At the edge of empire: conceptualizing Assyria's Anatolian Frontier ca. 700 BC

24 Pages·2008·0.59 MB·English

JOURNAI. OF
Anthropological
Archaeology
ACADEMIC
PRESS Journal of Anthropological Archaeology 21 (2002) 371-395
At the edge of empire: conceptualizing Assyria's
Anatolian Frontier ca. 700 BC
Bradley J. Parker
Deptir.tment of History, The Unioersily of Utah, 380 S 1400 E Rm 211, Salt Lnlce Cily, UT 54112-0311, USA
Received 23 May 2001; revision received 7 March 2002; accepted 14 March 2002
Introduction mental in shaping the physicality of modern na-
tion states (Baty, 1928; Boggs, 1930, 1937, 1940;
Frontier studies have long been the focus of Crocker, 191 9; Fawcett, 191 8; Fischer, 1949;
considerable debate in various fields and sub- Holdich, 1916). The fields of Classics and Ancient
disciplines of the humanities and social sciences. History also have a long "history" of frontier
As early as 1893, Fredrick Jackson Turner studies, with some of the major works in espe-
brought frontier studies to the forefront of re- cially Roman studies focused on the frontiers of
search in American History by arguing that the the empire (Dyson, 1985; Elton, 1996; Isaac, 1990;
political, economic and cultural underpinnings of Luttwak, 1976; Whittaker, 1994 for example).
American society were not the result of European Frontier studies reached the fields of social science
influence, but were rather the product of the and anthropology first through the study of the
American frontier experience. Expansion into the process of acculturation (Broom et al., 1954;
vast "empty" wilderness of the American west Spencer, 1961; Redfield et al., 1936) and more
insolated American pioneers from the European recently through the development and application
world while the rugged frontier lifestyle induced of Wallerstein's "World Systems Theory" (Wal-
the uniquely American qualities of individualism lerstein, 1974, 1980). In recent years world systems
and democracy. Although it has since been dis- theory has helped to direct frontier studies in these
credited on various grounds including, among disciplines by giving researchers a framework
other things, its ethnocentrism and linearity, within which to turn their focus from core polities
Turner's thesis sparked a debate that even today to peripheral regions (Chase-Dunn, 1988; Chase-
continues to inform approaches to frontiers in Dunn and Hall, 1991, 1992). This debate soon
many fields (Adelman and Aron, 1999; Cayton spread to archaeology as scholars attempted to
and Teute, 1998; Faragher, 1994; Jacobs, 1994; adapt the basic tenets of this model to the ancient
White, 1991). Frontiers are also, of course, fun- world (Algaze, 1989b, 1993; Blanton and Fein-
damental to geographers who took up their own man, 1984; Champion, 1989; Chase-Dunn and
discussion of the topic early in the twentieth cen- Hall, 1991; Kohl, 1987; Rowlands et al., 1987;
tury (Curzon, 1907; Fawcett, 191 8; Ratzel, 1897). Stein, 1999).
In fact, their definition of concepts such as mari- Despite nearly a century of research in these
time and aerial borders as well as their discussion and other fields, however, there is still very little
of political and ethnic boundaries was instru- consensus about classifications and comparative
frameworks within which to consider an interdis-
ciplinary study of frontiers. Prudence Rice re-
cently concluded that her review of various
0278-4165/02/$ - see front matter O 2002 Elsevier Science (USA). All rights reserved.
PII: SO278-4165(02)00005-3
372 B.J. Parker / Journal of An/hropological Arcltaeology 21 (2002) 371-395
approaches to the study of frontiers, borders, and archaeological data with textual sources from the
boundaries "has revealed the extent of disagree- Neo-Assyrian Iron Age (ca. 700 BC), I will char-
ment-and often outright contradiction~on- acterize a particular frontier region on the edge of
cerning the nature of these phenomena.. ." (Rice, an imperialistic state. It is my hope that this anal-
1998, p. 59). This statement illustrates that al- ysis will help bring this important frontier zone into
though many scholars are involved, in one way or focus and eventually lead to further studies that
another, in frontier studies, we have barely begun will illuminate the mechanisms contributing to the
to make systematic comparisons across time and dynamics of this and other frontier regions.
space or to develop overarching paradigms.
