in a recent review of gary feinman and joyce marcus’s ... · acting, in the words of joel mokyr...

199

C u r r e n t A n t h ro p o l o g y Volume 42, Number 2, April 2001� 2001 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204-2001/4202-0002$3.00

Initial SocialComplexity inSouthwestern Asia

The Mesopotamian Advantage1

by Guillermo Algaze

The emergence of early Mesopotamian (Sumerian) civilizationmust be understood within the framework of the unique ecologyand geography of the alluvial lowlands of the Tigris and Euphra-tes Rivers during the late 5th and 4th millennia b.c. The formergave Mesopotamian societies important advantages in agricul-tural productivity and subsistence resource resilience not pos-sessed by contemporary polities on their periphery, while the lat-ter gave them enduring transportational advantages. Thismaterial imbalance created opportunities and incentives thatmade it both possible and probable that early Mesopotamianelites would use trade as one of their earliest and most impor-tant tools to legitimize and expand their unequal access to re-sources and power. Given this, a still hypothetical (but testable)model is presented that accounts for the precocious socioeco-nomic differentiation and urban growth of southern Mesopota-mia in the 4th millennium as social multiplier effects inadver-tently set in motion by evolving trade patterns. This trade wasfirst largely internal, between individual southern polities ex-ploiting rich but localized ecological niches within the Mesopo-tamian alluvium during the Late Ubaid and Early Uruk periods.By the Middle and Late Uruk periods, however, inherently asym-metrical external trade between growing southern cities and so-cieties at their periphery in control of coveted resources gainedmore prominence. In due course, import-substitution processesfurther amplified the one-sided socio-evolutionary impact onsouthern societies of these shifting trade patterns. Unequal de-velopmental rates resulting from the operation of these processesover time explain why the earliest complex societies of south-western Asia appeared in southern Mesopotamia and notelsewhere.

g u i l l e r m o a l g a z e is Chair of the Department of Anthro-pology at the University of California, San Diego (La Jolla, Calif.92093-0532, U.S.A. [[email protected]]). Born in 1954, he was ed-ucated at the University of Puerto Rico (B.A., 1976) and the Uni-versity of Chicago (Ph.D., 1986). His publications include “TheUruk Expansion: Cross-cultural Exchange in Early MesopotamianCivilization” (current anthropology 30:571–608), The UrukWorld System (Chicago: University of Chicago Press, 1993), and“The Prehistory of Imperialism: The Case of Uruk Period Meso-potamia,” in Uruk Mesopotamia and Its Neighbors, edited by M.Rothman (Santa Fe: School of American Research Press, 2001).The present paper was submitted 11 v 00 and accepted 7 ix 00.

1. Earlier drafts of this article were read and criticized by EliseAuerbach (University of Chicago), Claudio Cioffi-Revilla (Univer-sity of Colorado), Alex Joffe (Penn State University), Joyce Marcus(University of Michigan), Robert McC. Adams (UCSD), JenniferPournelle (UCSD), Steve Rosen (Ben Gurion University), Michael

In a recent review of Gary Feinman and Joyce Marcus’sedited book Archaic States (1998), Robert McCormickAdams (2000) expresses surprise that contributors gen-erally neglected issues pertaining to the social evolu-tionary impact of interrelationships between human so-cieties and their natural ecosystems. “It is a sign of ourtimes,” he argues, “that questions like these, so seriouslyregarded a generation ago, are presently as widely dis-regarded as they appear to be in this otherwise state-of-the-art collection.” The absence of such questions partlyobscures our understanding of the very processes thatthe various contributors to Archaic States seek to clarify.Social process viewed in a vacuum is tantamount to so-cial determinism, which ultimately is as incorrect as thegeographical determinism that often tainted the work ofan earlier generation of anthropologists and geographers(Livingstone 1992). Socioeconomic systems invariablyoperate within geographical and ecological constraintson the social actions that are possible (given existingtechnologies) and probable (given earlier developmentaltrends [path dependency] and cultural perceptions of op-portunities and threats) at any one time. For this reason,history displays a wide range of results of the interactionof societies and their environment. At the same time,however, evolving environmental factors can also extendthe boundaries of both the possible and the probable byacting, in the words of Joel Mokyr (1990), as “focusingdevices” that powerfully influence the direction that so-cieties take in their search for technological innovations.New technologies, in turn, to quote Robert Heilbroner(1994[1967]:59), “cannot but impose certain social andpolitical characteristics upon the societies in which theyare found.”

To be sure, neither Mokyr nor Heilbroner advocates areturn to Marxian technological reductionism. Both wellunderstand that the present is shaped from the past bya combination of selective pressures that impose a mea-sure of directionality on human history, random acci-dents, and unpredictable innovations that add an ele-ment of indeterminacy to any attempt at historicalexplanation. Moreover, both are well aware that tech-nological innovation not only acts on society but is itselfoften contingent on socioeconomic forces (Scranton1994). However, their admonitions as to the social con-sequences of technological innovation and the existenceof an often close relationship between ecological contextand technological change do raise important issues forarchaeologists attempting to explain the evolution of

Rosenberg (University of Delaware), Eric Rupley (University ofMichigan), Gil Stein (Northwestern University), Donald Tuzin(UCSD), Tony Wilkinson (University of Chicago), and several anon-ymous CA referees. Each offered substantive criticism, importanteditorial comments, and crucial missing references. Their invalu-able contributions are acknowledged with gratitude, even thoughoccasionally I chose to disregard their advice. Remaining errors ofomission and interpretation are entirely my own. The manuscriptalso owes much to the editorial comments and grammatical cor-rections of my wife, Susan Becker Algaze. The accompanying map(fig. 1) was ably drafted by Valerie Batt.

200 F current anthropology Volume 42, Number 2, April 2001

complex societies. In addition to taking into account thedynamics of social and economic interactions within andbetween societies, such attempts must also explore whattechnological innovations the groups under investiga-tion developed in order to adapt to or, alternatively, tocircumvent or transform their evolving environmentaland geographic context.

This latter perspective is adopted here to explore theemergence of early Mesopotamian (Sumerian) civiliza-tion in the alluvial lowlands of the Tigris and EuphratesRivers of southern Iraq during the so-called Uruk period,roughly dated to 3800–3100 b.c. This emergence entaileda veritable revolution in human social and spatial or-ganization, culminating in new conceptualizations ofthe nature of the social order (Baines and Yoffee 1998)and the rise of the world’s earliest cities and states (Ad-ams 1966, 1981; Adams and Nissen 1972; Nissen 1988;Stone 1997). Of the myriad aspects of this phenomenon,this paper focuses on explaining why the earliest urban-ized state societies in southwestern Asia emerged in thealluvial lowlands of the Tigris and Euphrates fluvial sys-tem and not elsewhere. This question is now relevantbecause recent work across portions of southwesternIran, northern Iraq, northern Syria, and southeastern Tur-key shows that in those areas societies comparable intheir scale and level of organizational complexity tothose of 4th-millennium southern Mesopotamia aroseonly after the Sumerian emergence and, to a great degree,as a reaction to it (Algaze 1993).

By the beginning of the 3d millennium, for instance,we witness the emergence of a Proto-Elamite state thatextended over parts of Fars and Khuzestan provinces ofsouthwestern Iran and exerted considerable influenceover remote areas of the Iranian Plateau (Alden 1982,Lamberg-Karlovsky 1996). Equally pertinent and betterdocumented are the numerous competing Early BronzeAge city-states, each centered at a fortified capital ofsubstantial proportions, that coalesced across the well-watered plains of northern Syria, northern Iraq, andsoutheastern Turkey (hereafter Syro-Mesopotamia)sometime during the middle of the 3d millennium (Weiss1983, 1986, 1990; Wilkinson 1994). Recent reviews ofpertinent evidence by both Weiss (1983, 1986) and Wil-kinson (1994, 1997) show that these urban societies hadassociated settlement networks that were as complex,differentiated, and extensive as those that had existed insouthern Mesopotamia prior to the Sargonic period. Theyalso show that for most of the Early Bronze Age theagricultural potential of the high plains of Syro-Meso-potamia (based on extensive rain-fed cultivation of grain)was sufficient to sustain the urban systems that devel-oped there.

Best-known among the Early Bronze Age city-states ofSyro-Mesopotamia is ancient Ebla (Tell Mardikh), nearAleppo, where the chance find of palace archives helpsus better understand the internal social and economicstructure of the city and the extent of the regional ec-onomic and political control that it was able to exercise(Arcari 1988, Archi 1990, Astour 1988, Bonechi 1993,Matthiae 1988). These archives also reveal the degree to

which the organizational structure of Ebla (and presum-ably that of other comparable Syro-Mesopotamian cities)paralleled earlier and contemporary southern Mesopo-tamian prototypes. Economic texts, for instance, revealpatterns of palatial control of animal herds and of variousstrategic industrial activities (e.g., metals and textiles)that are immediately comparable to contemporary prac-tices in southern Mesopotamian city-states (Archi 1988).

If the potential to support regionally organized andurbanized state-level societies existed in a variety ofregions of southwestern Asia, how does one account forthe temporal primacy of social complexity in southernMesopotamia? I believe that this primacy was the resultof social and “technological” innovations that were se-lected for and promoted by the unique geographic andenvironmental framework of the Tigris-Euphrates allu-vial lowlands. This is a point already made by Adams(1978, 1981) but one that bears being revisited becausenew analytical methodologies allow for more precise in-terpretations of the available survey data from southernMesopotamia than were heretofore possible. At the sametime, new paleoenvironmental data and computer sim-ulations of ancient climates allow for more precise re-constructions of the ecology of the alluvial lowlands ofthe Tigris and Euphrates Rivers at the time Mesopota-mian civilization first crystallized than have been prac-ticable until now.

More specifically, I argue that the primacy of southernMesopotamia was in part due to the fact that southernsocieties had several important material advantages overpolities in neighboring areas. These included (1) a denserand more varied concentration of exploitable subsistenceresources, (2) higher and more reliable agricultural yields,and (3) an exponentially more efficient distribution sys-tem based on water transport. These advantages pro-moted the creation of inherently asymmetrical exchangepatterns among independent polities in the Mesopota-mian alluvium and between those polities and societiesin neighboring regions which, over time, produced im-portant organizational asymmetries between southernsocieties and contemporary polities. In particular, by thesecond half of the 4th millennium the various advan-tages listed had promoted the emergence and rapid dif-fusion of innovative mechanisms of commodity produc-tion, labor control, and information processing withinthe myriad competitive but mutually communicativepolities occupying the southern Mesopotamian alluviumand immediately associated areas (Susiana). These in-novations can be classed as “ideational” technologies,and more than anything else they help explain why theinitial locus of early state development in southwesternAsia was the alluvial lowlands of the Tigris and Eu-phrates Rivers.

In the sections that follow, I discuss the various ma-terial and ideational advantages underpinning the rise ofearly Mesopotamian civilization in the Uruk period.

algaze Initial Social Complexity in Southwestern Asia F 201

Is Geography Destiny?

While geography is certainly not destiny, southern Me-sopotamian societies, at their outset, did enjoy a varietyof environmental and geographical advantages over theirneighbors that increased the likelihood that self-aggran-dizing individuals and factions in the south would bemore successful in maintaining, expanding, and legiti-mizing unequal access to resources and power than theirpeers in surrounding areas. This made it probable thatcomplex societies would emerge earlier in southern Mes-opotamia than in other regions of southwestern Asia.

The first advantage of the southern societies over pol-ities in immediately adjoining areas was the greater re-silience with regard to subsistence failure of the Tigris-Euphrates lowlands resulting from a greater variety anddenser concentration of exploitable subsistence re-sources. As Adams (1978; 1981:11–14) has noted, south-ern Mesopotamian societies could easily exploit severalcomplementary ecosystems at the same time, including(1) the irrigable alluvial plain, which provided high once-a-year yields of subsistence crops, principally barley, (2)smaller irrigable areas near the rivers, providing for in-tensively cultivated gardens and orchards where multi-ple high-value crops could be produced throughout theyear, (3) fallow fields within and outside the irrigatedareas that provided extensive pasture lands for millionsof sheep and goats and some cattle, and (4) rivers, arti-ficial canals, freshwater marshes, and brackish lagoonsand estuaries that provided plentiful protein in the formof fish and fowl, as well as reeds and other useful prod-ucts.2 In contrast, save for a once-a-year crop of dry-farmed grain and ample pasture lands, the high plainsand intermontane valleys on the periphery of the Mes-opotamian lowlands offered no comparably varied, re-silient, or dense concentrations of subsistence resources.

If anything, the natural advantages of the south overthe north in resource variety and density would havebeen particularly pronounced through much of the LateUbaid and Uruk periods. Central to this advantage wasthe array of littoral resources that would have been avail-able to Mesopotamian societies at that time (Pournelle2000). Recent geomorphological work shows that a mid-Holocene maritime transgression dating to the late 5thand 4th millenia raised the level of the Persian Gulf tosome 2 m above its present elevation. This resulted inan intrusion northwards of the head of the Persian Gulfof up to 200 km in places (depending on local variationsof elevation and differential subsidence rates), thus plac-ing the resource-rich marshes and estuaries of the gulf’slittoral zone in close proximity to Uruk population cen-ters such as Eridu, Ur, Uruk, Girsu, and Umma (Sanla-

2. Dried fish used as rations were essential to the economy of Me-sopotamian city-states in the historical periods (Englund 1990), andthey also figure prominently in some of the Archaic Texts datingto the very end of the Uruk period and the immediately followingJemdet Nasr phase (Englund 1998:128–43).

ville 1989) (fig. 1).3 Not only were these resources par-ticularly accessible but they were especially dense andvaried at this time. Sedimentological evidence fromdeep-sea cores in the Persian Gulf indicates that summermonsoonal rains of Indian Ocean origin had a morenortherly track during the late 5th and early 4th millen-nia than at present, bringing intense monsoonal rains tolarge parts of the Mesopotamian alluvial plain and theentire length of the Persian Gulf (Sirocko et al. 1993,Petit-Maire, Sanlaville, and Yan 1995). The consequencesof this climatic phenomenon for littoral resources arewell understood: the resulting upwelling of nutrient-richwaters and increased water oxygenation would havecaused across-the-board biomass productivity increases(Reichart et al. 1997) both throughout the gulf and itsestuaries and within the mosaic of canals, marshes, andlagoons surrounding many Uruk population centers.

The second advantage of southern Mesopotamia overits neighbors was provided by the higher yields and re-liability of the southern agricultural base. While thisedge was to taper off somewhat after the Uruk period, itnever totally disappeared, and it naturally increased theaforementioned beneficial impact of greater resource va-riety and density. Modern agricultural data from a varietyof locales in southeastern Turkey, northern Syria, north-ern Iraq, and southern Iraq show that, under conditionsof controlled irrigation, the alluvial landscape of south-ern Mesopotamia is, on average, about twice as produc-tive per unit of land than the rain-fed agricultural re-gimes characteristic of northern societies (Weiss 1986:figs. 1 and 2; Wilkinson 1990a:42–46). Although thereare no exactly comparable ancient data regarding pro-ductivity per standardized land unit, inferences derivedfrom cuneiform records of late-3d-millennium date bear-ing on seed-to-yield ratios suggest that irrigated lands inthe south had an advantage in terms of productivity overnonirrigated lands elsewhere that was roughly in the 3:1 range.4 Finally, provided that labor existed to maintain,repair, and deepen canals, drain excess water from fields,and build levees, the reliability of irrigated agriculture

3. Some of these early centers, in fact, may owe their locationprecisely to their proximity to littoral resources at the time of theirfoundation. In reviewing factors that affect urban location, theeconomist Paul Bairoch (1988) argues that cities preferentially format the end of transport routes or at the juncture of two differenttypes of transport routes. The southernmost Uruk centers (Uruk,Eridu, Ur, Umma, and Girsu) were thus at the same time at theend of the vast north-south transport route created by the riversand within reach of the lateral routes created by extensive marshesand lagoons at the encroaching head of the Persian Gulf in the 4thmillennium. This privileged position naturally lessened their trans-portation costs and maximized their access to resources from var-ious areas.4. Cuneiform records dating to the Ur III period (last century of the3d millennium), for instance, suggest that yield-to-seed ratios of30:1 were normal for irrigated fields in southern Mesopotamia andthat higher ratios (up to 50:1) were possible (Jacobsen 1982, but seethe reservations of Powell 1985). While comparable data are notavailable from dry-farmed areas in Syro-Mesopotamia in antiquity,ratios for modern grain from a variety of Near Eastern locales cul-tivated under traditional dry-farmed agricultural practices suggestthat yield-to-seed ratios in the 7:1 to 9:1 range appear normal forwheat and barley (Russell 1988:tables 3, 15, 18).


Fig. 1. The ancient Mesopotamian alluvium during the late 5th and 4th millennia b.c., showing the locationof principal Uruk centers and major watercourses at the time (more precise assessments are the subject of on-going research by R. McC. Adams and J. R. Pournelle at the Mesopotamian Alluvium Project Laboratory, Uni-versity of California, San Diego). Location of probable marsh boundaries and Persian Gulf coastline deducedfollowing Sanlaville (1989). Map by and courtesy of J. R. Pournelle, following Adams (1981, 1999) and Pournelle(2000).

in the south was greater than that typical of the dry-farmed areas of the north, which are subject to substan-tial and unpredictable spatial and temporal variations inrainfall (Perrin de Brichambaut and Wallen 1963:figs. 2and 3; Cullen and deMenocal 2000; Turkes 1996).

The natural advantages in yield, reliability, and pre-dictability of the southern Mesopotamian landscapewould also have been particularly pronounced through-out much of the 4th millennium. A unique conjunctureof three related factors helps account for this. The firstfactor is climate. The results of new paleoenvironmental(Bar-Matthews, Ayalon, and Kaufman 1997, Blanchet,Sanlaville, and Traboulsi 1997, Frumkin et al. 1999) and

geomorphological (Hole 1994, 1997; Wilkinson 1999) in-vestigations from various Near Eastern locales can becombined with recent computer simulations of global(de Noblet et al. 1996, Harrison et al. 1998), latitudinal(Hoelzmann et al. 1998), and regional (Bryson and Bryson1997, 1999) climatic changes resulting from orbitally in-duced variations in the amount of solar radiation reach-ing the earth at about 4000 b.c. to reconstruct mid-Ho-locene climatic conditions across the Near East. Thesedata suggest that, when compared with that of the his-torical periods, the climate of the region at the 5th/4th-millennium transition was generally wetter, withwarmer winters and cooler summers. Further, as much


as a tenfold expansion (from present levels) in availablesurface water appears to have been the norm at the time(Hoelzmann et al. 1998:46). This meant that marginalareas of the Mesopotamian alluvium that are today un-productive because of insufficient water or lack of drain-age would likely have been integrated into fluvial net-works draining into the sea (Hoelzmann et al. 1998:47).Additionally, the already noted Persian Gulf sedimentcores indicate that parts of the southern Mesopotamianalluvium that today receive no summer precipitationwhatsoever would have been affected by summer mon-soonal rains. Such rains would have been absent fromthe Upper Mesopotamian plains at this time (Blanchet,Sanlaville, and Traboulsi 1997:fig. 2). The same cores(Sirocko et al. 1993, Nutzel 1976) suggest that conditionsof increasing aridity and extreme seasonality that moreclosely approach those prevailing at present started toappear by the second half of the 4th millennium andbecame fully established by the onset of the 3d millen-nium (Butzer 1995). This conclusion is supported by pal-ynological data from lake cores in the Zagros-Taurusmountain ranges (van Zeist and Bottema 1991).

The implications of these climatic data go well beyondthe issues already discussed in reference to littoral re-sources. The data have equally important implicationsfor the productivity of the southern Mesopotamian al-luvial plain through much of the Uruk period. Increasedwinter precipitation in the Tigris-Euphrates watershedas a whole would have increased river volume in thespring, making both canal and flood-recession irrigationeasier throughout the southern lowlands. Increased pre-cipitation would have fallen on the alluvium itself atthis time, including some rain in the now barren summermonths. Wetter winters with warmer average tempera-tures would have allowed winter grain crops to matureearlier than they do today, thus minimizing productionlosses due to heat. Cooler summers with cloud coverassociated with the monsoonal regime of the time wouldhave decreased soil moisture loss throughout the allu-vium, thus increasing the likelihood that some high-value crops could have been grown successfully in thearea well into the summer.

The second factor serving to heighten the natural ad-vantages in yield and reliability of the Mesopotamianalluvium during parts of the 4th millennium was thebehavior of the Tigris-Euphrates fluvial system at thistime. New and ongoing research by Robert McC. Adams(1999) and Jennifer Pournelle (1999) confirms earlier sug-gestions by Adams (1981:16–17) and others (e.g., Gibson1973) that the main channels of both the Tigris and theEuphrates have moved considerably over the course ofthe past 6,000 years or so. Using newly declassified sat-ellite images (CORONA), Adams and Pournelle havebeen able to correlate the location of previously surveyed4th-millennium sites in the south with relict river chan-nels that are visible in the new images but were notidentifiable with precision in previously available airphotos or LANDSAT images of the area. From this theyconclude that large portions of the main channels of theTigris and the Euphrates were actually located within a

short distance of each other (ca. 25–50 km) during the4th millennium and that the two rivers actually joinedat several locations within the Mesopotamian alluvialplain of the time, creating a single complexly inter-twined fluvial system (fig. 1). The rivers separated totheir present courses only after the Uruk period, pre-sumably as a result of the combined effects of environ-mental forces and anthropogenic processes that startedin the 4th millennium.5

Adams and Pournelle’s conclusions are very relevantto the issue of yield differences between alluvial Meso-potamia and its neighbors during the Uruk period. It isvery likely that throughout much of the 4th millenniumthe waters of the two nearby rivers commingled at floodstage in the northern parts of the alluvium, where theircourses came closest. Increasing this possibility is thefact that the rivers carried a greater overall amount ofwater at that time. The result would have been the cre-ation of larger areas than exist at present where varioustypes of high-value vegetables, fruits, and even some ce-reals (e.g., sesame and millet) might have been producedin late spring and possibly even early summer by meansof simple flood-recession irrigation (Sanlaville 1989:24;Pournelle 1999).

The third factor adding to the advantage in productiv-ity of the southern alluvial lowlands over neighboringareas was the effect of the Persian Gulf transgression onthe fluvial dynamics of the Tigris-Euphrates river sys-tem. One such effect was that, by reducing the overalllength of the rivers, the intrusion would have aggravatedthe rivers’ natural tendency to meander and overflowtheir banks (Hole 1994) at the same time that it raisedthe level of the water table in the surrounding alluvialplain (Sanlaville 1989:18). This would have increased thesize of the areas subject to flooding without the need forsubstantial human intervention at times of overflow.

The importance of flood-recession irrigation for theinitial crystallization of Mesopotamian civilizationshould not be underestimated. Using ethnographic andeconomic data from Senegal and elsewhere, ThomasPark (1992) has shown that recession irrigation is gen-erally the most productive form of agriculture known interms of units of labor invested, irrespective of the cropsbeing exploited. He argues that this practice representsa powerful spur for the evolution of social stratification

5. One of the long-term effects of the Persian Gulf transgressionnoted earlier would have been to increase overbank sedimentationin the areas between the two rivers through the 4th millenniumas a result of reduced river length, gradient, and velocity (Butzer2001). This natural process contributing to the buildup of sedimentsseparating the two rivers would have been compounded by the un-intended consequences of man-made irrigation efforts. As noted byPournelle (1999 and personal communication, 2000), by decreasingthe volume of water in the main river channel, Uruk-period irri-gation efforts would have reduced the river’s ability to carry its siltload, thus contributing to a further increase in sediments depositedin the surrounding alluvium. Additionally, because irrigation ef-forts at the time naturally focused on the intervening areas betweenthe Tigris and the Euphrates, those efforts must have heightenedlevee buildup along artificial channels in the intervening areas aswell as increased the silt loads deposited in preferentially floodedfields within those areas.


in societies practicing it because of the particularly highvariability of output that it creates between well-irri-gated, poorly irrigated, and unirrigated lands. Thus,flood-recession irrigation is likely to have been impor-tant in the initial development of population agglom-erations in the alluvium—particularly in its northernportion, where the rivers came closest to each other. Notcoincidentally, this is precisely the portion of the allu-vium in which Adams’s (1981) surveys document themost precocious development of Uruk urban centers(Early Uruk period).

