ISBN 978-4-902325-16-4
Occasional Paper 3
Linguistics, Archaeology
and
the Human Past
Edited by
Toshiki OSADA and Akinori UESUGI
Indus Project
Research Institute for Humanity and Nature
Kyoto, Japan
2008
CONTENTS
Preface
The spread of textile production and textile crops in India beyond the Harappan zone:
an aspect of the emergence of craft specialization and systematic trade
1
Dorian Q Fuller
Recent archaeological discoveries in Sindh, Pakistan
27
Qasid H. Mallah
Exploration in the Ghaggar Basin and excavations at Girawad, Farmana (Rohtak District) and
Mitathal (Bhiwani District), Haryana, India
77
Vasant Shinde, Toshiki Osada, M.M. Sharma, Akinori Uesugi, Takao Uno,
Hideaki Maemoku, Prabodh Shirvalkar, Shweta Sinha Deshpande,
Amol Kulkarni, Amrita Sarkar, Anjana Reddy, Vinay Rao and Vivek Dangi
Re-evaluating the linguistic prehistory of South Asia
159
Roger Blench
The spread of texitile production and textile crops in India beyond the Harappan zone
The spread of textile production and textile crops in India beyond the Harappan zone:
an aspect of the emergence of craft specialization and systematic trade
Dorian Q Fuller
Institute of Archaeology
University College London
London WC1H 0PY
Email: [email protected]
“Clothes make the man. Naked people have little or no influence on society”
- quotation attributed to Mark Twain, 19th century American novelist
ABSTRACT
This paper reviews the archaeological evidence for cotton and flax in South Asia. This is based primarily on archaeobotanical
evidence from seeds. This evidence indicates that both crops were established in the Indus region by the Harappan civilization
and spread elsewhere into India in post-Harappan/late Chalcolithic times. In addition some representative data from artefactual
evidence in the form of spindle whorls are considered for the Middle Ganges and Peninsular Indian regions, which suggests an
increase in spinning activities from the second half of the second millennium BC. This may indicate that spinning began slightly
before the introduction of cotton and flax crops, or else in the earliest stages the presence of cotton and flax is still lacking due to
taphonomic biases which have particularly affected evidence from early small scale production. In addition, a preliminary attempt
to gather historical lingustic evidence for these crops and for weaving in South Asia is provided, including clear evidence that the
advent of cotton and flax can be reconstructed for proto-South Dravidian associated with other terminologies for craft production
and social hierarchy. Some materials from Sanskrit and Munda languages are collated as a basis for further linguistic enquiry.
The spread of cotton beyond South Asia is briefly reviewed, including linguistic evidence from Southeast Asia and historical
and archaeological data from Africa. The spread and development of textile industries in South Asia can be associated with the
emergence of more complex societies in which specialized craft production and trade were of greater importance.
INTRODUCTION
received less attention, despite their inclusion by
Childe as part of his Neolithic revolution concept. In
Much effort by archaeologists and archaeobotanists
part this is a matter of archaeological preservation:
focuses on the origins of agriculture, the dome-
cereals and pulses by far outnumber other categories
stication of plants and animals. Gordon Childe
of plants in the archaeobotanical record (cf. Weber
(1936) highlighted these as central issues in the
1992; Zohary and Hopf 2000; Fuller 2002). It is
study of the Neolithic revolution. In the South Asian
nevertheless important to consider these crops, which
context, recent years have witnessed an expansion of
were aimed for raw materials of crafts, and what this
archaeobotanical research, much of it with a focus on
tells us about the nature of early agriculture, and how
agricultural origins (e.g. Kajale 1991; Saraswat 2004,
economies changed.
2005; Fuller 2002, 2006a). This has focused largely
In recent years Andrew Sherratt drew attention
on the domestication of food plants, especially staple
to the importance of seeing agriculture as about
cereals and to a lesser degree pulses. Fibre crops have
more than just subsistence (e.g. Sherratt 1995, 1999,
-1-
Dorian Q Fuller
2007) 1). In other words, cultivation was not just
tied into widening hinterlands of raw materials and
about getting enough to eat (for one community,
human demographic networks (McCorriston 1997).
for one year) but about getting excess, a storable,
In other words, the transformations towards more
surplus which could both be transmitted across
fibrous crop-plants and then more fibrous animal
time, to provide food in lean periods, but also
breeds contributed to the economic transformations
across space by being traded. Like Runnells and
of smaller-scale Neolithic societies towards urbanism
Van Andel (1988), Sherratt (1999) argues that the
and increases in the scale of economic networks.
development of regional exchange systems that tied
Textile crops are thus an important aspect of “Bronze
together communities was one of the factors that
Age Economics” (sensu Earle 2003).
promoted the emergence of food production, and
The production of textiles is an important part of
could be seen, for example in the early movement
craft production economies in two ways: first to do
of lithic raw materials. But he also suggested that as
with craft and second, with agriculture. In terms of
food production systems became more widespread
craft, textiles are labour-intensive and time-consuming
and intensified that there was a second major set
to produce. They require labour in spinning, as well
of revolutions, which involved new uses of animals
as weaving. In many traditional societies, textile
and new kinds of plants: animal secondary products
production was carried out as a domestic activity,
and long-lived perennial crops, like trees. Tree fruits
and women spent much ‘surplus’ time (i.e. when not
and nuts could be dried and traded or turned into
engaged in basic subsistence and cooking activities)
trade-able products like wine, while milk products
spinning (Barber 1991; McCorriston 1997). Weaving
extended the cycles of productivity in animal herds
is a highly skilled craft which must be learned, and
and provided longer-shelf-life products like cheeses.
different regional traditions of weaving are often
Thus certain non-staple plant foods could increase
distinctive and recognizable (cf. Barber 1999; Tuck
as commodities for trade, just as the products
2004). The production of textile crops, constitutes
of developing crafts. Some crops too may have
another important element of specialization, in as
undergone secondary transformations, such as flax
much as it is implies the use of land and agricultural
(Linum usitatisimum), which was most likely first
labour resources for species that will not be eaten, and
cultivated primarily for its edible oily seeds and later
thus implies additional surplus production beyond
used as a source of bast fibres and ultimately bred for
what is required to feed families and communities.
more fibre-productive varieties: which is indicated
It is therefore necessarily production for trade, as
by the derivative phylogenetic position of fibre-
fields of textile crops produce fibre far beyond what
selected flax cultivars (Allaby et al. 2005). Another
individual households are likely to use or have time,
important transition was the secondary products
and perhaps skill, to process. Craft crops therefore
revolution amongst animals (see also Sherratt 1981),
constitute an important early “cash-crop” (sensu
with the use of animal hairs, especially sheep wool,
Sherratt 1999), along with such things as valued
for fibre production and textile manufacture. As
trade-able fruits.
more wool-producing sheep breads evolved this
Between cash crop production and craft work, there
meant that agriculturally marginal lands, such as the
are additional labour costs in terms of processing.
hilly margins of Mesopotamia could be productive
Even before spinning the creation of fibres requires a
for wool. Meanwhile wool offered a less labour-
series of time-comsuming laborious operations. For
intensive source of fibres for weaving which could
flax this involves rippling and retting (soaking with
be supplied to emerging Bronze Age cities that were
partial fermentation of stems), followed by pounding
-2-
The spread of texitile production and textile crops in India beyond the Harappan zone
and combing to separate bast fibres (for a synopsis,
crops in South Asia, cotton and flax. I will then situate
see McCorriston 1997: 522-524). These can then be
these in their archaeological context by reference to
spun and weaved. With cotton it involves even more
the presence of spindle whorls, an artefactual indicator
steps: dehusking, seed removal (ginning), cleaning
for textile production. This archaeological picture
and smoothing, bowing or carding to separate the
will then be compared to the historical linguistics of
fine fibres, which are then rolled and ready to be
textiles in South Asia, as well as some related terms of
spun. Only then can weaving of the threads or yarn be
craft production.
done. Accounts of traditional cotton processing from
INDUS COTTON: ORIGINS TO
HARAPPAN EVIDENCE
different regions are similar, including those from
ancient India (Schlingloff 1974), China (Goodrich
1943; Schlingloff 1974: 85) and Africa (Spring and
Hudson 1995). In recent centuries much of this
Today there are four cultivated cotton species, two of
process has been aided by mechanical developments,
Latin American origin and two from the Old World
but traditional ginning was done by a labour intensive
(Wendel 1995). While the American cottons are
method involving a rolling pin (often metal) and a
perhaps the most important in modern production,
wooden board which was used to force out seeds.
the likely South Asian native was important in the
Bowing is done with a bow-like instrument in which
early development of textile production in the Indus
a vibrating cord helps to loosen to separate the fibres.
and South Asia, as well as in Indian Ocean trade in the
The implication of all this is that cotton production
Roman period. In the Old World there are two cotton
requires both the expenditure of more labour in
species, both closely related diploids, Gossypium
cultivation, beyond subsistence requirements, and
herbaceum, for which wild populations are identified
more expenditure of labour in the household in
for Southern Africa and tree cotton, G. arboreum.
processing, but with the result of a commodity by
Unfortunately for the archaeobotanist, we have not
which wealth can be accumulated, and fairly easily
yet developed methods for distinguishing the charred
transported, or sequestered by emerging elites.
seeds of herbaceum versus arboreum cotton, and their
seeds appear virtually identical even at an anatomical
Textile production may also play an important role
level.
as a technology of social differentiation. It is part of a
wider category of technologies by which the natural
Tree cotton, Gossypium arboreum L. is now
body is transcended and socialized to represent and
considered most likely of South Asian origin. A
reinforce aspects of the social order (for archaeological
weedy/wild form that is distributed in Southern
case studies in other contexts, see, e.g. Traherne
Sindh, and reported from dry hills of the Central
1995; Hill 1997; Chapman 2000; with theoretical
Deccan (Hutchinson and Ghose 1937; Santhanam
foundations in Douglas 1973, pp. 93ff.; Bourdieu
and Hutchinson 1974). The modern distribution
1984, pp. 175ff.; Shilling 1993, pp. 70ff.). Beads and
may not represent primary habitat as feral varieties
textiles are very often intertwined as dress, which are
may have spread together with the early cultivar
potential indicators of social identity and status, that
(Wendel 1995; Zohary and Hopf 2000). Climatic
we expect to become increasing important as societies
change, through aridification since the mid-Holocene,
become more complex (for an ethnographic example,
and habitat loss due to agriculture, especially in the
see Eicher 1998).
