You are on page 1of 22

Central Asian Survey

ISSN: 0263-4937 (Print) 1465-3354 (Online) Journal homepage: https://www.tandfonline.com/loi/ccas20

The demographic and agricultural development of


the Kokand oasis during the Russian Imperial era:
nomad immigration and cotton monoculture

Akira Ueda

To cite this article: Akira Ueda (2019) The demographic and agricultural development of the
Kokand oasis during the Russian Imperial era: nomad immigration and cotton monoculture, Central
Asian Survey, 38:4, 510-530, DOI: 10.1080/02634937.2019.1631754

To link to this article: https://doi.org/10.1080/02634937.2019.1631754

Published online: 08 Jul 2019.

Submit your article to this journal

Article views: 103

View related articles

View Crossmark data

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


https://www.tandfonline.com/action/journalInformation?journalCode=ccas20
CENTRAL ASIAN SURVEY
2019, VOL. 38, NO. 4, 510–530
https://doi.org/10.1080/02634937.2019.1631754

The demographic and agricultural development of the Kokand


oasis during the Russian Imperial era: nomad immigration and
cotton monoculture
Akira Ueda
Institute of Developing Economics, Chiba, Japan

ABSTRACT KEYWORDS
This study investigates oasis expansion in the precolonial period and Oasis; Kokand; cotton;
agricultural evolution in the colonial period using village-level historical statistics; Russian
statistics from the early twentieth century. This survey illustrates Central Asia; geographic
that the Kokand oasis in the Ferghana Valley initially appeared in information system
the central part, where the Sart population settled by the
seventeenth century. In the eighteenth and nineteenth centuries,
the Uzbek and Karakalpak tribes migrated to the valley
peripheries, while the Kyrgyz semi-nomads settled in the alluvial
fan. It is not surprising that such a migratory process created a
mosaic-like ethnic distribution. Furthermore, this study suggests
that the pattern coincided with a variety of agricultural practices,
and presents two contrasting models of cotton monoculture
under the Russian Empire. The main area occupied by the Sarts
presents a general model of canal-irrigated cotton planting in
Central Asia, while the Karakalpaks’ cotton planting using
groundwater in the periphery suggests multiple courses of nomad
sedentarization.

Introduction
This study investigates the expansion of the Kokand oasis in the precolonial period and its
agricultural evolution in the colonial period. The first part of this study, Ethnic Distribution,
reconstructs the spatial and demographic expansion of the oasis and the second, Agricul-
tural Development, argues that the variety of cotton planting in the colonial era reflected
this historically constructed spatial structure.
A series of nomadic migrations from the northern steppe into the southern oases is one
of the most important topics in the long-term history of Central Asia. The term oasis can be
defined as a type of water supply system in an arid region. The migration processes of
nomads have been frequently discussed in historical studies and anthropology (Khazanov
1994; Masanov 1995). These processes are closely connected with the rule of the Karakha-
nids, Chagatai Khanate, Timurid Empire, and Uzbek Khanates (Bartol’d 1963b, 234–292),
and the political and social history of Central Asia in the second millennium is usually
understood as being influenced by that process. Shioya’s (2014) and Abdurasulov’s
(2016) are the latest works on the political and social history of Central Asia concerning

CONTACT Akira Ueda Akira_Ueda@ide.go.jp


© 2019 Southseries Inc
CENTRAL ASIAN SURVEY 511

the relationship between oasis agriculture and ethnic problems in the Uzbek Khanates.
These studies are mainly based on narrative sources or field surveys, because there is
less quantitative information on nomads than on settled agricultural societies.
This study will use Russian colonial statistics as its main data source, but these statistics,
like other colonial statistics, have serious problems. S. Abashin (2012) has summarized the
three major problems with Russian colonial statistics on Central Asia. First, the colonial
authorities could not access all the information on local society. Second, there were
various visions and plans within the colonial authorities, and no single vision existed.
Third, the interactions between the colonial authorities and various groups in local
society affected the statistics. Abashin concluded that colonial statistics were very hetero-
geneous, diverse and ambiguous (147).
The recording of ethnicity in colonial statistics is the most problematic aspect. F. Hirsch
(2005) argued that the introduction of the ‘ethnic’ idea in Russian Central Asia was closely
connected with the academic context of the Russian Empire, and that this idea was formed
under the influence of Western ideas of ethnicity and nationality. The academic context of
scholars and self-references by the Central Asian native population interacted when the
statistics were compiled (101–144). In other words, the ‘ethnic’ information in the colonial
statistics was only occasionally in agreement with local forms of self-identification.
Because of this, historians cannot understand the reality of ethnicity by examining the stat-
istics. This article will mainly investigate the characteristics of population groups recorded
in colonial statistics. Although these groups had some continuity with earlier groups of the
same names, the research subject of this investigation is limited to population groups that
appeared in the colonial statistics. To distinguish between these two meanings, this article
encloses ‘ethnicity’ and references to ethnic groups in the statistics in quotation marks,
while ethnic terms that appear in earlier studies are used here without quotation marks.
Although these problems are obvious, the colonial statistics are the earliest (and almost
the only) quantitative data available on Central Asia. Regarding the economic collapse and
changes that occurred after the Russian Revolution, we must use the colonial statistics as
historical sources. This study uses the colonial statistics while taking the critiques of
Abashin and Hirsch into account.
Although many statistical data were gathered during the Russian colonial period, colo-
nial statistics were based on administrative divisions, and they did not coincide with his-
torical oases. In sum, the economic structure of local oases remained unclear, and the
economic histories of the colonial era were mainly concerned with the provincial level.
In the late twentieth century, scholars looked to technological breakthroughs to recon-
struct historical oases. Remote sensing and space archaeology have presented solutions.
Space archaeology in Central Asia has produced a series of results concerning historical
geography (Research Center for Silk Roadology 2003). In and after the 1960s, scholars of
remote sensing directly reconstructed the expansion or contraction of oases using satellite
photographs (Xie et al. 2014), but historians needed another way to connect written
sources with historical space.
In the twenty-first century, the academic use of geographical information system (GIS)
technology has become widespread, and historians of various topics have used it (Gregory
and Geddes 2014, ix–xix). Concerning Central Asia or the Russian Empire, some large pro-
jects to arrange historical GIS infrastructure, such as the Digital Silk Road Project of the
Japan National Institute of Information and UNESCO (http://dsr.nii.ac.jp), and the Imperiia
512 A. UEDA

Project of the Centre for Geographical Analysis, Harvard University (http://dighist.fas.