Although it never totally disappeared, the overallsouthern edge in yield and reliability per unit of landprovided by irrigation would have diminished somewhatafter the Uruk period for several reasons. First, as thegulf intrusion receded (lowering the water table acrossthe alluvium) and as the main channels of the riversseparated because of natural and man-made leveebuildup following the Uruk period (see n. 5), the extentof areas exploitable by flood-recession agriculture wouldhave been reduced. This would have necessitated theconstruction of ever larger and more expensive canals inorder to make up in grain production for the decline inhigher-value crops. Second, after centuries of irrigation,the higher water-table levels of the 4th millenniumwould naturally have resulted in varying degrees of sal-inization of parts of the alluvial plain (Jacobsen and Ad-ams 1958; Adams 1981:151–52). This largely unavoidableprocess would have been compounded by attempts toshorten fallow (Gibson 1974) and to bring ever more mar-ginal lands under cultivation as competing Early Dy-nastic polities tried to gain advantage over their peers byfurther increasing agricultural production within theirterritories.

At the same time that overall yields were eroding inthe south, developments outside of the Mesopotamianalluvium were helping increase agricultural yields acrossthe periphery, thereby contributing to a lessening of yielddifferences between the two areas after the Uruk period.As noted by Wilkinson (1994), the higher average south-ern yields per unit of land were eventually partly cir-cumvented in the Syro-Mesopotamian plains, at least inthe short term, by the intensive application of manureon cultivated areas near settlements and by an increasein the overall amount of land under cultivation (on thispoint, see also Weiss 1986). However, these forms of in-tensification required the formation of small territorialstates able to draw on agricultural resources from rela-tively broad areas through several tiers of dependent set-tlements, and there is no incontrovertible evidence forsuch states in those areas until the middle of the 3dmillennium b.c. Be that as it may, what is relevant hereis that throughout the 4th millennium the south stillhad a clear and uncontested advantage (per unit of land)in the yield, reliability, and resilience of its primary sub-sistence resources. Uruk elites could therefore extractrelatively larger surpluses per unit of labor than theircontemporary counterparts in the north.

The third advantage of alluvial Mesopotamia wastransportational efficiency. The cities of the alluvium

were, in effect, at the head of an enormous dendritictransportation system created by the north-to-south-flowing rivers. This allowed them to procure informa-tion, labor, and commodities from areas within the vastTigris-Euphrates watershed more efficiently than any po-tential upstream competitors or rivals away from therivers (Bairoch 1988:11, 14). To be sure, exports from thealluvium still had to move by means of relatively inef-ficient donkey caravans following the course of the riversnorthward, but the crucial edge of southern cities lay intheir ability to import needed commodities in bulk fromfaraway resource-producing areas in the surroundinghighlands at low cost, transporting them on rafts or boatsdownstream on the rivers, tributaries, and canals (BritishAdmiralty 1917:291–92; Potts 1997:122–38; Sauren1966). Of equal importance, the network of canals sur-rounding Mesopotamian cities and connecting themwith the main courses of the rivers allowed them tomove bulky agricultural commodities across their im-mediate dependent hinterlands with great efficiency(Weiss 1986:94). In contrast, societies in the Mesopota-mian periphery had to rely wholly on less efficient modesof overland communication, such as donkeys or carts,both for their long-distance exchange needs and for themovement of subsistence resources across their imme-diate hinterlands. The natural transportational advan-tages of southern cities help explain, in part, why thosecities could grow, on average, significantly larger andmore differentiated than their peripheral counterpartsthroughout the Chalcolithic and Bronze Ages.6

A Synergistic Cauldron

The developmental takeoff of southern Mesopotamia inthe 4th millennium b.c. can only be understood againsta background combining the extraordinary transporta-tional advantages offered by the Tigris-Euphrates fluvialsystem, the unique density and variety of the subsistenceresources provided by the southern alluvial environmentof the time, and the absence from that environment ofother necessary resources, such as metals, timber, andstatus-validating exotics. The synergy created by theseconditions spurred the creation of high levels of socialand economic differentiation, promoted unprecedented

6. Early Dynastic Warka, for instance, was somewhere between 400and 600 ha in extent (Finkbeiner 1991)—a size not reached by anycity in northern Mesopotamia until Nineveh grew to the enormoussize of 700 ha and became the center of the Neo-Assyrian empirein the 8th and 7th centuries b.c. (Stronach 1994). Wilkinson (1995)argues inductively, from pertinent survey evidence, that throughoutthe Bronze Age there was a 100-or-so-ha “natural” size limit forurban centers in Syro-Mesopotamia. He contends that the breakingof this limit that took place in Neo-Assyrian times was due to acombination of (1) more intensive coercive policies that increasedthe rate of surplus extraction from the countryside, (2) a resettle-ment program fueled by deportees from throughout the Near Eastthat brought lands previously too marginal for agricultural use intocultivation, and (3) a more flexible economy that succeeded in over-riding the until-then-normal limitations on population size in thenorth caused by the frictional effects of overland transport of bulkagricultural products.


population agglomerations within the alluvium, and se-lected for the creation of new forms of social organizationand technologies of social control that until then hadnever existed in human societies. To understand whythe ecological and geographical framework of Mesopo-tamia had such radical social consequences, we mustbriefly review the nature of urban development and itscauses.

Particularly pertinent to this review is the work of theiconoclastic urban expert Jane Jacobs, whose ideas aboutthe role of cities in human development have alreadyinfluenced some anthropologists interested in the dy-namics of urban growth in early civilizations (e.g., Kurtz1987) and are now gaining increasing currency amongmainstream economists (Nowlan 1997). Though Jacobsis best known for her often controversial views about thefuture of the urban experience in America (Jacobs 1961,1984), her work also includes a variety of more abstractcontributions about the nature of urban economies(1969, 2000) that are applicable to attempts to understandthe dynamics of early social complexity in Mesopotamia.In the more theoretical aspects of her work, Jacobs isultimately inspired by ideas first enunciated by HerbertSpencer well over a century ago. Spencer believed thata tendency toward increasing differentiation was inher-ent in all features of the universe, including the characterof human societies. He argued that as societies becamenaturally more diverse through time, relations of inter-dependency were created among their diverse social el-ements and that this produced emergent properties thatwere not predictable a priori (Spencer 1967[1876,1882]:8). In a similar vein but with a narrower focus, Jacobssees economic differentiation as central to processes ofsocial change and urban growth. For Jacobs, social dif-ferentiation results from economic differentiation, andsocial evolution ultimately depends on economic expan-sion. She is one of a small number of scholars who seehuman economies as sharing many characteristics withbiological ecosystems (see also Mokyr 2000 and Ziman2000 for in-depth discussions of similarities and differ-ences between biological and technological evolution).The most salient of these similarities are that (1) thedegree of resiliency and stability of the two types of sys-tems and their ability to expand are directly related tothe degree of diversity present (Jacobs 2000:22, 37); (2)expansion ultimately depends on capturing and usingexternal energy, principally light in the case of ecosys-tems and exogenous resources in the case of human so-cieties, and the more diverse means a system possessesfor using, modifying, and passing around energy/re-sources, the larger the cumulative consequences to thesystem as a whole (p. 47); and (3) development takes placeas part of a larger web of co-developments; the greaterthe internal diversity of the system, the more numerousand intricate the co-development relationships that willexist within it and the greater the number of emergentproperties that will be spawned (pp. 19–22).

Given these similarities, for Jacobs the developmentof social and ecosystemic complexity is studied by ex-ploring how diversity is created within the system, how

resources external to the system are incorporated intoit, and how developments that initially may have beenindependent join in inherently unpredictable and un-foreseen ways to cause further differentiation. Jacobs ex-plains that in both types of systems differentiation is anopen-ended and self-amplifying process, since each newdifferentiation constitutes the basis from which furtherdifferentiation can eventually emerge. Expansion occursas the processes generating diversity are repeated againand again.

More specifically, Jacobs argues that what determinesa settlement’s ability to grow is a positive feedback loopinitiated by its capacity to generate exports by combiningsome of its imports or preexisting resources with humanlabor and capital. This generates economic diversity atthe same time that it makes it possible for growing set-tlements to acquire more and different imports, some ofwhich can again be used to generate further exports. Thisprocess creates co-developments in the form of an in-creasingly large, skilled, and diverse workforce (i.e., hu-man capital), and this, in turn, creates the potential forfurther economic diversification by adding new types ofwork and new ways of working. As both work and di-versity expand, so does population density within theaffected settlements. This increase commonly takesplace at the expense of nearby rural populations, whichis why cities are always the economic and physical shap-ers of their hinterlands. Eventually, when their own mar-ket has developed sufficiently to justify producing someof the imports that had previously been acquired fromother places, expanding settlements commonly have ac-quired both the skills and capital to do so. In this manner,by replacing imports with local manufactures in a secondburst of economic expansion, growing centers generateeven greater diversity and employment, thereby settingthe stage for further economic and demographic growthand for an even sharper restructuring of settlement pat-terns in their immediate vicinity (Jacobs 1969, 2000).

While exchange between independent societies willalways have an impact on all of the polities involved,the beneficial multiplier effects of the two-step eco-nomic process just described will disproportionally affectsocieties in which imports consist mostly of raw re-sources while exports consist mostly of value-addedgoods. This is so because economic diversity, employ-ment, and skills expand doubly as a result of the needto process both the imports and the exports in thosesocieties. This, in turn, is significant for the developmentof sociopolitical complexity in such societies because thedivision of labor generates incentives for cooperation be-tween diverse individuals with common interests (Lich-bach 1996), as well as incentives for competition be-tween societal factions with conflicting or overlappinginterests (Brumfiel 1994).

The export-driven model for social and urban devel-opment that Jacobs proposes is particularly applicable tothe study of how and why early Mesopotamian civili-zation emerged. Cities and city-states were, in fact, themost typical and enduring political formations through-out the history of that civilization (Stone 1997, Yoffee


1995). Moreover, rich as the Mesopotamian alluviummay have been in agricultural, pastoral, and maritimeresources, it still lacked a substantial proportion of thenonsubsistence material requirements for the highlystratified social systems that emerged there by the Urukperiod (Algaze 1993). Thus, historically, the southern cit-ies imported from highland areas on the alluvium’s pe-riphery many resources that required a significant degreeof skilled processing before they could be incorporatedinto the economy. Most prominent among these labor-intensive imports were roofing-grade timber, wood, var-ious types of metal ores, various utilitarian, semipre-cious, and exotic stones, and bitumen, among others.Likewise, traditional exports from the southern cities,whether destined for other cities within the alluvium orfor distribution in faraway markets, were also almostentirely labor-intensive and consisted principally of sur-plus grain, leather products, dried fish, dates, processeddairy fats, and, most important, finished textiles (see Al-gaze 1993:63–74; Larsen 1987; van de Mieroop 1997:195–96; Yoffee 1981, 1995).

The production of textiles, traditionally the most im-portant export from the Mesopotamian alluvium, pro-vides us with a perfect case in point as to the applicabilityof Jacobs’s model to the Mesopotamian context. As prac-ticed in ancient Mesopotamia during historical times,textile production combined imports of wool from hin-terlands surrounding individual cities with a substantialamount of human labor and capital to create commod-ities that were both consumed domestically and exportedlong distances. Textual and iconographic evidence sug-gest, in fact, that textile production was already an im-portant urban industry in southern Mesopotamia duringthe second half of the 4th millennium. Many of the ear-liest pictographic tablets that appear by the end of theUruk period (the Archaic Texts), for instance, deal eitherwith the disbursement of finished textiles or with theacquisition of raw wool (Nissen 1986). By the same to-ken, depictions in seals and sealings of female workers(pigtailed figures) attending horizontal looms are com-mon in Uruk-period levels at a variety of sites (Amiet1972:nos. 673–74; 1980:nos. 319–20).

However, our most detailed evidence for the impor-tance of textile production to the economy of Mesopo-tamian cities comes from economic texts dating to the3d and early 2d millennia b.c. These documents showthat each major Sumerian city-state had a palace-organ-ized weaving establishment in which thousands of de-pendent women and children labored to process woolimported into the city from the surrounding countrysideinto finished fabrics and garments (Jacobsen 1970[1953],Maekawa 1980, Waetzoldt 1972). The texts reveal thatthe amount of time and human effort spent in the pro-duction of these fabrics, earmarked for both domesticconsumption and foreign markets, was astonishinglyhigh. According to Larsen (1987), simple fabrics took al-most a month to complete while particularly elaboratepieces, intended for export, took, on occasion, well overthree years. Larsen’s analysis of Old Assyrian texts ofearly-2d-millennium date clearly shows that the major-

ity of the textiles that merchant families based in thenorthern Mesopotamian city of Assur exported to north-ern Syria and Anatolia were actually of southern Me-sopotamian origin (i.e., “Akkadian”). Since the areaaround Assur was perfectly capable of supporting thenecessary wool industries, the fact that Assyrian mer-chants felt compelled to import textiles from farawaypolities in the south reveals how highly prized thesesouthern Mesopotamian products were in the areaswhere they marketed their goods. No doubt this waspartly the result of the superior workmanship and pres-tige of southern textiles, since the areas where thosetextiles were consumed, too, could have supported com-plex textile industries.

Once in place, the self-amplifying consequences of pat-terns of production and trade such as these are easy tovisualize. First, the exchange would have promoted con-siderable economic differentiation and associated co-de-velopments in the alluvium because of the need to pro-cess both imports and exports (Jacobs 1969, 2000). This,in turn, would have created incentives for the emergenceof factions within growing southern polities and in-creased the likelihood of competition within and be-tween them (Brumfiel 1994). Second, it would have stim-ulated the acquisition and maintenance of the(dependent) workers that, in a Mesopotamian culturalcontext, were needed to ensure the production of labor-intensive exports. Third, it would have spurred the ex-pansion and institutionalization of the bureaucratic su-perstructure needed to organize and maintain the labor,to record its production, and to redistribute the resultingcommodities across the landscape (Wright and Johnson1975). Fourth, it would have promoted the developmentof steeper settlement hierarchies within the alluvium,since an adequate flow of local agricultural resources wasrequired for export. Thus, the larger centers capable ofconducting the exchange would have taken steps to at-tract and appropriate, by whatever means necessary, theagricultural and pastoral production of rural communi-ties in their immediate hinterlands (Adams 1981:80–81).

What still needs to be clarified, however, is how ex-port-driven economies in southern Mesopotamiaevolved in the first place. What were the initial exportsthat brought in the initial imports, thus setting into mo-tion the export/import positive feedback loops describedby Jacobs? Here, I believe, is where the greater varietyand density of subsistence resources offered by the south-ern Mesopotamian landscape and the particularly une-ven distribution of some of these resources during the4th millennium came to serve as necessary conditionsfor the Mesopotamian takeoff. The considerable agri-cultural, pastoral, and maritime resources of the areaacted in effect as “unearned” imports, acquirable onlythrough human efforts (Jacobs 2000:54). Exports of initialstocks of diverse “starter” resources in different concen-trations throughout the Tigris-Euphrates lowlands werewhat allowed individual southern polities to begin todevelop export-driven economies as a result of intra-al-luvium trade before polities in surrounding regions coulddo so. In this manner, southern societies acquired a de-


velopmental head start over their neighbors which wasneither planned nor foreseen but was to have long-termconsequences for the developmental trajectories of bothgroups of societies.

The Evolution of Export-Driven Economies inSouthern Mesopotamia

Although detailed evidence is still lacking, Jacobs’smodel allows us to visualize a still speculative thoughultimately testable scenario to account for the preco-cious urban growth that took place in the Mesopotamianalluvium in the 4th millennium. In this scenario, thepositive feedback economic loops responsible for thatgrowth evolved over centuries as a result of exchangethat was at first largely internal between growing south-ern centers and later complemented and amplified by asignificant measure of external trade between those cen-ters and less developed polities in the Mesopotamianperiphery. For heuristic purposes, this evolving processmay be divided into a number of discrete stages, althoughsubstantial overlaps clearly existed between them.

As already noted, the initial impetus for the growth ofexport-driven southern economies with their attendingmultiplier effects would have been inadvertently pro-vided by burgeoning trade between polities exploitingthe various rich ecological niches that existed within theTigris-Euphrates alluvial lowlands during the Late Ubaidphase and the earlier part of the Uruk period (late 5th toearly 4th millennium). Each of these polities would havenaturally specialized in the production of a small numberof crops or commodities for which it had a competitiveadvantage owing to its location within the alluvial ec-osystem. Products traded in this initial stage would haveincluded (1) woven and dyed textiles, goat-hair products,leather goods, dairy fats, and other pastoral resourcesdistributed by polities situated at the margins of the bet-ter-watered parts of the alluvium, where they wouldhave enjoyed preferential access to pastoral and nomadicgroups producing these various commodities, (2) gardenand grain crops produced by polities in the northern por-tion of the Mesopotamian alluvial plain, where the com-bined flow of the Tigris and the Euphrates made irriga-tion agriculture and horticulture both more likely andmore profitable, and (3) dried, salted, and smoked fish,reeds, and other marsh or littoral resources preferentiallyproduced by polities near the Persian Gulf littoral.

A second stage in the process would have been markedby an emerging elite awareness of the social implicationsof the trade patterns in place until that point in time.This stage was characterized by a decrease in regionalspecialization within the alluvium as each competingpolity used the material surpluses and human skills ac-quired during the first stage to replace some imports fromnearby centers by creating their own productive capac-ities for those products, thus setting in motion the fur-ther growth spurt that accrues from the import-substi-tution mechanism discussed by Jacobs. In this stage,

competitive emulation would account for the diffusionof technologies and practices that were initially devel-oped by individual centers exploiting specialized nichesbut soon came to be perceived as highly advantageousby many of the centers in competition. Foremost amongthese quickly diffused activities would have been themanufacture of textiles, an industry present in everyMesopotamian polity of any consequence in the 3d and2d millennia b.c. Adding to the attraction of textiles isthe fact that they lend themselves particularly well tothe expression of the multiple statuses and local stylesthat are the natural outcomes of the processes of internaland regional differentiation described earlier (D. Tuzin,personal communication, 2000).

To be sure, some measure of trade with areas outsidethe Mesopotamian alluvium, allowing Mesopotamianpolities to acquire needed utilitarian and exotic resourcesand goods not locally available, must have existedthroughout this process. A case in point is the import offlint and obsidian obtained from areas on the peripheryof southern Mesopotamia, which is attested throughoutthe Ubaid period in the south (G. Wright 1969). This caseis important because it offers us what is perhaps theearliest example of import substitution within the al-luvium, as imports of raw flint and chipped stone toolswere partly replaced by locally manufactured clay sicklesthroughout the Late Ubaid and Uruk periods (Benco1992).

Nonetheless, external trade becomes a significant fac-tor in the overall economy of southern societies only ina third stage of the process, datable to the Middle andLate Uruk periods (ca. 3600–3100 b.c.). This stage wasfueled by the material and human capital accumulatedduring the earlier stages of largely internal exchange.Specifically, what made possible the intensification ofexternal trade at this time was the predevelopment anddiffusion of industries that were originally intended tosatisfy consumption requirements within the alluviumbut easily adapted for export markets outside it. Nodoubt, woven and dyed textiles were the principal com-modity fulfilling this double role in the 4th millennium,as they were later on in the historic periods. Other ex-ports of the time are more easily documented in thearchaeological record. These include bitumen fromsouthern Iraqi and southeastern Iranian sources(Schwartz, Hollander, and Stein 1999) and processed ag-ricultural (wine, unguents, aromatic oils?) and pastoral(animal fats) commodities contained within the varioustypes of Uruk ceramic containers (four-lugged, spouted,and pear-shaped jars) that are often found in many in-digenous Late Chalcolithic sites across the Mesopota-mian periphery (Algaze 1993:63–74; 1995; Badler, Mc-Govern, and Glusker 1996; Englund 1998:161–69).

This third stage of the process saw the establishmentof various types of southern outposts at strategic loca-tions of significance for transport across the Mesopota-mian periphery. These outposts served as collection andtransshipment points for the increasing amounts of pe-ripheral commodities imported into the alluvium in thelater part of the Uruk period and as distribution points


for the exports from the alluvium exchanged for them.The outposts need not be conceived as the product of asingle “master plan.” Rather, they are best considered asresulting from an organic process of action and counter-action, with individual Uruk polities scrambling to oc-cupy specific positions in order to secure access to thecritical lines of communication through which resourceswere obtainable and, of equal importance, to deny theirlocal rivals such access (Algaze 1993, 2001; but see John-son 1988–89, Oates 1993, Pollock 1992, Rothman 1993,Schwartz 1988, Surenhagen 1986, Stein 1999, and Roth-man 2001 for varying views about the nature and func-tion of the outposts).

By the later half of the Uruk period, regular access toexternal resources became necessary because of the needto supply the increasingly differentiated and dense urbanpopulations emerging in the southern alluvium. Importsof roofing-grade timber and copper (for the manufactureof cutting tools) must have been crucial at this point tosatisfy the substantial building requirements of rapidlygrowing Uruk cities. It has been estimated, for example,that somewhere between 3,000 and 6,000 m of roofing-grade timber would have been necessary to roof theLimestone Temple at Warka (Eanna V), dated to the LateUruk period (Margueron 1992).7 This is but a single struc-ture (albeit a large one: ca. 27 # 80 m p 0.21 ha) in acity that, at this point, occupied an area estimated atsomewhere between 200 and 250 ha (Finkbeiner 1991).Equally important would have been a steady supply ofprecious and semiprecious stones and base and exoticmetals, which, like some of the better-made textiles,were needed to mark and legitimize the ever more di-verse and unequal social statuses evolving within Urukcities. Substantial amounts of such exotic materialshave, in fact, been recovered at a variety of Late Uruk-period sites. A storeroom with various types of semipre-cious stones (including several kilos of lapis lazuli) andsome copper ingots, for instance, was found at JebelAruda, an Uruk outpost in northern Syria (Rouault andMassetti-Rouault 1993:435). Even more impressive wasthe hoard of exotic and semiprecious stones used as in-lays in furniture recovered from the Riemchengebaudestructure in the Eanna (Level IV) precinct at Warka,which may represent either a temple (Forest 1999) or astorehouse associated with a nearby temple (Nissen1988).8

While the great majority of imports into the alluviumthroughout the 4th millennium required some degree ofprocessing, it is likely that some finished goods made ofexotic materials were also being imported at this time.These would have been made by peripheral societieswith a long tradition of extracting and working those

7. The large range in Margueron’s estimate is due to the variouspotential ways in which the building could have been constructed(i.e., number of levels, whether the courtyard was roofed, and po-tential differences in beam interval).8. It is also very likely that many of the exotics contained in theso-called Sammelfund Hoard, which was discovered in a JemdetNasr–period level (Eanna III) at Warka, were heirlooms of Uruk-period date (Heinrich 1936).

materials in areas close to the pertinent raw materials(Kohl 1987a:16). Cases in point are provided by a smallnumber of elaborate obsidian vessels found in layers un-der the White Temple at Warka (Heinrich 1937:pl. 59)which closely resemble examples found in Late Chal-colithic tombs at the small northern Mesopotamian siteof Tepe Gawra (cf. Tobler 1950:pl. 53 B-C) and whichwere almost certainly imported as finished products.Also in the same category is at least one javelin or har-poon-head made of an unusual copper-silver alloy (over25% silver) found in the Riemchengebaude at Warka(Muller-Karpe 1991:109, fig. 3). This weapon is almostcertainly imported from the Taurus region of Anatolia,since weapons made of the same type of alloy are com-mon in Levels VIA and VIB at Arslantepe, a small LateChalcolithic chiefly center near Malatya (Palmieri,Hauptmann, and Hess 1997).

