World Archaeology

ISSN: 0043-8243 (Print) 1470-1375 (Online) Journal homepage: http://www.tandfonline.com/loi/rwar20

Canals versus horses: political power in the oasis

of Samarkand
Sebastian Stride , Bernardo Rondelli & Simone Mantellini
To cite this article: Sebastian Stride , Bernardo Rondelli & Simone Mantellini (2009) Canals
versus horses: political power in the oasis of Samarkand, World Archaeology, 41:1, 73-87, DOI:
10.1080/00438240802655302
To link to this article: http://dx.doi.org/10.1080/00438240802655302

Published online: 21 Feb 2009.

Submit your article to this journal

Article views: 308

View related articles

Citing articles: 7 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=rwar20
Download by: [Arizona State University Libraries]

Date: 24 November 2016, At: 16:59

Canals versus horses: political power in

the oasis of Samarkand
Sebastian Stride, Bernardo Rondelli and
Simone Mantellini

Abstract
At the heart of Central Asia, the Middle Zeravshan Valley and the city of Samarkand are
crisscrossed by a complex system of canals derived from the Zeravshan river (Fig. 1). In agreement
with the traditional theories proposed by soviet scholars and defended, in another context, by
Wittfogel, it is usually assumed that a strong, long-term association exists between the management
of this irrigation system and some form of centralized political power system.
By adopting an integrated approach, we will examine an alternative history of the emergence and
growth of Samarkand and the link between water management and socio-political power. In the rst
part we will focus on the archaeological evidence pertaining to the construction of the Dargom, the
primary canal of Samarkand and suggest that it is not necessarily the result of a short-term (or xed)
master plan or linked to a strong central political power. In the second part we will consider the nonirrigated grasslands surrounding the oasis and show that the socio-political structures of the oasis
cannot be understood without a more systemic approach.
The paper is based on the results of an ongoing project to survey the Middle Zeravshan Valley
systematically initiated by the Institute of Archaeology of the Academy of Sciences of Uzbekistan in
collaboration with various international teams (see acknowledgements).

Keywords
Samarkand; Middle Zeravshan Valley; Dargom canal; irrigation; nomads.

Introduction
Former Soviet Central Asia lies within the great semi-arid band of the northern
hemisphere. Its agricultural potential therefore depends to a large extent on irrigation,
derived from the rivers and streams descending from the mountains along its southern and
south-eastern borders. It is often perceived, almost intuitively, to be an area comparable to

World Archaeology Vol. 41(1): 7387 The Archaeology of Water

2009 Taylor & Francis ISSN 0043-8243 print/1470-1375 online
DOI: 10.1080/00438240802655302

74

Sebastian Stride et al.

Figure 1 The Middle Zeravshan Valley (LANDSAT image), overlapped on a DEM showing the
irrigated area (left). Simplied representation of the contemporary irrigation network (right).

the great cradles of civilization and in particular to Mesopotamia, with the Amu Darya
and Syr Darya rivers playing the role of the Tigris and Euphrates. Our knowledge of the
area is based to a large extent on the results of the Central Asian school of Soviet
archaeology, which was for decades one of the leading schools on a world-wide level
(Trigger 1989: 3239). The work of soviet archaeologists therefore shapes current research
and notably that of the history of irrigation,1 which was born with a multidisciplinary
perspective from the very start.
The best example of their achievements is the Multidisciplinary Archaeological
Expedition in Khorezm (XAKE`) created in 1937 by S. P. Tolstov. The expedition aimed
to reconstruct the long-term history of human occupation in the delta of the Amu Darya
in all its complexity. It included historians and archaeologists of course, but also
ethnologist, botanists, engineers, climatologists and artists among others. With direct
access to the upper spheres of Soviet power, no expense was spared: apart from camels and
tents the expedition was supplied with three aeroplanes to carry out systematic aerial
photographs upon which the reconstruction of the settlement pattern and in particular of
the irrigation system was later based.
The theoretical outline which S. P. Tolstov proposed for the history of Khorezm, and by
extension of Central Asia, was derived from the Marxist interpretation of the social
evolution of Western Europe, as acknowledged by Tolstov (1949: 25) and analysed by
Zejmal (1987). The prime mover that Tolstov postulated was class war and one of the key
pieces of archaeological evidence was the irrigation system. Particular emphasis was
placed on the Antique period (equivalent in Soviet terminology to Classical Antiquity)
when it was thanks to slavery that the great canals were supposed to have been build: the
construction of the great canals of Khorezm could only have been undertaken by a
centralized oriental despotism (Tolstov 1948: 49).
The interpretation of the irrigation systems by Soviet scholars is thus perfectly
compatible with Wittfogels thesis (Wittfogel 1957; see Barcelo, in Vea 1998: 7); what
changes is the prime mover postulated (Francfort and Lecomte 2002: 634). Both agree that
the construction of vast irrigation systems was linked to the emergence of states and to