harvard.edu/projects/imperiia/), are being conducted. Case studies on each oasis are
also progressing. For example, archaeological GIS studies on the Zarafshan Valley have
attempted to reconstruct the historical trade network (Dmitrieva, Suchilin, and Inevatkina
2012). F. Maksudov has used GIS to reconstruct historical geography and food circulation
and to manage archaeology in Central Asia (Hermes et al. 2018; Maksudov 2012; Maksudov
et al. 2010). Regarding the Ferghana Valley, the studies by Komatsu and Goto (2004, 2009)
reconstructed the general tendencies of ethnic distribution and the change of ethnic cat-
egories from the colonial period to the early Soviet period, but the 14 spatial units based
on uchastoks (portions) were too large to draw conclusions about the spatial structure of
oases in the Valley.
Concerning agricultural development, this study focuses on cotton planting. The
remarkable expansion of cotton cultivation in Central Asia was the most important econ-
omic phenomenon under the rule of the Russian Empire. In 1888, only 14,300 tons of raw
cotton were exported from Central Asia to Russia. In 1912, that amount had reached
224,357 tons a year, a 15-fold increase over 25 years (Aziatskaya Rossiya II 1914, 92;
Bartol’d 1963b, 319–335).
Among the statistical materials I could access, this study used a series of the earliest
village-level statistics, the ‘List of Population Points of Ferghana Province’ (published in
1909) as demographic data (Spisok 1909), land-tax data from 1899–1902 as agricultural
data (Materialy IV, 1912), and the village map with the 1926 census as a basic place-
name map (Inogamov 1948). Although these materials have several problems, as I
explain below, this study will try to integrate these data and reconstruct the demographic
and agricultural developments of the oasis.
I selected the Kokand oasis in Uzbekistan as the subject of my case study. It is in the
western part of the Ferghana Valley, and by the early twentieth century it was one of
the largest oases in Central Asia. The population in 1909 was approximately four
hundred thousand (Spisok 1909, 68). As capital of the Kokand Khanate, the Kokand oasis
expanded dramatically in the 1740s. Nalivkin (1886, 20) suggested that large areas of
the Kokand oasis were developed mainly in the 150 years prior to his study, in other
words, after the city of Kokand was constructed as a capital. Before the construction of
the Great Ferghana Canal in 1939, the Kokand oasis was not irrigated by the Syr River
system, which flows from east to west along the northern edge of the Ferghana Valley.
Instead, it was irrigated by the Sokh River, which flows into the valley from the southern
mountains. The area between the northern end of the Kokand oasis and the southern side
of the Syr River was a desert. This study adjusted the drainage map in Masal’skii (1913) to a
digital base map and compiled the drainage lines for GIS analysis.
Regarding the evolution of oases in the Ferghana Valley, Nalivkin (1886) already
described the expansion of the irrigation system of the Namangan oasis and the immigra-
tion of nomads from the northern steppe to the oasis. He stressed that large parts of the
irrigation system in the Namangan oasis had been constructed ‘in recent years’ by the Uzbeks
(23–31). The Namangan and Kokand oases, which started developing in the eighteenth
century, were two of the most recent oases in Central Asia. These oases are suitable for
the reconstruction of oasis evolution because of their newness. Other famous oases
such as Samarkand, Bukhara, Khorezm, and Tashkent had already formed a somewhat
developed structure two thousand years ago (Bartol’d 1963b, 172, 177–178; Bedrintsev
CENTRAL ASIAN SURVEY 513

and Korzhavin 1975, 138). This study will attempt to reconstruct the evolution of the
Kokand oasis more quantitatively and spatially than Nalivkin did with the Namangan oasis.
The latest work on the history of the Ferghana Valley, S.C. Levi’s 2017 monograph,
clearly argued for the ‘connected history’ of the Kokand Khanate by using descriptive his-
torical sources. Concerning the topic of this article, he analysed the relation between the
irrigation policy of the Khanate and the management of migrants and nomads (217–220).
He mainly refers to the largest canals in the Ferghana Valley around Namangan and
Andijan to clarify the development of agriculture under the rule of the khanate. This
study will focus on another oasis, including the capital city of the khanate, and will use stat-
istical data to illustrate the historical process of oasis evolution.

Brief history of the Kokand oasis


Although some archaeological ruins have been found near Kokand City, the Kokand oasis
is a relatively recently developed feature. Arab geographers in the tenth century, such as
Istakhri and Ibn Hawkar, described Kokand City (Khwaqant) in their works (Jo’raboev 2005,
205–206). Babur in the late fifteenth century, however, did not include Kokand among the
seven main cities of the Ferghana region (Andijan, Osh, Marg’iran, Isfara, Khujand, Kaniba-
dam and Kasan) (Babur-Nama I 1995, 37, 155). E. Mano pointed out that Kokand was not a
large city at that time (Babur-Nama III 1998, 52).
The Ming tribe of the Uzbeks established the Kokand Khanate in the early eighteenth
century. As capital of the khanate, Kokand expanded dramatically in the 1740s, and its pos-
ition as the capital had not changed by the end of the khanate. The khanate unified the
Ferghana Valley by the end of the eighteenth century and extended its power to Tashkent
and Turkistan in the steppe region in the early nineteenth century. In the mid-nineteenth
century, the power of the khanate declined, and the territory outside of the Ferghana
Valley was lost. The khanate was a Russian protectorate between 1868 and 1876. After
a period of prolonged political disorder, the Russian Empire annexed it in 1876, and the
city of Kokand became one of the five district (uezd) cities in Ferghana Province.1 The
administrative centre of the Ferghana region was transferred to the newly constructed
town of Novyi Margelan. After the Russian Revolution, the Russian Civil War created
severe famine throughout the Ferghana region. The Soviet regime eventually reconquered
most of Central Asia, and the Uzbek Soviet Socialist Republic, which included the Kokand
oasis, was formed in 1924. Under Soviet rule, large irrigation infrastructure projects such as
the Great Ferghana Canal were constructed (Teichmann 2007, 511–512), and the Ferghana
Valley became one of the most intensively developed cotton-producing areas in the world.

Ethnic distribution
Table 1 shows the outline of the ‘ethnic’ composition of the Kokand District recorded in the
‘List of Population Points of Ferghana Province’ (Spisok 1909). The Kokand District includes
population in the mountain cantons (volost’) and the Kanibadam oasis (Isfairam River
oasis), while the Kokand oasis excludes them.
This section attempts to reconstruct the ‘ethnic’ distribution in the Kokand oasis in the
1909 list. It compares the village map with the 1926 census data, published by Sh.I. Inogamov
in 1948, to the 1909 list, because the list does not include a village map. The 1926 census
514 A. UEDA

Table 1. The ‘ethnic’ composition of the Kokand District and the Kokand oasis in the ‘List of Population
Points of Ferghana Province’ in 1909.
Sart Uzbek Kyrgyz Tajik Karakalpak Kashgari Kipchak Others Total
Kokand Canton 257,277 107,455 14,888 53,151 10,477 0 573 4,409 448,320
57.4% 24.0% 3.3% 11.9% 2.3% 0.0% 0.1% 1.0% 100.0%
Kokand Oasis 254,194 80,430 1,587 5,330 7,921 0 573 2,512 352,547
72.1% 22.8% 0.5% 1.5% 2.2% 0.0% 0.2% 0.7% 100.0%
Source: Spisok naselennykh mest Ferganskoi oblasti [1909].

map of this area includes 569 settled points, including railway stations, while the list
recorded only villages. And some villages may have vanished, due to war and famine,
while new ones came into being because of the return and resettlement of refugees.
In sum, 83% (472) of the villages in 1926 were also recorded in 1909. According to the
1926 census, these 472 villages included 89% of the total households of 569 settled
points.
The number of missing settled points and population is not small, but we consider that
this absence is unlikely to skew the sample. First, the 97 points that we could not find in
the 1909 list are scattered throughout the whole area of the Kokand oasis. Second, the 472
villages we found had almost the same ‘ethnic’ composition as all 569 settled points. In the
1926 census, both groups were made up of 87–88% ‘Uzbek’ households, 10% ‘Tajik’, and
1% each of ‘Kirgiz’, ‘Karakalpak’ and others. In sum, the 472 samples are neither geographi-
cally nor ‘ethnically’ biased.
The 1909 list recorded only the ‘predominant’ group in each village. In this case, ‘pre-
dominant’ means the group comprising the relative majority, even if the ‘predominant’
group represented less than half of the village population. Nevertheless, the GIS map
formed of 472 samples shows the general tendencies of ‘ethnic’ distribution (Figure 1).2

Figure 1. The distribution of ‘ethnicities’ in the Kokand oasis in 1909. Source: Spisok naselennykh mest
Ferganskoi oblasti [1909]; Sh.I. Inogamov [1948] Etnicheskii sostav naseleniya i etnograficheskaya karta
Ferganskoi doliny v granitsakh Uzbekskoi SSR.
CENTRAL ASIAN SURVEY 515

Above all, Figure 1 shows the various ‘ethnicities’ concentrated in each part of the oasis. For
example, the centre of the oasis was made up of ‘Sart’-dominated villages, while the western
region was occupied mainly by ‘Uzbek’-dominated villages. The ‘Karakalpak’-dominated
villages were at the northern end. Small groups of ‘Kyrgyz’- and ‘Tajik’-dominated villages
were located on the alluvial fan in the southern part.
Previous studies of Central Asian history and ethnography show when each ethnic
group migrated to the Ferghana Valley. By comparing the ethnic distribution on the GIS
map with the chronological information, this study attempts to recognize the demo-
graphic development and spatial expansion of the oasis and clarify some ethnicity-
related problems in the statistics.