However important finished imports may have beento the overall economy of Uruk city-states in the earlyphases of the evolving exchange, it is clear that it didnot take long for further phases of import-substitutionprocesses to take hold in the south. A case in point isprovided by metals, which almost certainly were firstimported largely as finished products. While evidence ofmetalworking and metal-processing installations inUruk-period sites is still relatively rare, Moorey (1994:243) attributes this paucity to the use of simple but ef-fective technologies such as crucible smelting that havelargely escaped the attention of archaeologists lookingonly for more substantial installations such as furnaces.Nonetheless, the pictogram for a smith (showing a smelt-ing furnace with attached blowpipes) is already attestedin the earliest Archaic Texts (Uruk IV script), which dateto the very end of the Uruk period (Moorey 1994:243).Additionally, there are some archaeological indicationsthat metal artifacts began to be produced locally insouthern Mesopotamia during the Uruk period (still us-ing, of course, imported raw materials). At Warka, forinstance, both copper ores (Heinrich 1938:25) and a cop-per-processing installation (Nissen 1970:114) are re-ported in Uruk levels. Similarly, more recent excavationsat several intrusive Uruk outposts in the Upper Euphra-tes portion of Syria have yielded substantial and clearindications for metal processing in Mesopotamian so-cieties of the Uruk period. Tell Sheikh Hassan, for ex-ample, has provided evidence for smelting crucibles ina Middle Uruk-period level (Boese 1995:175, pl. 13a).From Late Uruk-period levels at Jebel Aruda, in turn,comes a hoard of eight copper axes (Rouault and Mas-setti-Rouault 1993:fig. 115), which almost certainlyserved as ingots (G. Stein, personal communication,2000).9 Finally, Habuba Kabira-sud, the largest intrusive

9. The axes in question were of more or less equal weight in spiteof variations in size and were found in a storeroom that also con-tained a variety of exotic stones and minerals (Rouault and Mas-setti-Rouault 1993:435), suggesting that they were being catego-rized as raw materials rather than as tools. Buttressing Stein’ssuggestion that the axes served as ingots is the fact that the Maritexts of early-2d-millennium date actually refer to copper ingots inthe form of axes (Yener 1980:32). Additionally, as Yener (pp. 34–35)


Late Uruk site on the Syrian Euphrates, has yielded ev-idence for the use of a fairly elaborate cupellation processto extract lead and silver from polymetallic ores im-ported into the site. These metallurgical activities arereported from several locations within the northeasternquarter of the site (Kohlmeyer 1997:447).

The Urban Takeoff of Southern Mesopotamia

How closely the emergence of social complexity insouthern Mesopotamia followed the developmental pathjust suggested is a subject for future research. What isclear is that, by the end of the 4th millennium, substan-tial developmental asymmetries had emerged betweensocieties in southern Mesopotamia and surrounding ar-eas. These asymmetries are most easily gauged by a com-parison of rates of population agglomeration in the twoareas at this time. Although detailed and systematic sur-vey coverage is still lacking for large portions of the northand northwestern periphery of alluvial Mesopotamia,there can be little doubt that throughout the 4th mil-lennium southern Mesopotamia had a substantial ad-vantage over areas on its periphery in the proportion ofits population living in urban agglomerations and theirimmediate vicinity. Adams’s calculations indicate that,depending on the area, somewhere between 40 and 70%of the total population in the Mesopotamian alluviumin the later half of the Uruk period lived in settlementsthat can be classed as towns or cities (i.e., 10 ha or larger;cf. Adams 1981:75, table 4). These results are now but-tressed by a recent reassessment of the same data byPollock (2001:table 5), who tries to correct Adams’s databy counting only sites likely to have been occupied si-multaneously during any one period. Her more conser-vative results suggest that the ratio of population livingin agglomerated settlements throughout the surveyedportions of the alluvium during the later half of the Urukperiod may have been even higher than Adams had orig-inally reckoned (ca. 69–77% depending on the area).

The most telling example of the disparities in popu-lation agglomeration between southern Mesopotamiaand peripheral areas in the 4th millennium is the cityof Uruk, which, as already noted, had reached the ex-traordinary size of 200–250 ha by the Late Uruk period.Adams’s (1981; Adams and Nissen 1972) surveys showthat Uruk was surrounded by an irregular array of de-pendent towns and villages located within a 13–15-kmradius of the city that accounted for an additional 280ha of occupied settlement. This, of course, almost cer-tainly substantially underestimates the actual popula-tion density of the Uruk countryside, since normal al-luviation patterns in southern Mesopotamia wouldnaturally obscure a large proportion of the smaller sites(Wilkinson 1990b). However one estimates the popula-tion of this megalopolis and its dependencies (see Post-gate 1994 for a discussion of the pitfalls of estimating

points out, the use of standardized tool or artifact forms as ingotsis well attested in a variety of ethnographic cases in Africa.

population densities in a Mesopotamian context), therecan be no doubt that the Uruk city-state constituted thelargest single agglomeration of population under a singleruler for its time not only in the Near East but in theworld. Moreover, the city of Uruk was hardly alonewithin the Mesopotamian alluvium in the Late Urukperiod. Multiple interacting urban sites (ca. 20–40� ha)existed across the surveyed portions of the southern Mes-opotamian alluvium at this time, all situated alongsidecanals and within relatively short distances of each other(Adams 1981). Additionally, at least two further Urukurban polities existed in the Susiana Plain of south-western Iran.

There is some debate about the nature of the politicalsystem within which these various polities existed. Flan-nery (1995) and Marcus (1998) explicitly assume that allpristine states emerge from a crucible of conflict as ear-lier regional chiefdoms are consolidated into a singlemuch larger and complex polity. They suggest that Urukwas the capital of a territorial state by the Late Urukperiod, pointing to the fivefold size differential betweenUruk and its nearest second-tier settlements as sup-porting evidence. This is a plausible interpretation of theavailable evidence from the Mesopotamian alluvium,which, unfortunately, outside of Uruk itself, still con-sists largely of surface survey data. Equally plausible atthis point, however, are interpretations that see theemergence of early Mesopotamian civilization as a casein which a number of rival statelets evolved in parallelas a result of common factors such as the multipliereffects of trade. From this perspective, the various urbancenters across the Mesopotamian alluvium in the LateUruk period are seen as independent polities ancestralto the competing city-states that characterized the areathroughout the 3d millennium and for much of its his-tory (Adams 1981, Baines and Yoffee 1998).

Of the two interpretations, I believe the latter is morelikely. In the first place, processes of urban growththrough the Uruk period appear to have resulted in thecreation of buffer zones between some of the emergingurban polities of the time. These uninhabited areas sug-gest competition and strife rather than an overarchingpolitical unity. The clearest instance of this pattern isthat documented by Johnson (1987) for the Susiana Plain.His surveys document the abandonment of all MiddleUruk–period villages along a 15-km-wide arc by the LateUruk period, separating what appear to have been tworival states, centered at Susa and Chogha Mish, respec-tively. These sites thus appear to have been independentof each other and almost certainly were also independentof contemporary polities in the Mesopotamian allu-vium.10 In the second place, later literary traditions, such

10. A further area of village abandonment can be observed betweenthe northern (Nippur-Adab) and southern (Uruk) regions of the Me-sopotamian alluvium surveyed by Adams (1981:compare figs. 12and 13). Whether this can also be interpreted as a buffer zone be-tween competing polities or sets of polities is, however, more am-biguous than the comparable case in Susiana because the area ofthe alluvium in question was one also affected by river coursechanges that may have contributed to the observed population shift(Adams 1981:63).


as the Sumerian King List (Jacobsen 1939), explicitly con-ceptualize the earliest Mesopotamian political configu-rations as composed of independent cities, each with itsown ruler. If a primordial state centered at Uruk thatheld sway over the whole of the Mesopotamian alluviumexisted in the Late Uruk period, it left no mark on laterliterary traditions that appear interested precisely in doc-umenting, however fancifully, the flow of political powerwithin Mesopotamia. Finally, political paramountcy isnot the only thing that can account for the size differ-ences between Uruk and smaller contemporary urbansettlements. Steinkeller (1999) has recently called atten-tion to a handful of Jemdet Nasr–period tablets showingindividual Sumerian cities sending resources to Uruk asritual offerings to Inanna. He sees these tablets as evi-dence for the operation of a precursor of the BALA dis-tribution system that would keep the religious institu-tions of Nippur supplied with offerings and resourcesduring the Ur III period, a millennium or so later. If thisis indeed true, it means that Uruk functioned as thereligious capital of Sumer through the 4th millennium,and, if later historical practice may be projected back-wards, this would have precluded a political dimensionfor the city at this time.11

Leaving aside the nature of political systems in theMesopotamian alluvium during the Late Uruk period,Adams’s data clearly show that by that time multipleinteracting urban settings had already characterized theMesopotamian alluvium for centuries. In fact, his sur-veys of the Nippur-Adab area of the alluvium show thatat least four separate 40�-ha centers existed along threeseparate branches of the rivers during the Early Urukperiod (Adams 1981:fig. 13). To this must be added thesite of Uruk itself, farther downstream, which was cer-tainly of comparable size at this point. These centers didnot exist in isolation. When the relevant survey data aretallied, it appears that complex three- and four-tieredsettlement structures already characterized the surveyedareas of the alluvium by the Early Uruk period. Thesedata indicate that the proportion of the population livingin urban agglomerations in the alluvium at that timewas already close to the 50% mark according to Adams’s(1981:75, table 4) calculations. If Pollock’s (2001:table 5)reassessment of the same data is preferred, that numberis as high as 80%.

When the settlement structure of the Mesopotamianalluvium throughout the various phases of the Uruk pe-riod is compared with that of contemporary societies inneighboring areas, the contrast could not be starker. Ad-mittedly, some individual Late Chalcolithic settlementsof substantial size did exist in the Upper Khabur andJebel Sinjar areas of Upper Mesopotamia, such as TellBrak at somewhere between 40 and 65 ha in extent(Oates and Oates 1997, Emberling et al. 1999) and Tell

11. Nippur, the acknowledged pan-Sumerian religious capital of theMesopotamian alluvium throughout the 3d millennium, was nevera key political player during that time.

Hawa at 30� ha (Wilkinson and Tucker 1995).12 How-ever, those settlements were not embedded in complexsettlement grids as southern centers were. Tell Hawa,for example, was surrounded in the Late Chalcolithicperiod by a corona of uniformly small village- or hamlet-sized sites (Wilkinson and Tucker 1995:fig. 35, top). Amore complex three-tiered settlement pattern structureappears in the vicinity of Hawa and the Sinjar area onlyafter the onset of contacts with the Uruk world (Wil-kinson and Tucker 1995:fig. 35, bottom). A similar caseappears to obtain for Tell Brak, which was clearly a sub-stantial indigenous settlement during most of the 4thmillennium (Area TW, Levels 19–13), before some sortof a southern Mesopotamian presence was introducedinto the site during the Late Uruk period (Area TW, Lev-els 12–11; Oates and Oates 1993, 1994, 1997). The LateChalcolithic phases of Brak have been described as doc-umenting a regionally integrated urban polity compa-rable to but even earlier than those of Uruk-period south-ern Mesopotamia (Oates and Oates 1997, Emberling etal. 1999, Lamberg-Karlovsky 1999). This claim is basedon the extent of the site in the Late Chalcolithic period,the fact that a small number of suburb areas where spe-cialized production took place surrounded the site at thetime, and the substantial nature of the architecture un-covered in the small exposures that have been practi-cable thus far in relevant levels (Area TW, 13–17 and18–19). Nonetheless, the available survey evidence fromthe immediate hinterland of Brak, though much less de-tailed than that available from the vicinity of Hawa,shows that during the Late Chalcolithic period the sitewas surrounded by uniformly small village- or hamlet-sized sites (Eidem and Warburton 1996:59–60).13 Unlesslarger dependent sites existed around Brak and Hawa thatare thus far undocumented, the bipolar regional settle-ment pattern that characterized both areas lacks thedepth and range of settlement size variability found tosurround contemporary southern Mesopotamian urbansites and is most parsimoniously interpreted as reflectinga chiefdom-level organizational structure. Finally, recentattempts to characterize the Late Chalcolithic settle-ment of Arslantepe (Period VIA) as a state-level society(Frangipane 1997, 2001) are also unconvincing becauseof the small overall size of the site (4 ha) and the absenceof detailed information about contemporary settlementpatterns in its vicinity.14

The fact remains that when we look at those areas of

12. Often-repeated claims that the site of Hamoukar in northeast-ern Syria was 90 ha in extent during the Late Chalcolithic period(Surenhagen 1986) are now discredited by recent work showing thatits actual extent was but 13 ha (Gibson 2000).13. The sites in question overwhelmingly appear to be 1.5 ha orless in total extent. The only exceptions to this are four sites con-taining Late Chalcolithic–period remains that vary in extent from2.3 to 2.8 ha (Sites 23, 25, 29, 32) and one site that is 5 ha in size(Site 38). However, the actual extent of the Late Chalcolithic–periodoccupation of these latter sites cannot be ascertained from the pub-lished data (Eidem and Warburton 1996:56–60, fig. 2).14. The one source available (Conti and Persiani 1993) lists numbersof sites but gives no indication of their extents or of the extent ofoccupation of the relevant period within them.


the Mesopotamian periphery for which quantifiable sur-vey evidence is available—such as portions of the Eu-phrates basin in southeastern Anatolia (Wilkinson1990a, Wilkinson in Algaze, Misir, and Wilkinson 1992,Algaze 1999, Algaze, Breuninger, and Knudstad 1994) andnorthern Syria (Wilkinson 1994, Danti 1997), portions ofthe Upper Tigris basin in southeastern Anatolia (Algazeet al. 1991), portions of the Upper Khabur basin of Syria(Stein and Wattenmaker 1990, Eidem and Warburton1996, Kouchoukos 1998), and portions of the Sinjar plainsof northern Iraq (Wilkinson and Tucker 1995)—we failto find evidence for settlement hierarchies across thenorth in the Late Chalcolithic period comparable in scaleand complexity to contemporary Uruk-period develop-ments in the south. Moreover, whatever large Late Chal-colithic settlements existed across the north were es-sentially isolated from one another. The compact andlong-lived systems of closely interacting urban settle-ments embedded in complex settlement grids whichcharacterized the south during the various phases of theUruk period simply did not exist in the less densely set-tled landscape of the Mesopotamian periphery at thattime, and such systems would not arise in the peripheryuntil the middle of the 3d millennium (Wilkinson 1994).

The Further Synergies of Density andPropinquity

Multiple consequences would have arisen from the dif-ferences in population density and distance between pol-ities typical for southern Mesopotamia and areas on itsperiphery throughout the 4th millennium. One of theconsequences of the greater density of population intowns, cities, and their immediate hinterlands of thesouth was the compounding of the natural advantagesin transport efficiency of the alluvial environment byfurther savings in transport costs and efficiencies in com-munication arising from the increasingly compact ar-rangement of the inhabitants of the area. This, in turn,must have spurred further economic differentation andits associated multipliers, because, as Adam Smith(1954[1776]:bk. 1, chaps. 1–3) explained more than 200years ago, gains in the efficiency of transportation andcommunication in human societies always act as a spurfor division of labor, economic growth, and technologicalinnovation. Additionally, greater density provided south-ern polities with a larger number of easily recruited la-borers (for both agriculture and construction), craft spe-cialists, soldiers, and potential colonists than could bemarshaled by their less densely agglomerated competi-tors in the periphery. When coupled with the higher pro-ductivity per unit of labor in the alluvium and the higherreliability of that production, the larger pools of deploy-able labor in the south meant that Uruk rulers couldextract and amass much larger surpluses than thoseachievable by even the most centralized of the northernpolities.

Equally important for southern societies were the con-

sequences of the greater concentration of polities thatexisted in the Mesopotamian alluvium throughout the700-year-or-so duration of the Uruk period. As ColinRenfrew and his colleagues have repeatedly argued (Ren-frew and Cherry 1986), the long-term presence of mul-tiple polities within relatively short distances of eachother invariably engenders important processes of com-petition, exchange, emulation, and technological inno-vation. The social evolutionary impact of these mutuallyreinforcing processes has been explained by RobertWright (2000:165–68), who notes that in situations whereantagonistic but mutually communicative polities exist,social and economic innovations that prove maladaptivein any one society are likely to be weeded out morequickly than in less competitive settings. Conversely,innovations that prove advantageous are more likely tospread quickly across the various polities in competition,thus accelerating the overall pace of change of the systemas a whole.

The Domestication of Human Labor and ItsAssociated Technologies

The Jacobs model of urban development clearly explainshow the geographic and environmental advantages of theMesopotamian ecosystem promoted patterns of com-modity production and exchange within 4th-millenniumsouthern societies that led to the creation of more ag-glomerated and differentiated settlements than existedelsewhere. The same model explains how larger markets,larger pools of labor, new skills and new forms of work,and repeated opportunities for cooperation and conflictwithin and between polities were created in the south.These conditions explain why the final and ultimatelymost important asymmetry that existed between Uruk-period societies and communities on their peripheryemerged. This was a disparity in the abilities of Meso-potamian and peripheral societies to resolve the inevi-table managerial conflicts that large-scale population ag-glomerations entail, to store and process information(writing and associated reckoning systems), and to com-mand and organize diverse labor pools. Here southernpolities held a definite “technological” edge over theircontemporaries that was, in effect, the sufficient con-dition for the crystallization of Mesopotamian civiliza-tion. This edge was made possible by important inno-vations in administrative technologies that took placethroughout the Uruk period (H. T. Wright 1978). Whilethis advantage would disappear in the 3d millennium asmany northern societies adopted forms of social controland writing systems ultimately derived from southernmodels (Postgate 1988), in the 4th millennium Uruk pol-ities were still more efficient than their neighbors inextracting labor from dependent workers and better atdeploying those workers where most needed, appropri-ating their production, and converting their output intosocially useful commodities.

Borrowing a page from V. Gordon Childe, we may use


the term “Labor Revolution” to describe this crucialcombination of socioeconomic advantages. Underlyingthis transformation was a conceptual shift in the waysome categories of human labor were looked at in south-ern Mesopotamian societies. Southern elites came toview and use fully encumbered laborers in the same ex-ploitative way that human societies, over the immedi-ately preceding millennia, had viewed and used the laborof domesticated animals. This represents a new paradigmof the nature of social relations in human societies. Isuspect that a comparable shift in the way in which hu-man labor is perceived (and exploited) is integral to allcases of early state formation, but in reality we do notknow exactly when this perceptual change occurred inthe Mesopotamian case. All we can say for certain is thatit had already taken place by the end of the Uruk period,as can be observed in the Archaic Texts. Scribal sum-maries detailing the composition of groups of foreign andnative-born captives used as laborers describe them withage and sex categories identical to those used to describestate-owned cattle (Englund 1998:176–81). Because theseparallels are repeated in numerous texts, they cannot beexplained away as either accidents or scribal idiosyncra-sies. Rather, it would appear that the two classes of labor(captive “others” and domestic animals) were consideredequivalent in the minds of Uruk scribes and in the eyesof the institutions that employed them. Early Near East-ern villagers domesticated plants and animals. Uruk ur-ban institutions, in turn, domesticated humans.

In practical terms, the perceptual shift just describedmeant that Uruk elites probably had a greater variety ofencumbered laborers at their disposal than did theirnorthern counterparts, that they could extract more en-ergy from those laborers, and, more important, that theywere better able to organize them in nontraditional waysso as to take advantage of increases in productivity andother economies of scale arising from the specializationof individuals according to principles of comparative ad-vantage (Snooks 1996:165). These qualitative advantageswould have been magnified further by the sheer force ofnumbers. While available documentation is not suffi-cient to quantify the contribution of slaves, prisoners ofwar, and other types of encumbered workers to the econ-omy of Uruk city-states with any degree of precision, itwould appear that their number was significant. Onerecently published Archaic Text fragment (Uruk IVscript), for instance, is a summary of several smaller in-dividual accounts and refers to a total of 211 male andfemale captive laborers (Englund 1998:178–79, fig. 66).Admittedly, there is no way to know how representativethis tablet is. However, a crude but perhaps useful mea-sure of the relative importance of slaves and encumberedworkers to the portion of the Uruk economy that wasbeing recorded by state scribes may be obtained by look-ing at the number of individual attestations of the per-tinent signs for the various categories of captive laborersin the Archaic Texts (5,820 complete and fragmentarytablets to date, including Uruk IV and III scripts). RobertEnglund’s (1998:70) comprehensive list of every knownattestation of each nonnumerical sign of this corpus

shows that the second-most-frequently mentioned com-modity was SAL (female slave), with 388 attestations (forcomparative purposes, barley, the most frequently notedcommodity, had 496 attestations). To this must be added153 attestations of laborers described as ERIM (captives)and 113 attestations of laborers described as KUR (maleslaves).

A less direct measure of the importance of encumberedlabor to the Uruk economy may also be available. Manyof the Archaic Texts record disbursements of textiles andgrain to individuals and presumably represent rationsgiven to some sort of fully or partly dependent workers(Englund 1998:178–79, fig. 67). Again, the contributionof such workers to the Uruk economy cannot be gaugedwith any precision, but if we presume that Nissen (1970,1976) is correct in assuming that the ubiquitous beveled-rim bowls were used as ration bowls, their incidence insouthern cities, where typically they are found inamounts that defy quantification, suggests that the num-ber of workers receiving state rations in return for theirlabor during the second half of the 4th millennium musthave been quite substantial indeed. Access to pools ofdependent labor of comparable scale does not appear tohave existed in indigenous societies on the northernMesopotamian periphery at this time. While redistri-bution of rations is attested to at some pertinent sites,most notably at Arslantepe VIA (Frangipane 1997) andTepe Gawra VIIIC (Rothman 1994), the much smallerquantities of mass-produced bowls found at those loca-tions suggests a much smaller number of individuals re-ceiving rations than that typical for southern sites.

Similarly, there is ample epigraphic (the Titles and Pro-fessions List [attested already in Uruk IV script]) andadministrative (seals, sealings, reckoning devices) evi-dence for both the extraordinary level of economic andsocial differentiation that had developed within Uruksocieties and the depth of the controlling bureaucracythat had emerged in southern Mesopotamia by that time.When measured against this evidence, bureaucratic andadministrative practices across the Mesopotamian pe-riphery during the second half of the 4th millennium arefound wanting. By the second half of the Uruk period,the south had greatly outpaced the periphery in admin-istrative technology—in spite of the fact that sealed claybullae, used as records of commodity shipments, hadactually been in use in northern sites at a much earlierdate (Halaf) than in any comparable southern context(von Wickede 1990, Lamberg-Karlovsky 1999). The“great leap forward” of the south over neighboring so-cieties in technologies of social control and in economiccomplexity is brought into relief by the fact that no pe-ripheral site has yet provided any evidence for the ex-istence of formal reckoning and writing systems com-parable in their complexity to those that had emergedin southern Mesopotamia and the Susiana Plain. This ismost clearly seen when we compare accounting proce-dures in use within Middle–Late Uruk–period levels ofsites such as Warka and Susa in the south (Nissen, Dam-erow, and Englund 1993, Amiet 1972, Dittmann 1986)and those in use in roughly contemporary Late Chal-


colithic-period levels at the sites of Arslantepe (PeriodVIA: Ferioli and Fiandra 1983[1988]) and Tepe Gawra(Levels XI–VIII: Tobler 1950, Rothman 1994) in thenorthern periphery. The thousands of sealings found dis-carded at various locations within the Period VIA com-plex at Arslantepe, for instance, commonly bear the im-pression or impressions of but a single seal. A similarcase obtains in the Late Chalcolithic levels of Tepe Ga-wra. At both sites, impressions of more than one seal ona single sealing surface are exceptional (Fiandra 1994:168; M. Frangipane, personal communication, 1999; M.Rothman, personal communication, 1999).

In contrast, contemporary procedures in Uruk citiesin the south and in Uruk outposts in the north regularlyexhibit the imprints of multiple seals, particularly in thecase of balls and bullae (Delougaz and Kantor 1996).These differences are relevant because the number ofimpressions of different seals in a single sealing gives usa glimpse of the number of agents and, possibly, wit-nesses involved in the transaction. Additionally, if Nis-sen (1977) and Dittmann (1986) are correct in seeing theseals as encoding information about the hierarchicalranking of specific authorizing individuals and the in-stitutions they worked for, then the much greater fre-quency of complex devices with multiple impressions ofdifferent superimposed seals in Uruk centers can betaken as a proxy for the greater number of levels of bu-reaucratic control and accountability that existed withinthose centers compared with the much smaller northernsites (Pittman 1993). The northern counterpart of theimpressed balls and numerical notation tablets of thesouth is but a few tallying slabs at Arslantepe VIA, withevenly sized indentations on their surfaces (presumablyrepresenting numbers), which appear to have served asmnemonic devices (Liverani 1988[1983]:figs. 1–4). Thesedevices could convey only a fraction of the informationthat could be transmitted by the more elaborate reck-oning systems of southern societies at this time, whichrelied on complex combinations of numbers and images(seal impressions). Moreover, unlike the more complexsouthern systems, which were capable of conveying in-formation at a distance and across time, the mnemonicsystems of the north were incapable of communicatingany information beyond their immediate institutionaland temporal context.