One reason such a consensus has yet to be
reached highlights the very nature of frontiers A question of definitions
while simultaneously underlining their appeal to
scholars of so many different disciplines: Frontier Traditionally the borders of the ancient polities
'
situations are by definition dynamic. Lying at the have been illustrated on maps in much the same
edge or between cultural spheres, frontiers en- way the boundaries of modern states are: either as
compass the ecological and geographic space a solid line or by the intersection of two con-
within which culture contact takes place. Fron- trasting colors. Indeed, as anyone who has at-
tiers are, therefore, constantly changing through tempted to cross a contentious border knows,
both space and time, and thus different geo- these "lines," more often than not, do act as
graphical or chronological views of a single barriers inhibiting the free flow of people and
frontier, not to mention comparative views be- goods between two or more polities. Even where
tween frontiers, can give sharply contrasting pic- modern borders either fail miserably, or are not
tures. Considering the potential for extreme intended to stop such cross-border flow, they can,
variability within and between frontiers, it is not nevertheless, be called upon to act as filters, or at
surprising that each discipline has developed its least obstacles, between regions. However, the
own set of theories and models for the study of idea of a "preclusive border" is a modern concept
frontier situations. associated with the administrative and political
Given this diversity both within and between boundaries of the modern nation state and, even
frontiers and within and between approaches and in the modern world, does not regulate all facets
analytic models in frontier studies, it is perhaps of life. Political borders are particularly inept at
best to begin by clarifying the basic goals that the governing linguistic boundaries, as the latest
study of frontiers should aim to address. First, it is controversy over the English language in France
imperative that we define the terminology with has shown. Religion is another aspect of life that,
which to classify the various manifestations of regardless of the recent efforts of states such as
frontiers. Second, using a unified terminology to Russia and China, has proven to be difficult for
aid in cross-frontier comparison, we should define political forces to constrain.
and categorize specific frontier situations. Third, In spite of the fact that these and other as-
in studying specific frontiers we should focus on pects of modern society are not always confined
defining the mechanisms behind andlor the factors by the political boundaries of modern nation
contributing to, frontier dynamics. Fourth, we states, the mental framework embodied by the
should attempt to understand how these dynamics concept of the modern preclusive border is often
affected both the core area and regions beyond the the bias that scholars bring to their study of the
frontier zone. And fifth, through the comparison ancient world. Luttwak, for example, in his piv-
of various specific frontier situations, we should otal study The Grand Strategy of the Roman
begin to build models within which to characterize Empire described the frontiers of the Roman '
and analyze common themes in frontier studies. Empire during the reign of Hadrian as "scien-
The object of this paper is to address the first of tific," saying that "the limits of empire were by
these problems through an in-depth examination then demarcated very precisely, on the ground,
of a specific frontier zone at a particular period in so that all could tell exactly what was Roman
history. I will approach the question of terminol- and what was not" (Luttwak, 1976, p. 60). Al-
ogy by discussing some previous attempts at cate- though the assumption that the Roman frontier
gorization and by proposing a working was a clearly defined line of separation has been
terminology based on definitions commonly vehemently contested by various classical schol-
utilized by geographers. Then, by combining ars (see especially Isaac, 1990, pp. 394-401 and
B. J. Parker I Journal of An~hropologicrrl A rcliaeology 21 (2002) 371-395 373
more recently Whittaker, 1994, pp. 62-70 and passes the more distinct terms "border" and
Elton, 1996), the fact that much of the debate at "frontier." "Border" and "frontier" are con-
a recent conference entitled "Shifting Frontiers in trasting terms. A "border" is a linear dividing line
Late Antiquity" focused on whether or not linear fixed in a particular space, which, according to
borders existed in antiquity shows that this issue geographers, is meant to mark the division be-
is by no means settled (Mathisen and Sivan, tween political and/or administrative units (Pres-
1996). cott, 1987, p. 12). A "frontier" on the other hand,
Another concept that imposes modern linear is a much more loosely defined area or transition
stereotypes on ancient frontiers is the so-called zone that lies between two political, administra-
"natural" boundary, an idea conceived by the tive or cultural units of varying degrees of com-
geographer Ratzel in 1895 and later developed by plexity or between one such unit and a hinterland
Curzon (1907) and Holdich (1916). The concept either where no such polities exist, or where such
of natural boundaries, which envisions the limits polities do not come into direct physical contact
of ancient states and empires to be set by geo- (Prescott, 1987, p. 36ff).