Sart
In this section, I will argue that the ‘Sarts’ included descendants of the indigenous group
from the Kokand oasis, but various other populations also formed part of this group. ‘Sart’
is known to be one of the most problematic historical terms of Central Asia. The Sarts were
one of the Central Asian ‘ethnic’ groups in Russian colonial statistics and were mostly inte-
grated into the Uzbek nation after 1924.
‘Sart’ originally meant ‘trader’, but Turkic and Mongol nomads used this term to indicate
Iranian-speakers in Central Asia. In the sixteenth century, the Uzbek conquerors con-
sidered the entire settled population to be Sarts (Bartol’d 1964, 527–529). After the
annexation of Turkistan, the Russian authorities used this term as an ‘ethnic’ category
suggesting a Turkic-speaking settled population. Prominent Russian orientalists worked
to create a definition of ‘Sart’. The famous orientalist and educator N.P. Ostroumov, who
studied under N.I. Il’minskii at the Kazan Theological Seminary, argued that the Sarts
had inherent ethnic characteristics and their own language. V.V. Bartol’d more cautiously
agreed that Sarts could be considered a distinct ethnic group only in the contemporary
context. On the other hand, the eminent Kazakh intellectual S. Lapin, who worked for
the colonial authorities at Samarkand, wrote that the so-called Sarts were merely a
mixture of the Uzbek and Tajik and had neither inherent ethnic characteristics nor their
own language (Abashin, Arapov, and Bekmakhanova 2008, 266–269).
Neither orientalists nor Russian government officials had precise criteria to distinguish
the Sarts from the Uzbeks. As Abashin (2012) pointed out, the lists of ethnicities in each
province of the Russian Imperial Governor-Generalship differed, and moreover, lists
created by the same section sometimes changed.3 In fact, both identities could have coex-
isted in the minds of natives, because the Sart identity meant a settled lifestyle, as opposed
to a nomadic one, and the Uzbek identity was tied to tribal lineage (Brower 1997, 128–
129).
By the 1920s, many scholars and officials, including the above-mentioned ones, had
taken part in the Sart argument and had argued about this problem according to
various meanings and nuances of the term in each region and period. For example, in
the colonial period, indigenous usages of the term varied among oases in Russian Turke-
stan (Bartol’d 1964, 527–529).4 Concerning ethnicity in the Kokand oasis, V.P. Nalivkin’s
(2011) contemporary description is useful. He was the compiler of a Sart-language diction-
ary and the author of a prominent ethnography on the Sarts of the Ferghana Valley. He
defined the Sarts as Muslims who settled in cities and villages and were descendants of
516 A. UEDA

the Turkic nomads and Tajik aborigines. Concerning their language, he stated that the des-
cendants of the Tajiks were also now Turkic-speaking (5).
The present village-level GIS study will reconstruct the spatial distribution of ‘Sart’-
dominated villages in the oasis and confirm their various origins. Komatsu and Goto
(2009, 105) estimated that the reconstruction of the distribution of the ‘Sarts’ by GIS
might identify the oldest settled area in the Ferghana region, but they could not prove
it, because the 14 spatial units in the Valley based on uchastoks were too large. The
present study, based on 472 villages in the Kokand oasis, argues that Komatsu and
Goto’s hypothesis is highly probable because the area where the ‘Sart’-dominated villages
are concentrated encircled the city of Kokand mentioned by Arab geographers in the
tenth century and Babur in the fifteenth century. In 1974–1975, Soviet archaeologists sur-
veyed the alluvial fan of the Sokh River to make a map of the ruins. The archaeological map
suggests that the oldest populated places in the first millennium were concentrated on
the eastern side of the Sokh River, an area suited to agriculture (Gorbunova and Shigin
1976, 526). The ‘Sart’-dominant area in the 1909 list included this area too. The southern
edge of the area where the ‘Sart’-dominant villages are concentrated spread along the
northern edge of the alluvial fan. From a geographical perspective, constructing irrigation
in this area, where the underground springs of the Sokh River appear, was easier than in
neighbouring lands. Thus, it is natural that the first irrigation work on the Sokh River was
conducted in this area.5
This information supports Komatsu and Goto’s hypothesis, but other information
suggests that various ethnic groups migrated to the ‘Sart’-dominant area on the GIS
map even under the reign of the Kokand Khanate. The ruler of Kokand in the early nine-
teenth century, Alim Khan (r. 1798–1810), assembled a new army of musketeers from thou-
sands of Ghalcha (Tajik mountain people). They immigrated to the Ferghana Valley from
mountain areas (Gubaeva 1983, 78; Kislyakov 1954, 50–51; Levi 2017, 82–86). Within the
area where the Sart-dominated villages were concentrated, two villages were named
after Ghalcha; however, dominant residents there were recorded as Sart in 1909 (Spisok
1909, 64–65). The case of Ghalcha is only one example that suggests the varied origins
of the Sart-dominated villages. This article will show other examples when discussing
the Uzbek and the Kashgaris. In sum, this section suggests that the population named
Sart occupied the centre of the oasis, and that this area included the oldest settled area
of the oasis, established in the first millennium. It also argues that some ancestors of
the residents of the Sart-dominated villages immigrated there in a later period, for
example, in the nineteenth century.

Uzbek
Earlier studies suggest that the Uzbek tribes migrated to the Kokand oasis in the late
seventeenth and eighteenth centuries (Gubaeva 1991, 33). Especially in the eighteenth
century, large numbers of the Uzbek population migrated to the Ferghana Valley from
Samarkand due to political disorder in the Emirate of Bukhara (Nalivkin 1886, 61).
S.C. Levi (2017, 122) argued that some irrigation construction under the Kokand
Khanate was intended to manage conflict among an ethnically diverse population. V.S.
Batrakov (1955, 177) argued that Uzbek immigrants settled peacefully in uninhabited
areas and did not infringe the vested interests of the indigenous population. Komatsu
CENTRAL ASIAN SURVEY 517

and Goto (2004, 113) pointed out that the Uzbek population in Ferghana Province
recorded in the 1897 census and the 1909 list was concentrated in the Kokand District.
They presented an open question: Did the Uzbek tribes concentrate around the capital
of the Khanate to hold on to political power,6 or did the Kokandis prefer to profess to
be Uzbeks, using the name of a ruling ethnic group? This study examines how their argu-
ments were connected with the spatial structure of the oasis.
The GIS map suggests that the Uzbek-dominated villages concentrated mainly in the
western and eastern peripheries of the oasis. These areas were irrigated by the
Begobat, Islam, Alatai and Turgai canals, among others, in the west and the Rishtan and
other canals in the east.
The Kokand Khanate devoted substantial efforts to constructing new irrigation canals
throughout its territory. Among these canals, the Yangi aryk and Uch-Kurgan aryk in the
Namangan District, the Sharikhan sai in the Margilan District, the Ulug-nahr in the
Andijan District (above, in the Ferghana Valley) and the Khan aryk in the Tashkent
region were well known. Bartol’d (1963b, 289–290) stressed the continuous irrigation con-
struction during the time of the Kokand Khanate (see also Ivanov 1958, 182; Tolstova 1959,
35). Irrigation systems in the Kokand oasis were also extended in the era of the khanate.
For example, Irdana Biy (r. 1751–52, 1753–69), Narbuta Biy (r. 1770–99) and Muhammad Ali
Khan (r. 1822–42) constructed canals from the Sokh River (Levi 2017, 39, 49, 87; Nabiev
1973, 165, 178).
Two villages named Khudayar Khan and Malla Khan (the names of the last rulers of the
khanate) were located on the eastern edge of the ‘Uzbek’-concentrated area of the
western region, alongside the sparsely populated area. Both villages were well-irrigated
by canal systems, which suggests that the spatial expansion of the Kokand oasis continued
until the last days of the khanate, especially in the Uzbek-dominated area (Materialy IV
1912, 78; Spisok 1909, 48).
Concerning Levi and Batrakov’s arguments, the GIS analysis indicates that canal con-
struction in the ‘Uzbek’-dominated areas of the oasis might have promoted ‘peaceful’
sedentarization of the Uzbek tribes living right beside the existing oasis. In contrast, the
example of the Kipchaks shows that the violent interventions of a nomadic group in
the oasis could cause serious problems. Concerning Komatsu and Goto’s arguments, the
clearly separated ethnic distribution shown in Figure 1 suggests that the former hypoth-
esis – that the Uzbek tribes were concentrated around the capital of the khanate – is more
likely.
Figure 1 suggests that some Uzbek-origin villages in certain parts of the oasis were not
recorded as being Uzbek-dominated in the 1909 list. R.N. Nabiev (1973, 165–166), quoting
the Ta’rikh-i Shahrukhi, presented a series of villages where the Ming tribe migrated before
the establishment of the Kokand Khanate. I found five of them (Targava, Pilla-khana, Tukai-
tepe, Pirtak and Tepe-kurgan) in the 1909 list. They are all in the central Sart-dominated
area, and the predominant residents of the five villages were recorded as Sarts.
In sum, the Uzbek tribes in the western region of the oasis were spatially concentrated
and recorded as Uzbek-dominated in 1909. On the other hand, the Uzbek-origin villages
surrounded by existing villages were recorded as Sart-dominated. This may have occurred
because the Ming tribe had become the minority in each village by 1909 or because they
changed their own identity. It is impossible for us to confirm which reason is more likely
because the Sart category vanished in the 1926 census, which recorded not only the
518 A. UEDA