The disparity in efficiency and complexity betweenthe accounting systems of the two areas (and by inferencethe divergence in the scale and complexity of their econ-omies) becomes even more marked by the very end ofthe Uruk period with the appearance of the earliest Ar-chaic Texts in the south (Uruk IV script: ca. 1,900 outof 5,820 tablets and fragments [Englund 1998:86]). Thistook place at a time that appears contemporary withArslantepe VIA on the basis of available radiocarbondates (Wright and Rupley 2001). The new pictographictablets of the final phase of the Uruk period not onlygave southern administrators a way to record inflows andoutflows of commodities in a form transmissible throughspace and time—as the earlier balls and numerical no-tation tablets had done—but allowed them to do so with

much greater precision. More important, because thenew pictographic tablets could also express nuances oftime, location, persons involved, and action effected,they allowed Uruk scribes to convey a range of economicinformation that was at the time unprecedented. More-over, the greater informational content and flexibility ofthe new tablets also allowed southern scribes to abstractand summarize detailed data about collections and dis-bursements of goods and labor in a form usable by them-selves at a later time, by higher-level supervisory officialsat any time, and by later generations of similarly trainedbureaucrats. A final difference is that, although rare, thenew lexical lists that also appear at this time (Uruk IVscript) gave southern Mesopotamian scribes the abilityto organize, categorize, and transmit information abouttheir material and social world beyond that contained inthe economic tablets that constituted the totality of ear-lier records (Nissen, Damerow, and Englund 1993, En-glund 1998).

In short, by the end of the Uruk period, cumulativeinnovations in the ways knowledge was gathered, pro-cessed, and transmitted through time and space providedsouthern Mesopotamian decision makers and the urbaninstitutions they worked for with a flow of varied andreliable economic data of the sort that is necessary forthe formation, maintenance, and effective expansion oflarge-scale economies and social groups. More impor-tant, these data allowed Uruk elites to deploy availablelabor and goods where they were needed most in orderto maximize their revenues, extend their power, andshore up the stability of the social system they were inthe process of transforming. This furnished southernMesopotamian societies with an important competitiveadvantage over contemporary neighboring polities, inwhich similar breakthroughs in accounting, accounta-bility, and classification appear to have been absent. Infact, indigenous systems of administrative control equiv-alent in their sophistication to those used in southernMesopotamia already in the second half of the 4th mil-lennium would not be attested in the Near East untilthe rise of the Proto-Elamite state in southwestern Iranat the beginning of the 3d millennium and would notappear in Syro-Mesopotamia for at least 600 years or soafter the end of the Uruk period—and when they ap-peared it was in the form of an imported technology thatrelied wholly on the cuneiform writing system and as-sociated accounting procedures elaborated centuries be-fore by southern societies (Postgate 1988).

Conclusions

In a recent review of the relationship between scale andcomplexity of social organization in cases of pristinestate formation, Gary Feinman (1998) argues that theform such states take at the outset is largely dependenton the particular combination of factors most central totheir creation. If this is the case, then geography, envi-ronment, and trade can be seen as the most importantfactors helping shape the initial nature of social com-


plexity in the Mesopotamian alluvium. Geography andenvironment would have been central not in a directdeterministic sense but rather in the sense that they pro-vided the opportunities and incentives that made it prob-able that self-aggrandizing individuals and groups woulduse trade as one of the most efficient strategies at theirdisposal in their quest to achieve wealth, status, andpower. Though we still lack much of the pertinent ev-idence for early trade in southern Mesopotamia in theill-understood Late Ubaid and Early Uruk periods, it isnot difficult to speculate on the impact that such tradewould have had on the formation of early southern Mes-opotamian cultural complexity. Trade would have beencrucial because its varied and self-amplifying social ram-ifications would have created a situation in which theparallel development of multiple independent centerswas a likely outcome. Each of these centers would havetaken advantage of its particular location within the al-luvial ecosystem to exploit a particular resource nicheor a particular transportation route in and out of thealluvium, both on principles of comparative advantage.The localization of resources across the greater Meso-potamian landscape and the resulting salient role oftrade, first internal and then external, in the processesof urban growth and state formation in the Tigris-Eu-phrates lowlands help explain why Mesopotamia becamea civilization characterized by competing urban-centeredstatelets throughout most of its history and why the ear-liest political development of Mesopotamian civilizationdiffers from that of other pristine civilizations (e.g.,Egypt, Monte Alban, Moche), in which warfare and con-flict appear to have been more important motors of socialevolution than trade and unitary territorial states werethe initial form of political organization (Marcus 1998).The same factors may also partially explain why the ear-liest Mesopotamian writing systems were almost en-tirely economic in their initial impetus, as opposed tothe early writing of other pristine civilizations, whichoften was more political or religious in function.

We can only speculate about the historical conse-quences of the early political balkanization of the Mes-opotamian alluvium, but many scholars see political di-vision as an important factor selecting for acceleratedsocial change. Patricia Crone (1989:161), for instance,argues that political fragmentation and interpolity com-petition were crucial for what she perceives as the uniquevitality of developmental rates in European polities ofthe late medieval and early modern eras as comparedwith those characteristic of other areas of the world atthat time: “Far from being stultified by imperial govern-ment, Europe was to be propelled forward by constantcompetition between its component parts.” So, too, pre-sumably, ancient Mesopotamia. However, the socio-ev-olutionary “critical mass” created by the compact andcompetitive social environment of the Tigris-Euphratesalluvial lowlands in the Uruk period was absent fromneighboring areas at the time. Complex as some indi-vidual Late Chalcolithic polities may have been duringthe 4th millennium, they were essentially isolated.Claims that such polities were comparable in scale to

the nascent city-states of Uruk Mesopotamia (Oates andOates 1997, Lamberg-Karlovsky 1999) or that they pos-sessed an equivalent level of organizational complexity(Frangipane 1997) incorrectly minimize the existenceand importance of scalar and organizational asymmetriesbetween southern and peripheral polities.

Arguments that minimize scalar differences gloss overthe fact that, while some individual Late Chalcolithicsites in the north did achieve substantial size, as a groupthe Late Chalcolithic polities were no match for con-temporary Uruk societies in the Mesopotamian allu-vium either in terms of intrasite scale (extent and ar-chitectural monumentality) and complexity (social andeconomic differentiation) or in terms of the density,depth, and extent of their associated regional settlementnetworks. The number of individual polities in closeproximity within southern Mesopotamia through theUruk period was simply unparalleled. Further, by thesecond half of the Uruk period the large, walled urbanpolity of Uruk and its corona of dependent towns, vil-lages, and hamlets were of a completely different orderof magnitude from anything the periphery could offer.Also without counterparts in contemporary peripheralsocieties were the scale and extent of the public struc-tures at the very center of Uruk at this time—the 8–9ha of walled-off ceremonial and administrative buildingsin the Eanna precinct and, in a separate part of the city,the massive Anu Ziggurat and associated temple, whichtowered over the site and the surrounding plain (Nissen1988).

Arguments that minimize organizational differencespay insufficient attention to the constraints and oppor-tunities of geography, the ways in which humans canexploit their environment to gain and expand their socialstanding, and the power of ideas. Such arguments com-monly start from the premise that there were no essen-tial differences in the material technologies available tothe two groups through the 4th millennium (Kohl 1987a,Stein 1999). This premise is correct only in the narrowsense that the transportational and craft technologiesavailable to both groups throughout the second half ofthe fourth millennium were comparable and misses themark in four important respects.

First, the premise should be rejected on principle, forit implies that economic changes caused by advances inmaterial technology are the sole force underpinning so-cial change. This reductionist viewpoint fails to take intoaccount the fact that technological innovation not onlyacts on society but is itself contingent on socioeconomicforces.

Second, it fails to take note of the geographical advan-tages in transportation of the southern alluvium overneighboring areas. If, as Jacobs argues, a key cause ofdevelopment is imports added to capital and labor whichproduce value-added exports, then southern Mesopota-mian societies had a substantial and irrevocable advan-tage over peripheral polities in that both labor and capital(agricultural, pastoral, and piscatorial surpluses andstatus-marking exotics) were more easily mobilized andredistributed in the south than elsewhere because of the


inherent superiority of southern water-borne transportsystems.

Third, it fails to acknowledge the environmental ad-vantages in productivity of the southern alluvium overneighboring areas, advantages that were particularly pro-nounced through the 4th millennium. These advantagesprovided the opportunities that elite groups exploitingthe alluvium’s rich and diverse ecological niches neededto amass the initial surpluses that set in motion theexport-driven economies that characterized southernMesopotamian polities throughout much of their his-tory. These groups neither understood nor foresaw theconsequences of the exchange patterns they initiated.Rather, they were merely attempting to do what elitesnaturally do in all human societies, namely, sanctionexisting social inequalities, legitimize their right to con-tinue demanding and receiving unequal access to re-sources, maximize the amounts and varieties of com-modities and labor at their disposal, and increase theirpolitical power. The richness of the ecological contextof the alluvial lowlands of the Tigris-Euphrates systemof the 4th millennium made such attempts both morelikely and more likely to succeed than would have beenthe case in less resilient environments.

Fourth, and most important, it overlooks the fact thatdifferences in ideologies of social organization and inassociated technologies of social control are as capableof creating significant developmental asymmetries be-tween societies as are imbalances in material technology(Goody 2000). By the second half of the 4th millenniumsouthern Mesopotamian societies enjoyed a variety ofrevolutionary organizational advantages not possessed atthat point by any of their contemporaries. These advan-tages partly fall in the realm of what Jack Goody (2000)refers to as “technologies of the intellect,” what MichaelMann (1986) terms “technologies of power,” and whatJohn Baines and Norman Yoffee (1998) refer to as “highculture.” The most salient of these advantages were theconceptual shift earlier described as the “Labor Revo-lution,” novel and much more efficient procedures ofinformation recording and processing, and economies ofscale arising from the specialized production of com-modities by dependent laborers organized according toprinciples of comparative advantage. When added to theaforementioned advantages in the number of potentiallaborers, compactness, and ease of transportation, theseorganizational innovations gave Uruk societies an im-portant edge over neighboring polities.

In other words, by the second half of the 4th millen-nium, cumulative gains in economic and social differ-entiation arising from centuries of asymmetricalexchange patterns and import-substitution processes hadcreated societies in the Mesopotamian alluvium thatwere not only substantially larger in scale and regionalreach than their contemporaries but intellectually betterprepared to deal with the complexities of supportinglarge-scale agglomerations of population as well. Theywere also more capable than their neighbors of resolvingthe inevitable managerial conflicts that such agglomer-ations entail and managing multiple contacts with far-

flung societies in control of coveted resources. The var-ious ideational technologies of social organization thatmade this possible were certainly selected for and thusmade more likely by the multiplier effects of the export-driven economies that had taken root in the rich anddiverse environmental framework that characterized thesouthern Mesopotamian alluvium during the formativestages of Sumerian civilization. As Jacobs (2000:63)notes, diversity expands naturally in inherently rich en-vironments, which are made richer by the diverse useand reuse of received energy.

As rules of historical development would have it, theinitial southern Mesopotamian edge in ideational tech-nologies of socioeconomic organization and control didnot and could not last long. With disparity in organiza-tional complexity came disparities in power, both realand perceived, and with them came opportunities for theexpansion of Middle and Late Uruk–period polities bymeans of colonial outposts at strategic locations acrossthe Mesopotamian periphery. In great part as a result ofprocesses of competitive emulation initiated across theMesopotamian periphery by the Uruk expansion, Subar-ians (Gelb 1944, Michalowski 1986), Hurrians (Gelb1944, Steinkeller 1998), Elamites (Carter and Stolper1984), and other 3d-millennium groups (Sollberger andKupper 1971, Grayson, Frame, and Frayne 1987, Hirsch1963, Lebeau 1990) were soon to master for themselvesthe fine art of creating truly effective social systems ofthe few, by the few, and for the few. This, too, was anopportunistic adaptive response on their part but onewhich was promoted largely by a social landscape andhistorical context that had been dramatically trans-formed by the earlier emergence of Sumerian societies.

Comments

b . brentjesEberhardstr. 3, D 10367 Berlin, Germany, 26 x 00

The main thesis of this paper—the importance of tradein the early phase of civilization in Mesopotamia andespecially the interrelation of society and ecosys-tem—can be accepted, but there are some details whichcan be questioned:

1. Algaze insists on the absolute primacy of southernMesopotamia as the nucleus of world civilization, butthe beginnings of the development in the Indus Valleyand farther east may not have been later than the late4th millennium b.c. and did not depend on Sumer.

2. Algaze stresses the importance of the availability ofnatural resources in southern Mesopotamia, and this iscorrect, but no less important were the shortages of rawmaterials—wood, stone, and metals—that forced the pur-suit of imports.

3. The role of climatic changes cannot be denied, butthe monsoon from the Indian Ocean, which was turnedmore to north than today, brought rain to the Indus


Valley, Baluchistan, and Afghanistan and not to Meso-potamia, far to the northwest. The Indian Ocean mon-soon that reached southern Arabia first moved west, wasdiverted to the north, over Ethiopia, crossed Yemen, andreached Oman and other countries from the south. Itcannot have been important for irrigation in Mesopo-tamia, which depended on precipitation over Asia Minorand northwestern Iran and received its winter rain fromthe northwest. If there really was summer rain in south-ern Iraq, it might have had sources in the northwest, too.

4. I doubt the inefficiency of caravans, given the trafficfrom Assyria to Asia Minor in the early 2d millenniumb.c.—and why they should have used their boats onlydownstream? It would have been inefficient to returnthe boats empty or on donkeys, and dragging or rowingis quite possible. The Mesopotamian (and Elamite) tradeto the east and the overland trade of the Indus civiliza-tion crossed high mountains and steppes for thousandsof kilometers up to the Ferghana Valley and to Chorezm.

5. It would have been useless to carry bitumen fromsouthern Mesopotamia to the north—the area of Mosulhad enough for export itself.

6. Algaze does not mention Uruk’s dual structure, withtwo central temples instead of one as usual in Sumer.Was it a combination of two former nuclei as somebelieve?

petr charvatOriental Institute, Academy of Sciences of the CzechRepublic, Prague, Czech Republic. 14 ix 00

This is certainly a very interesting assessment of theopportunities offered by the environment to the leadersof both northern and southern Mesopotamian commu-nities for the amplification of the potential to usher inmajor social transformations. The advantageous startingposition of the south is well argued, and I hardly believethat any modern scholar would contest the logic of Al-gaze’s conclusions. The crux, as I see it, is why.

What would have induced the southerners to apply somuch energy to embarking on a journey that is wellknown to us but absolutely original, new, and thereforepotentially dangerous for them? I am more and moreconvinced that the truly revolutionary changes andtransformations take place in human heads and, at that,in the beginning in a single one. A promising environ-ment will hardly fulfil its potential if the humans livingin it simply do not perceive this promise or prefer theirtraditional way of life to something that might be ad-vantageous but conflicts with traditional norms and cus-toms. Asked why they would not engage in agriculture,the San of the Kalahari simply replied, “Why should wework in the fields if there are so many mongongo nutsin the world?” This response is fairly typical for prestateand preindustrial societies. Humans had lived in south-ern Mesopotamia for millennia prior to the Late Urukperiod, and no traces of the multiplier effect proper tothe latter culture can be observed in the evidence it left(at Tell Awayli; see Forest 1996). My own studies (Char-

vat 1993, 1997) have pointed to the conclusion that theeconomic base of the Late Uruk elite was so diffuse andlacking in any visible signs of having been organizedintentionally from a single central point that it gives theimpression of the elite’s having adjusted to the existingcomplexities of the community’s subsistence activitiesrather than developing its own distinctive entrepreneu-rial culture.

There is, to be sure, one innovation that might havecaused the Late Uruk elites to turn towards exploitationof the full potential of their host landscapes, and this isa purely spiritual one. The belief that the fertility ofnature will be fully realized only if it is “triggered” bythe NA2 ceremony (probably a predecessor, or possiblyan ancestor, of the “sacred marriage rite”) performed bythe pontifical couple, the EN and the NIN, seems torepresent an invention of the (Late?) Uruk period, as Ishall be arguing in my forthcoming study of the icono-graphical sources for the emergence of statehood in theancient Near East (see also Charvat 1997). I believe thatthe Late Uruk phenomenon was a revolution of humansouls and minds rather than of hearts and stomachs.

claudio cioffi-revillaDepartment of Political Science, University ofColorado, Boulder, Colo. 80309-0333, U.S.A.([email protected]). 27 xi 00

The origins and early evolution of government—in Mes-opotamia, China, Peru, and Mesoamerica—are amongthe most fundamental puzzles in political science, al-though in recent decades such investigations havetended to gravitate towards the disciplines of archae-ology, anthropology, or anthropological archaeology. It istherefore intriguing and stimulating to see these puzzlesaddressed by modern scientific scholars such as Guil-lermo Algaze. Significant advances by archaeologists in-vestigating the origins of government in ancient socie-ties, especially when produced through comparative andinterdisciplinary methodologies, are crucial not only forarchaeological theory and research but also for politicalscience and allied social sciences.

As a political scientist who shares many of the sci-entific interests as Algaze and his colleagues in archae-ology, I must applaud his systematic effort to gather data,formulate hypotheses, and rigorously test theories of theorigin and early evolution of political complexity. Sincewe have the same basic epistemology, albeit not the samedisciplinary background, I shall focus my comments ona set of political science issues aimed at providing someplausible suggestions for further theoretical and empir-ical development.

I find the model developed by Algaze in this study,including his emphasis on the causal role of physicalecological variables, essentially correct but somewhatincomplete in a number of aspects. That is to say, “thestory” is still missing a set of critical events that shouldbe added for a deeper understanding of government for-mation. As some readers and Algaze himself might an-


ticipate, the missing elements are political, not environ-mental, although this is not to say that politicalscientists have done a good job of highlighting and ex-plicating these missing political pieces of the puzzle withthe necessary scientific precision.

In essence, environmental variables and events suchas those identified by Algaze (and Adams) are best seenas remote (or “ultimate”) causes of the origins of politicalcomplexity, whereas political variables and events, in-cluding mechanisms of decision making, are best seenas “proximate” causes of political complexity. The for-mation and maintenance of government in any givensociety constitute a process or sequence of contingentevents. Foremost among the political events embeddedin the process of government formation, whether in an-cient Mesopotamia or elsewhere, is what political sci-entists, economists, and sociologists call “collective ac-tion,” which is not a physical event although it may berelated to or ultimately caused by prior or contempora-neous physical events. In this context, collective actionis caused by the emergence of leaders (elites) and follow-ers (commoners) that jointly produce a system of gov-ernment. In turn, both leaders and followers emerge bya variety of processes or mechanisms, including decision-making calculations. For example, the actors involvedmay undergo a decision-making calculus of costs andbenefits (the so-called rational-choice model) or may de-cide to act through other mechanisms (such as deonticobligation or instinctive responses to social or physicalenvironmental stimuli). Although this is not the appro-priate venue in which to detail a collective-action pro-cess of government formation with the desirable rigor,suffice it to say that a system of government cannot formwithout leaders’ and followers’ solving a collective-ac-tion problem. Moreover, such actions sometimes suc-ceed and sometimes fail, as archaeologists have docu-mented for many cases.

Besides key ideas from the theory of collective action,including Lichbach’s (1995) important new work citedby Algaze, other key pieces of the puzzle of governmentformation include the classical choice-lottery distinctionfrom decision theory and game theory, the opportunity-willingness principle (Cioffi-Revilla and Starr 1995, Starr1978), and principles of political uncertainty (Cioffi-Re-villa 1998). For example, the first of these is used forparsing the complex process of government formationinto “states of Nature” (produced by lotteries) and “hu-man acts” (produced by decisions or games). Each ofthese, in turn, is modeled by specific variables andevents. The opportunity-willingness principle is used formodeling human acts subject to environmental con-straints, such as climate change, resource availability,and other significant conditions discussed by Algaze.Principles of political uncertainty specify the dynamicsproduced by probabilistic causes, often in counterintui-tive ways, such as the effects of multiple decision mech-anisms on leaders’ and followers’ choices in the emer-gence of political complexity.

The emergence of political complexity in Mesopota-mia between the 6th and 3rd millennia b.c.—the rise of

chiefdoms, states, and eventually empires—was a pro-cess of fundamental importance not just for archaeologyand political science but for virtually all areas of humanknowledge. Studies such as Algaze’s give us solid hopethat we may be close to some important new break-throughs in scientific explanation and understandingthrough greater interdisciplinary collaboration.

rene dittmannWestfalische Wilhelms-Universitat, Schlossplatz 2, D-48149 Munster, Germany ([email protected]).21 xi 00

Having worked on aspects of the Uruk period, I havefound it fascinating to read this stimulating and excel-lent paper. I have always viewed the Uruk horizon aspart of a longue duree development from the Ubaid tothe Early Dynastic and finally the Akkad and Ur III com-plexes. Therefore, as I see it the Uruk, especially the LateUruk complex, is to some extent the culmination of theUbaid development and in many respects the beginningof the processes that led to the Early Dynastic and lateraspects. Consequently I never use the term “colonies”for Uruk settlements outside Babylonia, because formost of Greater Mesopotamia the Uruk is simply theonly complex that existed at this period, with nothinglocal in the immediate neighbourhood (sites such as Ha-buba, Gebel Aruda, and earlier Tell Sheikh Hassan, TellBleibis, and others but of course not Godin, Hacinebi, orArslantepe). Our understanding of the network of Middleand Late Uruk sites, especially in Syria and southernTurkey, is focused on settlements along the Euphrates;almost nothing is known about its hinterland. We havesites—and very strange ones—that are very small, likefarms (for example, Shavi Hoyuk II, Nevala Cori, andsites in the Incesu Valley). Further, Hassek Hoyuk isreally not, in its structure, a settlement in the sense ofa village. In order to understand the Uruk network, then,we should look at each site and try to understand itsstructure and possible function within the system, andwe should investigate the hinterlands of the major riversystems. If we do this, I am convinced, we will find alot more variation in the Uruk system. I suspect thatAlgaze’s claim of an imbalance between southern andnorthern sites will be complemented to some extent inthe future by a deeper understanding of the complexitywithin the system. This definitely requires much furtherstudy.

Algaze’s suggestion that the Late Uruk period is al-ready characterized by a system of separate and possiblycompetitive city-states seems not impossible to me. Itis obvious from the glyptic material available at the mo-ment that we have many “conflict scenes” at Uruk butalmost no “economic scenes,” the latter being found al-most exclusively elsewhere (signs of a kind of dendriticeconomic network?) This could be a hint of what Algazesuggests in the sense that Uruk, as the religious andeconomic centre, may have tried to dominate smalleremerging centres at the end of the Uruk period by force


in order to prevent centrifugal processes within the sys-tem. The almost canonical representations and motifs ofthe Uruk glyptic on the whole seem to me to reflect aformer (Middle and beginning of Late Uruk) high levelof integration in the Uruk system not only economicallybut surely also ideologically. In this respect it would beunderstandable that the quality of the representationson Uruk artefacts is comparable in naturalism and re-alism only to that on Akkadian ones and that the motifshave an almost unlimited geographical distribution. TheUruk system seems to have had a panregional under-standing of the “four corners of the world,” and this, Isuggest, was at once the legacy of the long Ubaidiandevelopment and the romantic goal of the later, region-ally limited political systems. With the collapse of theUruk system, the iconological uniformity broke downas a result of the increasing development of local sym-bolic systems by the emerging—or consolidat-ing—neighbouring city-states. The form changed, butthe contents of the iconography of the Uruk period be-came an essential part of the Greater Mesopotamiansymbolic complex.

jonathan friedman and kajsa ekholmfriedmanDepartment of Social Anthropology, University ofLund, Box 114, 221 00 Lund, Sweden([email protected]). 22 xi 00

This substantial systematization of argument and dataon the initial emergence of southern Mesopotamia issure to take its place as a timely contribution to theissues. The author argues that the reason “civilization”emerged in the southern alluvium is that this zone pro-vided a set of comparative advantages that led to a syn-ergetic development that could not be matched in otherparts of the Near East. It reintroduces the once very pop-ular issue of ecology and environmental factors not inthe usual cultural materialist or determinist way but inone close to the general arguments that developed withincertain versions of structuralist Marxist and later globalsystemic anthropologies. Concentrating on recent andnot so recent data on resource diversity and stability,relative productivity, and ease and therefore productivityof water transportation, Algaze argues that the condi-tions for increasing density, differentiated economies,trade, and competition were quite special to this area.This statement of the approach is basically selective,expressed as distributions of constraints and, as hewrites, opportunities. The simplest form of the argumentis that the same strategy succeeds differentially in dif-ferent initial situations and that southern Mesopotamiais the most advantageous place for things to kick off.This assumes that the strategies, let us say, of accu-mulation, expansion, etc., are basically the same, so thatthe environment becomes crucial in locating the initialtakeoff. Is this what is meant here? In other parts of thearticle it is suggested that there are significant social andcultural differences as well, but it is not clear what their

role may have been. Some of these are translated intotechnologies, technologies of power, related to the formsof literacy, administrative organization, etc., but arethese regionally specific differences or effects of the de-velopment process itself? The argument is not clearlyspelled out here.

