EXCAVATING TEPE GAWRA

IN THE ARCHIVES OF THE

UNIVERSITY OF PENNSYLVANIA
MUSEUM
By Brian Peasnall and Mitchell S. Rothman

One of Iraqs Earliest Towns

TOP: Field chit.

BOTTOM: Gawra
1935 field crew.
From left, Cyrus
Gordon, H.A.
Schubart, Jr., E.B.
Bache, Charles
Bache, and E.B.
Mller, an architect.

hen does excavation not require the diggers

to get dirty? Such a riddle may belie most
peoples image of archaeology. In their imaginations, archaeology is for adventurers, like
the fictional Indiana Jones who traveled to
exotic lands to retrieve ancient relics. With huge crews, they dig
ancient mounds to bring back the most beautiful artifacts to fill
their museums cases. Here, however, we describe an unusually
clean kind of excavation, one that takes materials from archaeologys past and uses it to answer the most modern and scientific kinds of research questions.

W
34

VOLUME 45, NUMBER 3 EXPEDITION

TOP AND BOTTOM: UPM; MIDDLE: MITCHELL ROTHMAN, FROM TEPE GAWRA, 2002

MIDDLE: Greater
Mesopotamia in the
Late 5th and 4th
millennium B.C.

In the infancy of archaeology, excavators were often unaware

of the intricacies of site stratigraphy. They were frequently negligent in recording where artifacts were found. As a result, museum storerooms and basements house thousands
of largely ignored finds, and archives contain many
unread fieldnotes. Despite the technical problems
with these early excavations, they hold potential
treasures, not in gold, but in information. Many of
these sites were excavated in areas where we can no
longer dig, and many represent large horizontal
exposures of town plans, the cost of which few
modern excavators can afford.

range of time almost unknown during the 1930s at other sites.

The earliest level was never reached due to the start of the 1939
War in Europe.

TOP: UPM, NEG. #44793;BOTTOM: UPM NEG. #46201

TEPE GAWRA

Tepe Gawra in northern Iraq is one such site,

which, despite the date of its excavation (1932
1938), has proven to be a gem in the rough. Aside
from the richness of its artifacts, the quality of dig
records excavated from the Archives of the University of Pennsylvania Museum proved to be a
wonderful bonus. Those records served as the basis TOP: Tepe Gawra
of a re-analysis of the site and provided a powerful RIGHT: Ephraim Alvagor Speiser.
research tool for asking new questions.
Tepe Gawra is located about 18 miles northeast of Mosul in
There are a number of chronological
the piedmont zone adjoining the Assyrian Plains in northeastnomenclatures
for these periods. Of special
ern Iraq. It lies between the Tigris River and the first foothills of
interest here is the so-called Uruk period
the Zagros Mountains, by the entrance to one of the few histordated from about 4200 to 3000 B.C. The definition of the Uruk
ically documented passes onto the Iranian plateau through the
period is based on southern Iraqi sites. Contemporaneous
Jebel Maqlub. Gawra was certainly a transport link in trade for
northern and eastern Mesopotamian sites were traditionally
lapis lazuli and for other exotic goods from the Zagros highassigned to the Late Chalcolithic period. In an attempt to corlands and from the Upper Tigris basin into Mesopotamia proprelate the two chronologies scholars at an advanced seminar in
er. Sites like it supplied the heartland of Sumerian and
Santa Fe developed a new system , dividing the period into five
Babylonian cities with exotic goods in antiquity. They also served
sub-periods from LC1 beginning at 4400 B.C. to LC5 ending at
as centers of small societies at the edges of the hilly north.
3000 B.C. The results can be found in Uruk Mesopotamia and
Ephraim Avigdor Speiser, a young philologist originally from
its Neighbors.
Eastern Europe, found the site during a survey on the Assyrian
Of the twenty-one levels a level is one stratum of genersteppes and hills. After he finished his Ph.D. in 1924, a consorally contemporaneous buildings and open spaces the first
tium of institutions, including the University of Pennsylvania,
ten from the Akkadian to the late Early Uruk or late LC2 perithe American Schools of Oriental Research, and Dropsie Colod were completely exposed. The rest varied from two-thirds to
lege, sent him to Iraq to find ancient sites for excavation. In
under a half of the strata.
1929 Spieser wrote,I had practically completed my first season
in Iraq before I came upon Tepe Gawra. Gawra was intriguing
to him because, as he wrote,The mound in question is the oldWHAT DO WE WISH TO KNOW?
est site in Northern Iraq that has yet been dug.
When Speiser, in 1935, and a student of his, Arthur Tobler, in
The excavators of Gawra identified twenty-one stratified vil1950, published two volumes on the excavation of the Tepe
lages and towns, levels I to XX from top to bottom; the twentyGawra, they were primarily interested in culture history. Culture
first and lowest level simply referred to as either Area A or the
history was the dominant archaeological school or paradigm
Northeast Base. These early settlements dated primarily from
until the 1960s. Its focus was on constructing sequences of artithe sixth millennium to the early fourth millennium B.C., a
fact styles, religious buildings, technologies, and other cultural
WWW.MUSEUM.UPENN.EDU/PUBLICATIONS