dominant groups of the villages but also the populations of each group in the villages. Of
course, it is necessary to consider that the villages in the central area where the Uzbeks
migrated to before the Kokand Khanate was established were probably older than the
Uzbek-dominated villages in the western region. The relatively high proportion of
Uzbeks in the Kokand district in the 1909 list was even lower than the proportion of
people living in Uzbek-origin villages, because some of them were recorded as ‘Sart’-domi-
nated, and Uzbek-dominated villages were recorded only in the area of concentrated
Uzbek population.

Karakalpak
L.S. Tolstova noted that the oldest record of the Karakalpaks in Ferghana dates back to
1704. In 1723, when the Dzungar tribes attacked the steppe region, some groups of the
Upper (verkhnii) Karakalpaks around the city of Turkestan fled along the Syr River and
entered the Ferghana Valley. In the 1740s, when the Kazakh Khan Shygai ruled the Naman-
gan region (the northern part of the Ferghana Valley), some Karakalpaks migrated with
him and remained in the valley (Kozybaev 2000, 278–279; Tolstova 1959, 7, 17–20).
Although these records (or legends) support the supposition that migration occurred in
the early eighteenth century, the ancestors of the Karakalpak tribe in the Ferghana
Valley arrived mainly in the late eighteenth and early nineteenth centuries. Based on
local legends and archival documents, Tolstova argued that the fourth quarter of the
eighteenth century, under the reign of Narbuta Biy, was the peak of the Karakalpak
migration. In addition to the voluntary migration, some Karakalpaks were forced to
migrate as captives. For example, Alim Khan forced many Karakalpaks around Tashkent
to migrate to the Ferghana Valley (Tolstova 1959, 20–22).
The GIS map shows that the Karakalpak-dominated villages were concentrated in the
northern periphery of the Kokand oasis – to be precise, in a T-shaped region consisting
of an east–west line along the southern side of the Syr River and a north–south line
extending towards the centre of the oasis. The former area was not irrigated, while the
latter was irrigated by canals from the Sokh River, including the Karakalpak canal, Ak-
Tepe-sai and Kiyari-sai (Tolstova 1959, 33). The ‘Karakalpak’-dominant belt along the Syr
River may have originated with an influx of refugees in 1723 fleeing the Dzungar invasion.
It is not clear whether the Karakalpak group that migrated to the valley in the 1740s
reached the southern side of the Syr River.

Tajik and Kyrgyz


There were some famous Tajik settlements in the Ferghana Valley. Among them, Rishtan,
in the eastern region of the Kokand oasis, was one of the largest and best-known for its
long history. In the tenth century, Istakhri recorded that Rishtan was a large settlement
with two gates, a bazaar, a mosque and a square (Bartol’d 1963a, 214). On the other
hand, there is no information concerning the evolution of the small ‘Tajik’ and ‘Kyrgyz’
settlements on the alluvial fan. Besides that, it is noteworthy that the ‘Kyrgyz’ distribution
in 1909 might not stem from the distant past. Some sources suggest that more of the
Kyrgyz population lived in the oasis by 1862, when many Kyrgyz and Kipchak were
killed in the civil war (Nabiev 1973, 61–62).
CENTRAL ASIAN SURVEY 519

GIS data confirm that Rishtan was the largest of the ‘Tajik’ settlements in the oasis. Con-
cerning the small ‘Tajik’ and ‘Kyrgyz’ settlements on the alluvial fan, it is noteworthy that
irrigation construction on the alluvial fan was quite difficult due to conglomerate clay.
Even Soviet engineers in the 1970s evaluated the area where the ‘Kyrgyz’ had settled in
1909 as ‘not suited for irrigation’ (Ben’yaminovich and Tersitskii 1975: Ferganskaya
dolina fizicheskaya [map], Ferganskaya dolina, Srednee techenie i Chirchik-Akhangaran-
Keles [map]). Nevertheless, the ‘Kyrgyz’ population settled in the middle of the fan. The
reason for this distribution will be analyzed in the next section. The ethnic map of
1951–1954 also shows that the distribution of the Kyrgyz precisely coincided with the dis-
tribution shown in Figure 1 (Vinnikov 1959). This means that the ‘ethnic’ distribution of
1909 was still in place in the 1950s.

Kipchak and Kashgari


In addition to the five groups mentioned above, it is worthwhile discussing two unrec-
orded groups: the Kipchaks and Kashgaris/Uyghurs. In 1844, the Kipchaks, headed by
Musulmankul, seized power in the Kokand Khanate when they supported the enthrone-
ment of Khudayar Khan. Many Kipchaks entered the oasis and seized a large portion of
farmland and irrigation water from the Sarts. But they were massacred by the Sart popu-
lation and Khudayar Khan’s army in 1852, and the Kipchak survivors fled to the northern
piedmont (Levi 2017, 164–168; Nalivkin 1886, 163–180). After the Kipchak genocide of
1852, Khudayar Khan was forced to seek refuge in the Bukhara Khanate in 1858
because of the attack by his brother, who had allied with the Kipchaks and Kyrgyz.
After a series of civil wars, Khudayar Khan sat on the throne of the Kokand Khanate
again in 1865, and the relationship between Khudayar Khan and the Kipchak nobles
improved (Levi 2017, 189–190, 204).
The GIS map of the 1909 list confirms Nalivkin’s description. While the Kipchak-domi-
nated villages became concentrated in the northeast cantons of the Ferghana region
(outside of Figure 1), there was no predominantly Kipchak settlement within the
Kokand oasis. Figure 1 also suggests that the Kipchaks did not return to the Kokand
oasis even when ethnic tensions were reduced after 1865. The number of ‘Kipchaks’ in Fer-
ghana Province in 1914 was 68,193, but the number in the Kokand district was fewer than
1% (648 people) of the whole ‘Kipchak’ population in Ferghana Province (Statisticheskii
1917, 3).7 The only village named after the Kipchaks in the Kokand oasis was registered
as an Uzbek-dominated village in the 1909 list (Spisok 1909, 40).
The Kashgari population migrated to the region in a wave originating from Xinjiang.
Figure 1 does not include any Kashgari-dominated villages, but some village names
reflect the population’s origins in Xinjiang, for example Kashgar and Taglyk. Despite the
Xinjiang-related village names, the ‘ethnic’ majority in these villages was recorded as
Sart. This indicates that the migrants rapidly assimilated into the surrounding population
or that there were few, if any, differences between the Muslim populations of Ferghana
and Xinjiang. The latter hypothesis would not be surprising, because the languages of
both populations are nearly the same, and the Kashgaris – that is, the Muslim, settled
population in Xinjiang – were sometimes called Sart or Kashgar Sart.8
In short, the ‘ethnic’ distribution and periodic information for each group suggest that
the oasis originated in the central area where canal construction was easiest, and that later
520 A. UEDA