Trade is taken up as a factor in this argument, equiv-alent in a sense to geography and productivity, but thisis, we suggest, a misconstrual. Accumulation of wealthbased on trade is not a mere factor that can be listedalong with environmental and other conditions, sincethis underplays the systemic relation between trade,technology, slavery, expansion, and almost all of theother “factors” mentioned. Jacobs’s model of export-ledgrowth is not, of course, new or even specific to Jacobs,although she had the daring insight to apply it to ancienturban economies as well as to modern ones. As we un-derstand the use of this model, it entails that an elitehas a specific strategy of accumulation of wealth, status,or something that requires economic growth, a “some-thing” specifically social and cultural that is not speci-fied here. The model is based on modern Western capi-talism, and it is not clear that it is applicable to the earlycity-state without a further argument, which we havealso made, for some basic similarities. In any case, thisdynamic is not a mere factor that enabled the emergenceof Uruk but the very form of that development, one thatorganized geographical, technological, and other factorsin an expanding system.

For anthropologists a crucial question here would bethe actual organization of this dynamic—its sociocul-tural form. Quite a few years ago we suggested that thiskind of export/import-based economy might have grownout of an earlier regional system based on a strategy ofmonopolizing the circulation of prestige goods. While wedon’t necessarily want to defend such a model today, wewould stress the ultimate need at least to attempt a morespecific characterization of the social processes involved.The resource-differentiated southern plain may indeedhave afforded an arena for intensification of trade, butthere is evidence that more far-reaching trade was alsopresent from the earliest periods. An aspect of the dis-cussion that we find most interesting is the suggestionof a tendency toward import substitution in the takeoffperiod of urbanization. While this is a normal phenom-enon in modern capitalist economies in their protec-tionist phases, it may also, in this case, mark a majortransformation of items prestigious because of their for-eignness into more mundane practical ones, that is, theemergence of an increasingly commercial economy. Inorder to make any such suggestion, however, we needto know how the goods are distributed or not distributed.

The entire region including northern Mesopotamia ischaracterized by the early existence of regional trade, bytendencies toward the use of slavery, and by similaritiesin urban formation. Are we dealing here with a commonstructure that becomes differentially organized alongsomething like center-periphery lines, in which the cen-ters and even the centrality of regions shift over time?The fact that city-states develop throughout the entire


region raises a number of questions, such as the natureof initial conditions in these later contexts and the im-port of the explanatory model if we do not assume thatthis social organization simply diffused.

If this is the case, then it might be profitable to workout a dynamic model for the region, one that tends to-ward the formation of commercial city-state systems ofintensive wealth accumulation. The contrast with areassuch as Egypt and even South America suggested heremay have to do with the historical specificities of a vastregion between the Mediterranean and South and Cen-tral Asia in which trade may have been decentralizedfrom a very early period, with the result that pristineterritorial states would have been unable to establish amonopoly over the flows of wealth in the region—a sit-uation that could be overcome only by empire formation.Decentralized conditions of accumulation have been ar-gued to have been crucial in the development of WesternEurope.

Algaze makes crucial progress toward a more adequateunderstanding of Mesopotamian development, and ourquestions, typically anthropological, press for more spec-ificity if possible. Yes, by and for the elites, but what arethey up to?

a. bernard knappDepartment of Archaeology, University of Glasgow,Glasgow G12 8QQ, Scotland, U.K. ([email protected]). 17 xi 00

In this stimulating, nicely written essay, Algaze returnsfrom fieldwork in the periphery to synthesis in the core(“heartland”) of Mesopotamia, a subject area he knowswell and one in which, increasingly, he plays a leadinginterpretive role. At the outset he quotes R. McC. Adamsto the effect that contemporary archaeologists—in thethroes of a social archaeology—no longer see fit to ex-amine the interrelationship between natural ecosystemsand human beings. It seems odd to read this, coming outof the American anthropological tradition in archae-ology, where the environment has loomed large for anentire (processual) generation and continues to do so (atleast from my European vantage point). Perhaps, though,this is a perspective more specific to Near Eastern ar-chaeology as taught and studied in America; given thediversity of material culture and texts associated withsuch levels of social complexity as are found in Meso-potamia, the singularly ecosystemic approach of anthro-pological archaeology never worked very well anyway.Or, perhaps, Adams (and Algaze) might have benefitedfrom looking more closely at work being conducted else-where in Europe (the Old World), in particular at long-standing traditions in regional survey and rural archae-ology in the Mediterranean, where the dynamic betweenpeople and their environment, between tradition and in-novation, between urban and rural, has long held sway(e.g., Ward-Perkins 1964, Cherry, Davis, and Mantzour-ani 1991, Barker and Mattingly 1999–2000).

Employing an eco-environmental approach in the at-

tempt to understand past peoples and societies has longbeen integral to both archaeology and anthropology. Cri-tiques of environmental determinism stem in part fromrealizing that social patterns and human activities wereaffected by diverse and interrelated factors, not just ad-aptation or reaction to the environment. The inclusionof social theory in archaeology helped to bring about amore nuanced understanding of the ways in which theenvironment impacts and is impacted by human action.Social approaches also adopt an interpretive frameworkbased on the irreducibility of the human-environmentrelationship, itself framed and characterized by human-ecodynamics (McGlade 1995) rather than by some self-regulating, equilibrium-seeking system.

Eco-environmental approaches are widely regarded asbeing reductionist or deterministic, but often the cri-tique of reductionist models or assumptions is presentedas an either/or proposition: either you use reductionistmodels and underdescribe the reality of the past or youaccept the inability to know the past because of its ram-ified and situated nature. In fact reductionist models area first step towards a nuanced description of the past:they require a data set to which they can be applied, andregional survey often forms a critical component of cre-ating such a data set.

Algaze draws in part upon the rich regional survey dataaccumulated for the Mesopotamian lowlands by Adams,reassessed more recently by Susan Pollock. He alsodraws upon a wealth of other field data—site reports, claybullae, Archaic texts from Uruk and cuneiform records,iconographic evidence, satellite imagery, and more—tosubstantiate his main thesis, namely, that the develop-ment of a primary, complex state-level society in south-ern Mesopotamia during the mid-4th millennium b.c.should be understood as a technological innovation, onespurred by social and economic differentiation, unequalexchange, and import substitution. Within the “periph-ery” of northern Mesopotamia/Syria, southwestern Iran,and elsewhere, such developments came only later, dur-ing the 3d millennium b.c. The spatial and environ-mental situation of the alluvial south gave it pride ofplace in the region and, together with informational andmanagerial developments (“ideational technologies”),helped to ensure that Sumerian elites maximized all theresources at their disposal in establishing and perpetu-ating their own rule.

As we found in attempting to set the stage for a seriesof papers on archaeologies of landscape (Knapp and Ash-more 1999: 12–13), “ideational” is a term at once com-prehensive and vague; it has been equated with every-thing from the sacred landscape to landscapes of themind. Algaze’s use of the term seems to me much morecoherent and meaningful when applied to the technol-ogies and organizational advantages discussed in his pa-per and is certainly preferable to the alternativeshe mentions—technologies of the intellect (Goody),technologies of power (Mann), or “high culture” (Yoffee).

Algaze’s essay is a substantial and engaging contri-bution to contemporary archaeology. It succeeds in com-bining spatial, environmental, empirical, and social pa-


rameters in a thoroughly readable (and to me convincing)essay on the emergence of socially complex early Mes-opotamian culture, one that all archaeologists wouldbenefit from reading.

c. c . lamberg-karlovskyPeabody Museum, Harvard University, Cambridge,Mass. 02138, U.S.A. ([email protected]).7 xi 00

While disavowing geography as destiny, Algaze privi-leges three environmental/geographic “advantages” insouthern Mesopotamia: (1) “a denser and more variedconcentration of exploitable subsistence resources,” (2)“the higher and more reliable agricultural yields,” and(3) the efficiency of transport allowed by the river system.His discussion of these three items and the complexityof their interaction with the social order is the mostcompelling part of his thought-provoking essay. There isboth teleology and tautology in this argument: environ-mental benefits bring about social complexity while, inturn, social complexity is able to take increasing advan-tage of its environment. Environments present con-straints and restraints. The richness of an environmentdoes not ensure the transition to cultural complexity.Environments are filtered, transformed, and given theirmaterial reality by the beliefs and practices of a society.

The emphasis in this essay is on complexity and itsmechanisms, not on society. Algaze refers to the Eannaprecinct but does not credit it with responsibility forstructuring the social organization of the city of Uruk.The temples with their resident bureaucrats were thecenter of an administrative bureaucracy utilizing and insome cases inventing and/or perfecting what I havecalled a technology of social control—writing, seals, seal-ings, and standard units of measure—to manage laborand monitor the production, consumption, and redistri-bution of goods. Algaze speaks of “ideational” technol-ogies but nowhere expands upon the structures of beliefthat organized and gave meaning to both the social andthe natural environment.

The organization of people is effected by concepts, andthese differ in different societies. The organization oftribal societies and chiefdoms structures a political com-munity based on kinship, and this organization is en-tirely comprehensible. The structure of an emergingstate is very different; I do not know of a single one thatwas not formed by recourse to the supernatural. In south-ern Mesopotamia the gods owned the land, lived in thetemples, and presided as patrons over the city. The EN(chief bureaucrat), as a steward of the gods, was respon-sible for feeding both the gods and the people of thecommunity. The sacred Eanna precinct was home to bu-reaucrats whose authority was legitimized by the gods.The structure of the relationship that legitimized thesebureaucrats was as important a “feedback loop” and“synergy” as the relationship of exports to human laborand capital or of population agglomerations to settle-ment hierarchies discussed by Algaze. In this regard the

views of Sahlins (1974:68) on why state complexity didnot emerge in Polynesia are directly relevant:

Had the Polynesian gods been more demanding,temple building might well have generated new so-cial and political roles here as it did in Sumeria; butfor one reason or another they were too complacent.One suspects that they were too complacent pre-cisely because a perfectly satisfactory tribal systemexisted; differently put, one suspects that state struc-tures emerged among the Sumerians all the moreeasily in that no tribal organization had beendeveloped.

Gods should not appease, and bureaucrats must be de-manding if they are to establish exploitative central-place coercion. What Algaze refers to as “the Labor Rev-olution” has a distinct social setting—the subordinationof the population by divinely sanctioned rules writtenby the bureaucratic elite.

Algaze appears to find positive value in hierarchicalsettlement patterns, writing, and complexity. The per-ception of each in the past, however, may have beendifferent from what it is today. (1) In southern Meso-potamia settlement hierarchies went unnoticed. Thereare no words distinguishing city, town, and village; urusimply designated a place, without distinction as to itssize. In an effort to emphasize the correlation of settle-ment size and complexity Algaze notes that Uruk was,in its day, the largest community in the world. Not so!The Tripol’ye Culture in Ukraine, dated to the first halfof the 4th millennium, had several communities of equalor larger size, among them Dobrovody (250 ha), Majda-neckoe (270 ha), and Tal’janki (400 ha). While these com-munities were as large as the later Mesopotamian city-states, they did not endure as long or reach the samelevel of social complexity. By the middle of the 4th mil-lennium Uruk, at 250 ha, was a city of substantial com-plexity, while Tal’janki, at nearly twice its size, main-tained a level of complexity characteristic of village life.Settlement size, like hierarchy, is significant only in thecontext in which it is situated. It is not a universal at-tribute which alone signifies complexity. (2) Writingoriginated as a tool for controlling labor, production, con-sumption, and redistribution. The unanticipated conse-quence of its invention was the poetry, hymns, reportsof divination, and king-lists that were to appear some500 years later. (3) Valuing complexity merely articulatesa belief in cultural evolution, the placing of one stageupon another, often without accounting for the socialmechanisms responsible for the transitions betweenthem. Social complexity requires permanent institutionswhose authority emanates from specific ideologies. Insouthern Mesopotamia these institutions were the tem-ples, which preceded the Uruk period. To an earlier com-plexity the Uruk period added the invention of writing(seals, sealings, and standard units of value were earlierinventions), a process that Adams (1981) calls “hyperur-banization,” and a powerful ideology that sanctioned the


successful centralization of authority under the tutelageof the gods. The sacred precinct was set apart from thecommunity, and perhaps, as Strayer (1970:18) states forthe medieval European state, “a realization of this sep-arateness was an essential ingredient of state building.”

Finally, Algaze would have us believe that the peoplesof southern Mesopotamia “were intellectually betterprepared to deal with the complexities” than the inhab-itants of the periphery. I have long wondered why peopleof the periphery did not assimilate the complexity evi-dent in the south. Once exposed to writing, as they as-suredly were throughout the century-long Uruk and laterProto-Elamite expansions, why did virtually every dis-tinctive community on the periphery reject the use ofwriting? With so many different archaeological culturesexposed to the complexity of southern Mesopotamia, onecould argue that the rejection of complexity was a con-scious act. What was the reason for it?

Perhaps the gods of the north were, like the Polynesiangods referred to by Sahlins, simply less demanding. Per-haps, far from being intellectually less qualified to dealwith complexity, the peripheral peoples were smartenough to avoid its oppressive command structure for atleast another 500 years, when it was imposed upon themby military conquest initiated, naturally enough, fromthe south. In every instance the periphery initially re-jected the adoption of complexity even after direct ex-posure to it. In time, the development of the state madepossible such concentrated control of human resourcesthat no other type of social organization could avoid be-ing relegated to a subordinate role. Fundamental socialchange is rare, for it goes against the grain of social actorsand their organizations. There is no inexorable logic thatleads to social change or to social complexity. As Gellner(1988) has argued, most significant social change isbrought about through fortuitous openings. Individualsand/or institutions that attempt to impose fundamentalsocial changes are best characterized as adaptive failures.The people on the periphery chose to avoid complexityand, in so doing, avoided the cage of the state for anotherhalf-millennium. Perhaps southern Mesopotamia’s eco-logical circumscription had something to do with boththe loss of tribal ties and the harshness of its conceptionof the divine.

Complexity is a double-edged sword. Globalization isour modern complexity. Today in Pakistan, where I amexcavating a site, the newspapers debate whether thecountry should accept International Monetary Fundloans. Such loans, according to one argument, allow richWestern nations to keep Third World nations perma-nently in debt. In not taking those loans they avoid thenew complexity: globalization. Complexity is more thanan evolutionary stage of the “initial”; it is a moving tar-get that requires a more detailed understanding of the“social.”

joy mc corristonDepartment of Anthropology, Ohio State University,Columbus, Ohio 43210-1364, U.S.A. ([email protected]). 25 xi 00

Algaze addresses with new theoretical insight the long-standing question of why the Mesopotamian stateemerged early in the southern lowlands. He has onceagain found much to strengthen current archaeologicalapproaches with a new theoretical application from an-other discipline. His paper offers an original, nuanced,effective, geographically-environmentally situated ex-planation of Mesopotamian social development. In theirultimate assessment of Algaze’s fine contribution, schol-ars will ponder several basic points: How good is themodel, and how good are the data? I mention only a fewaspects of each.

Algaze’s article makes me want to read Jane Jacobs’seconomic work, for although I find his arguments com-pelling, it remains difficult for me to assess the under-lying plausibility of human economies’ developing in afashion comparable to ecosystem development, with itsexplicitly biological evolutionary mechanism of adap-tation through natural selection. Advocates of an ex-plicitly evolutionary archaeology (e.g., Dunnell 1980,O’Brian 1996, Rindos 1986) might reasonably suggestthat selection, an underlying mechanism in biologicalevolution, should be manifest also in human economiesif they are to be described by models that link culturaland biological evolution. In Mesopotamia, archaeologyhas depended heavily on settlement data to infer patternsof social and political complexity, but these patterns donot in themselves indicate mechanisms of change. Se-lection—or, indeed, any other mechanism of cultural de-velopment—will be manifest at the household level ofanalysis, where individual decision-making and its con-sequences for reproduction can be best deciphered fromdata that are of high resolution both chronologically andspatially. If Algaze’s thesis is testable, then householdarchaeology, albeit logistically difficult in southern Mes-opotamia, provides the scale and locus at which it canbest be tested. A household approach also serves to bal-ance emphasis on archaeologically and historicallyhighly visible elites in explanations of Mesopotamianstate building.

Algaze does an admirable job of building a geograph-ically patterned model that explains a nevertheless his-torically contingent developmental sequence. Tensionbetween developmental predictability and system his-tories must be faced in all ecosystem models and, indeed,in evolutionary theory itself (Gould 1986, 1994). As Al-gaze notes, in the Mesopotamian case geographic pre-dictability (propinquity and diversity) must be balancedwith significant historical contingency, specifically, dra-matic climate and environmental change during the 4thmillennium b.c. Nevertheless, his model, perhaps influ-enced by Jacobs’s interest in Spencer, leans toward a pro-gressivist and linear approach to economic developmentwith its emphasis on resiliency, expansion, and feedbackloops to explain the inevitability of trade imbalances and


Uruk social hierarchies. I would like to read the argu-ments for loss of complexity too. Why would not randomdramatic change, such as supposed loss of summer rain-fall, from the environmental advantages enjoyed in the4th millennium b.c. at least sometimes and in someareas have resulted in social homogenization and aheightened emphasis on subsistence production? Addi-tionally, where adjacent areas with comparable resourcesshow very different developmental trajectories, mightnot greater emphasis on historical contingencies explainintraregional variation?

There will always be hand-wringing over the data notavailable from southern Mesopotamia and disputesabout those that are. For example, high percentage es-timates of population concentrated in large settlementscannot but be affected by preservation factors that buriedsmaller sites during periods of aggradation and deflatedthem thereafter. I do not concur with Algaze’s suggestionthat millets and sesame were cultivated (implicitly withdecrue techniques all adopted from Africa in a summer-cropping regime): evidence to support this simply is notavailable from Mesopotamia, Arabia, or South Asia atthis date. All of us wish that more data were availablefrom early households. Textiles as exports and metals asimports have long histories of scholarly debate and shortarchaeological finds catalogues. In the case of Uruk tex-tiles, early weaving skills along might not account forthe value-added workmanship attested almost 2,000years later in Assyrian documents. I suggest that Urukweavers, who found flax in short supply, depended ratherheavily on wool as a novel basic textile fiber with animportant new characteristic: wool, in contrast to linen,readily accepts a dye. Many dyes, such as those fromsafflower, marine gastropods, and indigo, come fromwidespread sources linked only by Mesopotamian trade(McCorriston 1999, Edens 1994, Zarins 1989). Althoughsuch dyes became archaeologically visible only in the 3dmillennium, they offer the very raw resources that Al-gaze cites as critical energy captured for value-added ex-ports. Dyed textiles surely imply that wool emerged asthe major surplus textile fiber, a process with significantgeographical and ecological parameters (McCorriston1997). But these are mere details: Algaze has given usmuch to ponder and an important, coherent new solutionto the old problem of environmental determinism andthe Mesopotamian state.

hans j . nissenInstitute of Near Eastern Archaeology, Free Universityof Berlin, Huettenweg 7, D-14195 Berlin, Germany([email protected]). 20 xi 00

When I wrote my Grundzuge einer Geschichte der Fruh-zeit des Vorderen Orients (subsequently translated asThe Early History of the Ancient Near East, 9000–2000b.c. [1988]), I specifically used the wording “a history,”emphasizing that this was only one possible way of con-ceptualizing a history of the ancient Near East: I had inmind that either I or someone else would eventually

write that “other” history. I am happy to see that suchan attempt has been made on at least one of the mostcrucial questions which I tried to answer: why the Mes-opotamian urban civilization emerged where and whenit did. In particular, I welcome the attempt to focus ontwo “advantages” which I have always considered im-portant but not important enough to my argument tomerit a lengthy discussion. However, instead of openinga different line of argument, Algaze only reversespriorities.

Of his two advantages, fertility is for me one of theattractions for those new settlers who—wherever andwhenever they had entered the alluvium—had causedthe enormous population density which we observe inthe southern Mesopotamian plain by the end of the 4thmillennium b.c. The favorable situation for exchange ortrade, in contrast, has long been recognized but as es-sential not for the emergence but for the shaping of thisearly civilization. Though of great importance at one oranother point in the development, these advantages bothfail to offer any explanation for the initial ignition. Wher-ever these settlers had obtained their experiences, theywere acquainted with both agriculture and exchange.The new conditions offered by the alluvium certainlymade life easier and in the long run could have led tonew forms of organization, but it is difficult to imaginethat these conditions could have created the pressurenecessary to bring about changes of such magnitude aswe see happening between the Ubaid and the Late Uruk.

Arranging our information differently, I maintain thatrather than the confrontation with favorable conditionsit was the challenge of coping with the problems of thevast alluvial plain and the totally unprecedented popu-lation and settlement density that resulted in urban com-plexity. Fertility is not only one of the incentives foroccupying this plain in the first place but also responsiblefor the fact that on it less acreage was required to feedeach person. This in turn enabled settlements to drawcloser together, eliminating safety distances between thesettlements and thus creating new problems, which de-manded new solutions, and so forth.

Not denying that exchange/trade was an importantfactor, I am challenging the reasons for Algaze’s assertionthat it was one of the prime movers. On the one hand,I question two of the three points where he sees the maindifferences from the neighbouring areas: (1) Rather thanoffering a more varied concentration of exploitable sub-sistence resources, a vast plain with identical but close-ranged soil and climatic conditions has a more limitedrange of animal and plant species than the topographi-cally and climatically diversified regions of the neigh-boring areas. (2) While on the local level a division oflabor between fishermen, herders, and farmers wouldhave led to some exchange, I cannot see any path leadingto an exchange or even a trade system between the var-ious regions of the alluvium and the cities—especiallysince the near-identical soil and climatic conditionsthroughout the alluvium almost certainly resulted inidentical animal and plant populations. I simply cannot


see any reason for asymmetric relations’ arising betweenthe cities within the alluvium.

I could not agree more with what Algaze says aboutthe difference in scale between the Mesopotamian al-luvium and the neighboring regions—in particular be-cause I have stressed this point myself over the years(Nissen 1988, 1995, 1999). However, I have the feelingthat as much as Oates and Oates (1997), Lamberg-Kar-lovsky (1999), and Frangipane (1997) may have overstatedtheir argument in order to find a balance to the allegedtotal supremacy of southern Mesopotamia in earlierdays, Algaze also pushes too far. As I tried to show sometime ago (Nissen 1995), we are confronted with shiftingnetworks in which all parts respond to any changes any-where within their areas. Undoubtedly, the Mesopota-mian alluvium was driven to greater complexity at onepoint. But because of its limited and uneven resolution,our archaeological material tends to fool us when itseems to present us with evidence for continuing majordisparities.

joan oatesMcDonald Institute for Archaeological Research,University of Cambridge, Downing St., CambridgeCB2 3ER, England ([email protected]). 23 xi 00

This paper must be welcomed as an important, well-informed, and thoughtful contribution to the study ofstate formation in Mesopotamia. As someone who haslong been interested in the economic and social potentialof its irrigation economy, I am particularly pleased to seethe environmental dimension reintroduced. I have noargument with Algaze’s primary thesis, although itseems to me that the scale of the arable landscape con-stitutes a further limiting/enabling factor (compare CatalHoyuk) and that in this Brak and Hamouker are verysignificantly different from Warka. Indeed, it remains ob-vious that the complexity of the site of Warka in thesecond half of the 4th millennium b.c. is not paralleledelsewhere and that this precocity owes much to the po-tential of its landscape. My major criticism is that Algazeseems to have returned to an ecologically driven “centre-and-periphery” and that this view of the development ofcomplex society in wider Mesopotamia is misleading.On current evidence the background to the complex re-cording and administrative systems of Late Uruk Warkaand, indeed, to the stone and metal technology that cre-ated the outward symbols of its elite lies not in the southbut in the north (Duistermaat 2000; Esin 1989, 1994;Oates 1996). One has only to look at sites like Gobekli,Nevali Cori, and Cayonu to observe a level of ritual andtechnological development so far entirely absent in thesouth. This in no way invalidates Algaze’s general thesis,but it does bring into question his repeated assertionsabout “backward” and “less developed” polities in thenorth and their wider role.

