it should be 30 transactions, Nur-ili is either given his title
feasible to apply their terms to the associated (LUraqqi) or receives oil shipments for the perfume
architectural components of the complex, and workshop (bit raqqi).30The 30 texts constitute the
identify them. But it is again essential to consider sum of excavated (or at least published) records
the written and architectural evidence together, associable with Nur-ili's work in receiving and
and not merely to illustrate in general terms the treating oil.31 They cover the brief period of some
one with the other. three months, and thus represent a closed set of
The Middle Bronze Age palace at Mari provides particulars involving Nur-ili's workshop (Soubey-
ample opportunities for an investigation along ran in ARMTXXIII, pp. 415-20).
those lines. The palace archives allude by name to a His dealings with other businesses inside (and
number of areas within the palace; the palace in perhaps outside) the palace are limited to three
turn exhibits a variety of architectural peculiarities palaces: bit kuprim, the "bitumen depot/store-
and furnishings of potential significance. Efforts to house"; E.I.SAG, the depot for processed oil; and
locate the "Court of the Palms" and its adjacent E.fAddu-duri, the "household" of Addu-duri. The
shrine/papdaum (Parrot 1958: 57-Court 131; bit kuprim supplied him with the primary raw
Al-Khalesi 1978-Court 106; Margueron 1982: material for his industry: oil (i.GIS) or refined oil
360-63-Court 131?) have been inconclusive, in (i.GIS.BARA2-GA) [cf. ARMTXXIII: 492 and p. 416,
great part because stripping the area of its precious where the two qualities are distinguished (Soubey-
furnishings has eliminated for us its most charac- ran)]. His suppliers there are usually named: on
teristic features.28Nevertheless, other sectors of the one occasion mIli-asraya, on the others mBalumu-
palace, and the service quarters in particular, may namhe (ARMT XXIII: 479, 484, 490 from Room
well respond more receptively to the attributes 215), both well-attested figures in the Mari oil
assigned them in the archives. It would be irres- trade known from inventories elsewhere in the pal-
ponsible to ignore those indications when attempt- ace (see ARMT XI [Burke]). The E.I.SAGreceived
Fig. 3. a. Mari,Zimri-LimPalace Room 162 (after Parrot

1958: pl. 16:1). b. Mari, plan of eastern section of the
Zimri-Lim Palace (after Margueron 1982 II [plates]:
fig. 148).
Nur-ili's finished products, perfumed oils; for the

period covered by these texts, he delivered them in
several installments toward the end of the series. It
also supplied him with untreated oil.32 Whatever
this depot's functions, it apparently was not the
same storehouse as the bit kuprim (see Soubeyran
in ARMT XXIII: p. 418) as is certainly indicated,
too, by their different names (below). Finally, the
Addu-duri oil reserve is mentioned in only two
tablets, once as supplying oil (ARMT XXI: 112,
from Room 160), and once in a recapitulative
(ARMT XXIII: 482, from Room 215), written 16
days later and no doubt referring to the same
shipment. By far the majority of Nur-ili's external
transactions concern the bit kuprim, and therefore
we need to focus on this depot, and the perfume
workshop itself.
First, however, it is essential to review several
questions that bear directly on the extent to which
one can use the Mari texts. 1) Do the transactions
involving Nur-ili represent receipts or invoices? Do
variations in the terminology of those records
allow one to recognize them correctly? 2) Did the
supplier and the receiver keep their respective
documents with them (or, given the nature of those
documents, close to hand), in an office near their
b. workplace, or in an unrelated office? 3) Of equal
TABLE 1. Texts from the Nur-ili file, in chronological order
Nur-iliFiles from processedoil depot from bitumendepot from reserveof A-duri

[ ... sa E.I.SAG] [ ... sa bit kuprim] [ ... sa E fAddu-duri]
Room 160 113(req.) 111(inv.) 112(req.)
114(inv.) 116(inv.)
108(inv.)
Room 215 486 (inv.) 469 (inv.)
488 (inv.) 470-473 (inv.)
480 (inv.)
[482:recap.] [482:recap.]
483 (inv.)
492 (frag.)
to processedoil depot to bitumendepot to perfumeworkshop