semi-nomadic tribes settled in the peripheral area, where the conditions for irrigation were
less favourable. The geographical expansion of the oasis was closely related to the
migration of nomadic tribes.
Some immigrants settled not only in the peripheral belt but also in the central area of
the oasis. Almost all villages of immigrants in the central area, for example, the Uzbeks,
Kashgaris and Tajiks, were recorded as Sart-dominated in the 1909 list, even though his-
torical documents and village names preserved their origins. This means that the
mosaic structure of ‘ethnicities’ indicated in Figure 1 did not accurately show ethnic
origin. Isolated villages of migrants were easily assimilated by surrounding Sart villages
because, as seen above, ‘Sart’ did not refer to a specific tribal or geographical origin but
to a settled lifestyle. In areas where tens of villages of immigrants were concentrated,
people’s original identities were usually preserved. Such areas were naturally located on
the periphery of the oasis. Through these processes, the clear division of ‘ethnicities’
appeared when local officials interviewed residents. But small islands of ethnicity with
one or a few villages became invisible in the ‘ethnic’ distribution based on the 1909 list.
From this example, this study presents an abstract model of oasis evolution on an allu-
vial fan (Figure 2). It suggests that the original oasis was generally formed in the centre,
down from the edge of the fan, where irrigation conditions were best (I in Figure 2).
Next, a group of migrants formed settlements near the original oasis and constructed irri-
gation canals there, either on their own or with government assistance (II and III). This
process was repeated many times throughout the centuries (II′ , III′ , II′′ , and III′′ ). Finally, a
mosaic of ethnicities was formed in the oasis. On the alluvial fan itself, small settlements
of semi-nomadic people were formed.

Agricultural development
The next issue we consider is how the spatial structure of the oasis affected the develop-
ment of cotton planting. To answer this question, this study prepared four village-level GIS
maps. These maps show the average household (dvor) size in 1909 (Figure 3), the distri-
bution of cultivated areas in 1899–1902 (Figure 4), the ratio of cotton to the total area

Figure 2. A conceptual model of oasis evolution on an alluvial fan. Source: the author.
CENTRAL ASIAN SURVEY 521

Figure 3. Average household size in the Kokand oasis in 1909. Source: Same as Figure 1.

planted in 1899–1902 (Figure 5), and the amount of cotton planted per capita in 1899–
1902 (Figure 6).
Figure 3 relates the 1909 population data to the village list in Figure 1. Figures 4, 5, and 6
were compiled from the land-tax data. The Russian Empire collected 10% of the supposed
harvest value as ‘land tax’ from the farmlands of the settled population of Central Asia
(Materialy I, 1897, 4–5).9 Land tax was introduced in Ferghana Province in 1889 and occu-
pied the largest part of the budget revenues in the province. For example, in 1890, the land

Figure 4. Distribution of cultivated areas on the basis of data from 1899 to 1902, village by village.
Source: Materialy dlya statisticheskogo opisaniya Ferganskoi oblasti, rezul’taty nozemel’no-podatnykh
rabot, vup. IV, Kokandskii uezd [1912]; Sh.I. Inogamov [1948] Etnicheskii sostav naseleniya i etnografiches-
kaya karta Ferganskoi doliny v granitsakh Uzbekskoi SSR.
522 A. UEDA

Figure 5. Ratio of cotton planting to the total area planted, surveyed from 1899 to 1902. Source: Same
as Figure 4.

tax provided 90% of the budget revenues in the province, but the authorities did not have
precise data about land use, which was essential for accurate tax collection (Obzor 1893,
45–50). The land survey in Ferghana Province started in 1890 in the Andijan District.
Because of a limited number of surveyors, the survey in the Kokand District started only
in 1899 and finished in 1902. This survey was the first village-level land survey in the
Kokand oasis (Materialy I, 1897, 1, 31; IV, 1912, 15).

Figure 6. Cotton planting per capita, surveyed from 1899 to 1902. Source: Spisok naselennykh mest
Ferganskoi oblasti [1909]; Materialy dlya statisticheskogo opisaniya Ferganskoi oblasti, rezul’taty noze-
mel’no-podatnykh rabot, vup. IV, Kokandskii uezd [1912]; Sh.I. Inogamov [1948] Etnicheskii sostav nase-
leniya i etnograficheskaya karta Ferganskoi doliny v granitsakh Uzbekskoi SSR.
CENTRAL ASIAN SURVEY 523

The published report of the survey included only data on village-level planting on irri-
gated farmland, while maintaining the total amount of rain-fed planting at the canton
level. The entire amount of rain-fed cultivation in the Kokand district was only 2,496 desya-
tinas, in contrast with 148,684 desyatinas of irrigated planting (Materialy IV 1912, 204).10
The survey in the Kokand District was conducted in the middle of drastic social changes
in Russian Central Asia. Raw cotton exports from the Ferghana Valley were rapidly expand-
ing due to railway construction between Russia and Central Asia. In 1899, the Samarkand–
Andijan Railway was connected to the Trans-Caspian Railway. This meant that a direct rail
link from the Ferghana Valley to the port of Krasnovodsk on the Caspian Sea was estab-
lished. Raw cotton exports from Central Asia rapidly increased, as mentioned above.
Cotton-planted areas also expanded. The authorities of Ferghana Province estimated
cotton-planted areas in 1890 at 56,000 desyatinas; later, in 1902, they were estimated to
have reached 159,000 desyatinas (Aziatskaya Rossiya II 1914, 91; Obzor 1893, 8).
Rapid expansion of the cotton economy influenced various aspects of local society.
First, the growing cotton economy eroded food production in the Ferghana Valley. The
export of raw cotton caused an enormous influx of money, pushing up prices (Bartol’d
1963b, 290). Grain was imported into the valley from other parts of the Russian Empire,
but it could not stop the sharp rise in food prices. For example, the report of the land-
tax survey recorded the rise of wheat price from 0.23 rubles/pud in 1889 to 1.24 rubles/
pud in 1893 (Materialy I, 1897, 41).
The rapid changes in the economy and the development of usury in cotton areas desta-
bilized the social order. In 1910, the authorities understood that the emergence of landless
people related to the expansion of cotton cultivation was one of the causes of an increase
in crime (Statisticheskii 1912, 158–159). Public unrest in the Ferghana Valley had already
been observed in the 1890s. In 1898, the year before the land survey in the Kokand District
began, the Andijan uprising broke out. The rebels consisted of three groups. The first and
second groups were Kyrgyz nomads and Turkic peasants. The third group consisted of day
labourers, beggars, vagrants, etc. Many people of the third group were permanently sup-
ported by the Dukchi Ishan, the leader of the uprising (Andizhanskoe vosstanie 1898g 1938,
168). In 1894, local government officials and native influences entrusted the Dukchi Ishan
with the authority to maintain the social order, because it was severely disturbed under
Russian rule. Dukchi Ishan exercised this power by appointing his agents to uphold
Islamic law. By the time of the uprising, the Russian authorities and the Ishan’s jurisdiction
coexisted in the eastern part of the Ferghana Valley (Komatsu 1986, 596–598). Although
the Andijan uprising was quickly suppressed by the Russian authorities, its influence on
the local society and colonial authorities of Russian Central Asia persisted until the end
of Russian Imperial rule (Morrison 2012, 299–304).
This study has attempted to relate the village list of the tax survey to the village map of
the 1926 census. As a result, 346 villages were matched between the list and the map.
Figures 4, 5, and 6 were edited based on this data set.
Figure 3 shows that the distribution of average household size in the oasis was not
even. First, in the central area of the oasis, the average household in villages east of
Kokand City was apparently larger than in the area west of Kokand City. Second, the
eastern and western regions and the alluvial fan area had relatively average household
sizes. Third, the household size at the north end of the oasis along the Syr River was
obviously small. Comparing Figures 1 and 3, it is difficult to confirm any relation
524 A. UEDA