graphic constraints such as rivers or mountain Some scholars have further defined frontiers in
ranges (Fischer, 1949; Pounds, 1951; Prescott, terms of demographic shifts, arguing that a fron-
1987), has also proven to be elusive in the ar- tier is itself characterized by the influx of settlers
chaeological and historical records. Strabo makes or pioneers to a sparsely populated hinterland
a revealing comment about the role of the Eu- (Lamar and Thompson, 1981; Wynman and
phrates river as the frontier between the Roman Kroeber, 1957). In my opinion, using demo-
and Parthian empires when he says: "The graphic shift, especially one that is unidirectional,
boundary of Parthian power with the country as a defining feature creates a core-centered view
opposite (i.e., the Roman Empire) is the Eu- of frontier situations that is far too rigid. Instead,
phrates river. But parts within [Parthian terri- I believe that demographic shifts can and often do
tory] are held by the Romans and the phylarchs play an important role in frontier dynamics, but
of the Arabs as far as Babylonia; some of them this is not necessarily so. Furthermore, the char-
adhere more to the Parthians and others more to acter of such demographic shifts can take many
the Romans who are their neighbors" (Strabo forms. The archaeological and textual data pre-
16.1.28). This is one of many oft-quoted exam- sented below will show that the Northern frontier
ples from the Roman period used to illustrate of Assyria did not serve as a hinterland to which
that rivers served not as "natural" boundaries colonists voluntarily migrated, as was the case, for
but as corridors of transportation and commu- example, in the North American and South Afri-
nication. can frontiers during the eighteenth and nineteenth
Similar arguments regarding the incompatibil- centuries. Instead, demographic shifts took the
ity of modern concepts of borders, either "scien- form of forced resettlement into the frontier by
tific," "natural" or otherwise, can also be made in foreign populations and flight from the frontier by
reference to the Assyrian frontier. We shall see, indigenous peoples.
for example, that neither the Zagros nor the Turner (1920) and Lattimore (1962) both ar-
Taurus mountains acted as "natural" borders gued that frontiers are areas where pioneers have
during the Assyrian Iron Age, and that the Tigris opportunities to exploit vast reserves of natural
river was to the Assyrians, just as the Euphrates resources. Scholars have since argued that the
and other rivers like the Danube and the Rhine North American frontier, for example, was inex-
were to the Romans, a transportation corridor tricably linked to the European system of
rather than a boundary. In fact, in light of the merchant capitalism through the fur trade
archaeological and textual evidence discussed be- (Berkhofer, 1981; Cook, 1981; Hickerson, 1973),
low, the entire concept of the linear preclusive while the discovery of gold and diamonds is said
border must be abandoned as a modern stereo- to have done the same for Southern Africa (La-
type contrary to the nature of boundaries in the mar and Thompson, 1981). Other scholars have
premodern world. seen economic frontiers in, for example, the divide
It is for this reason that I have chosen to dis- between sedentary agrarian and nomadic pastoral
tinguish between the terms "boundary," "bor- subsistence systems in India (Eaton, 1993) and
der," and "frontier." Boundary is a general elsewhere. Although the exploitation of natural
expression for a divide between geographic, po- resources was certainly an important factor con-
litical, or cultural entities and therefore encom- tributing to frontier dynamics, and in many cases
374 B. J. Parker I Journal of Antlzropological Archaeology 21 (2002) 371-395
helped shape change in both the frontier and the If we consider the term "boundary" to describe
core polity, I do not see the extraction of wealth all categories of limits or divides, then this dis-
as the singular defining feature of frontier situa- cussion has highlighted the fact that there is
tions. Nor do I see the economic boundary be- considerable variation among various types of
tween different subsistence economies as a boundaries. Some of this variation can be cate-
dominant frontier characteristic. Like demog- gorized by more precisely defining the terms
raphy, economics are one part of the multi-fac- "border" and "frontier." If we follow geogra-
eted layering contributing to the makeup of pher's definition of "border" as a linear dividing
frontiers. line, and "frontier" as a transitional zone (Pres-
Borders and frontiers have traditionally been cott, 1987), then the above discussion demon-
seen as belts of separation between one or more strates that, in their most extreme manifestations,
polities or between a polity and a sparsely in- these terms can be considered opposites. Thus we
habited hinterland. However, recent trends in the might envision the various types of boundaries as
study of globalization and culture theory have appearing on a continuum between the static lin-
sharpened our focus on how culture contact is ear border on one end and the dynamic fluid
interpreted and negotiated by both local and frontier on the other. For the sake of discussion
expansionist groups (Appadurai, 1996; Clifford, let us call this continuum the "Continuum of
1997). Instead of conceptualizing contact areas as Boundary Dynamics" (Fig. 1). If we agree that the
zones of separation and distinction, scholars are Continuum of Boundary Dynamics is meant to
beginning to see such areas as zones of interac- characterize the diversity of boundary situations
tion and hybridization (van Dommelen, 1997, then we should envision the continuum starting
1998; White, 1991). Richard White's theory of with the most extreme variation of the static
the "middle ground" takes this idea one step border on the far left. This type of boundary is a
further by demonstrating that frontiers between closed, static, spatially restricted border or line of
previously distinct peoples often act as interac- separation. As we move from left to right the
tion zones that, under the right circumstances, boundary becomes more dynamic. At the far right
can produce a hybrid area or "middle ground" of the continuum is the most extreme variation of
composed of an outgrowth of various aspects of the fluid frontier. This type of boundary is an
each previously distinct culture (White, 1991. open, dynamic, spatially expansive frontier or
Also see van Dommelen, 1997, 1998). Doyle has zone of interaction. For ease of discussion I have
argued that it is often interaction taking place broken the continuum down into four stages from
outside the core polity, in the frontier, that in- left to right: static, restrictive, porous, and fluid.