Greater Indus valley and its hinterland could have
As a contribution to these issues, I will review the
wiped out the wild progenitor. In this regard it is
current archaeobotanical record for two major texilte
worth considering that during the wetter early to
-3-
Dorian Q Fuller
mid-Holocene that wild cotton extended across the
environment comparable to the Arabian peninsula at
Southern Arabian peninsula which must have acted
that time. There is no evidence that these early finds
as a bridge at some point uniting the wild ancestors
in Arabia or Nubia relate to early cultivation, and
of G. arboreum and G. herbaceum. In this regard
evidence for cultivation in Africa only begins from the
a mid-Holocene report of cotton fibres from the
Early Historic horizon (broadly speaking, the Roman
Arabian peninsula (Betts et al. 1994) is intriguing,
period) (cf. Rowley-Conwy 1989; Pelling 2005, 2007;
but requires further documentation. If the dating of
Clapham and Rowley-Conway 2006, 2007, in press)
evidence reported from Egyptian Nubia for the late
Tree cotton, as its name implies is naturally a woody
Fourth Millennium BC is accepted (for reservations
shrubby plant (Figure 1A). As such it might initially
on identification and archaeological dating , see
have been grown as a perennial fruit crop, along the
Zohary and Hopf 2000; Fuller 2002; full details of
lines of grapes or tree fruits such as dates, which are
the find in Chowdhury and Buth 1971, 2005), then
also documented as cultivars in the Indus region
wild cotton may have extended into what is today
from pre-Harappan times (Fuller and Madella 2001).
the Southern Sahara and formerly the Sahel, an
Cotton requires a long growing season, of ca. 200 days
Figure 1
Pictures of cotton and archaeological cotton. A. An illustration of Indian tree cotton, towering over a goat (from
Anonymous 1833). B. Leaves and flowers of a cotton, Gossypium herbaceum (after Sayre 1917). C. Drawing of cotton fruit and seed
in cross-section (after Engler 1937). D. SEM of charred archaeological cotton seed with preserved hairs from Hallur, ca. 900 BC
(after Fuller et al. 2004). E. SEM of cross section of seed coat from charred fragment from Hallur (by this author); F. “Cap-like”
structure from interior of cotton seed, charred example from Early Historic Ufalda, Garhwal (by this author).
-4-
The spread of texitile production and textile crops in India beyond the Harappan zone
Table 1 Archaeobotanical and textile remains of cotton from South Asia
Site
Mehrgarh,
Baluchistan
Evidence
Seeds (uncharred);
mieneralized thread in
copper bead
Period
Ceramic Neolithic, 6000-4500
BC
Reference(s)
Costantini 1983; Costantini &
Biasini 1985: 24;
Moulherat et al. 2002
Mohenjodaro
Cloth
Mature Harappan, 2600-2000
BC
Gulati and Turner 1929
Balakot, Sindh
Malavaceae pollen
type, comparable to
Gossypium
Mature Harappan,
2500-2000 BC
McKean 1983; also mentioned in
Dales 1986
Harappa
Seed(s) [Weber];
earlier textile reports
Mature Harappan(?), 2600-1900
BC
Weber 1999: 818
Kunal
Seed(s)
Saraswat & Pokharia 2003
Banawali
Seed(s)
Mature Harappan, ?25002000 BC, perhaps equivalent to
Harappa 3C(?), 2200-1900 BC
Mature Harappan (?=Harappa
3C),
2200-1900 BC
Sanghol
Seed(s)
Late Harappan, 1900-1400 BC
Saraswat 1997
Hulas
Seed(s)
Late Harapan, 1800-1300 BC
Saraswat 1993
Kanmer, Kacchh
Seed(s)
Late Harappan, 2000-1700 BC
Pokharia 2007 (in Kharakwal et
al. 2007)
Imlidhi Khurd,
Gorakhpur, Uttar Pradesh
Seed(s)
Period II, 1300-800 BC
Saraswat 2005
Waina,
Ballia,
Uttar Pradesh
Seed(s)
Period I, 1600-800 BC
Saraswat 2005
Sringaverapura, Dist
Allahabad, Uttar Pradesh
fibres
Late Ochre-Coloured Pottery,
1200-700 BC
Saraswat 1986
Hallur, Upper
Tungabhadra, Karnataka
Seeds & fragments
Early Iron Age, AMS direct date:
950-900 BC
Fuller et al. 2004; for dating:
Fuller et al. 2007
Sanghol, Ludhiana Dist.,
Indian Punjab
Seed(s)
Early Historic, Kushana, 200
BC- AD 300
Pokharia & Saraswat 1999
Charda
seeds
Period IIB, Early Historic, 200
BC- AD100
Chanchala 2002
Hund, Peshwar Dist.,
Pakistan
Seeds & fragments
Kushana through Mughal (all
periods), 200 BC-AD 1600
Author’s unpublished data;
Cooke 2002
Kausambi
Seed(s)
NBPW horizon, 550-250 BC
Chanchala 1995
Hulaskhera
Reported
indeterminate, appears
to cotton seed “cap”
Seed(s)
Iron Age/Early Historic, 600
BC- AD 250
Chanchala 1992, Plate 5, 24
Early Historic, 250 BC-AD 250
Anonymous, in Sankalia et al.
1960: 529-530
Nevasa (1954-1956
season)
Saraswat 2002
Kodumanal, Coimbatore
Dist., Tamil Nadu
Seeds & fragments
Early Historic/Late Megalithic,
300 BC- AD 300
Cooke et al. 2005
Perur, Coimbatore Dist.,
Tamil Nadu
Seed fragments
Early Historic/Late Megalithic,
300 BC- AD 300
Cooke et al. 2005
-5-
Dorian Q Fuller
Mangudi, Madurai Dist.,
Tamil Nadu
Seed fragments
Early Historic/Late Megalithic,
300 BC- AD 300
Cooke et al. 2005
Ufalda, Garhwal,
Uttaranchal
Seeds & fragments
Early Historic(?), AD 0-600
Author’s unpublished data
Singh Bhagwanpur,
Rupnagar Dist., Indian
Punjab
Seed(s)
Medieval, AD 800-1100
Vishnu-Mittre et al. 1984
Mangali Luduwala,
Haryana
Seed(s)
Sub-recent, AD 1500-1900
Willcox 1992
(6.5 months) with abundant water early in the season
processing, where bolls are ginned before the cotton
(equivalent to at least 50cm rainfall) and dry rain-
fibres are combed and spun. Thus the seeds indicate
free conditions for the last two months when the fruit
places of cotton processing for fibre. In general we
and seeds form so as not to damage the fibre from
expect these to be close to centres of cultivation,
dampness and mould (Burkill 1997; Robbins 1931:
although it may sometimes be the case that raw
497; Langer and Hill 1982: 262). It generally needs a
cotton is transported from areas of cultivation to
frost free environment, preferably with temperatures
centres, such as cities, that have larger available labour
above 21ºC (Langer and Hill 1982: 261; Burkill
populations.
1997). Cotton remains a major crop through much of
THE EVIDENCE OF FLAX
India, except for the eastern part of the country, where
rainfall continues too long (cf. Choudhary and Laroia
The flax plant (Linum usitatissimum L.) is an
2001).
Archaeobotanical recognition of cotton relies
important source for bast fibres for textile production
mainly on the preser vation of charred seeds or
(linen) as well as for an oily seed made edible with
seed fragments. As the fibre grows out of the seed
roasting. Once removed from the seed the oil goes
(Figure 1C), remnants of the fibre, which can often
quickly rancid and becomes inedible, and thus linseed
be preserved charred on seed surfaces are a give away
oil is better known for craft and non-culinary uses
(Figure 1D). In addition the layered cross section of
in the modern West, but in the seed it is a storable
the seed coat is distinctive (Figure 1E). In recent years
product of considerable nutritional value (see, e.g.
we have come to recognize a small structure, which
Seegler 1983). Use of the edible seed has probably
looks like a “cap” with a central circular pore (“belly
been more important in India than the fibre, as
button”), which comes from the inside of the cotton
numerous other fibre plants are available in this
seed where it attaches to the vasculature of the capsule
region, although fibre varieties are also cultivated
(Figure 1F; cf. Pelling 2007). This structure appears
(cf. Vavilov 1950 [1992]). While stands cultivated
to survive charring better than the rest of the seed and
for fibre are often harvested before seed production,
on its own indicates the former presence of cotton
and thus finds of seeds are more likely to result from
seeds. In addition textile fibres of cotton can usually
production for food (McCorriston 1997: 519), their
be identified in situations where textiles are preserved,
presence nevertheless raises the possibility of flax fibre
although given the importance of textiles as a trade
production in a region in prehistory, and so the South
commodity they provide no indication of areas of
Asian archaeobotanical record will be reviewed in this
textile production or cotton production. The seeds, by
paper. The evidence is summarized in Table 2.
The closest wild relative of flax is well established,
contrast, are expected to come from centres of cotton
-6-
The spread of texitile production and textile crops in India beyond the Harappan zone
Figure 2
Illustrations of the flax plant, Linum usitatissimum. A. Drawing of the flax plant in flower (after Berg and Schmidt
1958-1863). B. drawing and cross-section of flax capsule, at approximately twice the scale of A (after Berg and Schmidt 1858-1863).
C. SEM of the distinctive seed tip of flax (the author).
as Linum bienne Mill. It is distributed across
varieties were available to the Indus civilization. The
Meditarranean and steppic habitats in Southwest
earliest finds in South Asia come from Harappan
Asia, Europe and North Africa (Vavilov 1950 [1992];
period sites, (Table 2; Figure 3), including Nausharo
Zohary and Hopf 2000: 129). The wild distribution
(Costantini 1990) and Miri Qalat (Tengberg 1999),
coupled with archaeobotanical evidence from
and Balathal east of the Harappan orbit (Kajale 1996).