35

36

VOLUME 45, NUMBER 3 EXPEDITION

and politically that it could dominate less developed northern

and eastern societies.
Initially, scholars believed that Levels XII to VIII at Gawra
spanned the entire Uruk or LC15 period. As a small center on
the Tigris route between north and south, and as one of few
northern sites with wide horizontal exposures of its town plans,
Gawra would be a natural site to test Algazes hypothesis. It
would then be necessary to determine what Gawras economic,
social, and political institutions were during the periods contemporary with the southern Early, Middle, and Late Uruk
periods.
REANALYZING TEPE GAWRA
AND FINDING ANSWERS

Because it was dug more than 50 years ago, initially under

Speisers direction, using techniques that no current archaeologist would sanction, our question became how well the data
from Tepe Gawra could be trusted and utilized to assess its late
fifth to fourth millennium B.C. levels. Since no one could excavate in the Kurdish area of northern Iraq, the only way to investigate the site was to go back into the archives and collections at
the University of Pennsylvania Museum.
This effort proved more successful than any of us could have
hoped. The key to re-excavating Gawra proved to be chits, initially discovered in a sarcophagus in the University of Pennsylvania Museums subbasement by graduate student Sam Mild.
These chits were small, preprinted forms with a 10 by 10 grid
and spaces for recording object number, type, field number,
final storage location, and other comments. Speiser was not a
trained archaeologist, but a philologist. However, after the first
two seasons, Speiser brought in Charles Bache to direct excavation at Tepe Gawra. Bache, a descendant of Benjamin Franklin
whom the Pennsylvania Gazette describes as having two hobbies, tennis and archaeology, was well trained in the techniques
of a developing American anthropological archaeology school.
He understood stratigraphy and the imperative to make detailed
notes of finds. Using chits, Baches crew recorded the precise
three-dimensional position of over 5,000 artifacts recovered in
the third, fourth, fifth, and seventh seasons. They also often
described the nature of the fill, so it was possible to determine
the process by which artifacts found their way into a particular
place. Field locus sheets and grave sheets left another set of
invaluable data.
The chits proved very important because modern anthropological and historical archaeologists interpret the past through
contexts. We cannot assume that people in the past conducted
their lives as we do today. Which artifacts occur together in
what kind of space, open or in a building, is the key to differentiating between, for example, a rich or poor persons house, a