Canals versus horses

75

political despotism and both use the discovery of major irrigation canals to explain the
existence of the state: in one case it is because only a centralized state able to control the
workforce could have built them (e.g. Andrianov and Mukhamedzhanov 1980: 3742), in
the other because they would have caused the emergence of a despotic state. It is therefore
not surprising that, when Soviet criticism of Wittfogel went beyond the accusation of
bourgeois revisionism, it was centred not so much on the role of irrigation as on the causal
factor upon which the system was built.2
The weakness of the link postulated between irrigation and the state has been shown by
various scholars,3 and in this article we will not return to these questions. However, the
quality of the work of teams such as S. P. Tolstovs means that many of their ideas remain
commonly accepted as fact and justies a re-examination of the link between irrigation
systems and socio-political structures in the area.

The Middle Zeravshan Valley, Samarkand and the Dargom

The Middle Zeravshan Valley is a depression stretching in sub-meridian direction between
oshoots of the Turkestan and Zeravshan ranges, with a total extent of about 230km and
a maximal width close to 70km. The braided river beds and their out-spanning irrigation
systems form a cultivated oasis that can be considered as one of the agricultural heartlands
of Central Asia. It is surrounded by gently sloping loess foothills intensively exploited for
ovi-caprids and cattle pasture, as well as for dry farming in favourable years. The total
cultivated area of the Middle Zeravshan Valley measured on the satellite imagery (Fig. 1
left) is currently approximately 6600km2, making this area one of the more signicant
oases in Central Asia. Because agriculture in the Middle Zeravshan Plain depends on
irrigation, the settlement pattern is intimately linked to water management and to the
water landscape (Rondelli and Tosi 2006). The water landscape is formed by a complex
network of natural and articial watercourses crisscrossing the alluvial plain, creating a
series of jazirehs, each one of which is characterized by dierent ecological features (Fig. 1
right).
This landscape is in constant evolution, due both to natural phenomena such as erosion,
deposit of alluvium, oods and meandering and to articial interventions, either to
preserve an existing system or to build a new one. Its main arteries include the Zeravshan
(which is divided into two main channels Ak Darya and Kara Darya), the Bulungur and
Pajaryk on the right bank and the Dargom and Narpaj on the left bank. We will
concentrate our discourse on the Dargom, which provides water to the city of Samarkand
and the south and south-eastern part of the Middle Zeravshan Valley.
Despite its contemporary appearance, with numerous meanders, deeply cut into the
loess terraces (Plate 1), no one disputes the articial origin of the Dargom. It forms a
typical canal, deriving from the Zeravshan river at the height of the Dam of the First
of May, close to the Tajik-Uzbek border, cutting through the small water courses
descending from the Zeravshan mountain range and following as far as possible the
contour lines.
Its size, length (over 100km) and the area which it brings under irrigation (about
1000km2) enable it to be classied as a major engineering operation. It is quite clear that

76

Sebastian Stride et al.

Plate 1 The Dargom canal.

this irrigation system was (and is) associated with a hydraulically compact society, not a
hydroagricultural or a hydraulically loose society, and more precisely that it corresponds
to the most characteristic form of Wittfogels hydraulic society, according to the
classication of Price (1994: g. 1).
The construction of the Dargom has often been linked to the emergence of the city of
Samarkand as a major urban centre. Indeed, the size of the city from the moment of its
foundation or soon after (over 200ha intra muros) and its apparent dependence on water
from the Dargom seem to t in well with the idea of a large-scale state-sponsored
intervention.
Thus, the unication of the Middle Zeravshan Valley into one interlinked productive
unit, the creation of a vast irrigation network and the construction of Samarkand as a
capital city at the centre of the alluvial plain appear to be an excellent example of the link
between irrigation and political centralization. This is stated explicitly by Gentelle: Only a
state is capable, once the strategic position of the site on the commercial roads of Central
Asia has been recognized, of mobilising the enormous mass of labour necessary for the
construction of the whole (2003: 191).
Isamiddinov (2002: 30) proposes three main stages in the development of the irrigation
system in the Samarkand area. Starting in the Eneolithic period (fourth millennium BCE),
irrigation systems were based on small derivations from mountain streams (called sajs) and
underground springs and the total irrigated area is estimated at 5000ha. In the second
period, from the eight to the seventh century BCE, a vast network of canals was built,
including the Dargom and Bulungur, enabling the irrigation of some 70,000ha. Finally, by
the early centuries CE, in the third stage the system was completed by the extension of
canals to the west (with the construction of the Eski Angor, Pajaryk and Narpai canals),
which led to the maximum possible area being irrigated prior to the Soviet period.