between ‘ethnicity’ and household size.11 This study will return to household size distri-
bution after the analysis of cotton cultivation.
I divided the Kokand oasis into five areas (the central, eastern, and western regions, the
fan, and the north end) in Figure 3 and will use this division below. The fan and the north
end were obviously divided from the main body of the oasis. The border between the
western and central regions nearly coincides with the Uzbek and Sart border shown in
Figure 1. The border between the central and eastern regions runs through the less popu-
lated zone and the Uzbek cluster in the northeast.
Figure 4 suggests the cultivation amounts of the three major crops, while Figure 5 dis-
plays the ratio of cotton planting to the total amount of planted crops. The eastern region
was the area with the most intensive cotton cultivation in terms of both the amount and
ratio, while rice and wheat planting were uncommon. In the central area, crop composition
was more diverse, and a combination of wheat and cotton was common. Some villages in
the southern part of the central area grew rice as well.
The western region also had a combination of several crops. In comparison with the
central area and the eastern region, the western region can be characterized by a large
amount and a high ratio of rice cultivation. The population of the western region was
Uzbek-dominated.
In the central and eastern regions of the Kokand oasis, villages that had planted rice in
the Khanate era, including Gumayli, Dangara, Kirklar and Amirabad, planted cotton and
wheat in the colonial era (Materialy IV 1912, 22, 45; Nabiev 1973, 173). It is said that rice
planting requires more irrigation water than cotton. In the Aim Canton, in the eastern
part of the Ferghana Valley, for example, rapid expansion of cotton planting excluded
rice planting in the limited irrigated farmland during Russian rule (Penati 2014, 199). In
short, rice cultivation had spread even in the central and eastern regions of the oasis in
the Khanate era, but later it shrank and was concentrated in the western region. An exam-
ination of other parts of the valley suggests that the cotton boom and limited water
resources could have reduced rice production.12
The reason for a specific group to engage in cultivating rice during a cotton boom is
worth further consideration. Nalivkin (1886, 20) suggested that the Kokand oasis was
developed by land reclamation over 150 years. In general, rice cultivation tends to have
advanced towards the humid marginal areas of Central Asia, as demonstrated by
B. Penati’s (2010, 68–69) example in the colonial period. This reinforces the notion that
the conversion from a wetlands frontier to rice farmland was done by the Uzbek popu-
lation, who migrated there in the eighteenth century. In the preceding section, I argued
that the Uzbek migration to the western region occurred in relatively newly developed
settlements where irrigation conditions were inferior to those of the central area. Here, I
suggest that the western region included the frontier of agricultural development as
humid land.
The fan area was quite different from the three areas described above. Three settle-
ments can be identified in the middle of the fan in Figure 4. These settlements cultivated
neither rice nor cotton; wheat represented the largest portion of their crops. This situation
can be explained by inferior irrigation conditions. Two-thirds of the settlements were
‘Kyrgyz’-dominant, and the ratio of wheat planting was especially high. In the ‘Kyrgyz’
settlements in the fan, the agricultural pattern was like that of the Kyrgyz semi-nomads
in the piedmont. It mainly consisted of wheat. Semi-nomadic Kyrgyz preferred spring
CENTRAL ASIAN SURVEY 525

wheat on both irrigated and non-irrigated land in the winter quarters, because they
moved to mountain pasture land in the summer with their livestock. If Kyrgyz living in
the fan maintained mobile pastoralism, they might have preferred spring wheat over
autumn wheat, which occupied the main part of wheat planting by the settled population
on irrigated farmland (Ueda 2013, 37).
In considering the crop pattern shown in Figure 4, the north end should be divided into
two sub-areas, eastern and western. The agriculture of the western half of the north end is
like the central area’s crop composition of cotton and wheat. Figure 1 shows that both
areas were Sart-dominated. On the other hand, the eastern half of the north end had
the highest ratio of cotton planting in this map, while the amount was smaller than in
the villages of the central area and the eastern and western regions. The villages there
were the Karakalpak-dominated. The land tax report explained that the villages in this
area could plant cotton with plentiful groundwater without requiring canal irrigation.
The authorities applied land tax rates equal to canal-irrigated villages to these villages
because of their high production rates (Materialy IV 1912, 60–63). I was not able to identify
the reason the Karakalpaks were able to occupy such fertile land. Tolstova (1959, 17–18),
however, suggested that migration along the Syr River was the early phase of the Karakal-
pak migration to Ferghana that began in 1723. In other words, the formation of the Kar-
akalpak villages along the Syr River took place earlier than the rapid expansion of the
Kokand oasis in the 1740s.
Figure 5 summarizes some of the previous statements concerning cotton cultivation.
The most monocultural cotton planting areas were in the eastern region and the
eastern half of the north end of the valley.13 The land tax survey did not record any
cotton cultivation in the middle of the alluvial fan. Other areas (the central area,
western region, and the western half of the north end) had varied crops that included
cotton, wheat, and rice, among others. Among these areas, the western region’s Uzbek-
dominated villages can be characterized by rice cultivation. The average ratios of cotton
planting in all cultivated areas were 34% in the western region, 30.8% in the central
area, 42.7% in the eastern region, 18% in the fan, and 46.3% in the north end.
Figure 6 shows cotton planting per capita. This study focuses on two cotton-growing
areas: the eastern region and the eastern half of the north end. The former area obviously
had less cotton planting per capita than the latter.14 That difference suggests that there
were two types of cotton planting. Cotton in the eastern region was canal-irrigated,
with large areas under cultivation and labour-intensive cotton production by the Sart
farmers with relatively large households. Cotton cultivation in the north end was ground-
water-based, with a limited amount of planting and extensive cotton planting by the Kar-
akalpak farmers with small households.15
Figure 6 suggests that it is difficult to identify clear differences between the eastern
region, the central area and the western region. This suggests that the villages in the
central area and the western region had potential labour forces within villages for the
peak of cotton planting (sowing in spring, okuchka – heaping earth around the base of
the plants – in summer, picking in autumn). In the colonial era, the number of landless
labourers increased, and the cotton farming in the Ferghana Valley required many seaso-
nal labourers (Statisticheskii 1912, 97–98, 158–159). This means villages in the central and
western regions could provide seasonal labour in cotton fields in the eastern region.
526 A. UEDA

This study has discussed the economic activities of each ‘ethnicity’-dominant area, but
of course, the oasis served as an economic unit connected by several networks. First, the
weekly bazaar network connected each sub-area. The weekly bazaar network also con-
nected the Kokand oasis with the neighbouring oases. The physical area of the weekly
bazaar system extended over 70 km from east to west. Agricultural products were circu-
lated by native draymen (arbakesh) to the villages and cities (Sovremennyi kishlak 1927,
130–139).

Conclusions
The first part of this article illustrated that the expansion of the Kokand oasis was closely
connected to the migration waves of nomads. This is, of course, somewhat common-sense
in Central Asian history, but the early village-level statistics provide quantitative and spatial
evidence for this well-known argument. Furthermore, the example of the Kokand oasis
shows that the nomadic tribes generally settled in the outer peripheries of the existing
oasis, although some settled in the heart of existing oasis. Following these migration pro-
cesses, the colonial authorities recorded the mosaic-like ethnic distribution. In the colonial
survey, the migrants’ villages settled in the existing oasis were often recorded as Sart-
dominated, and therefore each ‘ethnic’ part on the map is probably more clearly distin-
guished than the actual distribution, but a major pattern remains. This type of ethnic
distribution structure might not be unique to the Kokand oasis. The conceptual image
in Figure 2 explains the historical evolution of oasis settlements in other regions.16
The second part illustrated the variety of oasis agriculture that generally coincided with
the distribution of population groups. In conclusion, I summarize the relationship between
nomadic migration in the precolonial era and cotton agriculture in the Russian colonial era.
With specific reference to this case study of the Kokand oasis, the settled population
referred to as ‘Sart’ played the main role in the development of cotton planting by occu-
pying the best-irrigated farmland. Some of the settled Uzbeks also engaged in cotton
planting around the Sart-dominated area, and other settled Uzbeks engaged in rice plant-
ing in former swamps. Settled Karakalpaks in the periphery used groundwater and
initiated very intensive cotton agriculture without canal irrigation. The semi-nomadic
Kyrgyz were excluded from cotton planting because of the lack of sufficient water
resources; thus, they mainly planted wheat.
This case study analysed two types of cotton-intensive economy. Relatively small plots
of farmland per capita and labour-intensive planting by the residents of the Sart -domi-
nated villages based on canal irrigation developed in the main area of the oasis. On the
other hand, large amounts of farmland per capita and extensive planting by the residents
of the Karakalpak-dominated villages based on fertile groundwater emerged in the periph-
ery. The former was related to relatively large households and the latter to small ones. It is
certain that the former example is more important in explaining the expansion of cotton
planting in Central Asia. The latter unique example, however, shows that there could be
multiple courses of nomad sedentarization.