[ana E.I.SAG] [ana bit kuprim] [ana bit raqqi]
Room 160 *107(rec./inv.)
Room 215 474 (rec.) [490:recap.] 475 (inv.)
477 (rec.) **(479:rec.
/ req.)
***(484:rec./req.)
485 (inv.)
487 (inv.)
363/357/358
beerdeliveries 359
[359':Room216]
Room 160:ARMTXXI
Room 215:ARMTXXIII
inv. = invoice (SUTIANur-ili) * rec. from Balumunamhe/inv. to Nur-ili
rec. = receipt(PN amhur) ** rec. from Ili-agraya/req.to Nur-ili
req. = requisition(sILA2
Nur-ili) *** rec. from Balumunamhe/req. to Nur-ili
recap.= recapitulative
import, but on a different level, do the findspots of ili (Nur-ili amhur], two texts, from Room 160
the Mari tablets reflect their original locations, or (ARMT XXI: 109, 115) record receipt of per-
did the looting of the palace (and perhaps post- fumed oil by unnamed individuals [-amhur],
excavation sorting) appreciably confuse the indi- locus unspecified. A third text, also from Room
vidual archives? We can reasonably address those 160 (ARMT XXI: 107), uses the formula with
issues here because of the restricted nature of the variation: -amhur .. . SUTIA Nur-ili: Nur-ili is the
Nur-ili files, and we can, in fact, reach a tentative receiver, in a significant redundancy of mots-clefs
resolution. (to be discussed below). Finally, ARMT XXIII:
The texts describing deliveries to the perfume 479 and 484, from Room 215, list both the supplier
workshop [ana bit raqqi] use two formulas:33SILA2 [mIli-asraya amhur, mBalumunamhe amhur] and
[PN] Nur-ili, "assignment to Nur-ili";34and SUTIA the receiver [SILA2Nur-ili].
[PN] Nur-ili, "received by Nur-ili," which is far At first impression, the SILA2PN and the SUTIA
more common.35 A third formula, which occurs at PN texts appear to be invoices or bills of sale,
times in conjunction with one of the other two, presumably handed over with the delivery-in this
acknowledges receipt from an individual, usually case, of oil-and kept by the receiver or the office
named: [PN] amhur, "I have received from [PN]."36 from which he depended (see Rouault [ARMT
Besides the texts (above) that record deliveries of XVIII: p. 174] for SILA2 PN). In contrast, texts with
scented oil to the processed oil storehouse by Nur- the formula PN amhur were receipts given to the
PN when he delivered his goods; he would later cashed them, in Balumunamhe's case, in the vicin-
cash them in for payment (ARMT XVIII, p. 175). ity of Room 160.
Thus Nur-ili, Balumunamhe, and his colleagues However, three PN amhur texts were also found
would keep the tablets that contained their names in Room 215: two issued to Balumunamhe, the
with amhur, attesting to their respective deliveries; third to Ili-asraya, far from the putative oil center
and Nur-ili or his controlling office held onto in Room 160. One is certainly a recapitulative
invoices reading SILA2 Nar-ili and SUTIA Ntr-ili. (ARMT XXIII: 490) compiled as the last text in
The distinction between the SILA2and the SUTIA the oil delivery series, and it mentions neither Nur-
texts is not obvious, although it must signify more ili nor perfumed oils; it can therefore be excluded
than an individual scribal preference. The issue is from consideration here. The other two texts,
further clouded by the use of either formula in the ARMTXXIII: 479 and 484, are atypical receipts in
receipts [PN amhur], specifically ARMT XXI: 107 that they also use the formula SILA2 PN. To explain
(SUTIA) from Room 160; and ARMT XXIII: 479, their presence in Room 215, one must understand
484 (SILA2)from Room 215. The subtleties implied the reason for which this formula was suitable to
by SUTIA may be explained by comparing No. 107 the texts.
with ARMT XXIII: 469, from Room 215. Those The distinction between SILA2and SUTIA may be
two texts, issued on the same day, seem to be the resolved by a second pairing of texts, again written
receipt and the invoice for the same shipment. on the same day:38ARMT XXI: 112 from Room
160, and ARMT XXIII: 475 from Room 215.
From Room 160:ARMT XXI: 107 (translatedby
Durand) From Room 160:ARMT XXI: 112 (translatedby
0.1.2 I.GIS BARA2!-GA / balumunamhe / amhur / Durand)
ana E raqqi / SUTIA/ nur-ili / ITI abim / U4 25- 12 DUG I.GIS / sa E fISKUR-dri / ina kunikkisa /
KAM. Z. 4' kanka / SILA2nur-ili / ITIhibirtim / U4 8 (?) KAM
[I received0.1.2.0filteredoil from Balumunamhe [12 oil jars from the "household"of Addu-dfiri,
for the perfume workshop;receipt by Nur-ili. sealed with her seal; assignmentto Nur-ili;/
MonthIV, the 25th."Z.4"' MonthV. the 8th.]
From Room 215: ARMT XXIII: 469 (translated FromRoom 215:ARMTXXIII:475 (translatedby
by Soubeyran) Soubeyran)
1 DUGI.GIS[O o] / sa ina DUGnaspakim / ina E I KUR I.GIS BARA2-GA / ana E raqqi / SUTIA /
kuprim / issaknu / SUTIA/ nur-ili / raqqi / GIR nur-ili/ ITIhibirtim/ U48-KAM / Z. 4'
dsu'en-muballit / ITIabim / [U4]25-K[A]M / Z. 4' [120 (qa) of refined oil for the perfumedoil
[one jar [ ] oil which was in the storejar of the workshop;receiptby Nur-ili;MonthV. the 8th.
bitumendepot;receivedby Nur-ilithe perfumer; "Z.4'"]
Sin-muballitwas comptroller.37
Month IV, the
25th."Z.4"'] These texts follow the format of the pair No.
107-469, but with the signal replacement of SUTIA
The differences between the two bills can be by SILA2in No. 112. Perhaps the unusual source for
explained by their reciprocal purposes. Number the twelve oil jars prompted a departure from the
107 specifies the precise amount of oil supplied by more common wording. Of all the deliveries made
Balumunamhe, because its quantity was of con- to Nur-ili during the months covered by these
cern to him when he redeemed this bill against texts, only No. 112 records a shipment released by
payment-whereas Nur-ili's only interest was the the "household" of Addu-duri, a circumstance that
actual delivery. On the other hand, Nur-ili's in- apparently required another step in bureaucratic
voice specified the source of the oil (which could procedure. Behind the mot-clef's translation as
come to him from several places), information that "assignment to" there seems to exist a transaction
was superfluous to Balumunamhe, whose regular involving a requisition and its approval. Receipts
reserve and work center this was. One might fur- No. 479 and 484, naming Ili-asraya and Balumu-
ther imagine from the two texts that the invoices/ namhe in conjunction with the formula SILA2PN
bills of sale remained within the area of Nur-ili's [Nur-ili] must correspond to the exceptional deliv-
activities, centered on Room 215; the receipts [PN eries, rather than disbursements from the standard
amhur] were under the deliverer's control until he depots. Their placement among the Room 215 files
would then result from a different office routine, in suggested that the broad distribution throughout
which the bill of requisition and its approval the palace of related records may represent a
remained in the hands of the receiver.39 deliberate effort to decentralize the bureaucracy,
One other problem to be resolved is the integrity thus reinforcing the absolute control of the resident
and significance of the tablets' findspots in the king. But the complicated system of invoices,
Mari palace. It was evident from the first readings receipts, and requisitions necessitated the keeping
of the tablets that the individual archives had to of records by all the related individual offices
some extent suffered either from hasty refiling or attached to the various palace workshops and
from looting in the final days of the palace (Parrot magazines. Even transactions as simple as those
1958: 80-81, citing Thureau-Dangin and Dossin). undertaken by Nur-ili were to require bookkeeping
The confusion has been most evident in the over- in every place with which he conducted business.
lapping of dossiers excavated in Rooms 160 and If the preceding interpretation of the office
215 (ARMT XXIII: i-iii), and in joins between formulas indeed reflects correctly the normal sys-
tablet fragments excavated in widely separated tem of invoices, receipts, and requisitions, the
areas. In the case of the Nur-ili file, three invoices association of the oil suppliers with Room 160, and
among the nine texts from Room 160 associable Nur-ili with Room 215 can now be related to the
with the perfumer fill a chronological gap in the architectural contexts in which those archives were
Room 215 sequence (ARMT XXI: 113 [V. 25], discovered.
114 [V. 28], and 108 [V. 29], "Z. 4'," correspond- The Nur-ili dossier is restricted, as we have seen,
ing to the gap in invoices between ARMT XXIII: to the archives from Rooms 160 and 215. Nur-ili's
484 [V. 24] and 485 [VI. 11]). A fourth invoice fits activities centered on two places, the perfumed oil
into the very beginning of the series.40Two receipts workshop and the "bitumen depot." He also de-
for Nur-ili's perfumed oils (ARMT XXI: 115 livered a finished product, scented oil, to a third
[VI. 11] and 109 [VII. 5]) evoke no visible classifica- locale, the depot for processed oil, which also
tion except perhaps that they fall late in the series supplied him with one-third of his base product
and seem random in contrast with the cluster of deliveries.42The association of these two or three
four Nur-ili receipts (the date is lost on a fourth) areas with Rooms 160 and 215 can be considered
from Room 215 dating to mid-V (ARMT XXIII: as follows.
474 [V. 7], 477 [V. 10], 478 [V. 13], [491, n.d.]).
The two texts ARMT XXI: 107 and 112 have Bit Raqqi: The Perfume Workshop
already been discussed. A ninth text (ARMT XXI:
116), another invoice, is undated. Of the 18 tablets that refer to an oil delivery to or
If any pattern emerges at all from the Room from a specified place, six concern the perfume
160 sample, it is that the first three delivery bills workshop. Five of these were found in Room 215:
listed here (Nos. 113, 114, 108) form a series miss- three are standard SUTIA PN invoices (ARMT
ing from Room 215-conceivably stored in one XXIII: 475, 485, 487) which we assume accom-
basket (along with many other tablets) referring panied the delivery. The other two, ARMTXXIII:
to transactions of other materials evacuated from 479 and 484, are the PN amhur texts that attest to
Room 215 and abandoned in 160 just before its further bureaucratic procedures formulated with
bearers reached the palace's main (and only exca- the mot-clef SILA2 [PN], "assignment to." The sixth
vated) gate a few rooms beyond it.41That explana- text is the receipt from Room 160, ARMT XXI:
tion could also extend to the two perfumed oil 107, whose reciprocal invoice (ARMTXXIII: 469)
receipts; however, the two reciprocal pairs (Nos. is in Room 215. If this PN amhur receipt was
107-469 and 112-475) cannot be explained in this indeed kept by the supplier, it and the five other
scheme, and would rather point to Room 160 as taibletsstrongly suggest that the bit raqqi was in the