SEALING PHOTO: CARL ROTHMAN; SEALING DRAWING: GEORGE GRENTZENBERG; FAR RIGHT: MITCHELL ROTHMAN

traits, and determining the spread of peoples and ideas largely

through migration and diffusion. The processual school that
dominates archaeology today looks at societies as a process of
evolution from past cultures. In order to explain change or evolution, this school focuses on human behavior and on social,
political, economic, and religious institutions.
As the culture historical school of archaeology faded, the
Gawra volumes became classics of a past era, which were much
less important than those of more recently excavated sites.
However, Gawra has again become a focus of attention because
of new questions raised about the origin of the earliest cities
and states on the southern alluvial plains of Mesopotamia and
resulting interactions between southern Mesopotamia and its
northern and eastern neighbors. The founding of state societies
meant rulers, bureaucrats, and complex economies with specialized production.
The two worlds, southern Uruk, and northern and eastern
Late Chalcolithic, were quite distinct stylistically. They also
appeared distinct culturally. The south, by the mid-fourth millennium, was organized as a set of cities with its dependent
towns and villages near the main channels of the Tigris and
Euphrates rivers. These are the city-states described in Robert
McCormick Adams Heartland of Cities. The north and east
were typified by smaller, more spread-out settlements with less
well-developed economic and political institutions.
Beginning in the mid-1980s, Guillermo Algaze, a student
from the University of Chicago, noticed an interesting pattern
in pottery style over the entire Greater Mesopotamian region.
As the fourth millennium B.C. progressed, more and more pottery of distinctly southern origin appeared in a particular set of
northern and eastern sites. By the Late Uruk period (LC5), the
last century and a half of the fourth millennium, these pottery
distributions appeared to form a dendritic pattern, with key
sites in the Assyrian steppes and then many branching lines
through more and more sites into the piedmont and hills. The
first choke points on the system were towns such as Habuba
Kabira and Jebel Aruda on the Euphrates, with entirely southern artifacts and presumably southern migrants.
Algaze put together this artifact style distribution based on
the fact that the south lacked raw materials such as metals, precious and semi-precious stones, tool-making obsidian, and logs
for large-scale construction of temples and palaces. The north
and east had these raw materials or were on routes to obtain
them. In his highly acclaimed work, The Uruk World System, he
hypothesized that the rulers of the first states and first cities had
organized a formal, international economic network to obtain
desired goods from the periphery through a kind of economic
colonial system. This system worked, Algaze theorized, because
the south was so much more highly developed economically

Table 1. Chronological Framework

church, factory, or government building. If you consider what

differentiates an industrial city from an agricultural village, it is
more than their respective sizes. An industrial city will have not
only factories, but also houses, that differentiate workers from
owners, products made, and items consumed. The sum of the
activities or functions (for example, craft production, religious
ritual, food preparation, sleeping areas, funerary practice, entertainment, government administration) gives us a clearer picture of what people did and how they organized their lives. In a
theoretical sense, functions are a key to interpreting how ancient
peoples constructed their institutions and saw their world.
For example, how would we know how a government
worked in the prehistoric past? Without written records, which
appeared at the end of the Uruk period, leaders used seals and
sealings. These clay locks were placed on sacks, boxes, jars, and
storeroom doors. They were stamped with a seal of authority
and were only to be
opened by people
authorized to do so.
Those who controlled raw materials
and foodstuffs had
access to large work
gangs, armies, craft
producers, and the
symbols of authority
and status. By correClay sealing from Level XIAB.
lating the presence of
seals or sealings in the functional contexts, it was possible to
determine what activities and goods leaders controlled.
To investigate these increasingly complex institutional
arrangements, we first needed to place each level of the developing site in time and space. We had to reconstruct the activity
areas and functions of the site for each period in succession.
Again, Gawra offers us the possibility to do this as so much of
each level was exposed.
However, Speiser, and even Bache, perceived the strata of the
mound to be in a layer cake form. The problem with this view
is that it ignores the fact that settlements do not spring up fully
formed. Occasionally, structures are modified or new structures are built while others are in the process of decay. Although
archaeological levels are treated as a single point in time, this
synchrony is only relative to the other levels that make up the
whole site.
Dating levels XII to VIII proved a problem. As we noted
above, most researchers believed that the so-called expansion of
southern (Uruk) cultures into their periphery happened at the
very end of the Uruk period (see table 1). Gawra VIII was
believed to be from that same period. Without absolute (C14)

B.C.

SAR

TEPE GAWRA

Middle
2000 Bronze

Old Babylonian
IV-VI

Early
Bronze
3,000

SOUTHERN MESOPOTAMIA
Old Terms

Akkadian
Early Dynastic

VII
LC5

Late Uruk

3,400 LC 4

hiatus

Late
Middle
Uruk

3,600
LC 3

VIII

late

IX-X

Early
Middle
Uruk

3800

LC 2
4000

XI/XA

Early
Uruk

early
4200
XIA/B
LC 1
XII

Ubaid
transitional

XIIA- XVI

Ubaid

XVII-XX

Halaf

4500
Ubaid

6700 Halaf

dates, we are left with the older relative dates, which compare
the style, mostly of pottery, with other sites on the assumption
that similarity or identity of style means contemporaneity. Gawra has a small but distinct set of forms. Luckily, other contemporaneous sites with C14 dates do exist. A recent re-working of
the Greater Mesopotamian region shows that Gawra VIII ended
at about 3750 B.C., not 3100; that is, at the beginning and not
the end of the period of contact and expansion. This was the
LC3 or Middle Uruk period. That period of contact was not 150
years long, but closer to 600 years, leading to new interpretations of regional change and development, as those explored in
Uruk Mesopotamia and its Neighbors.
First we reviewed the architects field drawings, the chits, and
archival photographs to redraw the town plans and place the
graves. In this effort, we found many mistakes in Speisers and
Toblers original publications (Bache died before he could prepare the second volume). We next put artifacts back into their