Neolithic and pre-Neolithic sites in Southwest Asia
From the post-Harappan horizon in the northwest,
indicates that this was a component of the earliest
finds come from Pirak (Costantini 1979).
agricultural economies in Southwest Asia, along
Flax or linseed is normally a winter crop in South
with wheat and barley (McCorriston 1997; Zohary
Asia requiring moderately high rainfall (>75cm) or
and Hopf 2000). A small phylogenetic analysis
irrigation during this period (cf. Weber 1991: 81). In
of flax, suggests a single domestication, and that
India this means either sowing immediately after the
domestication was initially for oilseed types (Allaby et
monsoons, in a region that have adequate rain levels
al. 2005). Fibre-adapted forms were developed later,
and water-retentive clay-rich soils, or broadcasting
but even so use of fibres had begun before the end of
into remnant standing water of har vested rice
the Neolithic in Southwest Asia, i.e. by ca. 7000-6000
paddies (McCorriston 1997: 524). Cultivation and
BC (Ryder 1965; McCorriston 1997: 519). It is
preparation is labour intensive (McCorriston 1997),
clear that flax was a significant fibre crop in the early
requiring weeding, the pulling up of plants for fibre
civilization of Egypt and Mesopotamia, and we can
(if grown for seed they can be cut below the capsules),
postulate that fibre-varieties had evolved by the end
rippling to remove seeds and capsules and then retting
of the Fourth Millennium BC. It is possible that such
(partial rotting in water for about 2 weeks) drying and
-7-
Dorian Q Fuller
Table 2 Archaeobotanical finds of linseed/flax (Linum usitatissimum) in South Asia
Site
Harappa
Evidence
Seed(s)
Period
Mature Harappan, 2600-1900
BC; and Late Harappan, 19001700 BC
Reference(s)
Weber 1999, 2003; personal
communication
Kunal
Seed(s)
Period 2, Early Harappan, 28002500/2300 BC [?]
Saraswat & Pokharia 2003
Miri Qalat, Makran
Seeds
Mature Harappan, 2500-2000
BC
Tengberg 1999
Nausharo, Baluchistan
Seeds
Mature Harappan, 2500-2000
BC
Costantini 1990
Balathal, Rajasthan
Seeds
Chalcolithic, 2500-2000 BC ?
Kajale 1996
Ojiyana, Bhilwara Dist.,
Rajasthan
Seed(s)
Ahar Culture, 2500-1500 BC
Pokharia & Saraswat 2004
Pirak, Baluchistan
Seeds
Late Harappan, 1950-1550 BC
Costantini 1979
Sanghol
Seeds
Late Harappan, 1900-1500 BC
(?)
Saraswat 1997
Babar Kot, Saurashtra
Seeds
Late Harappan, 2000-1700 BC
Reddy 1994, 2003
Rojdi, Saurashtra
Seeds
Late Harappan, 2000-1700 BC
Weber 1991
Loebanr 3, Swat
Seed(s)
Late Chalcolithic, 1700-1400
BC
Costantini 1987
Imlidhi Khurd, Gorakhpur,
Uttar Pradesh
Seed(s)
From Periods I & II,
2000(?)/1600-800 BC
Saraswat 2005
Narhan I, Gorakhpur Dist.,
Uttar Pradesh
Seed(s)
Period I, 1300-800 BC
Saraswat et al. 1994
Senuwar II, Rohtas District,
Bihar
Seed(s)
Period II, Chalcolithic, 1300600 BC
Saraswat 2004
Waina II,
Ballia Dist., Uttar Pradesh
Seed(s)
Period II, 800-500 BC
Saraswat 2005
Raja-Nala-Ka-Tila II,
Sonbhadra Dist., Uttar Pradesh
Seed(s)
Period II, 1300-700 BC
Saraswat 2005
Navdatoli, Maharashtra
Seed(s)
Jorwe Phase, 1500-1200 BC
Vishnu-Mittre 1961
Daimabad, Maharashtra
Seed(s)
Jorwe Phase, 1500-1200 BC
Kajale 1977
Hallur, Karnataka: Upper
Tungbhadra
Seed (fragments)
Early Iron Age, 1000-900 BC.
Two AMS dates from same
context.
Fuller et al. 2004; dating:
Fuller et al. 2007
Charda
seeds
Period I, 1000-600 BC; IIA,
600-200 BC
IV, AD 500-1000
Chanchala 2002
Paithan, Godavari river,
Maharashtra
Single seed
Period III, AD 300-700
Author’s unpublished data
Hund, Peshawar Dist., Pakistan
Seeds
Mughal period, AD 1100-1600
Author’s unpublished data;
Cooke 2002
-8-
The spread of texitile production and textile crops in India beyond the Harappan zone
Figure 3
The distribution of archaeological finds of cotton and flax seed in South Asia, indicated by broad time horizons (for
details of chronology and sources, see Tables 1 and 2). Sites numbered: 1. Mehrgarh; 2. Nausharo; 3. Pirak; 4. Miri Qalat; 5.
Mohenjodaro; 6. Balakot; 7. Hund; 8. Loebanhr 3; 9. Harappa; 10. Kunal; 11. Banawali; 12. Sanghol (indicating Late Harappan
and Early Historic evidence); 13. Hulas; 14. Balathal; 15. Ojiyana; 16. Kanmer ; 17. Babor Kot; 18. Rojdi; 19. Hulaskhera; 20.
Charda; 21. Imlidh-Khurd; 22. Narhan; 23. Waina; 24. Sringaverapura; 25. Kausambi; 26. Senuwar; 27 Raja-Nala-Ka-Tila; 28.
Navdatoli; 29. Paithan; 30. Daimabad;. 31. Nevasa; 32. Hallur; 33. Perur; 34. Kodumanal; 35. Mangudi; 36. Mangali/Luduwala;
37. Singh-Bhagwantpur; 38. Ufalda
THE ARCHAEOBOTANICAL DATA
FOR FLAX AND COTTON BEYOND
THE INDUS VALLEY
beating (bracking and scotching to remove the fibres
from the pith), and combing (“hackling”) to clean
away the pith fragments. After spinning and weaving
extra efforts are required to get the fibre to take and
The distribution of evidence for cotton beyond the
hold dyes.
Indus zone can be seen in Figure 3. As can be seen
all the early finds are in the Indus region, and only
-9-
Dorian Q Fuller
post-2000 BC finds are reported beyond this zone.
a threshold after which it becomes more likely to
Amongst those beyond the Indus valley it is those
recover these species archaeologically.
areas closest to the Indus that have evidence for the
More limited evidence for some other fibre crops also
first part of the Second Millennium BC, including
comes from the same horizon in the Gangetic zone.
Saurashtra, Rajasthan and the upper Ganges region.
This includes evidence for hemp (Cannabis sativa)
Further afield on the peninsula and in the middle
on the basis of both seeds and wood charcoal from
Ganges area finds are later, and all post-1500 BC.
Chalcolithic Senuwar, 1300-600 BC (Saraswat 2004).
On north peninsular sites, Chalcolithic evidence is
In addition fibres of ramie (Boehmeria cf. nivea), are
so far only available for Linum, with finds from the
reported from Narhan from the same peiord (Saraswat
Jorwe horizon, 1200-1500 BC. In the middle Ganges
et al. 1994: 287), This species may have been the first
both species are well represented in samples from the
important fibre cultivar of the Lower Yangzte region
Chalcolithic, which in this region can be placed from
in China, and is likely to be introduced to India (cf.
ca. 1300-800 BC. The only direct AMS date is from
Burkill 1966; Keng 1974).
Hallur, where both cotton and flax were found in the
SPINDLE WHORLS AND THE
EMERGENCE OF CHALCOLITHIC
CLOTH PRODUCTION
same rich sample of the early Iron Age, from which
cotton produced a date of 900-950 BC and another
seed from the same sample produced a date closer to
1000 BC (Fuller et al. 2007).
In the Harappan northwest the archaeobotanical
Artefactual evidence can also shed light on the
evidence for flax is more limited than that of cotton.
history of textile production in India outside the
This may be due in part to less robust seeds, as well
Indus valley. Spindle whorls, used for making thread
as cultivation of fibre varieties which are less often
from fibres, are a common archaeological find, often
allowed to set seed. It may also be due to differing
being made of ceramic. While a comprehensive
processing customs, as the time-consuming removal
review of the archaeology of spindle whorls in South
of cotton seeds (ginning ) may have been regularly
Asia is beyond the scope of the present contribution,
carried in settlement areas and domestic contexts
some representative patterns can be noted here, by
whereas flax retting may have taken place off site in
reference to published reports from some important
special locals where vats or pits were constructed for
excavations: Senuwar in the Ganges and Inamgaon
this process; as such flax might be less likely to come
in the Deccan. The evidence from the Southern
into contact with domestic fires, although waste from
Neolithic will also be considered.
In the middle Ganges region, I will use the data from
rippling flax before retting could be used as domestic
the Senuwar excavations (Singh 2004). This site spans
fuel.
Given that there are strong preservational biases
a well-dated sequence from a Neolithic phase that
against both species, the archaeobotanical picture can
starts ca. 2500 BC, during which native rice agriculture
only be taken to represent the very minimum period
was present prior to the introduction of non-native
of entry to a region. It may well be that we should
crops like wheat and barley (see Saraswat 2004).
regard the archaeobotanical evidence as indicating
By the end of this phase wheat and barley had been
the period during which cultivation became more
introduced, so a date of ca. 2200 BC can be inferred
widespread and use intensified rather than the initial
for the first influence from the Harappan zone to
introduction as such. There may be a matter of scale
the west, in this case in terms of staple crops. Indeed,
in which larger scale and more intensive use crosses
more recent evidence from Lahuradewa-IB, including
- 10 -
The spread of texitile production and textile crops in India beyond the Harappan zone
Se n u war : Sp in d le Wh o rls
I I. 1 3 0 0 - 8 0 0 B C
IB . 2 0 0 0 - 1 3 0 0 B C
IA . 2 5 0 0 - 2 0 0 0 B C
0
5
10
15
20
25
30
Figure 4 The evidence for spindles whorls in the three phases of Senuwar (data from Singh 2004). This shows clearly a minimal
presence in the early Neolithic (before 2000 BC) and a massive increase in evidence for textile production during Period IB (before
1300 BC).
an AMS date on barley and another AMS date from
as to what fibres were involved, although cotton and
a barley grain from Damdama confirm this diffusion
flax seem likely candidates. Their absence from the
(Saraswat 2005; Tewari et al. 2006). Dish-on-stand
archaeobotanical record until the Chalcolithic may
type vessels also first occur in this horizon and are
reflect scale of use. If so, then by Chalcolithic times
reminiscent of forms from the Greater Indus region.
the threshold had been crossed that leads to recurrent
During the late Neolithic at Senuwar (Seunwar-
archaeobotanical recovery, as several sites in the
IB) introduced crops become more important and
Middle Ganges region have evidence for flax and/
diverse, with the addition of pulses such as lentils
or cotton only during this period and consistently
(from the northwest?) and mungbean, which is small-
not from the earlier levels in these sites, despite the
grained (unlike Indus varieties at that time) and might
presence of spindle whorls in those levels. Such
thus derive from the peninsular region to the south
an argument, that there is a misleading absence of
(cf. Fuller and Harvey 2006). Finally the site has a
evidence assumes that it is more likely that techniques
Chalcolithic horizon characterized by metal finds
and raw materials (crops) were adopted together.
and a still greater crop diversity, including flax. Finds
Alternatively we might consider the possibility that
of spindle whorls divided into these broad phases
techniques developed first and created a demand for
are shown in Figure 4, in which it can be seen that
better raw materials. In other words, spinning was first
only a few perforated disc sherds are present in the
developed on the basis of some other, perhaps wild
lowest levels (Period IA). With such small counts
fibre source, and once the techniques were established
intrusion from later periods must be considered. In
there was a context in which better fibres from the
the Late Neolithic (Period IB) there is a substantial
fibre crops became desirable.
quantity of spindle whorls. This certainly indicates the
The evidence from Inamgaon in Maharashtra,
practice of spinning during this period. A comparable
suggests a similar pattern of increasing importance of
level continues through Period II. This suggests
spinning in the late Second Millennium BC (Figure 5).
that spinning (and presumably weaving practices)
The few spindle whorls from the early levels (Malwa
began in the later Neolithic during the early to mid-
Phase) are negligible, while quantities in the Jorwe
Second Millennium BC. There is no hard evidence
period are significant, and increase further in the late
- 11 -
Dorian Q Fuller
Figure 5
The evidence spindles whorls in the three phases of Inamgaon (data from Dhavalikar et al. 1988). This shows clearly a
minimal presence in the earlier Chaclolithic Malwa phases and a massive increase during the Jorwe period after 1500 BC.