Canals versus horses

77

Dating the construction of the Dargom is, however, extremely dicult4 because the
canal may well have partly reused ancient stream beds and previously existing canals
derived from sajs. Thus what needs to be dated is not only the construction of the canal
but also the date at which water from one source replaced water from another source. The
key to this lies in the type of alluvial deposits left by the Zeravshan and those left by
the mountain sajs (see Ivanitskij and Inevatkina (1999) for a detailed attempt to date the
Dargom to the rst half of the rst millennium BCE, basing themselves on this).
A preliminary reconstruction of the landscape, prior to the construction of the Dargom,
has been done by analysing SRTM data (Shuttle Radar Topography Mission), historical
maps, such as Fedchenko (1870), aerial photos and topographical maps from the 1950s
(1:10000 scale). This enabled the recognition of many palaeohydrological traces, including
underground evidence not visible in the eld or on the topographical maps. It is thus
possible to reconstruct the ancient north-western courses of the mountain sajs prior to the
construction of the Dargom.5
The reconstructed natural water courses shown in Figure 2 would have been able to
provide water to an important part of the southern Middle Zeravshan Valley plain. Each
of the sajs could itself be derived from and linked to the others, giving rise to protoDargoms, which were eventually incorporated into a single network by the derivation
from the Zeravshan and doubled by the construction of the Jangi Aryk (Fig. 3).
The location of sites from the Iron Age and Antiquity is perfectly compatible with a
system of small canals derived from the mountain sajs, whereas Samarkand itself is ideally
situated to exploit the water both from the sajs and from a series of natural springs, which

Figure 2 The original courses of the saj with possible sites of the Achaemenid period.

78

Sebastian Stride et al.

Figure 3 The complete irrigation system with sites of the Early Middle Ages.

could easily have provided sucient water for the emergence of the city and its initial
period of growth (Gentelle 2003: 2023; Isamiddinov 2002: 1718; Mukhamadzhanov
1969).
It is thus possible to postulate a progressive construction of the irrigation network over
the long term, without the existence of an initial master plan or a centralized political
decision. The fact that most sites and associated elds can be irrigated with water
proceeding from either of two potential sources the Zeravshan or the mountain sajs and
springs would have reinforced the overall resilience of the system.
The only sites that cannot be explained without recourse to the Dargom are those from
the Early Medieval Period (Fig. 3). It was during this period that most sites were occupied
for the rst time and it was also during this period that a number of key sites rose along
the Dargom, such as the administrative centre of Kar Kala (Mantellini and
Berdimuradov 2004: 115) and, at the start of the canal, the temple of Jartepe, probably
located on the ancient branch of the Dargom. It is also to this period that the best
comparative example dates (Fig. 4): the construction of Mashruqan canal by Shapur I in
Khuzistan, for which we have both archaeological and textual evidence (Christensen 1993:
1078).
The Sogdian period thus appears as either the period of construction of the Dargom
canal or the period during which the lands potentially irrigated by this canal were fully
occupied. It is therefore a period of key importance for understanding the relation between
the political structure and irrigation system of Samarkand. It is also a period during which
the archaeological landscape . . . expresses an extreme decentralization of public power

Canals versus horses

79

Figure 4 A comparison between the Mashruqan and the Dargom canals. On the right side the
Dargom canal (proposed reconstruction of the two main phases); on the left side the Mashruqan
canal (image from Alizadeh et al. 2004: 81, g. 13).