WWW.MUSEUM.UPENN.EDU/PUBLICATIONS

37

TOP: The White Room Building.

MIDDLE: Artifact distribution of the White Room Building, Level XII.

original contexts within buildings and open spaces. For example, one of the more important rooms in level XII is the so-called
White Room building. Its size and position near the main entryway to the mound made it appear important. By putting artifacts back into context, we were able to determine that it was a
domicile, but one of a very important person, perhaps a member of a chiefly lineage.
In this process, we often had to challenge the original excavators a priori assumptions concerning the nature of various features. A good example comes from Toblers discussion of the
burials. He interpreted many of them as sacrificial in nature,
solely on the basis of their proximity to structures thought to
have had a religious function. Two interpretive problems were
thereby created. Stratigraphic attribution of burials based on
this assumption often overruled sound principles of superposition, and new analyses showed that many buildings had functions other than those Tobler assigned them.

38

VOLUME 45, NUMBER 3 EXPEDITION

As part of the larger attempt to reveal the evolutionary trends

at Tepe Gawra from the fifth millennium to the very beginning
of the Uruk Expansion (LC3), Brian Peasnall in his appendix to
Tepe Gawra demonstrated how mortuary behavior helps archaeologists understand the past. Archaeologists tend to emphasize
economics and politics, because the artifacts we recover are
mostly relevant to those topics. From graves, however, we can get
a glimpse of how the ancients viewed their communities and
their concept of the divine. Cemeteries are really
communities of the dead, where differences in
wealth, religion, ethnicity, and the like are symbolized and preserved. For Gawra, based on architecture, town plans, and activity areas, we reconstructed a small center in the piedmont, serving farming
villages and pastoral nomads. Contrary to Algazes
original hypothesis, economic specialization and
political elaboration (complexity) were developing
before intensified interaction with the south. This
was not a backward periphery. However, the question remained, did the people at the time symbolize
their changing social structure in graves that we
archaeologists saw in economic and political artifacts?
People who were buried at Tepe Gawra were either placed in
built tombs made of libn or mudbrick, in simple pits, vessels,
pits with a small wall at their backs, mud plaster lined pits, or in
cists (round holes lined with stone). The goods placed with
them in their graves varied from a pot or necklace to rather rich
collections of gold, exotic beads, obsidian cores, lapis lazuli
seals, and the like. In fact, the graves represent the same sorts of
evolution of differences in status, power, and privilege that the
other analyses found. For example, through time the percentages of different kinds of internment varied until Level X. By
that level, adults were either buried in simple pits or in mudbrick tombs, the latter with the richest grave goods.
This recitation can only begin to highlight what our clean
excavation revealed. We did find reasons to challenge Algazes
theory of the underdevelopment of northern societies and the
dominance of southern city-states but confirmed the importance of this long-distance trade as one of a number of factors
in social and political development. We were able to draw a
richer picture of cultural change than was possible with the
original publications. Tepe Gawra: the Evolution of a Small Prehistoric Center in Northern Iraq, our final report, published by
the University of Pennsylvania Museum, was able to answer
questions undreamed of in Speisers philosophy. Brian Peasnall
is currently extending the archival analysis of Tepe Gawra back
into the periods before the Uruk.

TOP: UPM NEG. #44712; MIDDLE: AFTER ROTHMAN, FROM TEPE GAWRA, 2002; BOTTOM: CARL ROTHMAN

BOTTOM: Variation in pottery types from Tepe Gawra XII-VIII.

and two salvage excavations at sites on the Euphrates River in

Turkey near the Syrian border. His primary interest is in the origin of state-level society in Mesopotamia. Among his publications
are Chiefdoms and Early States in the Near East (with Gil
Stein), Uruk Mesopotamia and Its Neighbors: Cross-cultural
Interactions in the Era of State Formation, and Tepe Gawra: the
Evolution of a Small Prehistoric Center in Northern Iraq. The
latter was recently published by the University of Pennsylvania
Museum.