Table 3 Representative Spindle Whorls from the Southern Neolithic
Site
Budihal
Count
2
Phase(s)
Tr. 4, settlement, level
2(?).
Tekkalakota
12
Layers 2-4
Brahmagiri
IB: 1; II: 2
IB= Late Neolithic; IIMegalithic
Hallur
?
Phase I, period 2
(Layers 8, 9) Later
Neolthic phase III ,
1500-1300 BC
Phase II (layer 6); Early
Iron Age, Ca. 1000
BC
?
Sannarachamma (second
excavations)
21 possible
spindle
whorls,
Halakundi
1 perforated
mica schist
disk
Comments
Pre-1700 BC
radiocarbon
dates. Association
uncertain
Reference
Paddayya 1993, 2001
Nagaraja Rao &
Malhotra 1965
contexts not reported,
post-ashmound 17001000 BC
Probably Later
Neolthic, phase
III, 1500-1300 BC
Wheeler 1948
Nagaraja Rao 1971
More examples
from recent work:
only from post
1500 BC levels
Ansari & Nagaraja
Rao 1969
(on recent work, cf.
Boivin et al. 2005: 79)
Later Phase III(?)
[Black and Red
Ware present]
Indian Archaeology - A
Review 1959-1960: 72
Jorwe. This figure may be somewhat skewed by the
however, that the Jorwe and especially the late Jorwe
fact that a much greater site area and soil volume was
see increasing proportions of sheep and goat amongst
excavated for the Jorwe and Late Jorwe, but this does
the faunal assemblage (Thomas 1988; Pawankar and
not seem to account for the magnitude of difference.
Thomas 1997), so some use of animal fibres may
Although limited the evidence for flax begins in
also be involved, although wool is generally of little
the Jorwe period, although only at some other sites
significance in Peninsular India, and South India
in the region (see Table 2). It should also be noted,
today retains unimproved hairsheep breeds (Ryder
- 12 -
The spread of texitile production and textile crops in India beyond the Harappan zone
1984; Fuller 2006a: 26).
needed before we can conclude that this indicates that
Further evidence for a late Second Millennium BC
spinning techniques preceeded cultivation targeted at
start to spinning comes from the Southern Neolithic
fibre production.
(Table 3). Almost all spindle whorl finds come from
In terms of cultural context it should be noted that
Southern Neolithic Phase III, which dates from
this horizon is the same one that sees other changes
1800-1300 BC (for phasing see Korisettar et al. 2001;
towards increased crop diversity, craft diversity
Fuller et al. 2007). For several sites which have earlier
and possible craft specialization. This is indicated
levels, including Hallur, Sanganakallu and Brahmagiri,
in broadening crop and ceramic form repertoires,
whorls are absent from earlier levels. Total spindle
a process that definitely begins early in the Second
whorl numbers are low, and I would suggest that
Millennium BC (for the Peninsula, see Fuller 2005),
most of these actually come from the later half of this
and the addition of fruit tree-crops, indicated in
period, mainly after 1500 BC, although chronological
particular in the wood charcoal record from the
resolution is inadequate for most of these published
second half of the Second Millennium BC (see Asouti
finds. The earliest possible spindle whorls in the region
et al. 2005; discussion in Fuller 2006b). In addition
come from surface samples at Budihal (Paddayya
this is the period that sees the spread of copper objects
1993, 2001), a site that has Neolithic settlement
and probably copper-working in these zones (cf.
occupation through ca. 1700 BC (see dating evidence
Allchin and Allchin 1982).
summarized in Fuller et al. 2007), although some
THE HISTORICAL LINGUISTICS
OF SOUTH ASIAN CLOTH
PRODUCTION
small scale, or intermittent later Neolithic use is
possible, and the stratigraphic association of these
finds can not be linked clearly to the dated occupation
on the basis of evidence published so far. Although
some possible cotton fragments are present at
The beginnings of textile production and the
Sanganakallu from the end of Period III (ca. 1400
introduction of cotton and flax should be recognizable
BC) [not included in Table 1 due to the uncertainty
in historical linguistics, at least in a region in which
of identification], the only definitive evidence from
these developments happened later than other
this region is the 1000-900 BC cotton seeds from
linguistically-identifiable adoptions and where the
Hallur. Nevertheless it is worth noting the presence
species involved were not available wild. South
of Rubia cordifolia, an important traditional dye plant
India and evidence from the Dravidian languages
for cotton, at Sanganakallu from ca. 1400 BC (Boivin
meets these criteria. As shown above, the first textile
et al. 2005: 81), as this species would not have been
production indicated in the artefactual record comes
locally available but rather suggests transport to the
from the mid-Second Millennium BC (or perhaps
site from the Moist Deciduous woodland zones.
slightly earlier) and cotton and flax were certainly
Thus the evidence from both Ganges and the
cultivated by ca. 1000 BC. These developments post-
Peninsula suggest that the very beginnings of fibre
date the beginnings of subsistence agriculture and
spinning can be placed in the first half of Second
pastoralism. Historical linguistic reconstructions
Millennium BC, and perhaps slightly earlier in
for Dravidian suggests a Proto-Dravidian familiarity
the Ganges, but that there is a marked increase in
with domestic livestock (Fuller 2003; Southworth
spinning by the end of the Second Millennium BC.
2005) and with a number of indigenous wild trees of
Hard archaeobotanical evidence for fibre crops a
Peninsular India (Southworth 2005; Fuller 2006b,
slightly later still, although a larger sample size is
2007). As a slightly later stage, of “Late Proto-
- 13 -
Dorian Q Fuller
Table 4 Historical linguistic data relating to textiles in Proto-South Dravidian
Dravidian root
(*Proto-form, if given, from
Southworth 2005; entry no. from
Burrow & Emeneau 1984).
Suggested meaning
Attestations (from Burrow & Emeneau 1984)
PSDr. *cāl[DEDR 2475]
Cf. Skt. kōlika [CDIAL 3535]
“weaver, spider”
weaver caste name
Ta. Cālikan, cāliyan
Ma. cāliyan
Ka. sāliga, sāliya
Tu. tālye ‘weaver’; ‘spider’; sālye caste of weavers
Te. sāle; sālī˜du, sālevãˉdu ‘a weaver’
Ga. (S.2) sāle
Kuwi (S.) sāliesi
4
PSDr. *cēntr-ir
[DEDR 2809]
Ta. cēntiravar
Ka. jāda; jēda ‘a weaver of the Lingavanta sect’; ‘spider’
Tu. jāde, jādye ‘weaver’; ‘spider’.
Te. jēndra, dēndra ‘a caste of weavers’
Weaver
4 4
4
4
4
4
4
PSDr. *par-utti
[DEDR 3976]
4
Cotton
4
4
4
4
Ta. Parutti; pāram
Ma. parutti
To. pašty ‘wick’.
Ka. par‥ti, parti, patti
Kod. parati ‘cotton cloth’ Tu. parti
Te. p(r)atti
Go. (Ko.) part
Kui parti
Kuwi (Su.) pratti (Isr.) parti, (F.) parti
4
4
PSDr. *nūl[DEDR 3726]
cotton thread, or thread,
or yarn (from an older
terms for twisting/
spinning, cf. Kurux)
Ta. Nūl; nūrp-, nūrr- ‘to spin, compose (as a poem),
make a plot’
Ma. nūl; nūlkka ‘to spin’. Ko. nu·l thread; nurb(nurby-) ‘to twist’, ‘wring (neck)’.
To. nu·s; nu·sf- (nu·st-) ‘to join ends of thread by
rolling’.
Ka. nūl; nūlt- ‘to spin’; nūlige ‘spinning’; nuli ‘to twist’,
‘curl (whiskers)’, ‘roll (as cotton) between the hands’
Hal. nugulu ‘thread’
Kod. nu·lï ‘thread’
Tu. Nūlu; nūlodu ‘spindle’; nūpuni ‘to spin, twist’
Kor. (M.) nuglu
Te. nūlu; nulaka ‘a rough kind of rope or string’; nuli
‘entanglement in a thread’; nuliyu ‘to be twisted’;
nulincu, nul(u)cu, nul(u)pu, nulumu ‘to twist’
Kol. nuv,
Kin. nūl
Pa. nūl
Ga. (Oll.) nūl
Go. (many dialects) nūl ‘thread, string’
Konda nūlu; nuls- ‘to twist’
Pe. nūl; nōn- (nōt-) ‘to spin’, ‘twine’
Mand. nūl
Kui nūdu (pl. nūtka) ‘cotton yarn, thread’; nōlba (nōt-)
to twist strands together, spin thread; n. spinning.
Kuwi (Su. Isr.) lūlu, (F.) lūlū, (S) lōlu.
Kurux nõēnā ‘to wind or twist anything flexible’, ‘twist
grass or creeper into rope’.
4
4
4
4
4
4
4
- 14 -
4
4
The spread of texitile production and textile crops in India beyond the Harappan zone
3745 PSDr. #nec
Weave,
from older root,
PDr. #(n)ese, plaiting(?)