(Mars ak and Raspopova 1991: 188) with at least two realities coexisting: on the one hand,
rural castles sometimes compared to European feudalism and, on the other,
autonomous city-states in which the king was but a primus inter pares, hereditary
succession was not systematic and the city itself had full juridical status, and may even in
certain cases have minted its own coins (de la Vaissie`re 2004: 1513). The political maps
which have been proposed for the period6 clearly show that the hydrological boundaries
did not coincide with the political ones, presumably implying that there were sources of
social stratication and administrative control other than irrigation (Roeder in Hunt and
Hunt 1976: 403).
The idea of a fairly loose link between political structures and irrigation systems seems
supported by ethnographic evidence from both Chinese and former Soviet Central Asia.
Local specialists are able to build complex, long canals with extremely simple methods,
based on eye measurement, and often obtain better results than modern engineers using
state-of-the-art technology (Wawrzyn Golab 1951: 1923). Thus at the end of the
nineteenth century in what is nowadays southern Uzbekistan, two small agro-pastoral
villages derived water from the Surkhan Darya without diculty; Russian engineers,

80

Sebastian Stride et al.

however, had enormous diculties in successfully prolonging and maintaining the same
canal to bring water to the new town of Termez (Stride 2005: 4003).
This lack of involvement of the state seems in some cases to extend to the decision to
build the canal itself. Wawrzyn Golab thus describes the situation in eastern Turkestan in
the rst half of the twentieth century as one where it was in fact generally left to the local
headmen to decide if and when a canal was to be built. He adds that the government does
nothing: It neither grants funds nor supplies provisions, nor lends any technical assistance
whatever (Wawrzyn Golab 1951: 1956).
We would therefore argue that the main nodes of the settlement pattern, and in
particular Samarkand, emerged prior to the construction of the Dargom irrigation system
and that their localization was linked to the proximity of water from either mountain sajs
or springs but had nothing to do with a hypothetical centralized political entity. The
construction of the Dargom is likely to have been progressive and to have resulted from
consensual agreements and alliances between the dierent local populations which would
have beneted from it. There is no need to postulate an exceptionally large labour force
either for construction (see Francfort and Lecomte 2002: 632) or for maintenance (the
capacity and slope of the Dargom mean that it does not require a seasonal cleaning). And
there is certainly no evidence to link the construction of the Dargom canal to a strong
centralized state.
This said, the most compelling evidence against Wittfogels and Tolstovs models comes
from analysing the long-term relationship between the irrigation system and socio-political
structures in and around Samarkand.

Political power, water management and the nomads

The irrigated oases of Central Asia in general and the Middle Zeravshan in particular do
not exist in isolation from the rest of the landscape. They form part of the mixed zone, an
ecological zone characterized by comparatively small areas suitable for irrigated
agriculture, surrounded by much larger resource-rich pastoral areas such as desert, steppe
and/or mountains. This variety of landscapes encourages mixed economies (mostly
combining agriculture and mobile pastoralism), in which close interaction between
pastoralists and agriculturalists is the rule (Paul 2003: 312).
Economically it makes sense to combine dierent ways of landscape exploitation so as
to maximize the use of dierent ecological niches and potential synergies between pastoral
and agricultural economic systems. Indeed, nomads and sedentaries can interchange
services (for example, herds manuring elds after the crops), people and especially
products (food, clothing, artefacts, etc.) to the extent that some specialists have spoken
about the dierentiation between nomads and sedentaries as a kind of social division of
work on a vast scale or even as dierent professional groups coexisting and interacting
within the same economic system (Digard 1990: 101). Obviously, this co-evolutionary
interaction over the long term had a lasting eect on socio-political structures and in
particular on the assumed link between irrigation systems and the state.
In his analysis of China and Inner Asia, Bareld (1991: 29) denes four interactive
spheres of which the two extremes were the Eurasian steppe sphere, marked by seasonal
migrations, extensive economy, low population density and tribal political organization,