a proper archive. The evidence will be confirmed vicinity of Room 215 with its office. If it was not,
from the archaeological sector (below). Mari texts this first argument collapses. However, we can
which refer to records' inventory procedures show confirm that Room 215 was a dependent workshop
that the filing and storing of tablets followed a logic because of four additional tablets from the Room
impenetrable to outsiders (Sasson 1972: 55-66). 215 archive and a fifth from Room 216, next door.
Rouault (ARMT XVIII: p. 250 and 250 n. 100) Those were invoices for beer delivered to the
perfumed oil workshop.43 If the bit raqql can be as Room 122 (Parrot 1958: 287) and 221 (Parrot
identified at all solely from the archives, the 1958: 293). Four jars found in Room 221 had
complex formed by Room 215 and its environs capacities from 263 1. ? 20.5 (ar a) to 382.5 1. ? 30
seems a likely candidate. (jar d), for a total volume of well over 1250 1.).
Room 215 is a small room in the southern Margueron has argued elsewhere (1979) that the
central quarter of the Mari palace. It controlled Bronze Age Mesopotamian palaces did not contain
access to three long, narrow rooms, 216, 217, 218; workshops-the workshops were in the city out-
and could in turn communicate directly with the side the royal complex. Indeed, the other economic
(badly eroded) service quarters to the south and tablets from Room 215 refer to a variety of
the large central Court 131 to the north, by means activities: clothing and textiles, leatherworking,
of an uninterrupted corridor, 120, which follows a cabinetwork, and food supplies, as well as the
north-south axis. Room 215's location is ideally bureaucratic responsibilities of the official Asqu-
suited for deliveries and shipments both outside dum (ARMTXXIII: 250-51). However, they need
and inside the palace. It was furnished with two not be mutually exclusive. Perfume making by
small adjacent hearths that occupied the recessed distillation, a technique known in Mesopotamia at
north end of the room. Tablets were recovered least by the second half of the second millennium,
from the floor itself, and were thus presumably involved a simple apparatus and low heat. More
not from a collapsed upper story (Parrot 1958: important to the process was time: the mixtures
288-89). needed to sit for long intervals between heating
Although the features of Room 215 are not (Levey 1959: 36-41), and thus required ample
particularly instructive, Rooms 216, 217, and 218 storage facilities for the settling vessels, as Rooms
contained installations of an unusual type. All 216-218 would have afforded.44 Were the per-
three were lined with continuous low mudbrick fumed oil prepared simply by infusion, the neces-
benches allowing for only a narrow aisle between sary equipment would be even reduced: cloth for
them down the length of the room. In the upper squeezing and straining, and jars (Lucas 1948: 105-
surface of the benches were square or circular 7 for the Egyptian evidence; Leonard 1981: 96-99
depressions, some still preserving a rounded pot for Mycenaean perfumed oils and unguents; Shel-
base, but the majority filled with burnt wood. merdine 1985). On present evidence, those rooms
Parrot identified this wing as a workshop, insist- appear suitable for either process. The coincidence
ing that the burnt wood was fuel (charcoal) for of the area's physical plan and preserved fur-
the metalworking attested in the Mari archives nishings with the indications of the associated
(Parrot 1958: 288-92). Margueron has since cited archives makes the location of Nur-ili's workshop,
a number of arguments to refute Parrot (Mar- bit raqqi, in or near the Rooms 215-218 suite
gueron 1982: 336-38): excessively cramped quar- plausible.
ters, lack of ventilation for an industry producing
much heat and smoke, absence of relevant equip- Bit Kuprim: The Bitumen Warehouse
ment such as molds, tools, or even metal scraps,
and installations unsuitable as furnaces. He notes The bit kuprim, or "bitumen warehouse" of the
that while the excavation report makes only pas- Mari palace was furnished with large jars, nas-
sing mention of ceramic finds from those rooms, paku, in which at least two varieties of oil were
the publication photographs show numerous sherds stored; these in turn were transferred into unit jars
that the diggers piled up on the wall tops. The and delivered to Nur-ili's workshop (and we must
reports do not state that the "charcoal" came assume to other places). Such information is pro-
exclusively from inside the depressions. Parrot's vided by the Nur-ili dossier, which includes ten
charcoal is more likely to be burnt wood from invoices (SUTIA Nur-ili) citing the "bitumen ware-
roofing, interior furnishings, or accessories such as house" as the oil reserve supplying the goods.
wood potstands that burned when the palace was Two of the invoices come from the Room 160
destroyed. Margueron concludes that these rooms archive (ARMT XXI: 111 on Month IV. 26, at
were depots for medium-sized jars, rather than the the beginning of the series, 116, n.d.). The other
permanent storage vessels (which must be nas- eight were found in Room 215, along with two
paki) of the type found in neighboring areas such tablets citing the "bitumen warehouse" in reca-
pitulatives.45 The Nur-ili files would therefore ens," to ancient situations. He admits that this
point toward a location outside the immediate sector, which he calls "C," must have had a broad
environs of Room 215 for the "bitumen ware- range of functions, given the varying room types
house." The invoice-receipt pair (ARMT XXI: and their excavated contents: domestic (bathrooms,
107; ARMT XXIII: 469) with the oil invoice No. tableware), archival (scattered tablet finds, and the
469 from Room 215 citing the "bitumen ware- lot from Room 160), and storage (numerous vessels
house" as its source, suggests that it was in the from Room 162); he tentatively concludes that the
vicinity of Room 160. unit played an administrative role. It may indeed
The Nur-ili dossier is not the only one to be possible to define that role more closely.
mention that warehouse. An oil delivery receipt The principal evidence that relates "Sector C"
held by Ili-asraya and dated to another year, from with the bit kuprim/"bitumen warehouse" of
Room 134, accounts for oil deposited in the bit Zimri-Lim's palace rests with the final phase of
kuprim (ARMT XXI: 122: year Adad of Aleppo). Room 162 and its excavated remains. Room 162
But the room is also well attested for other was equipped, at the time of excavation, with at
functions. It is mentioned often as a repository for least 30 complete jars (and probably more) of
official weights (ARMT XVIII: 67-68: steward- medium to large capacities, discussed above, as
ship of Mukannisum [Rouault]), and for precious wine and oil containers. As a group, they invite
metals measured out with the weights.46The locale a broader perspective than as individual pieces.
apparently performed a variety of functions, but These vessels include a variety of types: handled
storage and supplying of bitumen are not noted.47 (a anse bifide), pitchers with trilobate spouts, local
Did the name bear no connection to bitumen? globular jars with restricted necks and no handles,
Was the depot known because of its bitumen- Habur jars, large storage vessels coated with bitu-
lined floors, a refinement prompted by the activi- men. In sum they form a mix of local and foreign
ties (pouring out oil?) there? Or was it understood pottery that nonetheless shares a common feature
that, since bitumen was also kept in the ware- (Parrot 1958: 32). The jars were exclusively con-
house, this specialized product distinguished it tainers for liquids, and fall into standardized
from others in the palace where precious metals, medium-large and large sizes: 10-12 1., and, for the
weights, and oil were also stored? The uncertain- bitumen-coated jars, 50. 1. They were discovered
ties make any identification for the bit kuprim still stacked in two lots, one in the room's west
tenuous, but the archaeological context of Room corner, the other against its south wall (Parrot
160 and its dependencies can help to clarify its use. 1958: photographs fig. 33, pl. 16:1). The arrange-
Room 160 resembles an open court in plan,48 ment recalls the rows of wine/oil jars in a Late
and is paved with baked bricks. It forms the Bronze warehouse at Minet-el-Beidha (Schaeffer
central element of a self-contained suite of rooms, 1939: pi. 9), and like it, certainly represents a
158-167, in the northeast corner of the Zimri-Lim deliberate effort: Room 162 was a depot. The
palace. It communicates with the palace's main different vessel types indicate that the products
(excavated) entrance to the west through Vestibule stored there were shipped in from the outside,
159 into Court 154; and with the central Court 131 both from the immediate area of Mari and from
and the remainder of the palace through Rooms the farther reaches of Mari's commercial network.
161 and 162 to the south. Parrot understood this The location of Sector C is eminently suitable for
suite with its 12 tightly interrelating rooms and receiving shipments from abroad and for redistrib-
central focus to be a quasi-autonomous hostel for uting them within the palace (see also Margueron
visitors. He stressed its convenient location just 1982: 333).49 I would propose more vigorously
inside the palace gate; its relative independence than Margueron that in Room 162 one can recog-
yet easy access to the rest of the palace; and nize an essential component of Sector C's function
its furnishings-several bathrooms, "kitchens" in as an initial processing station for palace supplies
Rooms 161 and 167-as indications of the unit's coming from the exterior.
intended function (Parrot 1958: 20, 21-33). Mar- The other rooms from Sector C have suffered
gueron challenges this interpretation (1982: 333- from relatively shallow preservation and from in-
34), contending that Parrot mistakenly applied trusive burials, and offer little more than the tab-
modern customs, especially regarding the "kitch- lets with which to recreate them. However, the
interior furnishings of the two northern rooms, dow arrangement allowed scribe and shipper or
167 and 165, do invite commentary. Room 167's artisan to keep each other informed of their re-
west wall is lined with a mudbrick bench 0.5 m spective duties without intruding on each other's
high, whose upper surface supports five shallow work space.
depressions roughly 0.5 m in diameter. Those de- If the area is to be identified with the bit
pressions contained burnt wood, which led Parrot kuprim, however, one could wish for a few more
to label them as hearths, the burnt wood as clues. The majority of the rooms had no objects,