Algaze writes that “multiple interacting urban settingsof comparable scale had actually characterized the Mes-opotamian alluvium for centuries before the Late Uruk

period.” This can equally be said of the northern plain,where the dating criteria are rather more secure thanthose used in the southern surveys. Indeed, Adams’s“Early Uruk” type fossils are not “early” in the sensethat the term is used in the north, where evidence forfour very sharply differentiated 4th-millennium-b.c.phases can be identified. This is not to criticize Adams,whose pioneering work is fundamental to our under-standing of state formation in the south, but only topoint out a fact of life known equally well to him—thatarchaeologists can only work with the evidence that isavailable.

Algaze cites a small survey in the immediate vicinityof Brak as indicating a settlement pattern that “lacks thedepth and range of settlement size variability found tosurround contemporary southern Mesopotamian urbansites”; in this he compares survey data that are hardlycomparable. Moreover, the sites indicated on Adams’s(mislabelled) figure 13 (1981) are identified by criteriathat are in northern terms “Middle Uruk” in date. Atthis period Brak is a site of some 100–160 ha, if oneincludes the area of the small sites which immediatelysurround it. Some 80 km to the east is 15-ha Hamoukar,with the estimated 30-ha Hawa only some 30–40 kmfarther east. “Middle Uruk” material is found also at thevery large sites of Tell Barri, Shaikh Names, and Farfara(106 ha), with distances measured from Brak of 9, 15, and28 km (Meijer 1986); unfortunately, at these sites theoverburden of later occupation makes size estimates im-possible. I should also add that the earliest phase of thetripartite Eye Temple was built in the indigenous “Mid-dle Uruk” phase, well before the southern Uruk intru-sion. An examination of the contemporary pattern rep-resented by Adams’s figure 13 reveals a generally similarbut far less dense pattern of small sites, one with farfewer moderately large sites and the very few 40� hasites situated at a distance of 80 km or more upstreamfrom Warka. I have no space to develop this point further,but it seems clear that in this “early urban” phase theevidence from the north is as compelling as that fromthe better-surveyed south; moreover, there is no evi-dence for a “complex, hierarchically organized” systemin the immediate vicinity of Warka. Even more signifi-cant is the “Early Uruk” phase at Brak, with possibleevidence of a city wall and certainly of a monumentalwalled compound on a site estimated at some 45 ha(Oates and Oates 1997, Emberling et al. 1999). At thesame time (ca. 4000 b.c.) there is occupation at Ha-moukar spread over approximately 300 ha (J. Ur, personalcommunication).

A second crucial question involves the role played bypeople of the ’Ubaid period in the north both in the de-velopment of the administrative techniques later to playan important role in the south but not as yet attestedthere and in the earliest organized acquisition of the re-sources of adjacent areas of Anatolia (e.g., Layer 7 atDegirmentepe, for which the evidence strongly suggeststhe presence of an ’Ubaid “outpost” [Esin 1985, 1989]).I agree that in none of these areas do we see a develop-ment of complexity that can be compared with Late


Uruk Warka, but it is here that Algaze’s ecological ar-guments prevail. My point is that before the Late Urukperiod the north played a more important role in thewider framework of social and economic developmentin Mesopotamia than is acknowledged in Algaze’s oth-erwise excellent paper.

Minor points: (1) It is unlikely that grain was a majorexport (Potts 1986). (2) Doubts about both manuring andthe degree of 3d-millennium salinization are expressedelsewhere (Powell 1985, Oates 1994); for example, thesite of Ur flourished in the same geographical positionfor some 5,000 years, hardly an argument for rapid lossof soil fertility. (3) Brak should be added to the Late Urukmetalworking sites (Oates and Oates 1997: fig. 16).

charles stanishDepartment of Anthropology, University of California,Los Angeles, Calif. 90095, U.S.A. ([email protected]). 22 xi 00

Since I am not a specialist in Old World archaeology, Iwill restrict my comments to what I believe are the im-portant theoretical implications of this paper. Algaze,citing a recent review by Robert McC. Adams, notes thatmuch of contemporary archaeological theory can becharacterized as “social determinism.” This body of the-ory locates the source of cultural change in social inter-actions between cognizant actors and plays down thedeterminative role of the physical environment. In con-trast, earlier neoevolutionary theories viewed the envi-ronment as a key agent in cultural change and humansas more passive agents that adapt to changes in theirenvironment. Algaze offers us a sophisticated approachthat represents a compromise between these two epis-temological positions. For him the natural environmentis important as the critical context in which humansinteract. Geography does not directly determine culture.Rather, it provides a suite of opportunities and con-straints for people interacting within a broad physicaland cultural landscape in space and time. From this per-spective, some geographical areas are more conducive tothe development of political and economic complexity,selecting, supporting, and/or promoting certain technol-ogies and social organizations. As Algaze notes, theseareas have “natural advantages” and are the loci of a“unique conjuncture” of resources.

In Algaze’s analysis, a unique combination of geo-graphical features in the alluvial lowlands of the Tigrisand Euphrates Rivers provided the initial “impetus,” the“initial developmental takeoff,” or the “head start” forthe emergence of complex society in the late 5th mil-lennium b.c. Building on the seminal work of Adamsand echoing the theoretical perspective of Elman Service,Algaze suggests that a mosaic of ecological niches wouldhave promoted economic specialization by individualcommunities. This specialization, in turn, promoted ex-port-driven trade economies inducing both cooperationand competition between these groups and ultimatelydriving an increase in political and economic complexity.

Once these regional economies were established, anew social dynamic was created that transcended manyof the constraints of geography. This new dynamic con-stitutes the social determinist component of Algaze’smodel. It is implied in concepts such as multiplier ef-fects, competitive emulation, self-aggrandizing elites,and the creation of socioeconomic asymmetries. Onecould argue that the adoption of information technolo-gies, a critical factor in his model, is best understoodwithin a social deterministic framework as well. Inshort, the nature of the physical environment (combinedperhaps with correspondingly high population densities)explains the initial development of political and eco-nomic complexity in the region. It furthermore consti-tutes the context in which all subsequent cultural evo-lution was nurtured and constrained. But a set of socialprocesses explains the emergence and maintenance ofmore complex political economies later on, a processthat cannot be reduced merely to a type of environmentaldeterminism.

In reality, most contemporary archaeological theoriesof the origin of complex society combine elements ofboth environmental and social determinism. In histori-cal terms, this represents a shift in thinking from view-ing cultural behavior strictly as the unconscious mani-festation of underlying evolutionary principles to theincorporation of conscious human agency into proces-sual models. The real challenge for anthropological ar-chaeology, in my view, is incorporating human agencyinto testable models of cultural evolution.

In response to this challenge, many of our colleagueshave too quickly abandoned comparative analysis andscientific logic. In spite of what some anthropologistshave suggested, the recognition of human agency in ev-olutionary processes does not automatically require a“particularistic” or historically contingent view of cul-tural change. The empirical fact is that the more closelywe look at regional sequences with more refined chro-nologies, the more we see striking parallels between cul-tural evolutionary patterns in different areas of theworld. There simply are a limited number of effectiveways to organize complex societies. Over time, peopleindependently and consciously arrived at these solutionsin a number of regions around the world. While I disagreewith Algaze on some specific theoretical points, I agreewith the general focus and epistemological assumptionsof his argument, and I commend his effort to provide atestable model of the evolution of complex society inMesopotamia using both social and environmental fac-tors. It is a welcome contribution to anthropologicaltheory.

t . j . wilkinsonThe Oriental Institute, 1155 East 58th Street, Chicago,Ill. 60637, U.S.A. ([email protected]). 20 xi 00

Algaze’s paper takes up a theme that is periodically ad-dressed in the literature on Mesopotamia (Gibson 1992,Hole 1994), and the author has made a welcome attempt


to factor the environment back into the development ofcivilization without taking a deterministic stance. HereI confine my comments to landscape productivity, re-gional asymmetries of population, and the significanceand development of transport systems.

Algaze highlights three advantages that southern Mes-opotamia had over neighboring regions. The first, greaterresiliency resulting from more varied resources, meansthat there is inbuilt insurance against subsistence fail-ure. Although this should result in greater system sta-bility, in certain complex systems it may lead to desta-bilization (McGlade 1999:464). In the Mesopotamiancase, rivers raised above flood basins by levees mightbreak their banks, resulting in instabilities that calledfor adroit administration.

In the context of land use productivity, irrigation re-sults in increased crop yield per unit area, but whetherthis also results in increased labor productivity has beenmuch debated. Hunt (2000:274–75) argues for traditionalrice cultivation that labor productivity can be higherwith more complex technologies (such as irrigation),thereby allowing for increased agricultural surpluses. Ifthis argument holds for irrigated cereals, the resultantsurplus could have fed the increased labor required forAlgaze’s industries or for increased public works such ascanal cleaning.

Transport has long been regarded as a fundamental ad-vantage that southern Mesopotamia had over neighbor-ing regions, and Algaze is correct to emphasize its im-portance. His assumption that the progradation of theGulf would have necessitated the “construction of everlarger and more expensive canal systems” requires fur-ther scrutiny, however. In part this is because the piece-meal construction and extension of canals can cause ma-jor engineering problems that entail considerablereexcavation (Doolittle 1990:153). Algaze, quite reason-ably and as have many before him, does not make a cleardistinction between rivers and canals. Thus Adams(1981) regards large-scale canal construction as taking offin the 3d millennium b.c., and Nissen (1988) for the sametime period states that mobilization of labor was fun-damental to the construction of canal systems. Never-theless, it remains unclear just how much of the channelsystem—up to 250 km in length—was actually excavatedand how much labor was required for such public works.If the canals were dug, their construction could haveexhausted the local labor supply that was normally en-gaged in cleaning and maintaining local canal networks.Such works would therefore have required the import oflabor from surrounding regions, thereby fueling furtherpopulation growth. An alternative mechanism for thedevelopment of the multiple channel systems charac-teristic of Sumer and Akkad is the enhancement of anatural process of bifurcation by local management. Thelow-gradient plains crossed by river channels raised bylevees provided ideal conditions for the abrupt shift oravulsion of channels that results in such anabranchingsystems (Verhoeven 1998:174; Jones and Schumm 1999;Morozova and Smith 2000). These conditions, enhancedby substantial within-channel sedimentation and down-

stream decrease in discharge, can result in abrupt chan-nel shifts that can be catastrophic to the inhabitants(Gibson 1973:454). However, localized cleaning andmanagement of the node of avulsion or bifurcation couldhave maintained flow in both channels rather than one,thereby substantially increasing the water supply andtransportation network. This would have provided thepotential for increases in land use and population. It isnot clear which of these mechanisms—large-scale canaldigging or the semimanaged avulsion of anastomosingchannels—produced the network that nourished theearly cities, but either would have encouraged or allowedfor increased population levels and would have contrib-uted to the growth of polities in the south. Algaze’s syn-ergistic cauldron may therefore benefit from the consid-eration of these two human-environmental interactivesystems.

Algaze correctly points out that water transport sys-tems provided an advantage to southern Mesopotamiaover peripheral areas such as northern Mesopotamia, butshould we, as he asserts, assume developmental asym-metries between north and south? Available settlementsurvey data suggest that during the Ubaid period esti-mated population densities in parts of the peripheralregions exceeded those in the southern alluvial plainsand maximum site size was probably similar to that inthe south. By the early 3d millennium b.c. southern set-tlements were considerably larger, although estimatedpopulation densities were less than is often assumed(Wilkinson 2000a:244, 249). Thus during the 4th millen-nium b.c. the balance of urbanization may have shiftedfrom the rain-fed north to the irrigated plains of thesouth. This shift may have resulted from processes dis-cussed by Algaze or, alternatively, it may have fueledfurther growth of settlement in the southern plains. Thequestion of regional asymmetries of settlement thereforeneeds to be addressed more explicitly by future studies.

By skillfully weaving together the physical environ-ment and social and economic factors in one interactivesystem, Algaze has made an important contribu-tion to the study of the development of the Sumeriancivilization.

Reply

guillermo algazeLa Jolla, Calif., U.S.A. 11 xi 00

I am grateful for the many insightful comments and clar-ifications raised by the various commentators, especiallythose of Friedman and Ekholm Friedman, who have cho-sen to examine in detail the speculative model that Iproposed to explain the evolution of the import/exporttrade patterns characteristic of early Mesopotamian civ-ilization. In brief, the model postulates three broad butoverlapping phases in this evolution: a first phase char-acterized by largely internal trade between southern pol-


ities exploiting varied ecological niches unique to 4th-millennium Mesopotamia along principles ofcompetitive advantage, a second phase in which import-substitution processes serve to homogenize the produc-tive capabilities of alluvial polities, and a third phasecharacterized by heightened competition between poli-ties that then possessed broadly comparable productivecapabilities. In the course of this latter phase, individualsouthern polities appear to have made efforts to openexternal markets for now widely diffused industries eas-ily adapted for export (principally the new dyed woolentextiles, as McCorriston insightfully notes) and to con-trol the flow of external resources brought into the south.Both steps can be seen as ways to legitimize evolvingsocial transformations and out-compete regional rivals.

Friedman and Ekholm Friedman agree in seeing themany ramifications of trade as central to processes ofsocial evolution in early Mesopotamia but call for a moredetailed characterization of the specific social processeswhereby the advantages in trade of Mesopotamian so-ciety were translated into technologies of power. Spe-cifically, they note the need for a thorough review of thesystemic relationships between trade, forms of labor or-ganization, slavery, information-processing technolo-gies, and external expansion during the Uruk period.While their points are well taken, we should rememberthat the level of specificity required to prove causal re-lationships between evolving trade patterns and specificsequences of social change is simply not achievable inthe Mesopotamian case for the period in which Urukcivilization first crystallized. There are three reasons forthis: (1) the lack of detailed textual documentation, (2)the fragmentary and unrepresentative nature of the ar-chaeological evidence at our disposal, and (3) our con-tinuing inability to subdivide the Uruk period into an-ything more precise than 200–300-year phases.

Falling, as I do, into the formalist camp in debatesabout the nature of premodern economies (Polanyi 1957),I found it gratifying that Friedman and Ekholm Friedmansee potential value to the use of the concept of importsubstitution to explain the socioeconomic processes inoperation in protohistoric societies. They properly ques-tion, however, whether the available evidentiary base isstrong enough to justify the use of concepts developedfor the explanation of modern capitalist economies inthe context of preindustrial societies. While this crucialquestion cannot be adequately addressed with the flaweddata at our disposal, the use of modern economic modelsin an ancient Near Eastern context is justified on twocounts. The first is that these models are not put forthas a prescription of conditions that must of necessityhave occurred; rather, as Knapp notes, they are intendedas hypotheses to be tested against the available archae-ological data, which all too often are fragmentary anddifficult to interpret. The second is that changes result-ing from the emergence of modern capitalism may havegreatly intensified and brought into sharp focus relation-ships of asymmetrical exchange leading to unequal de-velopment but by no means created those relationships,as Ekholm and Friedman (1979) themselves suggested

more than 20 years ago. The evolutionary sequence forthe initial development of import/export economies insouthern Mesopotamia proposed in this paper is emi-nently testable, and I hope that improved artifact-recov-ery methodologies and changed historical circumstancesin Iraq will soon allow a new generation of researchersto do exactly that.

Some of the commentators raise points of disagree-ment which I believe are more apparent than real andresult from inadequacies in the exposition of my ideas.Nissen, for instance, argues that environmental advan-tages fail to explain the Uruk takeoff. Instead, he seesthe challenge of large populations coping with an evolv-ing alluvial environment as a trigger for social complex-ity. In fact, our positions are not as dissimilar as Nissenbelieves, although I do think that he underestimates boththe degree of ecological variability of the southern al-luvial plain in the 4th-millennium b.c. and the economicpotential of marshes, lagoons, and estuaries. I see geo-graphical and environmental advantages and social prob-lems as causally and evolutionarily connected. In brief,my argument is that the inherent transportational ad-vantages of the southern Mesopotamian landscape, itsgreater fertility and reliability, and the unique densityand variability of its exploitable resources during the 4thmillennium created the conditions wherein unprece-dented population densities could emerge. Propinquity,in turn, is what created new social problems requiringresolution, and these problems were solved through theinvention of revolutionary “ideational” technologies,principally new forms of managing labor that empha-sized economies of scale and new forms of informationprocessing and record keeping.

Other commentators raise important issues not prop-erly addressed in my paper. Both Tony Wilkinson andJoan Oates, for instance, suggest that I erred in my char-acterization of the nature of Upper Mesopotamian so-cieties during the prehistoric periods. Oates notes thatby the Neolithic period (8th–7th millennium b.c.), anearly northern advantage in integrative institutions suchas cultic buildings/temples existed over the then sparselyinhabited south and that by the Late Neolithic and Halafperiods (6th–5th millennium b.c.) the north held an ad-vantage over the south (Early Ubaid) in site size and inthe complexity of administrative technology, as seenfrom the glyptic of the time. Wilkinson notes furtherthat the available survey data suggest that during theLate Ubaid period (late 5th millennium b.c.), estimatedpopulation densities in parts of Upper Mesopotamia ex-ceeded those in the southern alluvial plains and maxi-mum site sizes were probably similar in the two areas.

In retrospect, it seems to me that Oates and Wilkinsonare entirely correct when they argue that the northplayed a more important role in the wider framework ofsocial and economic development in (Greater) Mesopo-tamia before the Uruk period than is acknowledged inmy paper. Perhaps the time has finally come to abandonthe concept of the “pristine” state or civilization and toacknowledge that processes of social evolution are al-ways the results of regional and transregional patterns


of interaction (Kohl 1987b). The early social evolutionof Greater Mesopotamian polities before the 4th millen-nium is thus best conceptualized as a process of paralleldevelopment in which constant interaction between dif-ferent regions and cultural traditions fuels widespreadprocesses of competitive emulation. This, in turn, servesroughly to equalize the rate and direction of the evolu-tionary trajectories of otherwise very different politiesexploiting very different ecosystems. In this, the for-mative phases of Mesopotamian civilization appear sim-ilar to those of Mesoamerica, with its multiple but dis-tinct traditions of social complexity (the Maya, MonteAlban, Teotihuacan) developing independently but intandem as a result of comparable socio-evolutionary pro-cesses made widespread by extensive long-distance con-tacts (Marcus 1992).

The corrections to my characterization of social evo-lution in Greater Mesopotamia prior to the Uruk periodintroduced by Oates and Wilkinson serve to highlighteven further the abruptness of the Sumerian “takeoff”of the 4th millennium, when the balance of urbaniza-tion, sociopolitical complexity, economic differentia-tion, and ability to project power at a distance shifteddramatically to the irrigated plains of the south. Nothingbetter epitomizes this shift than the Middle–Late-Uruk-period expansion of southern populations into selectedportions of Upper Mesopotamia, although, as Dittmannnotes, we still have much to learn about the scale andvariability of that phenomenon.

Regrettably, some of the commentators fail to ac-knowledge the uniqueness of the Uruk revolution. Lam-berg-Karlovsky, for instance, notes that several sitescomparable to or larger than late-4th-millennium Urukexisted at the time in the context of the Tripol’ye culturein Ukraine. The even earlier Neolithic site of Mehrgarh,in Pakistan, estimated at 200 ha in extent (Jarrige andMeadow 1980), could also be added to his list. Oates,similarly, notes the existence of what she calls a 300-haoccupation of early Late Chalcolithic date (late 5th–early4th millennium) at Tell Hamoukar, in northeastern Sy-ria. These parallels are neither pertinent nor comparable.In trying to arrive at a cross-culturally valid definitionof what is urban, Fox (1977) has noted that settlementsize is one of the least important criteria and argued thatthe defining characteristics lie in the degree of internaldiversity and complexity within a settlement. Thus de-fined, Uruk was certainly unparalleled by a second halfof the 4th millennium. The Ukrainian sites never sur-passed the village level of social complexity. Mehrgarhis also largely undifferentiated, and its excavator explainsits anomalous size as the result of “horizontal stratig-raphy” created as a much smaller settlement periodicallyshifted its location within a small area over time. Sim-ilarly, the excavator of Hamoukar, McGuire Gibson (per-sonal communication, 2000), believes that a large por-tion of the 300-ha scatter of early Late Chalcolithicremains at that location is also largely the result of hor-izontal stratigraphy, possibly resulting from the use ofthe site as a center for the seasonal aggregation of no-madic tribal groups.

Oates also misses some of the uniqueness of the Urukphenomenon when she argues that the scale and densityof occupations in the north in the Late Chalcolithic pe-riod, spanning most of the 4th millennium b.c., werecomparable to or even exceeded those of the south. Sheargues, for instance, that Tell Brak was somewhere be-tween 100 and 160 ha in extent during the “MiddleUruk” phase. This figure presumes that the totality ofthe intervening area between the main mound and thering of satellites surrounding it was occupied. Recentwork in this area, however, found no evidence for suchan occupation (Emberling et al. 1999:16–17, 25) and sug-gests that a substantial portion of the intervening areaconsisted of large clay pits excavated in support of build-ing activities at the main site (Wilkinson 2000b:127).

Also problematic is Oates’s claim that the density ofLate Chalcolithic sites around Brak and the depth of thesite hierarchy in its vicinity surpassed those character-istic for Uruk-period southern Mesopotamia, particu-larly those typical for the northernmost portions of thealluvium surveyed by Adams. This should be qualifiedin two ways. First, the Nippur-Adab region of the allu-vium, where Oates notes the apparent absence of smalland medium-sized Uruk-period settlements, is preciselyone of the areas of southern Mesopotamia with thegreatest amount of recent alluvial deposition, which nat-urally tends to obscure the smaller sites (Adams 1981,Reichel 1997, Wilkinson 1990b). Absence of evidence isnot evidence of absence. I take it as a given that thevarious 40�-ha urban sites along parallel branches of theTigris-Euphrates fluvial system in the Nippur-Adab re-gion throughout the Uruk period did not exist in isola-tion from their hinterlands. Each must have been at theapex of a variegated settlement structure, even if thelower levels of that structure are now obscured by sed-iments deposited since the 4th millennium.

Second, while Oates is correct is noting that large siteswith Late Chalcolithic-period materials do exist in theUpper Khabur area at reasonably short distances fromBrak, the sites in question have not been surveyed in away that allows us to ascertain their nature or their ex-tent at the time. Clearly the time has come for an ex-tensive and coherent survey of the Jaghjagh branch ofthe Upper Khabur, where Brak is located, that would gobeyond the area covered by Eidem and Warburton (1996).Ideally, such a survey should follow the site recognitionand recording guidelines established by Wilkinson dur-ing earlier work in the Sinjar Plains of northern Iraq (Wil-kinson and Tucker 1995, Wilkinson 2000a) and now usedin ongoing surveys elsewhere in the Upper Khabur areaof Syria around Tell Beidar (Wilkinson 1998) and TellHamoukar (Gibson 2000). It may very well be that suchfuture work will support Oates’s contentions about thenature of the Late Chalcolithic settlement hierarchyaround Brak, but at present we have no reliable evidenceto consider beyond that discussed in my paper.