fluences and often precipitates change in the These terms are meant to aid us in our description
center (Doyle, 1986). A similar stand has been of frontier situations and thus describe various
taken by Chase-Dunn and Hall who see the regions along, rather than points on, the contin-
formation of complexity in regions peripheral to uum.
core polities to be the most likely location for the Beside highlighting the diversity of different
formation of new core polities (Chase-Dunn and boundary situations, the above discussion has also
Hall, 1991, 1993). shown that boundaries are, in fact, multi-dimen-
Although divisions between political entities sional. A complete analysis of any particular
are by far the most commonly discussed mani- boundary must therefore include various catego-
festation of borders and frontiers, Hugh Elton's
book Frontiers of the Roman Empire reminds us
that frontiers differ from borders not only because
they are zones rather than lines, but also because Boundaries 3
they include various categories of boundaries be- Border Frontier
yond the political and administrative. Elton em-
Sfatic Restrieti%,e Porous Fluid
phasizes cultural boundaries in the realms of
Geoeraphir I I
religion, ethnicity, linguistics, and material cul-
Pollllrnl - 1 -I
ture, demonstrating that political boundaries of-
ten do not correspond to cultural divisions. With D~n~~ra p-1hir -I
this in mind, Elton ultimately defines frontiers as C"l,"rnl -1 -I
overlapping zones of political, cultural and eco- F~ommlc -1 -I
nomic boundaries (Elton, 1996, pp. 3-9). Fig. 1. The continuum of boundary dynamics.
B.J. Parker / Journal of Antlzropological Archaeology 21 (2002) 371-395 375
ries of data. We have thus far considered political, geographically restrictive but culturally and eco-
administrative, geographic, demographic, and nomically porous. Furthermore, a "natural"
economic boundaries and in doing so it has be- boundary might be classified as geographically
come clear, not only that there are many layers or static or restrictive but politically, culturally, and
categories that might contribute to boundary sit- demographically fluid. It is from this point of
uations, but also that each of these categories reference and with these distinctions in mind that
might potentially be broken down into more nu- we will take the next step into the frontier, so to
anced sub-categories. For this reason I have in- speak, in an attempt to characterize the nature of
cluded five major categories of data to be Assyria's Anatolian frontier at the end of the
considered and evaluated on the Continuum of eighth century BC.
Boundary Dynamics: geographic, political, de-
mographic, cultural, and economic (Fig. 1). My
hope is that most more nuanced sub-types of data Defining the data set
will fall within one of these overarching catego-
ries. During the Mesopotamian Iron Age, from
In light of the above discussion, the frontier is about 1000 to 600 BC, the Assyrian empire was
transformed from a static line of exclusion to a the greatest power the world had yet known. For
dynamic zone of interaction where various cate- much of this period Assyria dominated the entire
gories of geographic, political, and cultural Near East from the Zagros mountains in modern
boundaries intersect and interact to form a unique Iran to the Eastern Mediterranean, and from the
matrix capable of changing, or at least influenc- Taurus mountains in Southern Turkey to the
ing, society, politics, and culture on both sides of Persian Gulf (Fig. 2). The Assyrian empire is well
the boundary. Furthermore, in using the Contin- known from references in the Bible. Perhaps the
uum of Boundary Dynamics it will hopefully be most famous of these references is to the Assyrian
easier to categorize, describe and compare various king Sennacherib who attacked Judah and be-
types of boundaries. For example, Luttwak's sieged Jerusalem during the reign of Hezekiah in
"scientific" boundary of the Roman empire 701 BC (Isaiah 36; 37. Also see Ussishkin, 1982
might be characterized as politically static and and Machinist, 1983).