charcoal, and the entire room as a kitchen (Parrot at least as reported by Parrot.50 Precious metals
1958: 24-26). The bench recalls the installations in and any metal vessels, or tools, would have disap-
Rooms 216-218, except that the one in Room 167 peared in the palace looting. But Parrot describes
is about 0.2 m higher, and the depressions connect one peculiar find as exceptional and puzzling: a
with the floor level by a conduit. Parrot did not large heap of bitumen (0.4 m thick, over an area
relate the two areas. Margueron raises the same of roughly 2.5 m x 1.5 m) that blocked the pas-
objections to Parrot's reading of Room 167 as he sage from Room 162 north into Room 161 (Parrot
does for Rooms 216-218: ventilation problems, 1958: 32). This cannot have been the bitumen's
absence of pottery-one would expect at least original state or location. It must have been stored
some pottery in a kitchen even if the majority of in solid blocks or chips, which melted during the
cooking vats were metal-the impossibility of palace fire. The bitumen otherwise fits in with
using the depressions as hearths because no air certain categories of "foreign" containers found in
could circulate, the likelihood that the "charcoal" Room 162, since, like them, it was a product
is secondarily-burnt structural timber, the absence shipped down the Euphrates by boat from north-
of any ash, and so forth (Margueron 1982: 222- western Syria (see n. 47). Perhaps the bitumen was
23). It would be reasonable to propose, in light processed in this sector, melted in the two hearths
of Margueron's observations on the features in in Room 161, to be painted on such vessels as the
Rooms 216-218, that the depressions in this room two large ones from the Room 162 depot. In any
also were supports for vessels, or that they had a case, the particularly violent fire in Sector C may
still undefined function. A window behind one of have been fueled precisely by the oil, bitumen, and
the bench depressions (Parrot 1958: 25, fig. 21) similar products stored there. Is this bitumen heap,
opened into Room 165. Parrot thought that the then, an index of the bit kuprim? Or is the bitumen
window was a service hatch to transferfood directly pavement of Vestibule 159, the entrance to Sector
from the kitchen into Room 165. Margueron (1982: C from the palace gateway, the salient and easily
222) disagrees, pointing out that all the smoke recognizable feature from which the nomenclature
from the hearths would have gone through the derives?s5
window, making Room 165's atmosphere unpleas-
ant. It is more likely that Rooms 165 and 167 E.I.SAG: TheProcessedOil Warehouse
were another storeroom or workshop and a scribe's
office, and that Margueron's Sector C was a de- The only feature in the archives to distinguish
livery point with a variety of functions distributed the E.I.SAG or "processed oil warehouse" from the
over several rooms. In Room 165 Parrot uncovered "bitumen warehouse" is that the former received
in situ three large-capacity jars. One was lined perfumed oil processed by Nur-ili in addition to
with bitumen and thus was for liquid storage supplying him with his raw material. Two receipts
(1958: 27-28). Another (?) jar contained a few attesting that Nur-ili [Nur-ili amhur] delivered
tablets, apparently unpublished. The activities in perfumed oil to this warehouse were found among
Rooms 165 and 167 apparently required a special- the Room 215 tablets (ARMTXXIII: 474,477; the
ized and restricted communication, which the win- four other perfume delivery receipts do not specify
dow provided. Such "ticket windows" have been the destination [from Room 215: ARMT XXIII:
proposed for the Mycenaean palace at Pylos, for 478,491; from Room 160:ARMT XXI: 109, 115]).
which a careful recent study of archival procedures In addition to the receipts, four SUTIA invoices and
based on tablet finds in the palace rooms suggests one SILA2requisition list oil deliveries from this
a bureaucratic routine similar to Mari's if on a warehouse: two of the invoices are from Room 215
more modest (hence more comprehensible) scale (ARMT XXIII: 486, 488). The other two (ARMT
(Palaima and Wright 1985: 261, n. 38). The win- XXI: 114, 108) and the requisition (ARMT XXI:
113), all from Room 160, are the three tablets that ceipts-those held by Balumunamhe and Ili-asraya
fill the gap in the Month V inventory otherwise (ARMT XXIII: 479, 484 and the recapitulative
filed in Room 215. 490)-back into Room 215 if processing them in
The dual function of the "processed oil ware- several interconnected offices were not part of the
house" as both supplier of crude oil and storage bureaucratic routine. The practice of transporting
facility for perfume would, as noted, preclude its tablet baskets from one office to another for book-
being identical with the bit kuprim (ARMTXXIII: keeping or inventory (Sasson 1972) must account
pp. 418-19 [Soubeyran]). The even distribution of for two offices, one in Room 215, the other in
E.I.SAG invoices between Rooms 160 and 215 does Room 160, holding similar documents. Finally, I
not lead to any conclusions. Since we have only have assumed that my reconstruction of the move-
the receipts made out to Nur-ili in exchange for ments of invoices and receipts is tenable, and that
specific volumes of scented oils and no invoices the invoices and receipts themselves have been
indicating in what sort of container they were correctly understood.
stored, we cannot speculate how those commodi- The association of Rooms 216-217-218 (con-
ties could be isolated and identified in the archae- trolled by the Room 215 records office) with the
ological record. One would expect receipts for perfumed oil workshop is a secure one. Corrobora-
Nur-ili's scented oils in the E.i.SAG, however; since tion provided by the beer delivery invoices on the
they have not yet appeared among the published one hand, and by the specialized furnishings of
Mari tablets, it is unlikely that "processed oil the rooms and their location in the palace complex
warehouse" can be located. on the other, support the initial indications of the
Nur-ili dossier from Room 215. It is less certain
SUMMARY that Margueron's Sector C, the northeast wing of
the palace centered on Room 160, is the bitumen
The situation surrounding both the textual and warehouse. Combining the archival evidence from
the archaeological references to Nur-ili's activities, Rooms 160 and 215 with the archaeological ele-
and to the areas where they took place, is subject ments of that wing provides conditions necessary
to many uncertainties, and consequently to a series to solve this problem, but not necessarily the
of assumptions. The first assumption is that his sufficient or exclusive conditions. The possibility
activities involved storerooms and workshops with- does exist. But reworking any of the three assump-
in the Zimri-Lim palace. Margueron (1979) would tions stated above would alter these conclusions on
challenge this view, but the delivery invoices do the bit kuprim and the rooms around Room 160,
not support his arguments. It does not seem prob- although their function inside the administration
able that deliveries to a workshop would be re- of the Mari palace is now clear. It would be
corded and filed deep within the Mari palace if profitable to submit other service sectors of the
the shop itself were outside the palace complex. Mari palace to similar scrutiny, especially the zone
Nor do the size and architectural variations of the to the west of the entrance gate (Parrot's quartier a
Middle Bronze building Parrot uncovered support I'ouest de l'avant-cour-Margueron's Sector H);
Margueron's contention. and, south of Court 131, the rooms surrounding
A second assumption, introduced at the begin- Room 134, which is the source of another major
ning of this discussion, involves the relationship tablet lot. The renewed vigor that cuneiform
between the closed set of tablets relating to Nur-ili scholars are now applying to the publication of the
and their findspots in Rooms 215 and 160. It is remaining Mari tablets should provoke further
easy to question the appropriateness of the ar- examination of the physical milieu that prompted
chive to Room 160: the tablets overlap-in fact the writing of those texts.
complete-the Room 215 archive in several in-
stances. We could picture Room 160 as a last stop CONCLUSIONS
along a logical evacuation route from the south
zone of the palace-with Room 215 at the back- This article does not lend itself to one general
to the north and only known gate. To explain conclusion. Its intent has been instead to open
Room 160 as a tablet dump would remove the perspectives by touching on a number of issues
anomalies in the Nur-ili dossier from Room 160. both indirectly and directly. Perhaps the most
However, it does not explain a movement of re- striking element that emerges from any such
applied study of pre-Hellenistic Syria, particularly ological and textual evidence affords the realistic
for the second millennium B.C.and especially dur- view. I have presented this article with the anticipa-
ing its initial centuries, is the insular quality of its tion that it will provoke others into examining
commercial-and by extension cultural-contacts. archaeological and textual evidence from a com-
References in the Middle Bronze Mari archives to bined front. The material itself appears static; but
Ugarit, to Crete, loom large in our imaginations. it is by asking questions from several angles that
We are thus quick to reconstruct steady com- we can elicit new responses.
merce with distant centers along the Mediterranean
coast and outside Mari's sociopolitical district. The
archaeological record, however, reflects far more ACKNOWLEDGMENT
limited transactions. Inland Syria existed on the
periphery of the Braudelian economie-monde I owe many thanks to J. M. Sasson, who has answered
represented by the second millennium Levant, and my questions with patience, wisdom and imagination,
coincided with its sphere of activity only for luxury but who must be spared any responsibility for the results
or prestige items (e.g., Caphtorite daggers), or for of his generosity; to my colleague L. Stephens for
the rare raw materials that it could not otherwise explaining the mysteries of solid geometry; and to L. De
obtain. Here too, the even balance of archae- Veaux for her research assistance.
NOTES
1To explain the continuing or recurring iconographic attempt is Al-Khalesi's search through archives and
features of ancient Near Eastern art, one may postulate rooms in the 18th century B.c. palace at Mari for its
that the discovery of earlier artifacts by later cultures "Court of the Palms" (Al-Khalesi 1978). At issue in
may have served in part as artistic inspiration. Buchanan those studies and others is whether the delicate balance