Canals versus horses

81

and Chinese society with sedentary lifestyles, intensive irrigated agriculture, high population density and a centralized bureaucratic government. Southern Central Asia ts with his
denition of Chinese Turkestan, which incorporates a number of dierent ecological zones
adapted to either/both nomadic or/and sedentary people. When compared to China or
Mesopotamia, irrigated agricultural oases such as the Middle Zeravshan Valley are
comparatively small. Not only is any given city limited in size by its agricultural hinterland,
but it entertains a complex relation with its pastoral hinterland. Furthermore, supraregional cohesion is fragile due to the dierent oases being isolated and lacking communications. The development of large irrigation-based systems is therefore exposed to strong
internal limits. The main external factor is, doubtless, the proximity of the Eurasian steppes
and the pastoral nomadism and tribal structures associated with this world. Like China,
southern Central Asia is in direct contact with the steppe world, however, unlike China,
much of southern Central Asia is suited to a pastoral nomadic lifestyle.
The socio-political elite is therefore permanently obliged to reach some form of
agreement with the nomadic world and, not surprisingly, this elite usually actually
originates in the steppe. The absence of references to irrigation is logical because real
power and social structures among the elite are decided outside the irrigated lands.
Various authors have thus argued that the revolt of Spitamenes against Alexander the
Great in 328 BCE, which started in the Middle Zeravshan Valley, was caused by the fact
that the Greeks were endangering the economic complementarity and the system of
interaction that existed between nomad and settled populations. This system had been
preserved by the Achaemenid power but was questioned by Alexander the Great who
wished to establish an urban state clearly separated from the nomadic world
(Mandelshtam 1977: 219). According to Holt (1988: 537), the revolt ended only when
Alexander reached an agreement with the local population and ceased attempting to
impose a frontier between nomadic and sedentary worlds.
The study of the political history of the Middle Zeravshan Valley in the last centuries
BCE and the rst centuries CE (Rapin 2007) conrms the key role played by the nomads, as
do the Chinese historical sources. The outer limits of the Middle Zeravshan Valley are
clearly marked in the landscape by the tombs of nomads (Fig. 5) and the most important
of these tombs (notably Kok Tepe) often occupy the citadel of important pre-existing rst
millennium BCE sites situated on the border of the irrigable plain (Kok Tepe, Sazagan).
Another excellent example is provided by the most famous ruler of Samarkand,
Tamerlane, whose magnicent gardens depended on the irrigation system we have
described but who ruled as a true semi-nomad. Ruy Gonzalez de Clavijo, the ambassador
of the king of Castilla, thus describes how his meetings take place not in the city of
Samarkand but in gardens surrounding the city, and more mentions are made of tents
than houses, as can be seen from the following characteristic extract:
On this plain the lord ordered many tents to be pitched for himself and his woman; and that
all his host, which was scattered in detachments over the land, should be assembled together,
each man in his place, and that their tents should be pitched. . . . After three or four days,
twenty thousand men were assembled round the tents of the lord. . . . Every division of the
horde is provided with all that the troops require and they are arranged in streets.
(Clavijo 1999: 270; in original text: VIII, 4)

82

Sebastian Stride et al.

Figure 5 Distribution of the main kurgans (tombs of nomads) to the south of Samarkand.

This impression is further conrmed by our surveys which oer little evidence of a dense
settlement pattern during this period.7
Finally a major German research project (SFB 586, B2) on Nomadic Rule in a
Sedentary Context and State Formation in Central Asia in the Sixteenth and Eighteenth
Centuries has shown how the Uzbek conquests of the main sedentary areas of Central
Asia such as the Middle Zeravshan Valley and the Surkhan Darya Province, led by
Mohammad Shajbani Khan in the early sixteenth century, did not result in the ruling
dynasty becoming an urban-based kingship despite ruling from Bukhara for centuries.
Indeed, Holzwarth describes eighteenth-century Bukhara as a state in which the nomadic
conquerors had institutionalized and consolidated their rule in a sedentary context but
where the ruling class and the upper strata of society remained based on a nomadic life
style and a segmentary social structure.
The military rule of the Uzbek tribal chiefs ended only in the mid-nineteenth century
when Amir Nasrullah nally strengthened the central government in Bukhara and started
relying for the rst time on a land-based bureaucracy in which non-Uzbeks, tied to the
sovereign by personal loyalty and not tribal relation, held key positions (Holzwarth 2006).
From Genghis Khan up until the Soviet period, the main way of legitimizing kingship was
based on descent from Genghis Khan, be it direct, by marriage or invented. Elite Turkic
and Turco-Mongol dynastic clans fought for control over the cities, which remained
centres of Irano-Islamic civilization representing a sedentary, oasis culture dependent on
irrigation agriculture and international trade (Subtleny 1989: 103). We would argue that
the pre-Mongol situation was probably not fundamentally dierent.
Thus, if one were to search for a common denominator among the main states that have
existed in Central Asia, it would be the nomadic pastoral factor. The vast majority of the

Canals versus horses

83

ruling dynasties are of nomad origin and retain key characteristics of this origin, notably
in the socio-political sphere. Indeed, one of the only cases when Samarkand was ruled by a
dynasty of non-pastoral nomadic origin was under the Samanids (ninth to tenth centuries
CE). They dened themselves as a wall against the steppe nomads (Golden 1990: 347, citing
Nakhshak); however, their policies against the nomads conversely created an everincreasing tribal Turkic presence in the heartland of their empire and eventually led to the
Karakhanid conquest of Central Asia (Golden 1990).