ARABIAN DESERT
Fig. 2. Map of the Assyrian empire at its maximum extent (ca. 680 BC).
376 B. J. Parker 1 Journal of Antlzropologiccrl Arclmeolog,~2 1 (2002) 371-395
However, the historical significance of the east of central Assyria, and the Taurus Mountains
Neo-Assyrian empire lies neither in the modern of southeastern Turkey to its north and north-
perceptions of Assyria nor in the influence the west, was a patchwork of polities of varying de-
Assyrians imposed on the creation of the early grees of political centralization. It contained large,
Judeo-Christian world. Rather, Assyria's impor- highly centralized states like Hanigalbat whose
tance in world history lies in the fact that the influence stretched from their heartland in the
Assyrian state that emerged during the Mesopo- Khabur plains of north Syria well into south-
tamian Iron Age represented an entirely new level eastern Anatolia, Bit Zamani around modern
of political development in the ancient world. The Diyarbakir, Shubria northeast of Diyarbakir and
Assyrians were the first true empire builders, and Kumme north of modern Zakho. The Northern
it was upon their legacy that later empires, such as highlands, known to the Assyrians as the lands of
those of the Persians and Romans, were built. Nairi and Habhu, also contained a wealth of other
Despite Assyria's importance in ancient his- less-centralized societies such as the Early Iron
tory, many gaps still remain in our current un- Age polities of the Upper Tigris River Valley and s
derstanding of this once great empire. Several of the towns and villages of Dirn~i,n the Bohtan and
the major excavations carried out in the Near East Garzan river drainages, and the Kashiiari moun-
over the past 150 years focused on Assyrian cities tains, the modern Tur Abdin (Fig. 3).
such as Ashur, Nimrud, Khorsabad, and Nineveh. Neo-Assyrian imperial expansion irrevocably
Beside bringing to light the grandeur of the altered this geo-political configuration. Only the
Assyrian palaces and other material remains of heartland of Kumme would survive, although it
the Assyrian empire, these excavations uncovered was reduced to an Assyrian vassal and much of its
the majority of the written evidence pertaining to former holdings were annexed to the province of
the Mesopotamian Iron ~ ~ ein,clu'd ing much of the Mashennu. Early in the Neo-Assyrian Impe-
the royal correspondence and many of the an- rial period, Hanigalbat was completely destroyed,
nalistic texts. Since then, smaller, but much more making way for Assyrian invasion of the entire
scientific excavations and surveys have continued Upper Tigris River region. Bit Zamani and the
to augment the steadily increasing corpus of data Upper Tigris River Valley were made into the
regarding this important period. Nevertheless, provinces of Amedi and Tushan, while Shubria
research on the Assyrian empire has thus far was eventually divided into the provinces of Up-
concentrated on the Assyrian heartland and some ummu and Kulimmeri. The lesser societies of
of the Western provinces. Until recently, few ar- Nairi and Habhu were either destroyed, used as
chaeologists and historians have focused their buffer areas or forced to unite against the Assyrian
attention on the peripheral regions of the empire. threat, coalescing into the empire of Urartu (Fig.
The nature of the written sources, which are in- 4). The kingdom of Urartu eventually expanded
herently biased toward the center, and the scarcity into an empire that engulfed the entire Eastern
of archaeological information from outside the half of Anatolia, the area of Lake Sevan (the
Assyrian heartland, has meant that little research modern republic of Armenia) and the northwest-
has focused on the frontier zones where imperial ern corner of Iran around Lake Urmia. Urartu
expansion and the interaction between the empire also held sway over many of the city-states of
and the periphery actually took place. north Syria and the Levant forming a political .
The historical record indicates that before the system that rivaled even Assyria in its magnitude
era of Neo-Assyrian imperialism from about 930 (Salvini, 1995; Zimansky, 1985, 1998). This gen-
to 610 BC, the arch of mountainous terrain era1 outline of the progress of Neo-Assyrian im-
formed by the Iranian Zagros mountains north- perial penetration into its Northern frontier shows
that southeastern Anatolia was indeed on the
1
cutting edge of Neo-Assyrian expansion.