(1966: 114), for example, suggests that deep Neo- in coordinating text and artifact has been achieved.
Assyrian foundation trenches uncovered seals whose 3The large "grain measures" contain roughly 4 1.; the
styles and motifs influenced some early first millennium small ones contain 21 (calculated for a 0.20 m high
B.C. artists. How satisfactory, then, to have confirma- example from Chagar Bazar [Mallowan 1936: fig. 14:13,
tion in an Old Babylonian period letter to a father from 4.27 1. + 0.34]. For volumetric estimates of vessels, see
his son, requesting old beads to be dug up from an Ericson and Stickel (1973).
ancient site (from Tell Harmal: cited in Oates 1979: 162; 4Bitumen is an effective sealant; Sams' experiments
first published by Goetze 1958). with vessels from Phrygian Gordion make him doubt
2Others, of course, also introduce arguments cham- the sealing qualities attributed to burnishing.
pioning the integration of texts and archaeological data, 5For a summary of criteria based on modern ethno-
and support them with practical applications. Glock graphic studies, see Henrickson and McDonald (1983:
(1983) correlated Late Bronze historical Taanach and 631-34). Caution must be exercised in identifying table-
its archaeological features; he applies similar techniques wares and serving vessels, whose use is based on
to pottery rather than to texts (Glock 1982). Kempinski individual ethnic/cultural habits to which no universal
(1983) pursues solely historical problems by relating the principles apply. One can cite the variety of vessels used
Hittite historical annals in particular to Middle Bronze to serve tea throughout the world as an example of such
IIB Syro-Palestine (see Gates 1976; 1981) for markedly entrenched and independent customs: cups with or
different results. Kempinski also cites Goetze, including without handles, small flared bowls, saucers, glasses, all
the caveat, "The final decision is up to the historian" fixed according to geographical and cultural limits that
(Goetze apud Kempinski 1983: opposite table of con- one cannot guess without specific references. So, too,
tents). But one can pursue other questions with similar with the manner of serving food, of which I observed a
methods. Few studies attempt to label architecturalunits recent example in southeastern Turkey: our local exca-
on the basis of associated archives. Tegyey (1983) indi- vation cooks judged as absolutely unsuitable the plates
cates in the title of the article that he will exploit we bought for serving individual portions at meals (in
written and excavated documentation together in his succession, so it was not a question of individual
identification of rooms in the Late Bronze Mycenaean preference). They served the meals in small bowls after
palaces, but he fails to do so. A far more ingenious their custom, and used the plates as potlids. Since the
plates were plastic, they buckled and cracked from the Age burial at Baghouz. One such jar (from Tomb Z 47)
cooking heat-an excellent example of Hally's "use contained two strainers; many other strainers still con-
alteration" (1983). For difficulties incurred when a potter tained fragments of straw. The jars range from 0.22 to
imitates a foreign vessel without understanding its 0.55 m high (so from single to double size). Du Mesnil
function, see Fontana (1973). du Buisson illustrated the example from Tomb Z 122
6See also Van Loon for an EB III brewery at (1948: pl. 69), unfortunately without scale; it is reprinted
Selenkahiye (Van Loon 1979: 108); the field report is not in Ayoub (1982) as type 52:5 (produced in Mesopotamia
sufficiently detailed to allow an outsider to evaluate its throughout the second millennium B.c.: Isin-Larsa to
contents. Kassite period), where it appears as 0.325 m high, in the
7Among the other items from the ceramic inventory lower range Du Mesnil du Buisson gave. Its capacity is
are 33 vessels from the courtyard, including lids 5.751. ? 0.05. Gerstenblith (1983: 63) unaccountably
(Dornemann 1981: fig. 9:10) and a jar stopper (Dorne- associates these vessels with wine; however, there is no
mann 1981: fig. 9:11) with many parallels elsewhere (at doubt that Du Mesnil du Buisson interpreted them
Terqa, Mount-Williams 1980: 64; fig. 14:35-37). Those correctly.
stoppers were used on bottles or jars for permanent "These are the only handled jars among the over 500
sealing, in contrast to the lids that were temporary vessels excavated in the Baghouz cemetery. A third
covers. There also was a strainer (Dornemann 1981: possible Syro-Palestinian import is a bowl with lug
fig. 9:2). From Room II came two large, two-handled handles from Tomb Z 147 (Du Mesnil du Buisson 1948:
craters with single tabular spouts (Dornemann 1981: pl. 79, type T: cf. Amiian 1970: pl. 25:1, from Middle
fig. 10:7). Bronze IIA Tomb 3162 at Megiddo [Megiddo XIV]).
8Capacities calculated according to the system set out 16One might ask if the entire Baghouz cemetery is a
by Ericson and Stickel (1973). Dornemann 1981: fig. 7:4: nomadic burial ground. Du Mesnil du Buisson found no
23.181. + 1.81; fig. 3:2 and 8.1: 89.71. + 7.0; fig. 8:3: contemporary settlement.
172 1. ? 13.4. 7The two Baghouz vessels (one is one-third taller than
9From Room III, Dornemann 1981: fig. 3:2-3; from the other) are described as "large" in the published
the courtyard, Dornemann 1981: fig. 8:2, 4. Fig. 8:2 has a report, but they are illustrated without scale. The Terqa
capacity of 325 1. ? 25.5; those illustrated from Room piece, from an Old Babylonian period house, has a
III are somewhat larger. capacity of 33 1. ? 2.3. The catalogued Mari examples
'lThe two spouted vessels from Room II (Dornemann are smaller, but of uniform size: M. 857 (the only
1981: fig. 10:7) and Room III (Dornemann 1981: fig. example with a published profile drawing) has a capacity
4:14) may relate to the older (late Early Bronze) beer jars of 10.16 1. + 0.79.
known from Anatolia and elsewhere (see Mellink 1969: 18Jarresa anse bifide are reported from Parrot's royal
69-76); a multispouted example comes from a private apartments in the northwest sector (Margueron 1982:
house near the Ishtar Temple at Mari (Parrot 1956: 212, 366 calls these apartments the main administrative
fig. 103, pl. 70: M. 674). quarter) and in the sector immediately east of the
" Contra Bouzek (1985: 15), who nonetheless is correct palace's north entrance.
in stressing that overland trade in pottery is limited 19Since no "Canaanite Jars" appear in inland Syria,
because of weight and fragility; hence imported Cypriot Aleppo must have included rebottling among its com-
pottery is concentrated in the coastal regions, with little mercial activities. For the Canaanite Jar (see Parr 1973),
penetration inland (except in Egypt, where the Nile but his derivation from Mesopotamian jar types seems
allows transport by ship). In a recent study of Cypriot unlikely. Canaanite Jars are illustrated and labeled in
pottery trade with the Levant, Gittlen (1981) includes Egyptian tomb paintings as transport containers for
references to the abundant previous scholarship. Levantine wine (see Davies and Faulkner 1947). Note, of
12Leonard (1981: 99) thinks the rhyta were stuffed with course, that Mari was dealing only with its allies and
wool and used to filter perfumed oil; they have also been vassals in this particular trade (see also Finet 1974/77:
described as wine-thieves (G. K. Sams, personal 123), and indeed in all trade except for extraordinary
communication). goods-such as tin-and luxury items.
13Although the Kas shipwreck is of overwhelming 20See Gelb (1982: 590): 1 SILA3or qa 1 quart =
richness, it does not produce a one-to-one container/ 1 liter; opinion on the capacity of the karpatu is unani-
product equivalence: analysis of the sediment in the mous, from Burke (1964: 74) to Bottero, Birot, Finet,
storage jars is revealing products as diverse as glass and finally Durand.
beads, amber, orpiment, and incense, all in identical 21The thorough looting by Hammurabi's soldiers in
vessels (Bass 1986). the Mari palace after Zimri-Lim's defeat would certainly
l4Du Mesnil du Buisson (1948: 51-52) commented not have spared food and drink along with more
that those jars accompanied practically every Bronze precious goods (Margueron 1984:41-44), all the more so
if enemy occupation of the palace is assumed. On 27, d.3) has a similar shape, although the strap handle
transactions recorded in the Mari archives concerning connects shoulder to neck. Alalakh and the 'Amuq are
empty jars (wine and oil jars are specified), see ARMT hardly more promising: Alalakh type 65 (Woolley 1955:
VIII: 80 (wine jars) and IX: 126 (oil jars), with commen- 325) is similar in all respects except for its flat base, but is
tary in ARMT IX para. 34 (Birot). Empty jars are not listed as occurring before Level IV/LB I, although
reusable, as Herodotus' comment on 5th century B.C. there may be earlier examples: the typology is based on
Egypt can attest: complete vessels (see Gates 1976: 28). Examples from
Ugarit with the typical shoulder loop handle (Schaeffer
I shallnow mentiona thingof whichfew of thosewho 1948: fig. 47D from Ras Shamra Tomb 57 [MB IIB],
sail to Egypt are aware.Twice a year wine is brought
which included a decorated version, fig. 48R), and from
into Egypt from every part of Greece, as well as from
Ruweise (Schaeffer 1948: fig. 76:3, from Tomb 8) all
Phoenicia, in earthen jars; and yet in the whole country
you will nowhere see, as I may say, a single jar. What
have ring bases; they are considered copies of Aegean
then, everyone will ask, becomes of the jars? This, too, I stone ware. Palestine provides more parallels with ring
will clear up. The mayor of each town has to collect the bases:jug type I from MB II Shechem (Cole 1984:pl. 29)
wine-jars within his district, and to carry them to and from Megiddo XII/MB IIA (Amiran 1970:pl. 34:7),
Memphis, where they are all filled with water by the but none was made in the characteristic technique of the
Memphians, who then convey them to this desert tract of Baghouz, Terqa, and Mari pieces.
Syria. And so it comes to pass that all the jars which 25Parrot(1959: 114) again proposes local manufacture
enter Egypt year by year, and are there put up to sale,
find their way into Syria, whither all the old jars have by resident foreign potters, but he cites parallels at
Lachish. An identical pitcher is illustrated among the
gone before them (Hdt.Bk 11I.6, trans. G. Rawlinson).
vessels from the Ebla/Tell Mardikh MB II "Tomb of the
2Habur Ware cannot account for all Syrian shipping Lord of the Goats" (Matthiae 1985: pl. 76a); it is the only
transactions in the early centuries of the second mil- real parallel that I have found. Versions of this jar occur