Concluding remarks
Current evidence does not enable us to link the construction of the irrigation system of the
Middle Zeravshan Valley to a specic type of socio-political structure and certainly not to
a strongly centralized state, as previous scholars have suggested. Indeed, the period of
maximum expansion of the settlement pattern, and therefore of the irrigation system upon
which it necessarily depended, dates to the Early Medieval Period, a period of political
fragmentation and expansion by segmentation and not nucleation. Furthermore over the
long term, all available data clearly point to a complex reality in which irrigated
agriculture is key from an economical perspective but the socio-political structure is
mainly dominated by an elite of pastoral nomadic origin.
Models such as Tolstovs or Wittfogels have been developed for and by specialists of
the great irrigated heartlands such as China or Mesopotamia.8 They have been applied to
cases such as the Middle Zeravshan Valley because most archaeologists imagine the
Middle Zeravshan Valley as a miniature Mesopotamia. This is due both to the nature of
archaeological evidence (which privileges evidence of permanent sites) and to the division
of our discipline between specialists of the pastoral nomadic world and specialists of the
sedentary agricultural world. Whereas the former emphasize the extreme rarity of pure
nomadic pastoralism, with most pastoralists being semi-nomadic and often part-time
agriculturalists (Khazanov 1984), the latter usually simply consider the pastoral nomadic
world as an exogenous factor.
The Middle Zeravshan Valley, just like Barelds Chinese Turkestan, is no miniature
Mesopotamia. In Central Asia, nomadic pastoralism and sedentary irrigated agriculture
coexist and societies exhibit non-uniform denitions of general institutions to begin with
(Frachetti in press: 25). In order to understand the link between irrigation systems and
socio-political structures, we therefore propose to characterize the cultural landscape of
the Middle Zeravshan Valley by the presence of two subsystems: the subsystem
of agricultural-sedentary rules with a population of sedentary agents and the subsystem
of pastoral-nomadic rules with a population of nomadic agents.
Both subsystems show a considerable degree of specialization within and division of
labour between them. The total system represents a combination of dierent methods of
landscape exploitation enabling an ecient use of dierent ecological niches. Finally, the
two subsystems do not only interact but emerge in a process of co-evolution, can be
understood only as such and should be conceived as a complex systemic whole.
The fact that the irrigation system of the Middle Zeravshan Valley is comparable to that
of other great rivers such as the Euphrates or the Huang He has socio-economic

84

Sebastian Stride et al.

Plate 2 Nomads surveying the canals: a symbolic image of water management in early twentiethcentury Central Asia.

implications but very few socio-political ones. Indeed, for millennia, the best peaches and
sweetest roses of Samarkand have been in the possession of men who depended not on
canals but on horses for their power (Plate 2).

Acknowledgements
The eldwork upon which this article is based has been carried out thanks to the support of
the Institute of Archaeology of the Academy of Sciences of the Republic of Uzbekistan
directed by Sh. Pidaev, by the Italo-Uzbek Archaeological Expedition (directed by
A. Berdimuradov and M. Tosi) and the Franco-Uzbek Archaeological Expedition (directed
by F. Grenet and M. Isamiddinov). Special thanks go to M. Isamiddinov. We would also like
to thank the Catalano-Uzbek (J.-M. Gurt, Sh. Pidaev) and Japano-Uzbek Expeditions
(A. Berdimuradov, T. Uno) for their assistance as well as F. Grenet, Y. Karev, C. C. Lamberg
Karlovsky, M. Tosi and two anonymous reviewers for their comments.

Sebastian Stride, Department of Prehistory,

Ancient History and Archaeology, University of Barcelona
[email protected]
Bernardo Rondelli, Department of Informatics,
Systems and Communication, University of Milano Bicocca
[email protected]
Simone Mantellini, Department of Archaeology,
University of Bologna
[email protected]

Canals versus horses

85

Notes
1 An excellent review of the main Soviet studies of irrigations systems in Central Asia (as
well as those of more recent research projects) is provided by Francfort and Lecomte
(2002).
2 As Andrianov said, State power was thus the main cause and not the result of the
success of the development of irrigation (1969: 65).
3 A particularly elegant demonstration is proposed in Francfort et al. (1989). The authors
created an expert system in order to analyse the deductive reasoning underlying major
archaeological theories (such as Wittfogel 1957) concerning the social evolution of
complex societies. They showed that the theories in question necessarily implied the
existence of irrigation not only in Mesopotamia but also in Iron Age Europe and
amongst Anthill societies (commented on in Francfort and Lecomte 2002: 6356).
4 Detailed reviews and references concerning the dierent dates proposed (ranging
from the eighth century BCE to the sixth century CE) are provided by Isamiddinov (2002:
207).
5 For a detailed description of the methodology see Rondelli and Tosi (2006). Previous
authors have already underlined the potential importance of these mountain saj prior to
the construction of the Dargom (Isamiddinov 2002: 19).
6 See, for example, map 5 in de la Vaissie`re (2004), according to which the Dargom passes
through at least two political units. See also the discussion in Grenet and de la Vaissie`re
2002: 166).
7 See also other surveys from the Surkhan Darya (Stride 2005: 3589) province or Herat
(U. Franke pers. comm.).
8 We would like to stress that we are not assuming that these models are correct in the
Chinese or Mesopotamiam contexts but simply stating that they should not be applied
to the Middle Zeravshan Valley.