Ta. ney to weave as clothes, string, link together;
neyvār the caste of weavers; neyvu weaving; necavu
weaving, act of weaving, texture, intertexture, web;
Ma. neyka to weave, plait mats; neyttu weaving; neyyal
weaving.
Ko. nec- (nec-) to weave; negc- (negc-) to make closewoven.
To. nic- (ni&cangle;-) to darn; nes- (nesQ-), ni·Q(ni·Q-) to weave.
Ka. nēў, nēўi, neyyu, nē, nēyu to weave, entwine; neyi,
nē, nēyu weaving, a web; nēўige, nēЎge, nēge, entwining
or being entwined; neysu, nēyisu to cause to weave;
nēўikāra, neygekāra, nēkāra weaver.
Kod. ne·y- (ne·yuv-, nejj-) to spin (thread); neyv
braiding, weaving.
Tu. neyuni to weave (as a spider); neyipini, nēpini,
nēyuni to weave, plait, braid; neyigè, nēgè texture;
neyigāre weaver.
Te. nēyu to weave; nēyincu to cause to be woven,
get woven; nẽˉta weaving, texture; nẽˉtakãˉdu, nẽˉtari
weaver; nẽˉta-purugu spider (see 4312).
Go. (Koya Su.) nēcc- to weave.
Konda ney- (-t-) to weave or thatch the roof with
leaves
Kui nehpa (neht-) to build a fence.
Kuwi (S.) neh’nai to interweave.
Kur. essnā (issyas) to weave, entwine into a fabric,
furnish or adorn any article with net-work or plaitwork.
Malt. ese to plait, do mat-work.
4
4
4
[DEDR 765] PSDr (?)
To card cotton; older
meaing suggested by
C.Dr./S-C.Dr. cognates
(Parji and Gondi) “to
weed” or “pick stones
from field”
Ta. e-kku (e-kki-) to pull with fingers (as cotton), to
scrutinize;
Ma. ekkuka to card cotton; ēkku carding cotton.
Ko. ek- (eky-) to scratch (oneself )
To. ök- (öky-) to scratch oneself.
Ka. ekku, yakku to divide, separate, dress cotton, card
wool; ekkike dressing cotton, etc.
Tu. ekkuni to gin.
Te. ēku to pick, beat, or clean (cotton); n. roll of
cleaned cotton prepared for the spindle; ēkudu
picking, beating, or cleaning cotton.
Pa. ēk- to pick and throw away stones and weeds from
field.
Go. (A. Y.) eh-, (Tr.) ehtānā, (Ph.) ahtānā to weed
Pe. ec- (-c-) to card cotton;
Kui ēspa (ēst-) to unravel.
4
PSDr. *tuu[DEDR 3393]
→ Skt. tūla- [CDIAL 5904, ‘cotton’]
→ Munda: Juang tula (Matson 1964),
≈? Kharia turai (Donegan and
Stampe 2004b), or tuday (Biligiri
1965)
→ Proto-Monic
*tŢ:[l]
Old Mon: tol
Modern Mon: tow
Nyakur: tual.L
[differs from Proto-Palaung-Wa *da:i
(Peiros & Starostin 2003)
Feather, soft hair,
sometimes derived
meaning cotton (loaned
to Indo-Aryan)
Ta. tūval ‘feather’; tuy ‘cotton’
Ma. tūval ‘feather’, ‘quill’, ‘painter’s brush’ toppa ‘wool’;
toppal ‘feather’
To. tu·fy ‘feather, bird’s tail’.
Ka. tippu r ‘bird’s wing or feather’; tuppur a ‘soft
¨ birds’, ‘soft hair of rabbits’; tuppa
¨ ru ‘wool’;
plumage of
¨
tuppata, tubata ‘wool’
Kod. toppïta ‘feather’
Tu. tuyi, suyi ‘feather’, ‘quill’
Kor. (M.) cippudu ‘feather’, ‘quill’
Te. tūnī˜ga, tūnĩga ‘dragon-fly’; truppudu ‘feather’,
‘hair’, ‘down’
Go. (Ma.) tō r(i) (pl. tōhku) ‘large feather’; (Mu.)
tokenj, (Ma.)˚tokonji ‘feather’
Mand. tūku ‘feather’
4
4
4
4
4
4
4
- 15 -
Dorian Q Fuller
[DEDR 1195] PSDr.(?) #katir
spinner’s spindle
Ta. katir
Ma. katir.
Ka. kadir, kadaru, kaduru
Tu. kadr; kaduring
Te. kaduru
Ga. kadur
PSDr.1 *ak-V-ce
[DEDR 3]
flax (or linseed) crop
Ka. Agase
Tu. agase-nār
Te. agise, agisiya, avise, avisiya
Commodity
Ta. carakku goods, articles of merchandise, gold, solid
worth, curry-stuffs, spices, medicinal substances.
Ma. carakku merchandise, cargo, different movables or
valuable articles as cloths, jaggery, or drugs.
Ko. cark spices for curry.
Ka. saraku, sarku goods, things, commodities,
merchandise, cargo; saraku esteem, regard, care.
Tu. Sarak; articles, goods, commodity; caraku
merchandise, articles, goods, cargo.
Te. saraku an article, commodity, thing, ornament,
jewel, trinket; care, heed, regard.
Konda sarku materials.
Kuwi (S.) harku thing, instrument, furniture, jewels;
hārka, pl. harkunga things (F.) harkū jewelry, thing;
(Isr.) harku implements.
←→ Old Indo-Aryan
atasiPSDr. *car-a-kk[DEDR 2353]
4
Dravidian” (sensu Southworth 2006) or Proto-South-
can be seen for flax and for one of the Sanskrit terms
Central Dravidian (PDr-2) (sensu Southworth 1988;
for cotton, which may originate in another Dravidian
Fuller 2003) terms for several native subsistence crops
term for feathers.
can be reconstructed, and this can be suggested to
The Indic languages contain another root word for
have a pretty good fit with the Southern Neolithic in
cotton, which may ultimately derive from a term
archaeological terms. At an even later stage, of Proto-
used by the Harappans. Hindi kapās, from a Prakrit
South Dravidian, several non-native crop names
kappāsa, from an earlier Sanskrit karpā’sa (Turner
can be reconstructed, including wheat and barley
1966: CDIAL 2877). The Persian term karvās also
(introduced archaeologically by ca. 1900 BC) as well
derived from this root. This is suggested to be a non-
as cotton, flax and some fruit trees (see especially
Indo-European substrate word (Mascia 1979; Fuller
Fuller 2007). As I have noted before, it is also to this
2003: 205), and based on its kar- prefix is amongst a
stage that a number of terms relating to emergent
group of terms that Witzel (1999, 2005) has referred
social hierarchy and craft specialization (including
to as “Para-Munda” or kubha-vipas, which have
metallurg y) can be reconstructed (Fuller 2006b,
broad Austroasiatic (or perhaps Austric?) structure
2007). Those terms relating to textiles and textile crops
(cf. Fuller 2007). This language is inferred to have
are collected in Table 4 (mainly following Southworth
been a major language of the Indus region during the
2005, with reference to Burrow and Emeneau 1984),
Harappan civilization, which would accord with the
together with terms of equal antiquity that relate to
great antiquity of cotton for this region.
craft specialization and trade with which we expect
The term for flax, atasi, which was also loaned
the development of South Indian textile industry to
to South Dravidian, also appears to be a substrate
be connected. Connections with other languages,
word (Mascia 1979; Fuller 2003: 205; Southworth
especially Indo-Aryan are indicated. And shared roots
2005), but in this case it is amongst those which are
- 16 -
The spread of texitile production and textile crops in India beyond the Harappan zone
Table 5 Indic terms relating to cotton processing with some Munda comparisons
Sanskrit term and cognates
Cotton processing stage/product
Skt. *vangaputa
Cotton pod (Turner 1966: CDIAL 11198)
Contrasts:
Munda: Pinnow 1959 #327 : Sora ə’di:-n; Kharia si’di䁑, (or sidij, Biligiri 1965)
Preparation of the cotton
Gin (n.)
Root: to gin
Comparisons: Sora rid, rәnid “cotton gin” (Donegan and Stampe 2004b)
“to tear off ” [an alternative to above]
Cylindrical roller of gin
Flat board of gin
Rough fibre from the gin (also, Masica 1979)
Comparisons: Remo rua “cotton”, Gorum ruj,
≈? Kharia tuRai “cotton ready for spinning” (Donegan and Stampe 2004b),but
tuday (Biligiri 1965); Bonda rŭa “cotton” (Bhattacharya 1968 #2280), but also
sũru (#2683).
Cotton bow (also, Southworth 2005: 226)
Root: “to bat”
“to strike” [an alternative to above]
“to beat” [an alternative to above]
Cotton bow
To smooth (or card)
Comparisons: no relationship to PSDr. Term (Table 4),
nor w/ Bonda tiŋ- “to card”, also “to pierce, shoot with an arrow” (Bhattacharya
1968 # 1367); Bonda jik “to chard cotton with hands (ibid. #1071).
“to pluck asunder” [used as an alternative to above]
Rolls of cleaned cotton
Comparisons: Munda pid-pid “sound produced with cotton bow” (Hoffman
1930-1938; Osada, pers. comm.);
pitlEd, reported for “to clean cotton” in Mundari and Santali (Donegan and
Stampe 2004a); Kharia pinuri, pue~ri “cotton lump prepared for spinning”
(Donegan and Stampe 2004b);
cf. PSDr. *par-utti (Table 4 above)
Spinning
Comparisons: PSDr.# katir (Table 4, above)
Bonda gurak’- “to spin”; gunurak’ “spindle” (Bhattacharya 1968 #915, 893)
Cotton thread
Comparisons: Juang sotorom “thread”, but also gola “thread” (Matson 1964);
Bonda sũru (Bhattacharya 1968 #2683).
Kharia sugtrom “thread” (Biligiri 1965)
Thread, warp; tántra [CDIAL 5663] “loom”
Comparisons: Juang tonti “weaver” (Matson 1964)
Bonda tãy- “weave” (Bhattacharya 1968 #1358)
Kharia tañ “weave” (Bligiri 1965; Pinnow 1959 #301); Santali teñ, Mundari
t䀘ŋ, Ho/Birhor teŋ, Turi teŋge:, Sora tañ, Gutob tai, Palaun te:ŋ, thă, Wa taiŋ; E.