lennium, a fact attested both by the diversity of networks in Alalakh IV as type 40 (Woolley 1955: 325 and pl. 111;
alluded to in contemporary archives, and by the archae- size is comparable, but the spout is rounded, not
ological record. Habur Ware is rare at Mari, absent at pinched; on dates for Alalakh typology, see n. 24); and in
Terqa (Kelly-Buccellati and Shelby 1977: 11)-where MB II Syro-Palestinian contexts (Amiran 1970:pl. 33:4,
one would expect to sample Mari's connections since from Megiddo XIV Tomb 3150, but one-third smaller
Terqa served as custom's station for Mari; it is absent at than the Mari pitchers, and red burnished). MB/LB Ras
Hadidi, very rare at Ebla (Matthiae 1977: 159 cites only Shamra is especially partial to trilobate spouts, but the
two examples). The Habur route ran along the Taurus overall resemblance to the Mari jars is not close (e.g.,
foothills, but did not penetrate south to the Euphrates to Schaeffer 1949: 276-77, fig. 119). Perhaps the best other
any significant degree. parallel is from the Qatna region, in the MB inventory
23M. 1584: 7.94 1. ? 062. The ethnographic record of Tomb I at Dnebi (Du Mesnil du Buisson 1930:
insists that wide mouths are suitable for liquid-storage pl. 31:6/25), if one can judge from the summary pub-
vessels (more than half the sampling in Henrickson and lished sketch. It would confirm the Ebla piece, and
McDonald 1983: 633), but they are not suitable for wine, localize the vessels in western inland Syria.
where air-tight storage is important. The wide-mouthed 26Oliveoil was imported from Carchemish and Aleppo
jars are claimed as part of the wine trade network by (ARMTIX: p. 268 [Birot]). All oil, regardless of origin,
[Hamlin] Kramer (1971: 70-73; 1977: 98) and Gersten- was carried in smaller jars and decanted on delivery into
blith (1983: 63-64). The association of those jars with large storejars called naspaki (ARMT VII: pl. 351
specific trade is always presented in general rather than [Bottero]; IX: p. 320 [Birot]). Although the naspakum is
applied terms; moreover, Gerstenblith's remark (1983: said in one instance to have a capacity of 20 karpatu-
83) that clay analysis on Habur Ware jars has invariably jars, it is also known to come in at least three sizes: 40,
indicated they were local to their findspots does not 150 and 160 1.(ARMTXXI: p. 192 n. 14 [Durand]). The
reconcile well with the notion of their use as long- capacity of the karpatu is, unfortunately, not indicated
distance containers. It is doubtful, however, that the (but see n. 38 for a possible volume of 10 qa or 10 1.).
Mari examples were analyzed, so the case is still open. 27I am very grateful to Dr. Yener for allowing me to
24Parrot suggested that these pitchers were locally read this forthcoming article in manuscript form.
made by foreign potters residing at Mari, for whom 28"Prancing lamassiu-statues," "rivetted railings," in
there is no other evidence. Comparanda are few. The ARMT XIII: 16, 40 (see Al-Khalesi 1978: 6-7). A
closest parallels are those cited by Du Mesnil du Buisson, candidate for the "sealed oil storehouse" dependent on
from Tomb I at Osmaniye (Du Mesnil du Buisson 1930: this court can be proposed for both locations, and is thus
pl. 31:9) and from Mishrife (Du Mesnil du Buisson 1930: not a determinate index.
pl. 31:10/167). Middle Bronze Carchemish is virtually 29Marguerondisclaims all responsibility for using the
unknown; but the spherical pitcher in Woolley (1969: pl. Mari texts by professing lack of competence in the
discipline; but a large number appear in translation Sasson reads more than seven wedges, certainly not 18,
(although they account for less than one-fourth of the probably 8. If these are indeed reciprocal texts, as is
total), nor is it difficult to grasp their basic elements (on strongly suggested, too, by the similar multiples of oil,
the usefulness of context archives to archaeologists, see then the oil jar must have had a capacity of 10 qa or 10 1.
Gibson 1972). He moreover contends that whereas the (like the wine jar).
archival data are discontinue, strictement ponctuelle, par 39Alternatively:Balumunamhe and Ili-asraya had al-
essence incomplete, the architectural remains provide ready cashed in these receipts for payments, and the
evidence that is continue, coherente, avec ses relations transaction was closed. The wide distribution through-
structurales (Margueron 1982: 330); a statement espe- out the Mari palace of oil receipts with their names in the
cially surprising in light of his own study of the Mari PN amhur formula can only be explained as completed
palace as an architectural complex that underwent a transactions, unless the system proposed by Rouault
sequence of confused transformations. Finally, he quali- (ARMT XVIII) is inadequate. On the other hand, the
fies efforts such as Parrot's to assign specific functions to fact that the various types of oil these men delivered
rooms "an artificial superposition of modern ideas on differed from one area of the palace to another (espe-
ancient realities" (Margueron 1982: 330). cially comestible vs. noncomestible) suggests that the
30I will not consider the few texts in which Nur-ili is archives do reflect discrete offices.
not referred to as a perfumer (e.g., ARMT XXI: 220, a 40ARMT XXI: 111, dated IV.26, to be placed between
list of disbursements mentioning a Nur-ili), since the ARMT XXIII: 469 (IV.25) and 470 (IV.27), and the day
crux of the argument here rests on his physical sphere of following the #107-469 exchange quoted above.
activity. 41The layout of the Mari palace would resolve the
3'With the exception of one text, from Room 5 bewilderment expressed by the editors of ARMT XXIII:
(ARMTIX: 277), a promissory note from Nur-ili for the ii regarding the distance separating Rooms 160 and 215
future delivery of perfumed oil in payment for oleander, in contrast with the close relationship between their
carobs, and sumlalu he had already received (the spices archives. The palace plan will be discussed in detail
were used in scenting oil). It is not surprising that this below.
note was found in another area of the palace, one 42On five occasions of the 16 deliveries recorded:
presumably dealing in those specialized goods. ARMT XXI: 113, 114, 108 from Room 160; ARMT
32Perfume deliveries listed from Room 215: ARMT XXIII: 486, 488 from Room 215.
XXIII: 474,477,478; from Room 160: ARMTXXI: 115, 43ARMT XXIII: 357-59, 363; 359' is from Room 216.
109. Oil deliveries from this storehouse to Nur-ili B. Lafont, who edited these texts, considers that the beer
ARMTXXI: 113, 114 Room 160. was again a base product to be flavored in the workshop.
33Mots-clefs in the Mari texts were generously inter- But it could also represent rations (see Sasson 1972: 61
preted for me by J. M. Sasson, who has patiently for artisans' ration deliveries). The beer deliveries date to
monitored my movements through the Nur-ili archives. I Z. 5', when Nur-ili, who is not named in them, was
have not found helpful the observations in Hamlin 1976: apparently no longer associated with the workshop.
42-49. 44It is regrettable that the only ceramic finds published
34From Room 215: ARMT XXIII: 479, 484; in both from this unit are a series of pot lids (Parrot 1959: 138-
cases the supplier is named; from Room 160: ARMT 40), but they are instructive; some were for medium-
XXI: 112, 113, where the source is specified (sa E.fAddu- sized jars (M. 1787, M. 1788, with diameters of 0.30 m
duiri,sa E.I.SAG). and 0.32 m), and others, most significantly (see below),
35From Room 215: ARMTXXIII: 469-73,480, 482- for spouted pitchers such as those from Room 162 (M.
83, 485: all for the perfume workshop from the bitumen 1792, M. 1793). Note, too, the large filter M. 1795
storehouse; 475: for the perfume workshop; 487: from (Parrot 1959: 137, fig. 94, pl. 37) found in corridor 120,
the oil presser [sa itti LU.I.SUR]; 488: from the E.I.DU10.GA onto which Room 215 and a number of others open; the
(variant for E.I.SAG). From Room 160: 107: for the filter would be admirably suited to Nur-ili's work.
perfume workshop; 108, 114: from the EL..SAG; 111, 116: 45ARMT XXIII: 469-73, 480, 483, 492 (frag.); (re-
from the bitumen storehouse. capitulatives) 482, 490; #490 names Bulumunamhe as
36It is the equivalent of itti[PN]immahru (ARMT supplier, but the [amhur] is supplied by the editor, and
XXIII: 477 p. 422); earlier translations (as in ARMT therefore may not be correct. The format is otherwise
XI), which translate PN in apposition to the subject, comparable to #482, listing oil supplied for a variety of
should be revised accordingly. purposes to a number of places; the delivery of oil "ana
37GIR may also imply that Sin-muballit was the bit kuprim" is therefore not exclusive to Room 215 here.
supplier or the delivery-boy (or shipping agent?): see 46ARMT IX: 176 (Birot, with commentary p. 312
Pack (1981: 52-55). para. 106), receipt from Room 5; ARMT XXI: 206,
38Whilethe transliteration for ARMT XXI: 112 reads silver disbursed: requisition from Room 160; p. 189 n. 7
U4-18 KAM, the translation gives "Mois V, le 8." J. M. (Durand); ARMT XXIII: 553: disbursement of silver,
receipt from Room 215; ARMT XXIII: 558: requisition Qatna pitchers (see n. 25) would obviate his suggestion.
for silver to four jewelers, from Room 215. Margueron offers a second hypothesis: Room 162 might
47Bitumen is listed as an import shipped in by boat, be the oil reserve associated with the Court of the Palms
hence from the northwest, perhaps around the Carche- (see n. 28); he admits there is no evidence for that
mish area, where it is still found (tablet from Room 24: suggestion.
ARMT XIII: 96; and Burke 1964: 76). 50Itis probably not a coincidence that the small finds
48The court may have been partly sheltered by an selected for the final publication (Parrot 1959) come
awning or a balcony, the stone blocks found on the floor mainly from the first palace rooms that were excavated.
by Parrot thus acting as bases for wooden poles or 5'Bitumen is used as sealant in the Zimri-lim palace
columns (Parrot 1958: 23, fig. 17). Margueron thinks the bathrooms such as Room 158, or in prominent places
room was roofed (therefore not a court: 1982: 310) such as the podium of Room 132 and the stairs leading
because of the tablets found there, which should not up to Room 66 from Throneroom 65; they were not
have been kept in the open. used as passageways. On the other hand, Room 159
49Margueron speculates further that the room could communicates directly with a bitumen-lined bathroom,
have been a depot for emptied jars, or for empty ones Room 158; the flooring might have been extended to
used for interpalace distributions (i.e., karpatu). But the include both rooms.
foreign vessels such as the Habur jars and the Ebla/
BIBLIOGRAPHY
Al-Khalesi, Y. Perspectives in Archeology, eds. S. R. Bin-