'
For the purpose of this paper I will refer to the This study focuses on three discrete areas along
Mesopotamian Iron Age as three roughly divided sub- the Tigris river in southeastern Anatolia. Al-
periods. The Early Iron Age, from approximately 1100-
though the archaeological components of this
900 BC, is the period directly prior to the expansion of
study are based on surveys conducted in the Up-
the Assyrian empire. The Assyrian Imperial period
per Tigris River region between 1988 and 1992
begins just before 900 BC and lasts until the fall of the
(Algaze, 1989a; Algaze et al., 1991), my analysis of
empire at the end of the seventh century. The Late Iron
Age, or Post-Assyrian period refers to the several the archaeological and especially the textual re-
hundred year period after the fall of the empire. cord is not limited to the specific survey areas, but
B.J. Parker I Journal of Anthropological Archaeologj~2 1 (2002) 371-395
Fig. 3. Map of southeastern Anatolia before Assyrian annexation.
instead focuses on the wider geographic region plain, in the vicinity of the modern border be-
within which they originally belonged.2 The first tween Turkey, Iraq, and Syria. The second study
study area is located in and around the Cizre area consists of two adjacent areas in the valleys
of the Bohtan and Garzan rivers; the third in-
cludes the U.m er Tigris River Vallev between the
The archaeological surveys analyzed in this paper were modern town & of ~i~an~d ~the figris-~atman
directed by Professor Guillermo Algaze to whom I owe a
confluence, and the Batman river valley between
great debt for allowing access to the data and for his help
the upper Batman dam and the Tigris (Fig. 5).
in the classification and analysis of the ceramics. Prelim-
inary reports of the surveys here discussed are published in
Algaze (1989a) and Algaze et al. (1991). For a more
complete discussion of the Iron Age in the Upper Tigris The nature of Assyria's Anatolian frontier
River region see also Parker (1997a, 2001). The textual
data discussed in this paper was extracted from the Neo- Geograplzic boundaries
Assyrian Text Corpus Project's database at Helsinki
University in Finland. I wish to thank Professor Simo
The physical geography of southeastern Ana-
Parpola and Dr. Robert Whiting not only for allowing me
tolia contrasts sharply with that of the Assyrian
access to the project's database, but also for many hours of
heartland in what is today Northern Iraq. The
consultation regarding the translation and interpretation
Assyrian heartland is characterized by steppe and
of the texts analyzed here. For up-to-date translations and
transliterations of most of the documents here discussed rolling hills, while Eastern Anatolia is dominated
see Parpola (1987) and Lanfranchi and Parpola (1990). by the Taurus mountains that, for all intents and
B.J. Parker 1 Journal of Anthropological Archaeology 21 (2002) 371-395 371
Turkey
Diyarbakir
River Region
Upper Tigris Wer Valky
Tigris
Mardin -
-- -__- - r
Iraq
,
, Syria
- _____- *- -
- -A - - Dam Projects I I
International Borders \
Fig. 5. Map showing the location of the survey areas analyzed in this paper (cf. Algaze, 1989a; Algaze et a]., 1991).
Assyria is traversed by a number of rivers in- Anatolian highlands. Despite the potential for
cluding not only the Tigris but also the Kosr and extensive agriculture in the plains of north Mes-
Zab rivers, its steppeland topography meant that opotamia, this region had few other natural re-
intensive irrigation agriculture was limited to a sources to offer. The opposite is true of the
few areas directly around the rivers or where an- Anatolian highlands where timber, metals, and
cient irrigation schemes could bring water to the semiprecious stones were abundant.' Even today
surrounding fields (Jacobsen and Lloyd, 1935; some of the richest copper mines in the Near East
Reade, 1978). Extensive dry farming agriculture in are located in the area of the Ergani Maden in
the plains of north Syria and Northern Iraq thus central Anatolia. Anatolia also contains large
produced dispersed rural settlement across nearly deposits of iron and was one of the main sources
the entire steppeland north of the 200mm isohy- of iron for the Assyrian empire (Maxwell-Hyslop,
ate (Weiss, 1986; Wilkinson and Tucker, 1995; 1974; Oppenheim, 1967). Maxwell-Hyslop has
Wilkinson and Barbanes, 2000). Adequate pre- pointed out that the mere existence of ore sources
cipitation to support rain-fed agriculture is much solved only part of the problem faced by ancient
less reliable in the Southern Assyria and south- metal workers. Since it takes over 8 tons of char-
central Syria. coal to smelt one ton of iron ore, only ore sources
In spite of the higher levels of precipitation in located in thickly wooded regions, like the Taurus
the Anatolian foothills, the lack large tracts of mountains, would have made mining and smelting
arable land that characterize lowland Mesopota- ore viable (Maxwell-Hyslop, 1974, p. 143). The
mia, meant that highland economies emphasized Assyrian dependence on this area for their supply
animal husbandry rather than extensive agricul- of metals and other raw materials absent from the
ture. Topography and the lack of adequate top- Mesopotamian lowlands is exemplified by the
soil, as well as the short growing season, all played Assyrian booty and tribute lists which describe
a role in making large parts of Eastern Anatolia large quantities of copper, iron, and other valued
unsuitable for cultivation (Zimansky, 1985). commodities extracted from enemies and vassal
The elevation differential and its corresponding states on the Northern periphery of the empire.