Austro-Asiatic: Khasi tha:in, Nicobarese tәñә, Bahnar/Boloven/Niahon/Alak
tañ, Lave tăñ,
Khmer p Ţ nţañ (Pinnow 1959 #301)
Weaving; weaver
Skt. Parikarma
Skt. lothinī, lodhanī
Skt. Root: luñc, luth
Skt. vilup
Skt. kanaka
Sky. oronī
Skt. rūta
H. rūī, <Pk. rūa
Skt. piñjana
Skt. Root: pij
Skt Root: sphut
Skt. vihan
H. dhanukī, dhanuhī, dhunkī
Skt. pramrd
H. pīnnā, pīmjnā
Skt. vikrs
Skt. pūnikā
H. pīnī, piunī,
Skt. kartana
Skt. Root: krt
Skt. sūtra
H. sūtī
Skt. tántu [CDIAL 5661]; H.
tãtī “weaver” [CDIAL 5666]
Skt. vāya
Skt. Root: ve,
RV. vayī (Turner 1966: CDIAL
11298)
H. kaprā
Cloth
Comparisons: Juang kote (Matson 1964)
Bonda kŢdi “clothe worn by men” (Bhattacharya 1968 #713), differs from nŢƢ ri
“cloth worn by woman (#1622), mp ŢƢ “cloth” (#2210).
Nahali kupra (Kuiper 1962 #323)
- 17 -
Dorian Q Fuller
relegated to “Language X”. While I have previously
This term refers to the activity of weaving, which
hypothesized that “Language X” might be associated
is one form or another is likely to be universal and
with the Ganges Neolithic, the range of plant taxa
Palaeolithic, rather than to any particular product,
found in this language suggests instead that it accords
such as cotton or flax.
with some part of the Greater Harappan zone. It may
THE SPREAD OF COTTON
BEYOND SOUTH ASIA
well be that Harappan language was itself already a
mixed language, combining Language X (of unknown
affinity) and the “Austric-oid” kubha-vipas. As both
this term and the preceding cotton term appears to
The eastward spread of cotton appears to be tracked
be substrate loan words, it is not possible to use the
by historical linguistic data. As indicated in Table
linguistic evidence to suggest their antiquity, except
4, one set of Indic cotton terms derives from a
that they appear to the South Asian, as they are
South Dravidian term originally denoting “feathers”,
absent from Iranian, and pre-Indo-Aryan. This fits
Proto-South Dravidian *tuu- , Sanskrit tūla-. This
with lost substrate language(s) in the northwestern
appears to be the source of cotton terms in some
subcontinent and with the known archaeological
Munda languages (e.g. Kharia turai), and some
antiquity of both these crops as at least Harappan or
Southeast Asian languages, including Monic (Old
older as cultivars in the Indus region.
Mon tol, Modern Mon tow, Nyakur tual. L (Peiros
In addition to terms for the fibre plants themselves,
and Starostin 2003). This differs from another set
we are able to identify Sanskrit terms for some of the
of related terms, which are derived instead from the
processes involved in processing them, especially for
other Sanskrit term karpā’sa (Turner 1966: CDIAL
cotton, as well as some equivalent words in Munda
2877). Loans are found in some Munda languages
and Dravidian languages. The Sanskrit terms are
(Karia and Juang kapas, Gorum and Remo kapa, Gta
identified by Schlingloff (1974) on the basis of early
kopa, Mundari ka’dsom: from Donegan and Stampe
Jain and Buddhist texts, as well as some modern
2004b) and in several Southeast Asian Austroasiatic
(Hindi) terms. Consideration in terms of historical
branches: Old Khmer krəpa:s, Proto-Viet-Muong
linguistics to track these as cognates or loans in various
*k-pa:lh, Proto-Katuic, Proto-Banharic and Proto-
languages is needed, as is work on the equivalent
Pearic *kə-pa:jh (Peiros and Starostin 2003; cf. Osada
terms in Dravidian or Munda languages, although
2006: 163-164). This root is also borrowed into
a few terms are collected here. The table offered
Austronesian languages such as Malayan and Batak
is therefore only a starting point for such research
(Osada 2006: 163). The distinct etyma borrowed
(Table 5). Further compilation of alternative or
into Mon and Khmer would suggest that cotton
cognate terms in other languages, especially amongst
arrived in Southeast Asia after the divergence of
Dravidian and Munda languages, and possible loans
the Mon-Khmer family, although the divergence of
amongst Southeast Asian languages is needed. Of
these is generally considered much earlier than the
note are several terms that are shared between Indic
likely arrival of cotton (cf. Diffloth 2005), which is
and some Munda languages, as well as a few shared
presumably during or after the Early Historic period
with Proto-South Dravidian. One widespread term
of trade between India and Southeast Asia starting
for weaving/weaver (# tan) is perhaps originally
from the end of the First Millennium BC. To the
Austric(oid), as it is widespread in Munda languages,
north, the first cotton fabrics apparently reached
Eastern-Austroasiatic, and appears related to the
China as “tribute” from Java in AD 430, while cotton
Sanskrit tántu, perhaps then an earlier substrate term.
fabrics from Gangetic India were sent as “tribute” to
- 18 -
The spread of texitile production and textile crops in India beyond the Harappan zone
the Chinese court at the start of the Sixth Century
process of establishment of cotton cultivation across
(Goodrich 1943). Nevertheless during the Tang
the southern frontiers of the Roman world, indicated
Dynasty and until the 13th Century AD, cotton does
by finds of seeds in the Southern Egyptian Oases
not appear to have been a widely known product in
(Dakleh: Thanheiser 1999; Khargeh: A.J. Clapham,
China (Laufer 1919: 490-492; Goodrich 1943).
unpublished; cf. Pelling 2005: 406; Clapham and
The westward diffusion of cotton is rather better
Rowley-Conwy, in press), and in Southern Libya,
tracked through archaeology, and is a process that
the kingdom of the Garamantes (Pelling 2005), as
takes place mainly in the Roman era (less than 2000
well as in Nubia (Clapham and Rowley-Conwy, in
years ago). There is no evidence that cotton came
press). It is tempting to link this new region of cotton
to be grown in the Mediterranean region or Egypt
cultivation to the spread of Gossypium herbaceum
in the Bronze Age or early Iron Age. Indeed, during
originating in sub -Saharan Africa, which was
the Roman period, cotton textiles were one of the
tentatively identified from desiccated capsule remains
desired products from Indian trade ports, as indicated
at Qasr Ibrim (Rowley-Conwy 1989; cf. Clapham and
in the Periplus Maris Erythraei, a First Century AD
Rowley-Conwy, in press; Wild et al. 2007), but it is
Roman mariner’s travel guide, written in Greek
also possible that this represents the introduction of
probably in Egypt (see Casson 1989). Cotton is also
tree cotton from India. In the latter connection, one
indicated as an import in Papyrus Vinod (Casson
wonders whether the Nubian term (Nobiin) koshmaag
1990; Sidebotham 1991). It is suggested that Indian
(cf. Fuller and Edwards 2001) might also be derived
imported textiles had Z-spun thread, which dominate
ultimately from the Sanskrit karpā’sa ?
the archaeological textile record at the Roman era
CONCLUSION
port of Berenike on the Red Sea coast (Wild and
Wild 1998, 2001, 2005). Cotton is rarely attested in
papyrological records from Roman Egypt (Bagnall
The evidence reviewed in this paper allows us to
1993: 33, n. 123). but is first recorded as a local
assess the role of textile production in the “Neolithic
cultivar in the Second Century AD in the Khargeh
revolution” in the South Asian context. When
Oasis, and a Fourth Century AD cultivar in the
defining the “Neolithic Revolution”, Childe (1936)
Dakhleh Oasis (Winter and Youtie 1944; Bagnall
considered textile production, together with ceramics,
1993). Already in the First Century AD, Pliny in
as one of the technological hallmarks of the Neolithic.
his Naturalis Historia described the cotton crop and
A “self-sufficing economy”, i.e. food-production, based
indicated cultivation in Nubia and parts of upper
on domesticated plants and animals was his main
Egypt (see Clapham and Rowley-Conwy in press).
focus, but he suggested that the breeding of woolly
Archaeological finds suggest that cotton cultivation
animals and cultivation of fibrous plants would have
and fibre-processing was established in the Meroitic
also made textiles part of this transition. It is clear
Kingdom of Nubia, as indicated by finds of desiccated
that in South Asia, and I suspect elsewhere, this was
seeds and capsules from Qasr Ibrim (Rowley-Conwy
not the case, in terms of a strict sense of Neolithic
1989; Clapham and Rowley-Conwy 2006, 2007, in
beginnings. However, neither is pottery, as pre-
press), as well as quantities of textiles from Lower
ceramic food production is clearly in evidence in
Nubia (Crowfoot and Griffiths 1934; Bergman 1975;
Southwest Asia, Pakistan, and parts of the Americas
Crowfoot et al. 1977: 46; Crowfoot 1979; Mayer-
(e.g. Bar-Yosef and Meadow 1995; Crown and Wills
Thurman and Williams 1979; Adams 1986: 507;
1995; Smith 1992; Burger 1992; Jarrige et al. 2006;
Wild et al. 2007). This must be seen as part of wider
see discussion in Fuller 2006a: 60). Pottery, however,
- 19 -
Dorian Q Fuller
remains an important development in the technology
production.
of food processing (grinding and pulverizing tools
could be cited as another important technology). It
Acknowledgements
is clear that together with food production, changes
This paper was prepared while a visiting fellow with
in processing which allowed the more intensive
Professor Y.-I. Sato at the Research Institute for
extraction of nutrients and the more complex cultural
Humanity and Nature and supported by a Japanese
transformation and combination of raw foods, was
Society for the Promotion of Science fellowship.
an important and recurrent feature of the changes
I thank Ruth Pelling for teaching me more about
that occurred between the pure foraging of the
cotton through her research on Saharan material.
Palaeolithic and the development of economies
Thanks to Alan Clapham for discussion and updating
that supported hierarchical, complex societies and
me on Eg yptian evidence. I especially grateful to
“civilization”. If the Neolithic is re-conceived as an
Toshiki Osada for commenting on my linguistics
extended period of directional transformations in
section and providing me with additional linguistic
human economy and social organization, which
source materials, and sharing his knowledge of Hindi
provided the necessary basis for the development of
and Munda. Any errors remain my own.