Notes
1. Concerning the political history of the Kokand Khanate, see Levi (2017).
CENTRAL ASIAN SURVEY 527

2. Figure 1 includes not only the Kokand oasis, irrigated by the Sokh River, but also a neighbour-
ing oasis in the west, irrigated by the Isfara River. This study does not investigate this neigh-
bouring oasis.
3. Concerning the place in the history of Russian official statistics that the 1909 list occupied, see
Abashin (2012).
4. Concerning the ‘Sart’ problem in general, see Abashin (2004).
5. The average yearly rainfall in the Kokand oasis is about 100 mm. Agriculture in this area is
almost impossible without irrigation (Ben’yaminovich and Tersitskii 1975, 44).
6. The Kokand oasis was under the direct control of Kokand Khan, unlike other oases in the
Ferghana Valley (Nabiev 1973, 108).
7. Of course, this number can be also questioned. It can be assumed that the Kipchak population
in the Kokand oasis wanted to hide their origins by identifying themselves as members of
other groups after the Kipchak genocide in the nineteenth century, but we have not yet
found any documents suggesting that this occurred.
8. Concerning the Kashgar/Uyghur population in the Ferghana Valley, see Ueda (2017).
9. Concerning the tax policy of the Russian authorities in Central Asia and the case of the Samar-
kand region, see Morrison (2008, 88–118).
10. An important revision of the land tax system was introduced during the land survey. How to
assess rainfed cultivation in the land tax system was a controversial matter. B. Penati (2010,
66–67, 73–74) pointed out that the tax on rainfed cultivation had not been clearly legislated
by 1900, and the 1900 law concerning tax collection on rainfed cultivation was fully enforced
only after 1908. The revision of the related provision (paragraph 255 of the Turkestan Statute)
in 1900 occurred during the survey in the Kokand District. Of course, the ongoing survey could
not fully respond to the revision.
11. A one-way analysis of variance significant at the 5% level suggested that the differences
among the ethnic groups are not statistically significant.
12. Concerning the rice plantation of Khudayar Khan in Er-mechet’ and the water problem, see
Nabiev (1973, 180–181, 185–187).
13. The optimized Hot Spot Analysis tool in ArcGIS identifies statistically significant spatial clusters
of high values (hot spots) and low values (cold spots). This tool confirmed with 99% certainty
that two hot spots of cotton cultivation ratio existed in the east and the north.
14. Average cotton cultivation per capita was 0.206 desyatina in the western region, 0.345 desya-
tina in the north end, 0.136 desyatina in the central area, 0.167 desyatina in the western region,
and 0.167 desyatina in the fan.
15. Extensive characteristics in the Karakalpaks’ cotton planting is a provisional argument,
because seasonal labour on cotton farmland was quite common. The Karakalpaks, if they
wished, could employ seasonal labourers from neighbouring villages, the piedmont
nomadic area, and even from Chinese Xinjiang.
16. For example, the case of the sedentarization of nomads on the lower Nile Delta in the twen-
tieth century shows the dynamics of oasis expansion on the alluvial fan, like the model in
Figure 2 (Kato, Tsumura, and Iwasaki 2013, 24–31).

Acknowledgments
The author is grateful to the editor, Alexander Morrison, and the anonymous reviewers for their valu-
able comments and comprehensive recommendations. The author is also grateful to Adham Ashirov
and Hisao Komatsu for their research assistance, and to Ichiro Iwasaki and Tomohiko Uyama for their
helpful feedback on earlier versions of this work.

Disclosure statement
No potential conflict of interest was reported by the author.
528 A. UEDA

Funding
This work was supported by Japan Society for the Promotion of Science [grant no. 17J01899].

References
Abashin, S. N. 2004. “Naselenie Ferganskoi doliny: k stanovleniyu etnograficheskoi nomenklatury v
kontse XIX – nachale XX veka.” In Ferganskaya dolina: Etnichnost’, Etnicheskoe protsessy,
Etnicheskie Konflikty, edited by S. N. Abashin, and V. I. Bushkov, 38–101. Moskva: Nauka.
Abashin, S. 2012. “Empire and Demography in Turkestan: Numbers and the Politics of Counting.” In
Asiatic Russia: Imperial Power in Regional and International Contexts, edited by T. Uyama, 129–149.
London and New York: Routledge.
Abashin, S. N., D. Arapov, and N. Bekmakhanova, eds. 2008. Tsentral’naya Aziya v sostave Rossiiskoi
imperii. Moskva: Novoe literaturnoe obozrenie.
Abdurasulov, U. 2016. “The Aral Region and Geopolitical Agenda of the Early Qongrats.” Eurasian
Studies 14: 3–36.
Bartol’d, V. V. 1963a. “Turkestan v epokhy mongol’skogo nashestviya.” In Akademik V.V. Baltol’d
Sochneniya I. Moskva: Vostochnoi Literatury.
Bartol’d, V. V. 1963b. “Istoriya kul’turnoi zhizni Turkestana.” In Akademik V.V. Baltol’d Sochneniya II (1),
167–433. Vostochnoi Literatury: Moskva.
Bartol’d, V. V. 1964. “Sart.” In Akademik V.V. Baltol’d Sochneniya II (2), 527–529. Moskva: Nauka.
Batrakov, V. S. 1955. “Kharakternye cherty sel’skogo khozyaistva Ferganskoi doliny v period
Kokandskogo khanstva.” Trudy Sredneaziatskogo gosudarstvennogo universiteta im. V.N. Lenina,
LXII, 113–139, Tashkent.
Bedrintsev, K. N., and B. D. Korzhavin, eds. 1975. Irrigatsiya Uzbekistana, tom I. Tashkent: Fan.
Ben’yaminovich, E. M., and D. T. Tersitskii, eds. 1975. Irrigatsiya Uzbekistana, tom II. Tashkent: Fan.
Brower, D. 1997. “Islam and Ethnicity: Russian Colonial Policy in Turkestan.” In Russia’s Orient: Imperial
Borderlands and Peoples, 1700–1917, edited by D. Brower, and E. J. Lazzerini, 115–137. Bloomington
and Indianapolis: Indiana University Press.
Dmitrieva, Y. A., A. A. Suchilin, and O. N. Inevatkina. 2012. “Razrabotka struktury arkheologicheskoi
GIS: Kul’turnoe nasledie Zarafshanskoi doliny.” Kratkie soobshcheniya instituta arkheologii,
Rossiiskaya akademiya nauk institut arkheologii 226: 122–133.
Gorbunova, N. G., and A. E. Shigin. 1976. “Raboty Ferganskoi ekspeditsii.” In Arkheologicheskie otkry-
tiya 1975 goda, 526–527. Moskva: Nauka.
Gregory, I. N., and A. Geddes, eds. 2014. Toward Spatial Humanities, Historical GIS & Spatial History.
Bloomington: Indiana University Press.
Gubaeva, S. S. 1983. Etnicheskii sostav naseleniya Fergany v kontse XIX– nachale XX v. po dannym topo-
nimii. Tashkent: Fan.
Gubaeva, S. S. 1991. Naselenie Ferganskoi doliny v kontse XIX – nachale XX v. Tashkent: Fan.
Hermes, T. R., M. D. Frachetti, E. A. Bullion, F. Maksudov, S. Mustafokulov, and C. A. Makarewicz. 2018.
“Urban and Nomadic Isotopic Niches Reveal Dietary Connectivities Along Central Asia’s Silk
Roads.” Nature, Scientific Reports 8 (5177): 1–11. doi:10.1038/s41598-018-22995-2.
Hirsch, F. 2005. Empire of Nations: Ethnographic Knowledge & the Making of the Soviet Union. Ithaca:
Cornell University Press.
Inogamov, Sh. I. 1948. Etnicheskii sostav naseleniya i etnograficheskaya karta Ferganskoi doliny v gran-
itsakh Uzbekskoi SSR. Tashkent: Institut istorii i arkhaeologii akademii nauk UzSSR.
Ivanov, P. P. 1958. Ocherki po istorii Srednei Azii (XVI – seredina XIX v.). Moskva: Izdacheristvo
Vostochnoi literatury.
Jo’raboev, O. 2005. “Qo’qon.” In O’zbekiston milliy entsiklopediyasi, 11-jidd, 205–207. Toshkent:
O’zbekiston milliy entsiklopediyasi’.
Kato, H., H. Tsumura, and E. Iwasaki. 2013. “GIS as a Tool for Researching the Socioeconomic History of
Modern Egypt.” Journal of Asian Network for GIS-Based Historical Studies 1: 22–32.
Kislyakov, N. A. 1954. Ocherki po istorii Karategina. Stalinabad: Tadzhikgosizdat.
CENTRAL ASIAN SURVEY 529