I
climate and precipitation rates, combined with the The above discussion shows that within the
extremely rugged character of the Taurus foot- category of "geographic boundaries" there are
hills, obviously means that the Taurus range several sub-categories that contribute to the geo-
formed a formidable geographic boundary be- graphic make-up of Assyria's Anatolian frontier.
tween these two regions. However, the Taurus In a sense, the geography of Northern Mesopota-
range was not only a topographic boundary
between Assyria and its Anatolian periphery, it
Yener has shown that although much of this area i?
was also the frontier between the resource-poor currently devoid of trees, this deforestation is a relative11
Mesopotamian lowlands and the resource-rich recent phenomenon (Yener, 1982).
380 B.J. Parlcer 1 Journal of Antltropological Archaeology 21 (2002) 371-395
mia and southeastern Anatolia fits well with ge- the last half of the eighth century BC during and
ographer's definition of a "natural" boundary slightly after the second phase of Assyrian inter-
(Fischer, 1949; Pounds, 1951; Prescott, 1987). This vention in the Upper Tigris River region.
region is marked by an extreme variation in to- Assyria's first major campaign aimed at the
pography, a topography that surely restricted annexation of portions of southeastern Anatolia
movement into and within this frontier zone. took place in 882 BC. It was in this year that the
However, two aspects of natural boundaries that Assyrian king Ashurnasirpal, responding to what
are often not considered by geographers are also Assyrian scribes referred to as a "rebellion" that
revealed by this discussion. One of these divided threatened Assyria's allies at the city of Damam-
the two regions and another brought them to- musa, took military control of large parts of the
'
gether. To begin with, the variation in precipitation Upper Tigris River region (Parker, 2001). The
and land type meant that there was a stark differ- textual and archaeological records indicate that
ence in the subsistence patterns between Northern Ashurnasirpal fully intended to permanently oc-
Mesopotamia and southeastern Anatolia. Al- cupy parts of the Tigris basin, for it was at this
though we can assume that this difference in sub- time that the Assyrians established the first in
sistence strategies led to cultural differences in what would become a series of fortified centers
terms of foodways and domestic production, this along the Upper Tigris river. The most important
assumption is difficult to test with the available of these strongholds was the city of Tushan, which
data. The distribution of natural resources is the also served as the capital of a province of the same
second sub-category that surely affected the nature name. Ashurnasirpal describes in some detail the
of Assvria's Anatolian frontier. The fact that re- construction of this city, saying that he sur-
sources such as timber and metals that were not rounded it with a city wall, constructed a palace
obtainable in Northern Mesopotamia were readily and large storage facilities there, and traversed the
available in southeastern Anatolia played an im- Tigris with a bridge of rafts to give the inhabitants
portant role in Assyria's foreign policy (see below). ready access to the fertile farmland on the north
Although the physical geography of Assyria's bank of the river (Grayson, 1991a, p. 202). He
Anatolian frontier must be considered to have followed this construction three years later by
been very restrictive, other aspects of the geogra- establishing two more fortified towns (Assyrian
phy of this region certainly contributed to both Tidu and Sinabu) at regular intervals east of the
centrifugal and centripetal forces shaping this city of Tushan (Grayson, 1991a, pp. 257-262).