1978 The Court of the Palms: A Functional Inter- ford and L. R. Binford. Chicago: Aldine.
pretation of the Mari Palace. Bibliotheca 1983 In Pursuit of the Past: Decoding the Archae-
Mesopotamica 8. Malibu, CA: Undena. ological Record. London: Thames and
Amiran, R. Hudson.
1970 Ancient Pottery of the Holy Land. New Bouzek, J.
Brunswick, NJ: Rutgers. 1985 The Aegean, Anatolia and Europe: Cultural
ARMT Archives royales de Mari: textes. Paris: Interrelations during the Second Millen-
Geuthner. nium. Studies in Mediterranean Archaeology
Ayoub, S 29. Goteborg: Studies in Mediterranean
1982 Die Keramik in Mesopotamien und in Archaeology.
den Nachbargebieten (von der Ur III-Zeit Braidwood, R. J.
bis zum Ende der kassitischen Periode). 1940 Report on Two Sondages on the Coast of
Miinchener Vorderasiatische Studien II. Syria, South of Tartous. Syria 21: 183-221.
Mainz. Buchanan, B.
Baramki, D. 1966 Catalogue of Ancient Near Eastern Seals
1973 The Impact of the Mycenaeans on Ancient from the Ashmolean Museum I: Cylinder
Phoenicia. The Mycenaeans in the Eastern Seals. Oxford: Clarendon.
Mediterranean. Nicosia. 192-97. Burke, M. L.
Bass, G. IF. 1964 Lettres de Numusda-Nahrariet de trois autres
1986 A Bronze Age Shipwreck at Uluburun (Kas): correspondantsa Idiniatum. Syria 41: 67-103.
1984 Campaign. American Journal of Arch- Cole, D. P.
aeology 90: 269-96. 1984 Shechem I: The Middle Bronze IIB Pottery.
Berthoud, Th. Winona Lake, IN: Eisenbrauns.
1979 Etude par l'analyse de traces et la modelisa- Crawford, H. E. W.
tion de la filiation entre minerai de cuivre et 1973 Mesopotamia's Invisible Exports in the
objets archeologiques du Moyen Orient (IVe Third Millennium B.C. World Archaeology
et IIIe millenaires avant notre ere.) Doctoral 5:232-41.
thesis: Paris VI. Davies, N. deG., and Faulkner, R. O.
Bikai, P. 1947 A Syrian Trading Venture to Egypt. Journal
1978 The Pottery of Tyre. London: Aries and of Egyptian Archaeology 33: 40-46.
Philips. Delougaz, P., Hill, H. D., and Lloyd, S.
Binford, L. R. 1967 Private Houses and Graves in the Diyala
1968 Archaeological Perspectives. Pp. 5-32 in New Region. University of Chicago Oriental In-
stitute Publications 88. Chicago: University nium B.C. Syrian City. Ann Arbor, MI:
of Chicago. University Microfilms International.
Dornemann, R. H. 1981 Alalakh Levels VI and V: A Chronological
1979 Tell Hadidi: A Millennium of Bronze Age Reassessment. Syro-Mesopotamian Studies
City Occupation. Annual of the American 4/2: 11-50.
Schools of Oriental Research 44: 113-51. Gelb, I. J.
1980 Tell Hadidi: An Important Center of the 1982 Measures of Dry and Liquid Capacity. Jour-
Mitannian Period and Earlier. Pp. 217-32 in nal of the American Oriental Society 102:
Le moyen Euphrate: zone de contacts et 585-90.
d'changes, ed. J.-C1. Margueron. Leiden: Gerstenblith, P.
Brill. 1983 The Levant at the Beginning of the Middle
1981 The Late Bronze Age Pottery Tradition at Bronze Age. American Schools of Oriental
Tell Hadidi, Syria. Bulletin of the American Research Dissertation Series 5. Winona Lake,
Schools of Oriental Research 241: 29-47. IN: Eisenbrauns.
1983 The Archaeology of the Transjordan in the Geva, S.
Bronze and Iron Ages. Milwaukee, WI: 1982 Archaeological Evidence for the Trade be-
Milwaukee Public Museum. tween Israel and Tyre? Bulletin of the Ameri-
Disson, G. can Schools of Oriental Research 248: 69-72.
1970 La route de l'etain. Revue d'Assyriologie Gibson, McG.
64: 97-106. 1972 The Archaeological Uses of Cuneiform
Du Mesnil du Buisson, R. Documents: Patterns of Occupation at the
1927 Les ruines d'el-Mishrife au nord-est de Homs City of Kish. Iraq 34: 113-23.
(Elmese).Syria 8: 13-33; 277-301. Gittlen, B. M.
1930 Compte-rendu de la quatrieme campagne de 1981 The Cultural and Chronological Implications
fouilles a Mishrife-Qatna. Syria 11: 146-63. of the Cypro-Palestinian Trade during the
1948 Baghouz: L'ancienne Corsote. Le tell archa- Late Bronze Age. Bulletin of the American
ique et la necropole de l'age du bronze. Schools of Oriental Research 241: 49-59.
Leiden: Brill. Glock, A. E.
Durand, J.-M. 1982 Ceramic Ethnotechniculture. Pp. 145-51 in
1983 Relectures d'ARMT VIII, 11: ARMT VIII, Studies in the History and Archaeology of
89 et le travail du metal a Mari. Mari Annales Jordan I Bath: Pitman.
de Recherches Interdisciplinaires 2: 123-39. 1983 Texts and Archaeology at Tell Ta'annek.
Ellis, R. S. Berytus 31: 57-66.
1977 Mountains and Rivers. Pp. 29-34 in Moun- Goetze, A.
tains and Lowlands. Essays in the Archae- 1957 Alalakh and Hittite Chronology. Bulletin of
ology of Greater Mesopotamia, eds. L. D. the American Schools of Oriental Research
Levine and T. C. Young, Jr. Malibu, CA: 146: 20-26.
Undena. 1958 Fifty Old Babylonian Letters from Harmal.
Ellison, R. Sumer 14: 73-74.
1984 The Uses of Pottery. Iraq 46: 63-68. Hachmann, R. et al.
Ericson, J. E., and Stickel, E. G. 1970 Kamid-el-Loz-Kumidi: Schriftdokumente
1973 A Proposed Classification System for Ce- aus Kamid-el-Loz. Saarbrucker Beitrdge zur
ramics. World Archaeology 4/ 3: 357-67. Altertumskunde 7. Bonn: Habelt.
Finet, A. Hally, D. J.
1974/77 Le vin a Mari. Archiv fur Orientforschung 1983 Use Alteration of Pottery Vessel Surfaces:
25: 122-31. An Important Source of Evidence for the
Fontana, B. L. Identification of Vessel Function. North
1973 The Cultural Dimensions of Pottery: Ce- American Archaeologist 4: 3-26.
ramics as Social Documents. Pp. 1-13 in Hamlin, C.
Ceramics in America, ed. I. M. G. Quinby. 1976 Cuneiform Archives as Data: Reliability of
Charlottesville, VA: University of Virginia. the Mari Archive for Agricultural Recon-
Gadd, C. J. struction. Ann Arbor, MI: University Micro-
1940 The Tablets from Chagar Bazar and Tell films International.
Brak, 1937-1938. Iraq 7: 22-61. Heltzer, M.
Gates, M.-H. C. 1977 The Metal Trade of Ugarit and the Problem
1976 Alalakh-Tell Atchana, Levels V and VI: A of Transportation of Commercial Goods.
Re-examination of a Mid-Second Millen- Iraq 39: 203-11.
Henrickson, E. F., and McDonald, M. M. A. 1982 Recherches sur les palais mesopotamiens de
1983 Ceramic Form and Function: An Ethno- l'Age du Bronze, I (text); II (plates). Series:
graphic Search and an Archaeological Appli- Institut franqais d'archeologie du Proche
cation. American Anthropologist 85: 630-43. Orient. Bibliotheque archeologique et his-
Kelly-Buccellati, M., and Shelby, W. R. torique 107. Paris: Geuthner.
1977 Terqa Preliminary Report No. 4: A Typol- 1984 Les derniers moments du palais de Mari.
ogy of Ceramic Vessels of the Third and Art & Fact 3: 41-44.
Second Millennia from the First Two Sea- Matson, F. R.
sons. Syro-Mesopotamian Studies 1/6: 171- 1965 Ceramic Ecology: An Approach to the Study
236. of Ancient Cultures in the Near East. Pp.
Kempinski, A. 202-17 in Ceramics and Man, ed. F. R.
1983 Syrien und Paldstina in der letzten Phase der Matson. Viking Fund Publications in An-
Mittelbronze II-B Zeit. Wiesbaden: Harras-
thropology 41. New York: Wenner-Gren
sowitz.
Foundation.
Kramer, C. [Hamlin]
P.
1971 The Habur Ware Ceramic Assemblage of Matthiae,
Northern Mesopotamia: An Analysis of its 1977 Ebla, un impero ritrovato. Turin: Einaudi.
Distribution. Ann Arbor, MI: University 1985 I tesori di Ebla. Rome: Laterza.
Microfilms International. Meijer, D. J. W.
1977 Pots and People. Pp. 91-112 in Mountains 1980 The Excavations at Tell Selenkahiye. Pp. 117-
and Lowlands. Essays in the Archaeology of 26 in Le moyen Euphrate: zone de contacts et
Greater Mesopotamia. Malibu, CA: Undena. d'echanges, ed. J. Margueron. Leiden: Brill.
Leemans, W. F. Mellink, M. J.
1968 Old Babylonian Letters and Economic His- 1969 A Four-Spouted Krater from Karatas. Ana-
tory. A Review Article with a Digression on dolu 13: 69-76.
Foreign Trade. Journal of the Economic and Merrillees, R. S.
Social History of the Orient 11: 171-226. 1962 Opium Trade in the Bronze Age Levant.
Leonard, A., Jr. Antiquity 36: 287-92.
1981 Considerations of Morphological Variation Monchambert, J.-Y.
in the Mycenaean Pottery from the South- 1983 La ceramique de fabrication locale a Ougarit
eastern Mediterranean. Bulletin of the a la fin du Bronze Recent: Quelques exemples.
American Schools of Oriental Research 241: Syria 60: 25-45.
87-100. Moorey, P. R. S.
Levey, M. 1985 Materials and Manufacture in Ancient
1959 Chemistry and Chemical Technology in An- Mesopotamia: The Evidence of Archaeology
cient Mesopotamia. Amsterdam: Elsevier. and Art. Oxford: British Archaeological
Liverani, M. Reports.
1973 Memorandum on the Approach to Histori- Mount-Williams, L.
ographic Texts. Orientalia 42: 178-94. 1980 Terqa Preliminary Reports No. 8: Object
Lloyd, S. Typology of the Third Season: The Third
1938 Some Ancient Cities in the Sinjar District. and Second Millennia. Syro-Mesopotamian
Iraq 5: 123-42. Studies 3/2: 35-97.
Lucas, A. Muhly, J. M.
1948 Ancient Egyptian Materials and Industries. 1977 The Copper Ox-hide Ingots and the Bronze
London: Arnold. Age Metals Trade. Iraq 39: 73-82.
Mallowan, M. Oates, J.
1936 The Excavations at Tell Chagar Bazar and an 1979 Babylon. London: Thames and Hudson.
Archeological Survey of the Habur Region, Oppenheim, A. L. (with L. F. Hartman)
1934-35. Iraq 3: 1-85. 1950 On Beer and Brewing Techniques in Ancient
Maniatis, Y., et al. Mesopotamia. Supplement to the Journal of
1984 Punic Amphoras Found at Corinth, Greece: the American Oriental Society 10.
An Investigation of their Origin and Tech- 1967 Lettersfrom Mesopotamia. Chicago: Univer-
nology. Journal of Field Archaeology 11: sity of Chicago.
205-22. Pack, M. D.
Margueron, J. 1981 The Administrative Structure of the Palace at
1979 Existe-t-il des ateliers dans les palais orien- Mari (Ca. 1800-1750 B.c.). Ann Arbor, MI:
taux de l'Age du Bronze? Ktema 4: 3-25. University Microfilms International.
Palaima, T. G., and Wright, J. C. Age Pylos. American Journal of Archaeology