References
Alizadeh, A., Kouchoukos, N., Wilkinson, T. J., Bauer, A. and Mashkour, M. 2004. Human
environment interactions on the Upper Khuzestan Plains, southwest Iran. Recent investigations.
PaleOrient, 30(1): 6988.
Andrianov, B. V. 1969. Drevnie orositelnye sistemy Priaralja (v svjazi s istoriej vozniknovenija i
razvitija oroshaemogo zemledelija). Moscow: Nauka.
Andrianov, B. V. and Mukhamedzhanov, A. R. 1980. Irrigatsija i ejo rol v sotsialnoe`konomicheskoj istorii drevnej i srednevekovoj Srednej Azii,. Obshchestvennye Nauki v Uzbekistane,
11: 3542.
Bareld, T. J. 1991. Inner Asia and cycles of power in Chinas imperial history. In Rulers for the
Steppe: State Formation on the Eurasian Periphery (eds G. Seaman and D. Marks). Los Angeles:
University of Southern California.
Christensen, P. 1993. The Decline of Iranshahr: Irrigation and Environments in the History of the
Middle East, 500 B.C. to A.D. 1500. Copenhagen: University of Copenhagen, Museum Tusculanum
Press.
Clavijo, Ruy Gonzalez de 1999. Embajada a Tamorlan (ed. and intro. F. Lopez Estrada). Madrid:
Clasicos castalia.

86

Sebastian Stride et al.

Digard, J. P. 1990. Les relations nomades-sedentaires au Moyen-Orient: elements dune polemique.

In Nomades et sedentaires en Asie centrale (ed. H. P. Francfort). Paris: CNRS, pp. 97111.
Fedchenko, A. 1870. Topographical sketch of the Zarafshan Valley. Journal of the Royal
Geographical Society of London, 40: 44861.
Frachetti, M. In press. Dierentiated landscapes and non-uniform complexity among Bronze Age
societies of the Eurasian steppe. In Monuments, Metals and Mobility: Trajectories of Complexity in
the Late Prehistoric Eurasian Steppe (eds B.K. Hanks and K. Lindu). Cambridge: Cambridge
University Press.
Francfort, H. P. and Lecomte, O. 2002. Irrigation et societe en Asie centrale des origines a` lepoque
achemenide. Annales, 57(3): 62563.
Francfort, H. P., Lagrange, M.- S. and Renaud, M. 1989. Palame`de: Application des syste`mes experts
a` larcheologie des civilisations urbaines protohistoriques. Paris: CNRS.
Gentelle, P. 2003. Traces deau: Un geographe chez les archeologues. Paris: Belin.
Golden, P. B. 1990. The Karakhanids and early Islam. In The Cambridge History of Early Inner Asia
(ed. D. Sinor). Cambridge: Cambridge University Press, pp. 34370.
Grenet, F. de la Vaissie`re, E. 2002. The last days of Panjikent. Silk Road Art and Archaeology, 8:
15590.
Holt, F. L. 1988. Alexander the Great and Bactria: The Formation of a Greek Frontier in Central Asia.
Leiden: Brill.
Holzwarth W. 2006. The Uzbek state as reected in eighteenth century Bukharan sources. Asiatische
Studien, 60(2): 32153.
Hunt, R. C. and Hunt E. 1976. Canal irrigation and local social organization. Current Anthropology,
17(3): 389411.
Isamiddinov, M. 2002. Istoki gorodskoj kultury samarkandskogo Sogga. Tashkent: Iz. Narodnogo
nasledija.
Ivanitskij, I. D. and Inevatkina, O. N. 1999. Periodizatsija i e`tapy razvitija vodosnabzhenija
Afrasiaba (po materialam raskopok magistralnogo kanala v 19961997 gg. Istorija Materialnoj
Kultury Uzbekistana, 30: 96103.
Khazanov, A. 1994. Nomads and the Outside World, 2nd edn. Madison, WI: University of Wisconsin
Press.
Mandelshtam, A. M. 1977. Rannie kochevniki v drevnej istorii Srednej Azii. In Le plateau Iranien et
lAsie Centrale des origines a` la conquete islamique. Paris: CNRS, pp. 21523.
Mantellini, S. and Berdimuradov A. E. 2004. New investigation in the Sogdian Fortress of Kar
Kala. In Ancient Civilizations from Scythia to Siberia, Vol. 11(12). Leiden: Brill, pp. 10731.
Mars ak, B. I. and Raspopova V. I. 1991. Cultes communautaires et cultes prives en Sogdiane. In
Histoire et cultes de lAsie centrale preislamique. Paris: CNRS, pp. 18796.
Mukhamadzhanov, A. R. 1969. K voprosu o vodosnabzhenii Afrasiaba. Afrasiab, 1: 294300.
Paul, J. 2003. The state and the military: a nomadic perspective. In (ed.): Militar und Staatlichkeit
(ed. I. Schneider). Orientwissenschaftliche Hefte, 12, Mitteilungen des SFB Dierenz und
Integration 5. Halle, pp. 2568.
Price, D. H. 1994. Wittfogels neglected hydraulic/hydroagricultural distinction. Journal of
Anthropological Research, 50(2), pp. 187204.
Rapin, C. 2007. Nomads and the shaping of Central Asia (from the early Iron Age to the Kushan
period). In After Alexander: Central Asia before Islam. Proceedings of the British Academy 133.
Oxford: Oxford University Press for the British Academy, pp. 2972.