An important issue raised by some of the commen-tators is that of the role of religion in the transformationsattending the Uruk revolution. A full examination ofreligion as a factor in the rise of early Mesopotamian


civilization was never my intention in this paper. Rather,my goal was to use Jacobs’s ideas to gain insight into theprecociousness and unique vitality of processes of urbangrowth and socioeconomic differentiation in Uruk-pe-riod Mesopotamia. I am quite aware that my discussionsof the reasons underpinning the Uruk takeoff do not ad-dress the crucial issue of how the new ideological con-structs of the states emerging in alluvial Mesopotamiaat the time came into being. In any event, I fully concurwith the comments of Charvat and Lamberg-Karlovskyas to the centrality of religion in the transformations thatgave rise to Mesopotamian civilization. The central po-sition that cultic structures occupy at the very core ofUruk cities, as seen at Warka, allows for no other inter-pretation. Further, Lamberg-Karlovsky is entirely correctwhen he argues that ideologies about the relationshipbetween the rulers and the supernatural affect the re-lationship between the rulers and the ruled and that thisrepresents a synergy as important as the socioeconomicand demographic ones highlighted in my discussion.

Returning to topics more directly addressed in my pa-per, I would like to expand on a comment by Brentjesthat struck me as particularly important. This is the is-sue of the dual structure of the city of Uruk, with itsparallel ceremonial areas, Eanna and Kullab. Here againis a case where understanding the importance of earlytrade in Mesopotamia may help us better understand thespecific trajectory of early Mesopotamian urbanism andthe form of early Sumerian civilization. Adams (personalcommunication, 2000) has suggested that early Meso-potamian twin cities such as Uruk and Kish may resultfrom a conflation of two initially independent towns ad-jacent to but on opposite sides of a major canal or riverbranch. The growth of such (Ubaid) towns, initiallyfounded along natural crossing points, into major (Urukand later) cities can be understood in terms of the self-amplifying mechanisms discussed by Jacobs. Adams ar-gues that twin-town-type settlements would have ex-panded in tandem with improved crossing facilitiesbetween them such as multiple boat ferries. In so doingthey would have grown further as they attracted an evenlarger proportion of long-distance traffic from hinter-lands on either side, with their embankments becomingdocks for transfer of shipments from boats to caravansand vice versa.

I would like to close by turning to issues raised byCioffi-Revilla and Stanish regarding the role of agencyin human history. Cioffi-Revilla argues that the envi-ronmental, geographic, demographic, and socioeconomicvariables that are the focus of my paper and of compa-rable efforts by like-minded colleagues studying the or-igins of early states and civilizations elsewhere fail toaddress the proximate causes of the phenomena we seekto comprehend, which he sees as entirely political andhistorically contingent. He is certainly correct. However,because of the nature of the evidence at our disposal forearly “pristine” civilizations across the world, we re-grettably will never be able to reconstruct in detail thevarious strategies used by elites in the earliest states thatappeared across the world to convince commoners in

their societies that their interests were cotermi-nous—what Cioffi-Revilla refers to as the “collective-action problem.” One way to approach this conundrumwould be a systematic analysis of the available ethno-historic record of the modern era to explore how thecollective-action problem was solved in documentedcases where chiefdom-level societies were transformedinto states. This analysis could then serve as a source oftestable hypotheses applicable to analyses of social dy-namics at the time the world’s earliest states and civi-lizations emerged.

Agreeing with Cioffi-Revilla on the need for expla-nations that take into account the decisions of consciousactors, Stanish argues that the principal challenge in an-thropological archaeology today is finding a way to in-corporate the vagaries of human agency and historicalcontingency into testable and cross-culturally applicablemodels of social evolution. While I am less optimisticthan he about our ability to identify agency in the ar-chaeological record, I believe that an approximation inthe aggregate can indeed be achieved. In my paper I triedto meet Stanish’s challenge by suggesting that sometypes of human actions, such as the will to accumulatestatus, wealth, and power, are predictable across bothcultures and time (Hayden 1995:20–23) and that somegeographical areas are more conducive to achieving thesegoals than others. This naturally results in differentiallevels of probability for the development of initial socialcomplexity across the landscape. Southern Mesopotamiain the late 5th and 4th millennia b.c. proved to be par-ticularly fertile ground for precisely such a development.

References Cited

a d a m s , r . m c c . 1966. The evolution of urban society. Chi-cago: University of Chicago Press.

———. 1978. Strategies of maximization, stability, and resiliencein Mesopotamian society, settlement, and agriculture. Proceed-ings of the American Philosophical Society 122:329–35.

———. 1981. Heartland of cities. Chicago: University of ChicagoPress.

———. 1999. Archaeological surveys of the Mesopotamianplains: Discoveries and issues still in prospect. Paper deliveredat the Oriental Institute, Chicago, March 27.

———. 2000. Scale and complexity in archaic states. LatinAmerican Antiquity 11:187–93.

a d a m s , r . m c c . a n d h . - j . n i s s e n . 1972. The Uruk coun-tryside. Chicago: University of Chicago Press.

a l d e n , j . 1982. Trade and politics in Proto-Elamite Iran. cur-rent anthropology 23:613–28.

a l g a z e , g . 1993. The Uruk world system: The dynamics ofexpansion of early Mesopotamian civilization. Chicago: Uni-versity of Chicago Press.

———. 1995. “Fourth-millennium trade in Greater Mesopotamiaand the question of wine,” in The origins and ancient historyof wine. Edited by S. Fleming, P. McGovern, and S. Katz, pp.89–96. New York: Gordon and Breach.

———. 1999. “Trends in the archaeological development of theUpper Euphrates Basin of southeastern Anatolia during theLate Chalcolithic and Early Bronze Ages,” in Archaeology ofthe Upper Syrian Euphrates: The Tishrin Dam area. Edited by


G. del Olmo Lete and J.-L. Montero Fenollos, pp. 535–72. Bar-celona: Editorial Ausa.

———. 2001. “The prehistory of imperialism: The case of Urukperiod Mesopotamia,” in Uruk Mesopotamia and its neigh-bors: Cross-cultural interactions and their consequences in theera of state formation. Edited by M. Rothman. Santa Fe:School of American Research Press.

a l g a z e , g . , r . b r e u n i n g e r , a n d j . k n u d s t a d . 1994.The Tigris-Euphrates Archaeological Reconnaissance Project:Final report of the Birecik and Carchemish Dam survey areas.Anatolica 20:1–96.

a l g a z e , g . , r . b ru e n i n g e r , c . l i g h t f o o t , a n d m .ro s e n b e r g . 1991. The Tigris-Euphrates Archaeological Re-connaissance Project, 1989–90. Anatolica 17:175–240.

a l g a z e , g . , a . m i s i r , a n d t . w i l k i n s o n . 1992. San-liurfa Museum/University of California excavations and sur-veys at Titris Hoyuk, 1991: A preliminary report. Anatolica 18:33–60.

a m i e t , p . 1972. Glyptique susienne. Memoires de la Delega-tion Archeologique Francaise en Iran 43. Paris: Geuthner.

———. 1980. La glyptique mesopotamienne archaıque. Paris:Centre National de la Recherche Scientifique.

a r c a r i , e . 1988. “The administrative organization of the cityof Ebla,” in Wirtschaft und Gesellschaft von Ebla. Edited byH. Waetzoldt and H. Hauptmann, pp. 125–29. Heidelberg: Hei-delberger Orientverlag.

a r c h i , a . 1988. “Zur Organization der Arbeit in Ebla,” inWirtschaft und Gesellschaft von Ebla. Edited by H. Waetzoldtand H. Hauptmann, pp. 131–38. Heidelberg: HeidelbergerOrientverlag.

———. 1990. “The city of Ebla and the organization of the ruralterritory,” in The town as regional economic center in the an-cient Near East. Edited by E. Aertz and H. Klenglel, pp. 15–19.Leuven: Leuven University Press.

a s t o u r , m . c . 1988. “The geographical and political structureof the Ebla empire,” in Wirtschaft und Gesellschaft von Ebla.Edited by H. Waetzoldt and H. Hauptmann, pp. 139–58. Hei-delberg: Heidelberger Orientverlag.

b a d l e r , v. r . , p . e . m c g o v e r n , a n d d . l . g l u s k e r .1996. Chemical evidence for a wine residue from Warka (Uruk)inside a Late Uruk period spouted jar. Baghdader Mitteilungen27:39–43.

b a i n e s , j . , a n d n . y o f f e e . 1998. “Order, legitimacy, andwealth in ancient Egypt and Mesopotamia,” in Archaic states.Edited by G. Feinman and J. Marcus, pp. 199–260. Santa Fe:School for American Research Press.

b a i ro c h , p . 1988. Cities and economic development. Chi-cago: University of Chicago Press.

b a r - m a t t h e w s , m . , a . a y a l o n , a n d a . k a u f m a n .1997. Late Quaternary paleoclimate in the eastern Mediterra-nean region from stable isotope analysis of speleothems atSoreq Cave, Israel. Quaternary Research 47:155–68.

b a r k e r , g . , a n d d . m a t t i n g l y. Editors. 1999–2000. Thearchaeology of the Mediterranean landscape. 5 vols. Oxford:Oxbow. [abk]

b e n c o , n . 1992. Manufacture and use of clay sickles from theUruk Mound, Abu Salabikh. Paleorient 18:119–34.

b l a n c h e t , g . , p . s a n l a v i l l e , a n d m . t r a b o u l s i .1997. Le Moyen-Orient de 20,000 ans BP a 6,000 ans BP: Essaide reconstitution paleoclimatique. Paleorient 23:187–96.

b o e s e , j . 1995. Ausgrabungen in Tell Sheikh Hassan. Vol. 1.Vorlaufige Berichte uber die Grabungskampagnen 1984–1990und 1992–1994. Saarbrucken: Saarbrucker Druckerei undVerlag.

b o n e c h i , m . 1993. Repertoire geographique des textes cunei-formes 12/1: I nomi geografici dei testi di Ebla. Wiesbaden: Dr.Ludwig Reichert.

b r i t i s h a d m i r a l t y, n a v a l i n t e l l i g e n c e d i v i s i o n .1917. A handbook of Mesopotamia. Vol. 1. Oxford.

b ru m f i e l , e . m . 1994. “Factional competition and politicaldevelopment in the New World: An introduction,” in Fac-tional competition and political development in the New

World. Edited by E. M. Brumfiel and J. W. Fox, pp. 3–14. Cam-bridge: Cambridge University Press.

b ry s o n , r . a . , a n d r . u . b ry s o n . 1997. “High-resolutionsimulations of regional Holocene climate: North Africa andthe Near East,” in Third millennium b.c. climate change andOld World collapse. Edited by H. Nuzhet Dalfes, G. Kukla,and H. Weiss, pp. 565–94. Berlin: Springer Verlag.

———. 1999. Holocene climates of Anatolia as simulated witharchaeoclimatic models. Turkiye Bilimler Akademisi ArkeolojiDergisi 2:1–14.

b u t z e r , k . w. 1995. “Environmental change in the Near Eastand human impact on the land,” in Civilizations of the an-cient Near East. Edited by J. Sasson, J. Baines, G. Beckman,and K. Rubinson, pp. 123–51. New York: Scribner.

———. 2001. “Geoarchaeological implications of recent researchin the Nile Delta,” in Egypt and the Levant: The 4th-millen-nium background. Edited by E. van den Brink and T. E. Levy.London: Leicester University Press.

c a r t e r , e . , a n d m . w. s t o l p e r . 1984. Elam: Surveys ofpolitical history and archaeology. Berkeley: University of Cali-fornia Press.

c h a rv a t , p e t r . 1993. “Uruk IV society: Evidence for thescript,” in Sulmu IV: Everyday life in the ancient Near East(Papers presented at the International Conference, Poznan,19–22 September 1989). Edited by J. Zablocka and St. Zawa-dzki, pp. 53–62. Poznan: Wydawnictwo Naukowe UAM. [pc]

———. 1997. On people, signs, and states: Spotlights on Sume-rian society, c. 3500–2500 b.c.. Prague: Oriental Institute. [pc]

c h e r ry, j . f . , j . l . d a v i s , a n d e . m a n t z o u r a n i . 1991.Landscape archaeology as long-term history: Northern Keos inthe Cycladic Islands from earliest settlement to moderntimes. (Monumenta Archaeologica 16.) Los Angeles: UCLA In-stitute of Archaeology. [abk]

c i o f fi - r e v i l l a , c l a u d i o . 1998. Politics and uncertainty:Theory, models, and applications. Cambridge: Cambridge Uni-versity Press. [cc]

c i o f fi - r e v i l l a , c l a u d i o , a n d h a rv e y s t a r r . 1995.Opportunity, willingness, and political uncertainty: Theoreticalfoundations of politics. Journal of Theoretical Politics 7:447–76. [cc]

c o n t i , a . m . , a n d c . p e r s i a n i . 1993. “When worlds col-lide: Cultural developments in Eastern Anatolia in the EarlyBronze Age,” in Between the rivers and over the mountains:Archaeologica Anatolica et Mesopotamica Alba Palmieri De-dicata. Edited by M. Frangipane, H. Hauptmann, M. Liverani,P. Matthiae, and M. Mellink, pp. 361–413. Rome: Universita diRoma “La Sapienza.”

c ro n e , p . 1989. Pre-industrial societies. Oxford: BasilBlackwell.

c u l l e n , h . m . , a n d p . b . d e m e n o c a l . 2000. North At-lantic influence on Tigris-Euphrates streamflow. InternationalJournal of Climatology 20:853–63.

d a n t i , m . d . 1997. “Regional surveys and excavations,” inSubsistence and settlement in a marginal environment: Telles-Sweyhat, 1989–1995, preliminary report. Edited by R. Zett-ler, pp. 85–94. Philadelphia: MASCA.

d e l o u g a z , p . , a n d h . j . k a n t o r . 1996. Chogha Mish I:The first five seasons of excavations, 1961–1971. Edited by A.Alizadeh. Chicago: Oriental Institute.

d e n o b l e t , n . , p . b r a c o n n o t , s . j o u s s a u m e , a n d v.m a s s o n . 1996. Sensitivity of simulated Asian and Africansummer monsoons to orbitally induced varitions in insolation126, 115, and 6 kBP. Climate Dynamics 12:589–603.

d i t t m a n n , r . 1986. “Seals, sealings, and tablets: Thoughts onthe changing pattern of administrative control from the LateUruk to the Proto-Elamite period at Susa,” in Gemdet Nasr:Period or regional style? Edited by U. Finkbeiner and W. Rol-lig, pp. 332–66. Wiesbaden: Ludwig Reichert.

d o o l i t t l e , w. e . 1990. Canal irrigation in prehistoric Mex-ico: The sequence of technological change. Austin: Universityof Texas Press. [tjw]

d u i s t e r m a a t , k i m . 2000. “A view on Late Neolithic sealingpractices in the Near East: The case of Tell Sabi Abyad, Syria,”


in Administrative documents in the Aegean and their NearEastern counterparts. Edited by M. Perna. Naples. [jo]

d u n n e l l , r . c . 1980. Evolutionary theory and archaeology.Advances in Archaeological Method and Theory 3:35–99. [jm]

e d e n s , c . 1994. “On the complexity of complex societies:Structure, power, and legitimation in Kassite Babylonia,” inChiefdoms of early states of the Near East. Edited by G. Steinand M. Rothman, pp. 209–24. Madison: Prehistory Press. [jm]

e i d e m , j . , a n d d . w a r b u r t o n . 1996. In the land of Nagar:A survey around Tell Brak. Iraq 58:51–64.

e k h o l m , k . , a n d j . f r i e d m a n . 1979. “‘Capital’ imperial-ism and exploitation in ancient world systems,” in Power andprogaganda. Edited by M. T. Larsen, pp. 41–58. Copenhagen:Akademisk Vorlag.

e m b e r l i n g , g . , j . c h e n g , t . l a r s e n , h . p i t t m a n , t .s k u l d b o e l , j . w e b e r , a n d h . t . w r i g h t . 1999. Exca-vations at Tell Brak 1998: Preliminary report. Iraq 61:1–41.

e n g l u n d , r . k . 1990. Organisation und Verwaltung der Ur-III Fischerei. Berlin: Dietrich Reimer.

———. 1998. “Texts from the Late Uruk Period,” in Mesopota-mien: Spaturuk-Zeit und Fruhdynastische Zeit. Edited by P.Attinger and M. Wafler, pp. 15–236. Freiburg Schweiz: Univer-sitatsverlag/Gottingen: Vandenhoeck and Ruprecht.

e s i n , u fi k . 1985. “Some small finds from the Chalcolithic oc-cupation at Degirmentepe (Malatya) in eastern Turkey,” inStudi de paletnologia in onore di Salvatore M. Puglisi. Editedby M. Liverani, A. Palmieri, and R. Peroni, pp. 253–63. Rome.[jo]

———. 1989. “An early trading centre in eastern Anatolia,” inAnatolia and the Near East: Studies in honor of Tahsin Ozguc.Edited by K. Emre, B. Hrouda, M. Mellink, and N. Ozguc, pp.35–41. Ankara. [jo]

———. 1994. “The functional evidence of seals and sealings ofDegirmentepe,” in Archives before writing. Edited by P. Feri-oli, B. Fiandra, G. G. Fissoro, and M. Frangipane. Torino: Scrip-torium. [jo]

f e i n m a n , g . 1998. “Scale and social organization: Perspectiveson the archaic state,” in Archaic states. Edited by G. Feinmanand J. Marcus, pp. 95–134. Santa Fe: School of American Re-search Press.

f e i n m a n , g . , a n d j . m a r c u s . Editors. 1998. Archaicstates. Santa Fe: School of American Research Press.

f e r i o l i , p . , a n d e . f i a n d r a . 1983 (1988). Clay sealingsfrom Arslantepe VIA: Administration and bureaucracy. Origine12:455–509.

f i a n d r a , e . 1994. “Discussion,” in Archives before writing.Edited by P. Ferioli, E. Fiandra, G. G. Fissore, and M. Frangi-pane, p. 168. Rome: Centro Internazionale di Recerche Ar-chaeologiche Anthropologiche e Storiche.

f i n k b e i n e r , u . 1991. Uruk: Kampagne 35–37, 1982–1984, Diearchaologische Oberflachenuntersuchung. Mainz am Rhein:Philipp von Zabern.

f l a n n e ry, k . v. 1995. “Prehistoric social evolution,” in Re-search frontiers in anthropology. Edited by C. R. Ember and M.Ember, pp. 1–26. Englewood Cliffs: Prentice-Hall.

f o r e s t , j . d . 1996. Mesopotamie: L’apparition de l’etat,VIIIe–IIIe millenaire. Paris: Mediterranee. [pc]

———. 1999. Les premieres temples de Mesopotamie. British Ar-chaeological Reports International Series 765.

f o x , r . 1977. Urban anthropology: Cities in their cultural set-tings. Englewood Cliffs: Prentice-Hall.

f r a n g i p a n e , m . 1997. A fourth millennium temple/palacecomplex at Arslantepe-Malatya: North-south relations and theformation of early state societies in the northern regions ofsouthern Mesopotamia. Paleorient 23:45–73.

———. 2001. “The development of a powerful early state centrenorth of Taurus: Continuity and discontinuity in Greater Mes-opotamia,” in Uruk Mesopotamia and its neighbors: Cross-cultural interactions and their consequences in the era ofstate formation. Edited by M. Rothman. Sante Fe: School ofAmerican Research Press.

f ru m k i n , a . , i . c a r m i , a . g o p h e r , d . c . f o r d , h . p .s c h w a r c z , a n d t . t s u k . 1999. A Holocene millennial-

scale cycle from a speleothem in Nahal Qanah Cave, Israel.The Holocene 9:677–82.

g e l b , i . j . 1944. Hurrians and Subarians. Chicago: Universityof Chicago Press.

g e l l n e r , e r n e s t . 1988. Plough, sword, and book. Chicago:University of Chicago Press. [ccl]

g i b s o n , m c g . 1973. “Population shift and the rise of Mes-opotamian civilization,” in The explanation of culture change:Models in prehistory. Edited by C. Renfrew, pp. 447–63. Lon-don: Duckworth.

———. 1974. “Violation of fallow and engineered disaster inMes-opotamian civilization,” in Irrigation’s impact on society. Ed-ited by T. E. Downing and McG. Gibson, pp. 7–20. Tucson:University of Arizona Press.

———. 1992. “The origin and development of Sumerian civiliza-tion and its relation to environment.” International Center forJapanese Studies, Tokyo, International Symposium no. 6, pp.1–27. [tjw]

———. 2000. Hamoukar: Early city in northeastern Syria. Orien-tal Institute News and Notes 166:1–8, 18–19.

g o o d y, j . 2000. The power of the written tradition. Washing-ton, D.C.: Smithsonian Institution Press.

g o u l d , s . j . 1986. Evolution and the triumph of homology, orWhy history matters. American Scientist 74:60–69. [jm]

———. 1994. In the mind of the beholder. Natural History 103:14–23. [jm]

g r a y s o n , a . k . , g . f r a m e , a n d d . f r a y n e . 1987. Assyr-ian rulers of the third and second millennia b.c. (to 1115 b.c.).Toronto: University of Toronto Press.

h a r r i s o n , s . p . , d . j o l l y, f . l a a r i f , a . a b e - o u c h i ,b . d o n g , k . h e r t e r i c h , c . h e w i t t , s . j o u s s a u m e ,j . e . k u t z b a c h , j . m i t c h e l l , n . d e n o b l e t , a n d p .v a l d e s . 1998. Intercomparison of simulated global vegeta-tion distributions in response to 6 kyr BP orbital forcing. Jour-nal of Climate 11:2721–41.

h a y d e n , b . 1995. “Pathways to power: Principles for creatingsocioeconomic inequalities,” in Foundations for social ine-quality. Edited by T. Douglas Price and Gary M. Feinman, pp.15–86. New York: Plenum.

h e i l b ro n e r , r . l . 1994 (1967). “Do machines make history?”in Does technology drive history? Edited by M. R. Smith andL. Marx, pp. 54–65. Cambridge: MIT Press.

h e i n r i c h , e . 1936. Kleinfunde aus den archaischen Tempel-schichten in Uruk. Leipzig: Otto Harrassowits.

———. 1937. Die Grabung im Planquadrat K XVII. VorlaufigerBericht uber die von der Deutschen Forschungsgemeinschaftin Uruk-Warka unternommenen Ausgrabungen 8:27–55.

———. 1938. Grabungen im Gebiet des Anu-Antum-Tempels.Vorlaufiger Bericht uber die von der Deutschen Forschungsge-meinschaft in Uruk-Warka unternommenen Ausgrabungen 9:19–30.

h i r s c h , h . 1963. Die Inschriften der Konige von Agade. Archivfur Orientforschung 20:1–82.

h o e l z m a n n , p . , d . j o l l y, s . p . h a r r i s o n , f . l a a r i f ,r . b o n n e fi l l e , a n d h . j . p a c h u r . 1998. Mid-Holoceneland-surface conditions in northern Africa and the ArabianPeninsula: A data set for the analysis of biophysical feedbackin the climate system. Global Biochemical Cycles 12:35–51.

h o l e , f . 1994. “Environmental instabilities and urban origins,”in Chiefdoms and early states in the Near East. Edited by G.Stein and M. S. Rothman, pp. 121–52. Madison: PrehistoryPress.

———. 1997. “Evidence for Mid-Holocene environmental changein the western Khabur drainage northeastern Syria,” in Thirdmillennium b.c. climate change and Old World collapse. Ed-ited by H. Nuzhet Dalfes, G. Kukla, and H. Weiss, pp. 39–66.Berlin: Springer Verlag.

h u n t , ro b e r t c . 2000. Labor productivity and agriculturaldevelopment: Boserup revisited. Human Ecology 28:251–77.[tjw]

j a c o b s , j . 1961. The life and death of great American cities.New York: Random House.


———. 1969. The economy of cities. New York: Vintage.———. 1984. Cities and the wealth of nations. New York: Ran-

dom House.———. 2000. The nature of economies. New York: Modern

Library.j a c o b s e n , t . 1939. The Sumerian king list. Chicago: Univer-

sity of Chicago Press.———. 1970 (1953). “On the textile industry at Ur under Ibbi-

Sin,” in Towards the image of Tammuz. Edited by W. L.Moran, pp. 216–30. Cambridge: Harvard University Press.

———. 1982. Salinity and irrigation agriculture in antiquity: Di-yala Basin Archaeological Projects, report on essential results,1957–58. Malibu: Undena.

j a c o b s e n , t . , a n d r . m c c . a d a m s . 1958. Salt and silt inancient Mesopotamian agriculture. Science 121:1251–58.

j a r r i g e , j . - f . , a n d r . h . m e a d o w. 1980. The antecedentsof civilization in the Indus Valley. Scientific American 243:122–33.

j o h n s o n , g . a . 1987. “The changing organization of Uruk ad-ministration in the Susiana Plain,” in The archaeology ofwestern Iran. Edited by F. Hole, pp. 107–39. Washington, D.C.:Smithsonian Institution Press.