Komatsu, H. 1986. “Andijan houki to Ishan (The Andijan Uprising and Ishan).” Toyoshi-Kenkyu 44 (4):
589–619. (in Japanese).
Komatsu, H., and Y. Goto. 2004. “The Ferghana Project: Central Asian Studies with GIS.” In Islamic Area
Studies with Geographical Information Systems: New Horizons in Islamic Studies, edited by A. Okabe,
103–121. London and New York: Routledge.
Komatsu, H., and Y. Goto. 2009. “Chuuou Azia no doutai wo yomu (To Analyse the Dynamics of
Central Asia).” In Chiiki kenkyuu no tame no GIS (GIS for Area Studies), edited by T. Mizushima,
and M. Shibayama, 95–112. Tokyo: Kokon-Shoin. (in Japanese).
Kozybaev, M. K., ed. 2000. Istoriya Kazakhstana s drevneishikh vremen do nashikh dnei v pyati tomakh,
tom 3. Almaty: Atamura.
Khazanov, A. 1994. Nomads and the Outside World. 2nd ed. Madison: University of Wisconsin Press.
Levi, S. C. 2017. The Rise and Fall of Khoqand: Central Asia in the Global Age, 1709–1876. Pittsburgh:
University of Pittsburgh Press.
Maksudov, F. 2012. “Bati Türkistanda Karahanli kentlerinin gelişimi üzerine (On Development of the
Karakhanid Settlement in Western Turkistan). IIB International Refereed Academic Social Science
Journal International Congress on Culture and Society Special Issue, 2012 3 (5): 339–351. (in Turkish).
Maksudov, F., A. Mosca, B. Rondelli, and S. Stride. 2010. El projecte SilkRoDE: una proposta de gestió de
l’arqueologia a l’Àsia central. Tribuna d’Arqueologia, 2008–2009, 299–317.
Masal’skii, V. I. 1913. Rossiya, polnoe geograficheskoe opisanie, tom 19, Turkestanskii krai. Sankt-
Peterburg: Izdanie A.F. Devriena.
Masanov, N. E. 1995. Kochevaya tsivilizatsiya kazakhov: osnovy zhiznedeyatel’nosti nomadnogo
obshchestva. Almaty: Sotsinvest.
Morrison, A. S. 2008. Russian Rule in Samarkand 1868–1910: A Comparison with British India. Oxford:
Oxford University Press.
Morrison, A. S. 2012. “Sufism, Pan-Islamism and Information Panic: Nil Sergeevich Lykoshin and the
Aftermath of the Andijan Uprising.” Past & Present 214 (1): 255–304.
Nabiev, R. N. 1973. Iz istorii Kokandskogo khanstva: Feodal’noe khozyaistvo Khudoyar-khana. Tashkent:
Fan.
Nalivkin, V. P. 1886. Kratkaya istoriya Kokandskago khanstva. Kazan: Tipografiya Imperatorskago
Universiteta.
Nalivkin, V. P. 2011. Tuzemtsy ranishe i teper’: etnograficheskie ocherki o tyurko-mongol’skom naselenii
Turkestanskogo kraya. Kazan: Imperatorskii Universitet.
Penati, B. 2010. “Swamps, Sorghum and Saxauls: Marginal Lands and the Fate of Russian Turkestan
(c. 1880–1915).” Central Asian Survey 29 (1): 61–78. doi:10.1080/02634931003765548.
Penati, B. 2014. “Life on the Edge: Border-Making and Agrarian Policies in the Aim District (Eastern
Fergana), 1924–1929.” Ab Imperio 2014: 193–230.
Research Center for Silk Roadology, ed. 2003. A Study on Cities, Settlements and Archaeological Sites in
the Silk Road Region Using Satellite Photos, Silk Roadology 17. Nara: Research Center for Silk
Roadology (in Japanese).
Shioya, A. 2014. “Povorot and the Khanate of Khiva: A New Canal and the Birth of Ethnic Conflict in
the Khorazm Oasis, 1870s–1890s.” Central Asian Survey 33 (2): 232–245. doi:10.1080/02634937.
2014.916077.
Teichmann, C. 2007. “Canals, Cotton, and the Limits of the De-Colonization in Soviet Uzbekistan,
1924–1941.” Central Asian Survey 26 (4): 499–519. doi:10.1080/02634930802018240.
Tolstova, L. S. 1959. Karakalpaki Ferganskoi doliny: istoriko-etnograficheskii ocherk. Nukus:
Karakalpakskoe Gosudarstvennoe Izdatel’stvo.
Ueda, A. 2013. “How did the Nomads Act During the 1916 Revolt in Russian Turkistan?” Journal of
Asian Network for GIS-Based Historical Studies 1: 33–44.
Ueda, A. 2017. “Population Change and Migration in the Ferghana Region in the Late 19th and Early
20th Centuries.” In The Family in Central Asia: New Perspectives, edited by S. Roche, 361–377. Berlin:
Klaus Schwarz Verlag.
Vinnikov, V. P. 1959. “Sovremennoe rasselenie narodov i etnograficheskikh grupp v Ferganskoi
doline.” In Sredneaziatcheskii etnograficheskii sbornik II, Trudy instituta etnografii, tom XLVII, 382–
409. Moskva: Izdatel’stvo akademii nauk SSSR.
530 A. UEDA

Xie, Y., J. Gong, P. Sun, and X. Gou. 2014. “Oasis Dynamics Change and its Influence on Landscape
Pattern on Jinta Oasis in Arid China from 1963a to 2010a: Integration of Multi-Source Satellite
Images.” International Journal of Applied Earth Observation and Geoinformation 33: 181–191.

Archive and statistical sources


Andizhanskoe vosstanie 1898g. 1938. Krasnyi Arkhiv 88 (3): 123–181.
Aziatskaya Rossiya, II. 1914. S. Peterburg: Pereselencheskoe upravlenie.
Babur-Nama (Vaqayi’), I. Critical Edition Based on Four Chaghatay Texts with Introduction and Notes by
Eiji Mano. 1995, III. Translation and Annotation with Introduction and Indexes by Eiji Mano. 1998.
Kyoto: Shokado.
Materialy dlya statisticheskogo opisaniya Ferganskoi oblasti, rezul’taty nozemel’no-podatnykh rabot,
vyp. I, Andizhanskii uezd. 1897, vyp. II, Margelanskii uezd. 1899, vyp. III, Namanganskii uezd. 1910,
vup. IV, Kokandskii uezd. 1912, vup. V, Oshskii uezd. 1910. Novyi Margelan i Skobelev: Tipografiya
Ferganskoi oblastnoi pravleniya.
Obzor Ferganskoi oblasti za 1890 god. 1893. Tashkent: Tipografiya Ferganskoi oblastnoi pravleniya.
Sovremennyi kishlak Srednei Azii, vyp. 6, Isfarinskaya volost’ (Ferganskoi oblasti Uzbekskoi SSR). 1927.
Tashkent: Sredneaziatskoe byuro TsK VKP(b).
Spisok naselennykh mest Ferganskoi oblasti. 1909. Skobelev: Ferganskoi oblastnoi statisticheskii
komitet.
Statisticheskii Obzor Ferganskoi oblasti za 1910 god. 1912. Skobelev: Ferganskoi oblastnoi statistiches-
kii komitet.
Statisticheskii Obzor Ferganskoi oblasti za 1914 god. 1917. Skobelev: Ferganskoi oblastnoi statistiches-
kii komitet.

You might also like