“complexity”, then textile production should indeed
be included in this. Rather than being formative of the
Notes
Neolithic, textiles appear to have been transformative,
1) See also Sherratt’s on-line ArchAtlas: http://www.
in that the development of, and increasing scale of,
archatlas.dept.shef.ac.uk/OriginsFarming/Farming.
textile production and the growing of textile crops,
php
2) Or Woman.
was part of a wider process of craft specialization
and commodification that was necessary to the
References
development of larger polities, states and cities. It
Adams, W.Y. (1986) Ceramic Industries of Medieval
is clear that in South Asia textile production was
Nubia, parts I and II, Memoirs of the UNESCO
earliest in the northwest and was well-established for
Archaeological Survey of Sudanese Nubia, vol. I.
the development of Harappan urbanism. Elsewhere,
University Press of Kentucky.
textile production is later. Rather than seeing this as
Allaby, R.G., G.W. Peterson, D.A. Merriwether, and Y.-B. Fu
“fall out” from the Harappan civilization, however, I
(2005) Evidence of the domestication history of flax
would suggest that this indicates the internal drives
(Linum usitatissimum L.) from genetic diversity of
in early village (Neolithic) societies in various parts
sad2 locus. Theoretical and Applied Genetics 112: 58-
of India as they developed more complex economic
65.
interdependencies, internal hierarchical social
Allchin, B., and F. R. Allchin (1982) The Rise of Civilization
competition, and moved towards “Bronze Age”
in India and Pakistan. Cambridge University Press,
economies (which in many parts of India are rather
Cambridge.
Anonymous (1833) Das Pfennig-Magazin der Gesellschaft zur
associated with the Iron Age). Thus by combining
Verbreitung gemeinnütziger Kenntnisse 11 Mai 1833 -
Childe and Twain : man 2) makes and remakes
Nr. 2. Retrieved from internet: http://ngiyaw-ebooks.
himself through the technological developments
de/ngiyaw/pfennigmagazine.htm, on 27 July 2007.
of the past, but clothes too re-make the man, and
Ansari, Z.D. and M.S. Nagaraja Rao (1969) Excavations at
documenting the beginnings and intensification of
Sanganakallu - 1964-65. Deccan College, Pune.
textile production may be as important to tracking the
Asouti, E., D.Q. Fuller and R. Korisettar (2005) “Vegetation
early transformations of society as the origins of food
context and wood exploitation in the Southern
- 20 -
The spread of texitile production and textile crops in India beyond the Harappan zone
London.
Neolithic,” in U. Franke-Vogt and J. Weisshaar (eds.)
Burrow, T. and M. B. Emeneau. (1984) A Dravidian
South Asian Archaeology 2003. Linden Soft, Aachen.
Etymological Dictionary, second edition. Clarendon
pp.336-340.
Press, Oxford.
Bagnall, R.S. (1993) Eg ypt in Late Antiquity. Princeton
Casson, L. (1989) The Periplus Maris Erythraei. Text with
University Press, Princeton.
Introduction, Translation, and Commentary. Princeton
Bar-Yosef, O. and R . Meadow (1995) “The Origins of
University Press, Princeton.
Agriculture in the Near East,” in T.D. Price and A.B.
Gebauer (eds.) Last Hunters-First Farmers. New
Casson, L. (1990) New Light on Maritime Loans: P. Vindob.
Perspectives on the Prehistoric Transition to Agriculture.
G. 40822. Zeitschrift für Papyrologie und Epigraphik
School of American Research Press, Santa Fe. pp.39-
84: 195-206.
Chanchala, S. (1992) The fruit and seed remains from ancient
94.
Hulaskhera, District Lucknow, U.P. (c. 700 B.C.–500
Barber, E.J.W. (1991) Prehistoric textiles. Cambridge
A.D.). Pragdhara 2: 65–80.
University Press, Cambridge.
Chanchala, S. (1995) Some seed and fruit remains from
Barber, E.J.W. (1999) The Mummies of Urumchi. Norton
Kausambi, District Allahabad, U.P. (ca 600 B.C.-450
Bellinger, New York.
B.C.). Geophytology 24: 169-172.
Berg, C., and C.F. Schmidt (1858-1863) Darstellung und
Beschreibung sämtlicher in der Pharmacopoea Borusica
Chanchala, S. (2002) “Botanical remains,” in D.P. Tewari (ed.)
aufgeführten offizinellen Gewächse. Arthur Felix,
Excavations at Charda. Jarun Prakashan, Lucknow.
Leipzeig.
pp.166–94.
Bergman, I. (1975) Late Nubian Textiles. Scandinavian Joint
Chapman, J. (2000) “Tension at funerals: social practices
Expedition to Nubia, volume 8. Scadinavian University
and the subversion of community structure in later
Books, Copenhagen, Oslo, Stockholm.
Hungarian prehistory,” in M.-A. Dobres and J. Robb
(eds.) Agency in Archaeology. Routledge, London. pp.
Betts, A., K. van der Borg, A. de Jong, C. McClintock and M.
169-195.
van Stryndonck (1994) Early Cotton in North Arabia.
Childe, V.G. (1936) Man Makes Himself. Watts and Co.,
Journal of Archaeological Science 21: 489-499.
London.
Bhattacharya, S. (1968) A Bonda Dictionary. Deccan College
Choudhar y, B. and G. Laroia (2001) Technolog ical
Post-Graduate and Research Institute, Pune.
developments and cotton production in India and
Biligiri, H.S. (1965) Kharia. Phonolog y, Grammar and
China. Current Science 80: 925-932.
Vocabulary. Deccan College Postgraduate and
Chowdhury, K.A. and G.M. Buth. (1971) Cotton seeds from
Research Institute, Pune.
Boivin, N.L., R. Korisettar, and D.Q. Fuller (2005) Further
the Neolithic in Egyptian Nubia and the origin of
research on the Southern Neolithic and the Ashmound
Old World Cotton. Biological Journal of the Linnean
Tradition: The Sanganakallu-Kupgal Archaeological
Society 111: 303-312.
Chowdhury, K.A. and G.M. Buth (2005) Plant remains from
Research Project. Journal of Interdisciplinary Studies in
excavation of terraces of the Nile at Afyeh, Nubia and
History and Archaeology 2: 59-86.
Egypt. Purattatva 35: 154-159.
Bourdieu, P. (1984) Distinction. A Social Critique of the
Clapham, A. and P. Rowley-Conwy (2006) Rewriting the
Judgement of Taste. Har vard University Press,
History of African Agriculture. Planet Earth Summer
Cambridge, Mass.
2006: 24-26 [http://www.nerc.ac.uk/publications/
Burger, R . L. (1992) Chavin and the origins of Andean
planetearth/]
civilization. Thames and Hudson, London.
Burkill, I. H. (1966) A dictionary of the economic products
Clapham, A. and P. Rowley-Conwy (2007) “New Dsicoveries
of the Malay peninsula, second edition. Ministry of
at Qasr Ibrim, Lower Nubia,” in R. Cappers (ed.)
Agriculture and Co-operatives, Kuala Lampur.
Fields of Change. Proceedings of the 4th International
Burkill, I.H. (1997) The Useful Plants of West Tropical Africa,
Workshop for African Archaeobotany. Barkhuis &
Groningen University Library, Groningen.
second edition. vol. 4. Royal Botanic Gardens, Kew,
- 21 -
Dorian Q Fuller
Clapham, A . and P. Rowley- Conwy (In press) “ The
Southwest,” in W.K. Barnett and J.W. Hoopes (eds.)
archaeobotany of cotton (Gossypium sp. L) in Egypt
The Emergence of Pottery: Technology and Innovation
and Nubia with special reference to Qasr Ibrim,
in Ancient Societies. Smithsonian Institution Press,
Egyptian Nubia,” in A. Fairbairn and E. Weiss (eds.)
Washington, D.C. pp.241–256.
Dales, G.F. (1986) “Some fresh approaches to old problems
Ethnobotanist of Distant Pasts: Essays in Honour of
in Harappan archaeology,” in J. Jacobson (ed.) Studies
Gordon Hillman. Oxbow Books, Oxford.
in the Archaeology of India and Pakistan. Oxford and
Cooke, M. (2002) Investigating Changing Agricultural
IBH, Delhi. pp.117-136.
Production and Patterns of Subsistence at Ancient Hund,
Dhavalikar, M.K., H.D. Sankalia and Z.D. Ansari (eds.) (1988)
in the Vale of Peshawar, Pakistan. BSc Dissertation,
Excavations at Inamgaon. Vol. 1, pt. i. Deccan College
Institute of Archaeology, University College London.
Postgraduate and Research Institute, Pune.
Cooke, M., D.Q. Fuller and K . Rajan. (2005) “Early
Historic Agriculture in Southern Tamil Nadu:
Diffloth, G. (2005) “ The contribution of ling uistics
Archaeobotanical Research at Mangudi, Kodumanal
palaeotology to the homeland of Austro-asiatic,” in L.
and Perur,” in U. Franke-Vogt and J. Weisshaar (eds.)
Sagart, R. Blench, and A. Sanchez-Mazas (eds.) The
South Asian Archaeology 2003. Linden Soft, Aachen.
Peopling of East Asia: Putting together archaeology,
pp.329-334.
linguistics and genetics. Routledge, London. pp.79-82.
Costantini, L. (1979) “Plant remains at Pirak,” in J.-F. Jarrige
Donegan, P.J. and D. Stampe (2004a) “Comparative Munda
and M. Santoni (eds.) Fouilles de Pirak, vol. 1.
(mostly North).” Accessed from internet 30 July 2007,
Diffusion de Boccard, Paris. pp.326-333. .
http://www.ling.hawaii.edu/faculty/stampe/aa.html
Costantini, L. (1983) “The beginning of agriculture in the
Donegan, P.J. and D. Stampe (2004b) “South Munda Cognate
Kachi Plain: the evidence of Mehrgarh,” in B. Allchin
Huntng Files.” Accessed from internet 30 July 2007,
(ed.) South Asian Archaeolog y 1981. Cambridge
http://www.ling.hawaii.edu/faculty/stampe/aa.html
Douglas, M. (1973) Natural Symbols, revised paperback
University Press, Cambridge. pp.29-33.
edition. Penguin Books, London.
Costantini, L. (1987) “Appendix B. Vegetal remains,” in
G. Stacul (ed.) Prehistoric and Protohistoric Swat,
Earle, T. (2003) Bronze Age Economics. The Beginnings
Pakistan. Instituto Italiano per il Medio ed Estremo
of Political Economies. Westview Press, Boulder,
Orientale, Rome. pp.155-165.
Colorado.
Costantini, L. (1990) “Harrapan agriculture in Pakistan: the
Eicher, J.B. (1998) “Beaded and Bedecked Kalabari of
evidence of Nausharo,” in M. Taddei (ed.) South Asian
Nigeria,” in L.D. Sciama and J.B. Eicher (eds.) Beads
Archaeology 1987. Instituto Italiano per il Medio ed
and Beadmakers. Berg, Oxford. pp.95-116.