———. 1988–89. Late Uruk in Greater Mesopotamia: Expansionor collapse? Origini 14:595–611.

j o n e s , l . s . , a n d s . a . s c h u m m . 1999. “Causes of avul-sion: An overview,” in Fluvial sedimentology 6. Edited by N.D. Smith and J. Rogers, pp. 171–78. International Associationof Sedimentologists Special Publication 26. [tjw]

k n a p p , a . b . , a n d w. a s h m o r e . 1999. “Archaeologicallandscapes: Constructed, conceptualized, ideational,” in Ar-chaeologies of landscapes: Contemporary perspectives. Editedby W. Ashmore and A. Bernard Knapp, pp. 1–30. Oxford: Black-well. [abk]

k o h l , p . l . 1987a. “The ancient economy, transferable tech-nologies, and the Bronze Age world-system: A view from thenortheastern frontier of the ancient Near East,” in Centre andperiphery in the ancient world. Edited by M. Rowlands, M. T.Larsen, and K. Kristiansen, pp. 13–24. Cambridge: CambridgeUniversity Press.

———. 1987b. The use and abuse of world systems theory: Thecase of the pristine West Asian state. Advances in Archaeolog-ical Method and Theory 11:1–36.

k o h l m e y e r , k . 1997. “Habuba Kabira,” The Oxford encyclo-pedia of archaeology in the Near East. Edited by E. Meyers, pp.446–48. New York: Oxford University Press.

k o u c h o u k o s , n . 1998. Landscape and social change in LatePrehistoric Mesopotamia. Ph.D. diss., Yale University, NewHaven, Conn.

k u r t z , d . v. 1987. The economics of urbanization and stateformation at Teotihuacan. current anthropology 28:329–40.

l a m b e r g - k a r l o v s k y, c . c . 1996. Beyond the Tigris andEuphrates: Bronze Age civilizations. Jerusalem: Ben GurionUniversity of the Negev Press.

———. 1999. “Households, land tenure, and communication sys-tems in the 6th–4th millennia of Greater Mesopotamia,” inUrbanization and land ownership in the ancient Near East.Edited by M. Hudson and B. Levine, pp. 167–201. Cambridge:Peabody Museum of Archaeology and Ethnology.

l a r s e n , m . t . 1987. “Commercial networks in the ancientNear East,” in Centre and periphery in the ancient world. Ed-ited by M. Rowlands, M. T. Larsen, and K. Kristiansen, pp.47–56. Cambridge: Cambridge University Press.

l e b e a u , m . 1990. Esquisse d’une histoire de la Haute Mesopo-tamie au debut d l’Age du Bronze. Zeitschrift fur Assyriologie80:241–96.

l i c h b a c h , m . i . 1996. The cooperator’s dilemma. Ann Arbor:University of Michigan Press.

l i v e r a n i , m . 1988 (1983). Fragments of possible counting andrecording devices. Origini 12:511–21.

l i v i n g s t o n e , d . n . 1992. The geographical tradition: Epi-sodes in the history of a contested enterprise. Oxford: BasilBlackwell.

m c c o r r i s t o n , j . 1997. The fiber revolution: Textile extensi-

fication, alienation, and social stratification in ancient Meso-potamia. current anthropology 38:517–49. [jm]

———. 1999. “Syrian origins of safflower production: New dis-coveries in agrarian prehistory of the Habur Basin.” Proceed-ings of the International Symposium on the Origins of Agri-culture, Aleppo, Syria. Edited by A. B. Damania, J. Valkoun, G.Willcox, and C. O. Qualset, pp. 39–50. Aleppo, Rome, and Da-vis, Calif.: ICARDA, IPGRI, FAO, and GRCP. [jm]

m c g l a d e , j . 1995. Archaeology and the ecodynamics of hu-man-modified landscapes. Antiquity 69(262):113–32. [abk]

———. 1999. “Archaeology and the evolution of cultural land-scapes: Towards an interdisciplinary research agenda,” in Thearchaeology and anthropology of landscape. Edited by Peter J.Ucko and Robert Layton, pp. 458–82. London: Routledge. [tjw]

m a e k a w a , k . 1980. Female weavers and their children in La-gash: Presargonic and Ur III. Acta Sumerologica 2:81–125.

m a n n , m . 1986. The sources of social power. Vol. 1. Cam-bridge: Cambridge University Press.

m a r c u s , j . 1992. Dynamic cycles of Mesoamerican states. Re-search and Exploration 8:392–411.

———. 1998. “The peaks and valleys of ancient states: An exten-sion of the dynamic model,” in Archaic states. Edited by G.Feinman and J. Marcus, pp. 59–94. Santa Fe: School of Ameri-can Research Press.

m a r g u e ro n , j . c . 1992. Les bois dans l’architecture: Premieressai pour une estimation des besoins dans le bassin mesopota-mienne. Bulletin on Sumerian Architecture 6:79–96.

m a t t h i a e , p . 1988. “On the economic foundations of theEarly Syrian culture of Ebla,” in Wirtschaft und Gesellschaftvon Ebla. Edited by H. Waetzoldt and H. Hauptmann, pp.75–80. Heidelberg: Heidelberger Orientverlag.

m e i j e r , d . w. 1986. A survey in northeastern Syria. Istanbul:Nederlands Historisch-Archaeologisch Instituut te Istanbul. [jo]

m i c h a l o w s k i , p . 1986. “Mental maps and ideology: Reflec-tions on Subartu,” in The origins of cities in dry-farming Syriaand Mesopotamia in the third millennium b.c. Edited by H.Weiss, pp. 129–56. Guilford, Conn.: Four Quarters Publishing.

m o k y r , j . 1990. The lever of riches: Technological creativityand economic progress. New York: Oxford University Press.

———. 2000. “Evolutionary phenomena in technologicalchange,” in Technological innovation as an evolutionary pro-cess. Edited by J. Ziman, pp. 52–65. Cambridge: CambridgeUniversity Press.

m o o r e y, p . r . s . 1994. Ancient Mesopotamian materials andindustries: The archaeological evidence. Oxford: ClarendonPress.

m o ro z o v a , g . s . , a n d n . d . s m i t h . 2000. Holocene avul-sion styles and sedimentation patterns of the SaskatchewanRiver, Cumberland Marshes, Canada. Sedimentary Geology130:81–105. [tjw]

m u l l e r - k a r p e , m . 1991. “Aspects of early metallurgy inMesopotamia,” in Archaeometry ’90. Edited by E. Pernicka andG. A. Wagner, pp. 105–16. Basel: Birkhauser Verlag.

n i s s e n , h . j . - j . 1970. Grabung in den Planquadraten K/L XIIin Uruk-Warka. Baghdader Mitteilungen 5:101–92.

———. 1976. “Zur Frage der Arbeitsorganisation in Babylonienwahrend der Spat-uruk-Zeit,” in Wirtschaft und Gesellschaftin Alten Vorderasien. Edited by J. Harmatta and G. Komar-oczy, pp. 5–14. Budapest: Akademiai Kiado.

———. 1977. “Aspects of the development of early cylinderseals,” in Seals and sealings in the ancient Near East. Editedby McG. Gibson and R. Biggs, pp. 15–24. Malibu: Undena.

———. 1986. The Archaic Texts from Uruk. World Archaeology17:317–34.

———. 1988. The early history of the ancient Near East. Chi-cago: University of Chicago Press.

———. 1995. “Kulturelle und politische Netzwerke des 4. und 3.vorchr. Jahrtausends im Vorderen Orient,” in Beitrage zur Kul-turfeschichte Vorderasiens (Festschrift Boehmer). Edited by U.Finkbeiner, R. Dittmann, and H. Hauptmann. Mainz: Zabern.[hjn]

———. 1999. Geschichte Altvorderasiens. Munchen: Oldenbourg.[hjn]


n i s s e n , h . - j . , p . d a m e ro w, a n d r . k . e n g l u n d . 1993.Archaic bookkeeping. Chicago: University of Chicago Press.

n o w l a n , d . 1997. “Jane Jacobs among the economists,” inIdeas that matter: The worlds of Jane Jacobs. Edited by M. Al-len, pp. 111–13. Owen Sound, Ont.: Ginger Press.

n u t z e l , w. 1976. The climate changes of Mesopotamia andbordering areas. Sumer 32:11–24.

o a t e s , j o a n . 1993. Trade and power in the fifth and fourthmillennia b.c.: New evidence from northern Mesopotamia.World Archaeology 24:403–22.

———. 1994. Comment on: The structure and dynamics of dry-farming states in Upper Mesopotamia, by T. J. Wilkinson. cur-rent anthropology 35:510–11. [jo]

———. 1996. A prehistoric communication revolution. Cam-bridge Archaeological Journal 6:165–76. [jo]

o a t e s , j . , a n d d . o a t e s . 1993. Excavations at Tell Brak1992–93. Iraq 55:155–99.

———. 1994. Tell Brak: A stratigraphic summary. Iraq 56:167–76.

———. 1997. An open gate: Cities of the fourth millennium b.c.(Tell Brak 1997). Cambridge Archaeological Journal 7:287–307.

o ’ b r i a n , m . l . Editor. 1996. Evolutionary archaeology. SaltLake City: University of Utah Press. [jm]

p a l m i e r i , a . , a . h a u p t m a n n , a n d k . h e s s . 1997. Themetal objects from the “royal” tomb dating to 3000 b.c. at Ar-slantepe (Malatya): A new alloy (CU-AG). Arkeometri Sonu-clari Toplantisi 13:115–22.

p a r k , t . 1992. Early trends toward class stratification: Chaos,common property, and flood recession agriculture. AmericanAnthropologist 94:90–117.

p e r r i n d e b r i c h a m b a u t , g . , a n d c . c . w a l l e n . 1963.A study of agroclimatology in semi-arid and arid zones of theNear East. Geneva: World Metereological Organization.

p e t i t - m a i r e , n . , p . s a n l a v i l l e , a n d z . w. y a n . 1995.Oscillations de la limite nord du domaine des moussons afri-caine, indienne, et asiatique, au cours du dernier cycle clima-tique. Bulletin de la Societe Geologique de France 166:213–20.

p i t t m a n , h . 1993. “Pictures of an administration: The LateUruk scribe at work,” in Between the rivers and over themountains: Archaeologica Anatolica et Mesopotamica AlbaPalmieri Dedicata. Edited by M. Frangipane, H. Hauptmann,M. Liverani, P. Matthiae, and M. Mellink, pp. 235–46. Rome:Universita di Roma “La Sapienza.”

p o l a n y i , k . 1957. “The economy as instituted process,” inTrade and markets in the early empires. Edited by K. Polanyiet al., pp. 243–69. Chicago: Henry Regnery.

p o l l o c k , s . 1992. Bureaucrats and managers, peasants andpastoralists, imperialists and traders: Research on the Urukand Jemdet Nasr periods in Mesopotamia. Journal of WorldPrehistory 6:297–336.

———. 2001. “The Uruk period in southern Mesopotamia,” inUruk Mesopotamia and its neighbors: Cross-cultural interac-tions and their consequences in the era of state formation. Ed-ited by M. Rothman. Santa Fe: School of American ResearchPress.

p o s t g a t e , j . n . 1988. “A view from down the Euphrates,” inWirtschaft und Gesellschaft von Ebla. Edited by H. Waetzoldtand H. Hauptmann, pp. 111–20. Heidelberg: Orientverlag.

———. 1994. How many Sumerians per hectare? Probing theanatomy of an early city. Cambridge Archaeological Journal 4:47–65.

p o t t s , d . t . 1986. Rethinking some aspects of trade in theArabian Gulf. World Archaeology 24:423–40. [jo]

———. 1997. Mesopotamian civilization: The material founda-tions. Ithaca: Cornell University Press.

p o u r n e l l e , j . 1999. The Mesopotamian heartland revisited:Tentative identification of 4th–3rd millennium Tigris and Eu-phrates channels in Sumer and Akkad. MS, Department of An-thropology, University of California, San Diego.

———. 2000. The littoral origins of Near Eastern civilization.MS, Department of Anthropology, University of California, SanDiego.

p o w e l l , m . 1985. Salt, seed, and yields in Sumerian agricul-

ture: A critique of the theory of progressive salinization. Zeit-schrift fur Assyriologie 75:7–38.

r e i c h a r t , g . j . , m . d e n h u l k , h . j . v i s s e r , c . h . v a nd e r w e i j d e n , a n d w. j . z a c h a r i a s s e . 1997. A 225 kyrrecord of dust supply, paleoproductivity, and the oxygen mini-mum zone from the Murray Ridge (Northern Arabian Sea). Pa-leogeography, Paleoclimatology, Paleoecology 133:149–69.

r e i c h e l , c . 1997. Changes in the plain levels of Babylonia andthe Diyala region from 6000 b.c. onwards: A view from exca-vated sites. Paper presented at the 64th Rencontre Assyriolo-gique, Venice.

r e n f r e w, c . , a n d j . f . c h e r ry. Editors. 1986. Peer polityinteraction and socio-political change. Cambridge: CambridgeUniversity Press.

r i n d o s , d . 1986. The evolution of the capacity for culture: So-ciobiology, structuralism, and cultural selection. current an-thropology 27:315–31. [jm]

ro t h m a n , m . 1993. “Another look at the ‘Uruk expansion’from the Tigris Piedmont,” in Between the rivers and over themountains: Archaeologica Anatolica et Mesopotamica AlbaPalmieri Dedicata. Edited by M. Frangipane, H. Hauptmann,M. Liverani, P. Matthiae, and M. Mellink, pp. 163–76. Rome:Universita di Roma “La Sapienza.”

———. 1994. “Seal and sealing findspot, design audience, andfunction: Monitoring changes in administrative oversight andstructure at Tepe Gawra during the fourth millennium b.c.,”in Archives before writing. Edited by P. Ferioli, E. Fiandra, G.Fissore, and M. Frangipane, pp. 97–119. Rome: Scriptorium.

———. Editor. 2001. Uruk Mesopotamia and its neighbors:Cross-cultural interactions and their consequences in the eraof state formation. Santa Fe: School of American ResearchPress.

ro u a u l t , o . , a n d m . g . m a s s e t t i - ro u a u l t . 1993.L’Eufrate e il tempo: Le civilta del medio Eufrate e della Ge-zira siriana. Milano: Electa.

ru s s e l l , k . 1988. After Eden: The behavioral ecology of earlyfood production in the Near East and North Africa. British Ar-chaeological Reports International Series 391.

s a h l i n s , m a r s h a l l . 1974. Stone Age economics. London:Tavistock. [ccl]

s a n l a v i l l e , p . 1989. Considerations sur l’evolution de labasse Mesopotamie au cours des derniers millenaires. Paleo-rient 15:5–27.

s a u r e n , h . 1966. Topographie der Provinz Umma, nach derUrkunden der Zeit III. Dynastie von Ur. Vol. 1. Kanale undBewasserungsanlagen. Bamberg: K. Urlaub.

s c h w a r t z , g . 1988. Excavations at Karatut Mevkii and per-spectives on the Uruk/Jemdet Nasr expansion. Akkadica 56:1–41.

s c h w a r t z , m . , d . h o l l a n d e r , a n d g . s t e i n . 1999. Re-constructing Mesopotamian exchange networks in the 4th mil-lennium b.c.: Geochemical and archaeological analyses of bitu-men artifacts from Hacinebi Tepe, Turkey. Paleorient 25:67–82.

s c r a n t o n , p . 1994. “Determinism and indeterminacy in thehistory of technology,” in Does technology drive history? Ed-ited by M. R. Smith and L. Marx, pp. 143–68. Cambridge: MITPress.

s i ro c k o , f . , m . s a r n t h e i n , h . e r i e n k e u z e r , h .l a n g e , m . a r n o l d , a n d j . d u p l e s s y. 1993. Century-scale events in monsoonal climate over the past 24,000 years.Nature 364:322–24.

s m i t h , a . 1954 (1776) The wealth of nations. New York: E. P.Dutton.

s n o o k s , g . d . 1996. The dynamic society. London: Routledge.s o l l b e r g e r , e . , a n d j . - r . k u p p e r . 1971. Inscriptions roy-

ales sumeriennes et akkadiennes. Paris: Editions du Cerf.s p e n c e r , h . 1967 (1876, 1882). The evolution of society: Selec-

tions from Herbert Spencer’s Principles of sociology. Edited byR. L. Carneiro. Chicago: University of Chicago Press.

s t a r r , h a rv e y. 1978. “Opportunity” and “willingness” as or-dering concepts in the study of war. International Interactions4:363–87. [cc]

s t e i n , g . 1999. Rethinking world systems: Diasporas, colonies,


and interaction in Uruk Mesopotamia. Tucson: University ofArizona Press.

s t e i n , g . , a n d p . w a t t e n m a k e r . 1990. “The 1987 TellLeilan regional survey: A preliminary report,” in Economy andsettlement in the Near East: Analyses of ancient sites and ma-terials. Edited by N. Miller, pp. 8–18. Philadelphia: MASCA.

s t e i n k e l l e r , p . 1998. “The historical background of Urkeshand the Hurrian beginnings in northern Mesopotamia,” in Ur-kesh and the Hurrians: Studies in honor of Lloyd Cotsen. Ed-ited by G. Buccellati and M. K. Buccellati, pp. 75–98. Malibu:Undena.

———. 1999. “Archaic city seals and the question of early Baby-lonian unity,” in Thorkild Jacobsen Memorial Volume. Editedby Tzvi Abusch, pp. 1–12. New Haven: American OrientalSociety.

s t o n e , e . 1997. “City-states and their centers: The Mesopota-mian example,” in The archaeology of city-states. Edited by D.L. Nichols and T. H. Charlton, pp. 15–26. Washington, D.C.:Smithsonian Institution Press.

s t r a y e r , j o s e p h r . 1970. On the medieval origins of themodern state. Princeton: Princeton University Press. [ccl]

s t ro n a c h , d . 1994. “Village to metropolis: Nineveh and thebeginnings of urbanism in northern Mesopotamia,” in Nuovefondazioni nel vicino oriente antico: Realta e ideologia. Editedby S. Mazzoni, pp. 84–114. Pisa: Giardini Editori.

s u r e n h a g e n , d . 1986. “The dry-farming belt: The Uruk pe-riod and subsequent developments,” in The origins of cities indry-farming Syria and Mesopotamia in the third millenniumb.c. Edited by H. Weiss, pp. 7–43. Guilford, Conn.: Four Quar-ters Publishing.

t o b l e r , a . j . 1950. Excavations at Tepe Gawra. Vol. 2. Phila-delphia: University Museum.

t u r k e s , m . 1996. Spatial and temporal analysis of annual rain-fall variations in Turkey. International Journal of Climatology16:1057–76.

v a n d e m i e ro o p , m . 1997. The ancient Mesopotamian city.Oxford: Clarendon Press.

v a n z e i s t , w. , a n d s . b o t t e m a . 1991. Late Quaternaryvegetation of the Near East. Wiesbaden: Dr. Ludwig ReichertVerlag.

v e r h o e v e n , k . 1998. “Geomorphological research in the Mes-opotamian flood plain,” in Changing water courses in Babylo-nia: Towards a reconstruction of the ancient environment inlower Mesopotamia. Edited by Hermann Gasche and MichelTanret, pp. 159–245. Ghent: University of Ghent/Chicago:Oriental Institute. [tjw]

v o n w i c k e d e , a . 1990. Prahistorische Stempel Glyptic inVorderasien. Munich: Profil Verlag.

w a e t z o l d t , h . 1972. Untersuchungen zur neusumerischenTextilindustrie. Rome: Centro per le Antichita e la Storiadell’Arte del Vicino Oriente.

w a r d - p e r k i n s , j . b . 1964. Landscape and history in centralItaly. Oxford: Clarendon Press. [abk]

w e i s s , h . 1983. Excavations at Tell Leilan and the origins ofnorth Mesopotamian cities in the third millennium b.c. Paleo-rient 9:39–52.

———. 1986. “The origins of Tell Leilan and the conquest ofspace in third millennium north Mesopotamia,” in The originsof cities in dry-farming Syria and Mesopotamia in the thirdmillennium b.c. Edited by H. Weiss, pp. 71–108. Guilford,Conn.: Four Quarters Publishing.

———. 1990. Tell Leilan 1989: New data for mid-third-millen-nium urbanization and state formation. Mitteilungen derDeutschen Orient-Gesellschaft 122:193–218.

w i l k i n s o n , t . j . 1990a. Town and country in southeasternAnatolia. Vol. 1. Settlement and land use at Kurban Hoyukand other sites in the Lower Karababa Basin. Chicago: Orien-tal Institute.

———. 1990b. Early channels and landscape development aroundAbu Salabikh: A preliminary report. Iraq 52:75–84.

———. 1994. The structure and dynamics of dry-farming statesin Upper Mesopotamia. current anthropology 35:483–520.

———. 1995. “Late Assyrian settlement geography in UpperMesopotamia,” in Neo-Assyrian geography. Edited by M. Liver-ani, pp. 139–59. Rome: Universita di Roma “La Sapienza.”

———. 1997. “Environmental fluctuations, agricultural produc-tion, and collapse: A view from Bronze Age Upper Mesopota-mia,” in Third millennium b.c. climate change and Old Worldcollapse. Edited by H. Nuzhet Dalfes, G. Kukla, and H. Weiss,pp. 67–106. Berlin: Springer Verlag.

———. 1998. “Tell Beydar survey,” in The Oriental Institute:1997–1998, annual report, pp. 19–28. Chicago: OrientalInstitute.

———. 1999. Holocene valley fills of southern Turkey and north-western Syria: Recent geoarchaeological contributions. Quater-nary Science Reviews 18:555–71.

———. 2000a. Regional approaches to Mesopotamian archae-ology: The contribution of archaeological surveys. Journal ofArchaeological Research 8:219–67. [tjw]

———. 2000b. “Remote sensing and geographical informationsystems,” in The Oriental Institute: 1999–2000 annual report,pp. 123–30. Chicago: Oriental Institute.

w i l k i n s o n , t . j . , a n d d . j . t u c k e r . 1995. Settlement de-velopment in the north Jazira, Iraq: A study of the archaeolog-ical landscape. London: British School of Archaeology in Iraq.

w r i g h t , g . a . 1969. Obsidian analyses and prehistoric NearEastern trade: 7500–3500 b.c. Ann Arbor: University of Michi-gan, Museum of Anthropology.

w r i g h t , h . t . 1979. “Toward an explanation of the origin ofthe state,” in The origin of the state: The anthropology of po-litical evolution. Edited by R. Cohen and E. Service, pp. 49–68.Philadelphia: Institute for the Study of Human Issues.

w r i g h t , h . t . , a n d g . a . j o h n s o n . 1975. Population,exchange, and early state formation in southwestern Iran.American Anthropologist 77:267–89.

w r i g h t , h . t . , a n d e . ru p l e y. 2001. “Calibrated radiocar-bon age determinations of Uruk-related assemblages,” in UrukMesopotamia and its neighbors: Cross-cultural interactionsand their consequences in the era of state formation. Editedby M. Rothman. Santa Fe: School of American Research Press.

w r i g h t , r . 2000. Nonzero: The logic of human destiny. NewYork: Pantheon Books.

y e n e r , k . a . 1980. Third millennium b.c. interregional tradein southwest Asia with special reference to the Keban Regionof Turkey. Ph.D. diss., Columbia University, New York, N.Y.

y o f f e e , n . 1981. Explaining trade in the ancient Near East.Malibu: Undena.

———. 1995. Political economy in early Mesopotamian states.Annual Review of Anthropology 24:281–311.

z a r i n s , j . 1989. “The early utilization of indigo along thenorthern Indian Ocean rim,” in South Arabian archaeology.Edited by C. Jarrige, pp. 469–83. Paris: CNRS. [jm]

z i m a n , j . 2000. “Evolutionary models for technologicalchange,” in Technological innovation as an evolutionary pro-cess. Edited by J. Ziman, pp. 3–12. Cambridge: Cambridge Uni-versity Press.

in a recent review of gary feinman and joyce marcus’s ... · acting, in the words of joel mokyr...

Documents