Canals versus horses

87

Rondelli, B. and Tosi, M. 2006. GIS and Silk Road studies: monitoring landscape and population
changes at Samarkand and in the Middle Zeravshan Valley. In Reading Historical Spatial
Information from around the World: Studies of Culture and Civilization based on Geographical
Information Systems Data (ed. T. Uno). Kyoto: International Research Centre for Japanese Studies,
pp. 45989.
Stride, S. 2005. Geographie archeologique de la Province du Surkhan Darya (Ouzbekistan du sud,
Bactriane du nord). PhD thesis, University of Paris I.
Subtelny, M. 1989. Baburs rival relations. Der Islam, 66(1): 10218.
Tolstov, S. P. 1948. Drevnij Khorezm, opyt istoriko-arkheologicheskogo issledovanija. Moscow:
MGU.
Tolstov, S. P. 1949. Periodizatsija drevnej istorii Srednej Azii v srednie veka (VIXV vv.). Kratkie
soobshchenija Instituta Istorii Materialnoj Kultury AN SSSR, 28: 1829.
Trigger, B. G. 1989. A History of Archaeological Thought. Cambridge: Cambridge University Press.
Vaissie`re, E. de la 2004. Histoire des marchands Sogdiens. Paris: Colle`ge de France.
Vea, L. 1998. Una arqueologa gigantesca: El estudio de la santiguas sociedades hidraulicas en las
republicas centroasiaticas de la extinta URSS. Barcelona: Bellatera, Universitat Auto`noma de
Barcelona.
Wawrzyn Golab, L., 1951. A study of irrigation in East Turkestan, Anthropos, 46(12): 18799.
Wittfogel, K. 1957. Oriental Despotism: A Comparative Study of Total Power. New Haven, CT: Yale
University Press.
Zejmal, E. V., 1987. K periodizatsii drevnej istorii Srednej Azii (seredina I tysjatcheletija do n. e`.
seredina I tysjacheletija n. e`.). Tsentralnaja Azija: Novye pamjatniki pismennosti i iskusstva (eds
B. B. Piotrovkij, G. M. Bongard-Levin). Moscow: Nauka, pp. 14957.

Sebastian Stride (PhD in Archaeology at Paris I University) has been working in Central
Asia since 1995. He is a member of the Franco-Uzbek and Catalano-Uzbek
Archaeological Missions and currently teaches Asian Studies and Archaeology at
Barcelona University.
Bernardo Rondelli (PhD in Archaeology at Bologna University) is a member of the ItaloUzbek archaeological expedition. He is currently based at the University of Milan-Bicocca
for a post-doctoral research fellowship on knowledge representation and material culture
studies.
Simone Mantellini is a PhD candidate in Archaeology at Bologna University and the main
topic of his research concerns water management in Antiquity. As a member of the ItaloUzbek Archaeological Mission he is studying the relationships between settlement
dynamics and irrigation systems in the Samarkand oasis.