frontier region during the Mesopotamian Iron This string of fortresses is still conspicuously
Age. visible along the south bank of the Upper Tigris
river. Three large mounds with extensive Assyrian
Political boundaries occupation grace the landscape of the Tigris basin
east of Diyarbakir. It is obvious from the surface
Assyria's intervention in the Upper Tigris morphology of these sites that each of them con-
River region of southeastern Anatolia can be tained a monumental gateway flanked by large
broken down into three main phases: the annex- towers strategically facing north toward the river.
ation of Nirdun (the Upper Tigris River Valley) Extensive research on the historical geography
and its conversion into the province of Tushan of southeastern Anatolia has made the direct
during the reign of Ashurnasirpal (883-859 BC); attribution of these sites with fortresses mentioned
'
the conquest of Ulluba (the Cizre plain and in Assyrian texts virtually certain (Karg, 1999;
surrounding area) and its conversion into the Kessler, 1980; Liverani, 1992; Parker, 1998,2001 ).
province of the Mashennu during the reign of This reconstruction strongly supports the location
Tiglath-pileser 111 (744-727 BC); and the subju- of the city of Tushan at the site of Ziyaret Tepe 3
gation of Shubria (the Taurus foothills south of about 10 km east of the modern town of Bismil; the
the Murat river) and its division into the provinces city of Tidu at the site of uGtepea bout 10 km west
of Upummu and Kullumeri during the reign of of Bismil; and the site of Sinabu at modern Mur-
Esharhaddon (680469 BC, see Figs. 3 and 4). atta? (previously known as Pornak) about 30 km
Since nearly all of the royal correspondence west of Bismil (Kessler, 1980; Liverani, 1992; Par-
originating in, or pertaining to, the Upper Tigris ker, 1998,2001). The modern city of Diyarbakir is
River region dates to the reigns of Tiglath-pileser located atop the ancient town of Amedi (Fig. 6).
111 (744-727 BC) and Sargon I1 [721-705 BC Recent archaeological surveys at Ziyaret Tepe
(Parpola, 1981)], this study focuses primarily on confirm that occupation at this site vastly
B. J. Parker. / Journal of Anthropological Archaeology 21 (2002) 371-395 38 1
the Masha
\
Fig. 6. Map of southeastern Anatolia with three of Assyria's key eighth century frontier provinces highlighted. The dots
show the location of known Assyrian frontier fortifications. From left to right the fortifications in the province of
Tushan are; Sinabu, Tidu and Tushan. The fortification at the Tigris-Bohtan confluence (between Dirru and Habhu) is
inferred from Assryian letter NL 67 (see Parker, 1997b, 2001). The ancient name of this site is unknown. The fortified
center on the Tigris in the province of the MaSennu was known to the Assyrians as ~abire~Tuh. e name of the fort in
Kumme is unknown.
expanded during the Neo-Assyrian Imperial chariots, cavalry, and specially trained troops for
period (Algaze, 1989a; Algaze et al., 1991; Matney, a swift strike to the east. This elite force reached
1998). The site grew from a relatively small village the valley of the Garzan river (Assyrian Dirru)
of only a few hectares centered on the high mound within a day and over the course of the next sev-
during the Early Iron Age, to a large center of over eral days ravaged the villages and towns of this
32 ha, which included the high mound and a vast and the neighboring Bohtan River Valley [Assyr-
lower town, during the Assyrian Imperial period ian Habhu, Fig. 6 (Parker, 2001; contra Radner
(Matney, 1998; Parker, 1998). Magnetometry and Schachner, 2001)l. In contrast to his efforts to
surveys of portions of the lower town at Ziyaret consolidate his military gains in the Upper Tigris
Tepe (Matney and Somers, 1999) have revealed River Valley, Ashurnasirpal made no attempt to
what appear to be substantial fortifications in the
form ofwalls and towers and several other mon-
umental structures. Surface survey also recovered
several terracotta clawed forepaws that presum-
ably belonged to monumental lions that guarded
each entrance to the city (Fig. 7). Excavations at
the site of uGtepe (Assyrian Tidu) have also re-
vealed substantial structures including fortifica-
tion walls over 3 m thick that enclosed the citadel
during the Assyrian Imperial period (Koroglu,
1998).
Ashurnasirpal was also the first Assyrian king
to venture into the second study area: the Garzan
and Bohtan river valleys. After consecrating his Fig. 7. Terracotta lion forepaw discovered in surface
new palace at the city of Tushan in 879 BC, survey at Ziyaret Tepe (Assyrian Tushan [photo cour-
Ashurnasirpal selected an elite force of heavy tesy of Matney]).