FOR FURTHER READING

Adams, Robert McCormick. The Heartland of Cities.

Chicago: University of Chicago Press, 1981.
TOP: Libn tombs 111, 114, and 110.
RIGHT: Sample of grave goods from
Tomb 110.

TOP TO BOTTOM: UPM NEG. #NC35-44598; MITCHELL ROTHMAN; MICHAEL ROSENBERG; CARLENE FRIEDMAN

Our article should show, if nothing else, the incredible information

lying in archives and collections. You
do not even need to get too dirty or worry about visas or foreign diseases to recover it.
Brian Peasnall has spent the last 10
years investigating the emergence of
settled village life in southeast Anatolia. Additionally, he has conducted
extensive archival research on Tepe
Gawra at the Museum. Peasnall is
currently a research associate in the
Museums Near East section where he is developing a digital catalog of the materials excavated from the site of Ur during the
1920s and 1930s by Sir Leonard Woolley.

TOP: Peasnall with one

of the workmen during
the 1997 excavation of
Demirkoy, a small
Neolithic-period village
in southeastern Turkey.
BOTTOM: Rothman with
a local mullah.

Mitchell Rothman is an associate professor of archaeology and anthropology at Widener University in Chester
and a Consulting Scholar at the University of Pennsylvania Museum. He
has been doing archaeology in the
Middle East since a survey in Khuzistan, Iran, in 1974. In subsequent years,
he has been on excavation crews at
Tal-i Malyan in Iran, and Gordion in
Turkey, and on survey teams in southeastern Turkey. Rothman has led a
survey in highland eastern Turkey

Algaze, Guillermo. The Uruk World System. Chicago:

University of Chicago Press, 1993.
Eichler, Barry. Ephraim Avigdor Speiser. In The Oxford
Encyclopedia of Archaeology in the Near East, 5, E.
Meyers, pp. 6869. Oxford: Oxford University Press, 1997.
Peasnall, Brian. Appendix: Burials at Tepe Gawra from
Levels VIIIXIAB. In M. S. Rothman, Tepe Gawra: The
Evolution of a Small, Prehistoric Center in Northern Iraq.
Philadelphia, PA: University of Pennsylvania Museum of
Archaeology and Anthropology, 2002.
Postgate, Nicholas, ed. Artifacts of Complexity: Tracking the
Uruk in the Near East. Oxford: British School of
Archaeology in Iraq, 2002.
Rothman, Mitchell S. Tepe Gawra: The Evolution of a
Small, Prehistoric Center in Northern Iraq. Philadelphia,
PA: University of Pennsylvania Museum of Archaeology
and Anthropology, 2002.
Rothman, Mitchell S., ed. Uruk Mesopotamia and its
Neighbors: Cross-cultural Interaction in the Era of State
Formation. Santa Fe, NM: SAR, 2001.
Rothman, Mitchell S. Seal and Sealing Findspots, Design,
Audience and Function.: In Archives Before Writing, edited
by P. Ferioli, E. Fiandra, G. Fisore, and M. Frangipane, pp.
97121. Rome: Universit di Roma, 1994a.
Rothman, Mitchell S. Sealings as a Control Mechanism in
Prehistory: Tepe Gawra XI, X, and VIII. In Chiefdoms and
Early States in the Near East, edited by G. Stein and M. S.
Rothman, pp.103120. Madison, WI: Prehistory Press,
1994b.
Speiser, E. A. Traces of the Oldest Cultures of Babylonia
and Assyria. Archiv fr Orientgesellschaft 5 (1928/29):
16164.
Speiser, E. A. Preliminary Excavations at Tepe Gawra.
Annual of ASOR 9 (1929): 1794.
Speiser, E. A. Excavations at Tepe Gawra 1. Philadelphia,
PA: University of Pennsylvania Museum of Archaeology
and Anthropology, 1935.
Tobler, Arthur. Excavations at Tepe Gawra 2. Philadelphia,
PA: University of Pennsylvania Museum of Archaeology
and Anthropology, 1950.

WWW.MUSEUM.UPENN.EDU/PUBLICATIONS

39