1985 Ins and Outs of the Archive Rooms at Pylos: 88: 260.
Form and Function in a Mycenaean Palace. 1985 The Perfume Industry of Mycenaean Pylos.
American Journal of Archaeology 89: 251-62. Studies in Mediterranean Archaeology:
Parr, P. J. Pocketbook no. 34. G6teborg: P. Astr6m's.
1973 The Origin of the Canaanite Jar. Pp. 173-81 Stubbings, F. H.
in Archaeological Theory and Practice: 1951 Mycenaean Pottery from the Levant. Cam-
Essays Presented to Professor W. F. Grimes, bridge: Cambridge University Press.
ed. D. E. Strong. London: Seminar Press. Tegyey, I.
Parrot, A. 1983 Archaeology and Interpretation (A Survey of
1956 Mission archeologique de Mari I Le temple the Relation of Linear B Tablets to Their
d'Ishtar. Paris: Geuthner. Archaeological Context). Acta Classica Uni-
1958 Mission archeologique de Mari 1:1. Le pa- versitatis Scientiarum Debreceniensis 19:
lais: architecture. Paris: Geuthner. 17-22.
1959 Mission archeologique de Mari 11:3. Le Thureau-Dangin, F., and Dunand, M.
palais: monuments et documents. Paris: 1936 Til Barsip. Paris: Geuthner.
Geuthner. Van Loon, M.
1964 Les fouilles de Mari: Treizieme campagne 1979 1974 and 1975 Preliminary Results of the
(printemps 1963). Syria 41: 3-20. Excavations at Selenkahiye near Meskene,
Postgate, J. N. Syria. Annual of the American Schools of
1978 An Inscribed Jar from Tell al-Rimah. Iraq Oriental Research 44: 97-112.
40: 71-75. Waetzoldt, H.
Pulak, C., and Frey, D. A. 1981 Zur Terminologie der Metalle in den Texten
1985 The Search for a Bronze Age Shipwreck. aus Ebla. Pp. 363-78 in La lingua di Ebla.
Archaeology 38: 18-24. Atti del convegno internazionale [Napoli,
Renfrew, C. 21-23 aprile 1980], ed. L. Cagni. Naples:
1973 The Emergence of Civilisation. London: Oriental Institute.
Methuen. Waetzoldt, H., and Bachmann, H. G.
1977 Production and Exchange in the Early State 1984 Zinn- und Arsenenbronzen in den Texten aus
Societies, The Evidence of the Pottery. Pp. Ebla und aus dem Mesopotamien des 3.
1-20 in Pottery and Early Commerce; Char- Jahrtausends. Oriens Antiquus 23: 1-18.
acterization and Trade in Roman and Later Winter, I. J.
Ceramics, ed. D. P. S. Peacock. London: 1979 On the Problems of Karatepe: The Reliefs
Academic Press. and Their Context. Anatolian Studies 29:
Saade, G. 115-51.
1979 Ougarit, metropole cananeenne. Beirut: Im- 1982 Art as Evidence for Interaction: Relations
primerie Catholique. between the Assyrian Empire and North
Salonen, A. Syria. Berliner Beitrage zum Vorderen Orient
1966 Die Hausgerate des alten Mesopotamiens. I: 355-82.
Annales Academiae Scientiarum Fennicae, Woolley, C. L.
Ser. B: Tom. 144. Helsinki. 1955 Alalakh: An Account of the Excavations at
Sasson, J. M. Tell Atchana in the Hatay, 1937-1949. Ox-
1966 A Sketch of North Syrian Economic Rela- ford: Oxford University.
tions in the Middle Bronze Age. Journal of 1969 Carchemish Part II: The Town Defenses.
the Economic and Social History of the Oxford: Oxford University.
Orient 9: 161-81. Yener, K. A.
1972 Some Comments on Archive Keeping at 1983 The Production, Exchange and Utilization of
Mari. Iraq 34: 55-67. Silver and Lead Metals in Ancient Anatolia:
Schaeffer, C. F. A. A Source Identification Project. Anatolica
1939 Ugaritica I. Paris: Geuthner. 10: 1-15.
1948 Stratigraphie comparee et chronologie de Yener, K. A., and Ozbal, H.
l'Asie occidentale (IIIe et IIe millenaires). 1987 Tin in the Turkish Taurus Mountains: The
London: Oxford University. Bolkardag Mining District. Antiquity 61:
1949 Ugaritica II. Paris: Geuthner. 220-26.
Shelmerdine, C. W.
1984 Perfume in International Trade at Bronze

Bronze Age Syria PDF

Uploaded by

Copyright:

Available Formats

Bronze Age Syria PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bronze Age Syria PDF

Uploaded by

Copyright:

Available Formats

Dialogues between Ancient near Eastern Texts and the Archaeological Record: Test Cases from

Bronze Age Syria

TEROA TPR 4:18

Fig. 2. Assorted Middle Bronze containers. From Bag-

The occurrence of animal bones in unusual con- Metal

Fig. 3. a. Mari,Zimri-LimPalace Room 162 (after Parrot

Nur-ili's finished products, perfumed oils; for the

TABLE 1. Texts from the Nur-ili file, in chronological order

Nur-iliFiles from processedoil depot from bitumendepot from reserveof A-duri

to processedoil depot to bitumendepot to perfumeworkshop

Al-Khalesi, Y. Perspectives in Archeology, eds. S. R. Bin-

Palaima, T. G., and Wright, J. C. Age Pylos. American Journal of Archaeology

You might also like