Engler, A. (1937) Syllabus der Pflanzenfamilien, 12th edition.
Estremo Oreintale, Rome. pp.321-332.
Gebrüder Bornträger, Berlin.
Costantini, L. and L.C. Biasini. (1985) Agriculture in
Fuller, D.Q. (2002) “Fifty Years of Archaeobotanical Studies
Baluchistan between the 7th and 3rd Millenium B.C.
in India: Laying a Solid Foundation,” in S. Settar and
Newsletter of Baluchistan Studies 2: 16-37.
Crowfoot, G.M. and F.L. Griffiths (1934) On the Early Use
R. Korisettar (eds.) Indian Archaeology in Retrospect,
of Cotton in the Nile Valley. Journal of Eg yptian
Volume III. Archaeology and Interactive Disciplines.
Manohar, Delhi. pp.247-363.
Archaeology 20: 5-12.
Crowfoot, E., J.M. Plumley and W.Y. Adams. (1977) Qasr
Fuller, D.Q. (2003) “An Agricultural Perspective on Dravidian
Ibrim, 1976. Journal of Egyptian Archaeology 63: 29-
Historical Linguistics: Archaeological Crop Packages,
47.
Livestock and Dravidian Crop Vocabulary,” in P.
Crowfoot, E. (1979) “Textiles” in R.D. Anderson and W.Y.
Bellwood and C. Renfrew (eds.) Examining the
Adams, “Qasr Ibrim, 1978”. Journal of Eg yptian
farming/language dispersal hypothesis. McDonald
Archaeology 65: 39-40.
Institute Monographs. McDonald Institute for
Archaeological Research, Cambridge. pp.191-213.
Crown, P.L. and W.H. Wills (1995) “Economic intensification
Fuller, D.Q. (2005) Ceramics, seeds and culinary change in
and the origins of ceramic containers in the American
- 22 -
The spread of texitile production and textile crops in India beyond the Harappan zone
Hill, J.D. (1997) “The end of one kind of body and the
prehistoric India. Antiquity 79: 761-777.
Fuller, D.Q. (2006a) Agricultural Origins and Frontiers in
beginning of another kind of body’? Toilet instruments
South Asia: A Working Synthesis. Journal of World
and ‘Romanization’ in Southern England during the
Prehistory 20: 1-86.
first century AD,” in A. Gwilt and C. Haselgrove (eds.)
Reconsturcting Iron Age Societies. New Approaches to the
Fuller, D.Q. (2006b) “Silence before sedentism and the advent
British Iron Age. Oxbow Books, Oxford. pp.96-107.
of cash-crops: a status report on early agriculture
in South Asia from plant domestication to the
Hoffman, J. (1930-1938) Encyclopedia Mundarica. Patna
development of political economies (with an excursus
Hutchinson, J.B. and R.L.M. Ghose (1937) The classification
on the problem of semantic shift amongst milets
of cottons of Asia and Africa. Indian Journal of
and rice),” in T. Osada (ed.) Proceedings of the Pre-
Agricultural Science 7: 233-257.
Symposium of RIHN and 7th ESCA Harvard-Kyoto
Jarrige, J.-F., C. Jarrige, and G. Quivron (2006) “Mehrgarh
Roundtable. Research Institute for Humanity and
Neolithic: the updated sequence,” in C. Jarrige and V.
Nature, Kyoto. pp.175-213.
Lefèvre (eds.) South Asian Archaeology 2001. Editions
Recherche sur les Civilisations, Paris. pp.129-141.
Fuller, D.Q. (2007) “Non-human genetics, agricultural
origins and historical linguistics in South Asia,” in
Kajale, M.D. (1977) On the botanical findings from
M.D. Petraglia and B. Allchin (eds.) The Evolution
excavations at Daimabad, a Chalcolithic site in
and History of Human Populations in South Asia:
Western Maharashtra, India. Current Science 46: 818-
Inter-disciplinary Studies in Archaeology, Biological
819.
K a j a l e , M . D . ( 1 9 9 1 ) “ C u r r e n t s t a t u s o f In d i a n
Anthropolog y, Linguistics and Genetics. Springer,
palaeoethnobotany : introduced and indigenous
Doetinchem, The Netherlands. pp.393-443.
Fu l l er, D.Q. a n d D. N . E d wa r d s ( 2 0 0 1 ) Me d i e va l
food plants with a discussion of the historical and
Plant Economy in Middle Nubia : Preliminar y
evolutionary development of Indian agriculture and
Archaeobotanical Evidence from Nauri. Sudan and
agricultural systems in general,” in J.M. Renfrew (ed.)
Nubia 5: 97-103.
New Light on Early Farming - Recent Developments
in Palaeoethnobotany. Edinburgh University Press,
Fuller, D.Q. and E. Harvey (2006) The archaeobotany of
Edinburgh. pp.155-189.
Indian pulses: identification, processing and evidence
Kajale, M.D. (1996) Palaeobotanical Investigations at
for cultivation. Environmental Archaeology 11: 241-
Balathal: Preliminary Results. Man and Environment
268.
21: 98-102.
Fuller, D.Q. and M. Madella (2001) “Issues in Harappan
Archaeobotany : Retrospect and Prospect,” in S.
Keng, H. (1974) Economic Plants of Ancient North China as
Settar and R. Korisettar (eds.) Indian Archaeology in
mentioned in Shih Ching (Book of Poetry). Economic
Retrospect, vol. II: Protohistory. Manohar, New Delhi.
Botany 28: 391-410.
Kharakwal, J.S., Y.S. Rawat and T. Osada (2007) “Kanmer: A
pp.317-390.
Fuller, D.Q., N. Boivin, and R. Korisettar (2007) Dating the
Harappan site in Kachchh, Gujarat, India,” in T. Osada
Neolithic of South India: new radiometric evidence
(ed.) Linguistics, Archaeology and the Human Past,
for key economic, social and ritual transformations.
Occasional Paper 2. Indus Project, Research Institute
for Humanity and Nature, Kyoto. pp.21-46..
Antiquity 81: 755-778
Fuller, D.Q., R. Korisettar, P.C. Venkatasubbaiah and M.K.
Korisettar, R ., P.C. Venkatasubbaiah and D.Q. Fuller
Jones (2004) Early plant domestications in southern
(2001) “Brahmagiri and Beyond: the Archaeology
India: some preliminary archaeobotanical results.
of the Southern Neolithic,” in R. Korisettar and S.
Vegetation History and Archaeobotany 13: 115-129.
Settar (eds.) Indian Archaeology in Retrospect, vol. I:
Prehistory. Manohar, New Delhi. pp.151-237.
Goodrich, L.C. (1943) Cotton in China. Isis 34: 408-410.
Kuiper, F.B.J. (1962) Nahali. A Comparative. N. V. Noord-
Gulati, A.N. and A.J. Turner (1929) A note on the early
Hollandsche Uitgevers Maatschappij, Amsterdam.
history of cotton. Journal of the Textile Institute 20:
Langer, R.H.M. and G.D. Hill (1982) Agricultural Plants.
1-9.
- 23 -
Dorian Q Fuller
Pawankar, S.J. and P.K. Thomas (1997) Fauna and subsistence
Cambridge University Press, Cambridge.
Laufer, B. (1919) Sino-Iranica. Chinese contributions to the
pattern in the Chalcolithic culture of Western India,
history of civilization in Ancient Iran. Anthropological
with special reference to Inamgaon. Anthropozoologica
Series Vol. XV, No. 3. Field Museum of Natural
25-26: 737-746.
Peiros, I. and S. Starostin (2003) “Austro-Asiatic etymology
History, Chicago.
Masica, C.P. (1979) “Aryan and Non-Aryan Elements in
(Tower of Babel Database),” accessed from internet on
North Indian Agriculture,” in M.M. Deshpande and
25 July 2007, http://starling.rinet.ru/cgi-bin/ main.
cgi?root=config&map=austric
P.E. Hook (eds.) Aryan and Non-Aryan in India.
Pelling , R . (2005) Garamantean agriculture and its
Center for South and Southeast Asian Studies,
University of Michigan, Ann Arbor. pp.55-151.
significance in a wider North African context: the
Matson, D. M. (1964) A Gramatical Sketch of Juang, a Munda
evidence of plant remains from the Fazzan Project. The
Journal of North African Studies 10: 397-411.
Language. PhD Dissertation, University of Wisconsin.
Pelling , R . (2007) Agriculture and Trade amongst the
University Microfilms, Ann Arbor.
Mayer-Thurman, C.C. and B.B. Williams (1979) Ancient
Garamantes: 3000 years of archaeobotanical data
Textiles from Nubia. Art Institute of Chicago,
from the Sahara and its margins. PhD Dissertation,
Chicago.
University College London.
Pinnow, H.-J. (1959) Versuch einer historischen lautlehre der
McCorriston, J. (1997) The Fiber Revolution. Textile
E xt ens i f i c ati o n , A l i enati c ati o n , a n d S o c i a l
Karia-sprache. Otto Harrassowitz, Wiebaden.
Stratification in Ancient Mesopotamia. Current
Pokharia, A.K. and K.S. Saraswat (1999) Plant economy
during Kushana period (100-300 AD) at acient
Anthropology 38: 517-549.
Sanghol. Pragdhara 9: 75-122.
McKean, M.B. (1983) The palynolog y of Balakot, a pre-
Pokharia, A.K. and K.S. Saraswat (2004) “Plant resources
Harappan and Harappan age site in Las Bela, Pakistan.
at Ojiyana, Rajasthan.” National Seminar on the
PhD Dissertation, Southern Methodist University.
Moulherat, C., M. Tengberg, J.-F. Haquet, and B. Mille (2002)
Archaeolog y of the Ganges Plain, Joint Annual
First Evidence of Cotton at Neolithic Mehrgarh,
Conference of the Indian Archaeological Society, Indian
Pakistan: Analysis of Mineralized Fibres from a
Society of Prehistoric and Quaternary Studies, Indian
Copper Bead. Journal of Archaeological Science 29:
History and Culture Society, Lucknow, 2004. Abstracts.
Reddy, S.N. (1994) Plant Usage and Subsistence Modeling: An
1393-1401.
Nagaraja Rao, M.S. and K.C. Malhotra (1965). Stone Age Hill
Ethnoarchaeological Approach to the Late Harappan
Dwellers of Tekkalakota. Deccan College, Pune.
of Northwest India. PhD Dissertation, University of
Wisconsin.
Nagaraja Rao, M.S. (1971) Protohistoric Cultures of the
Reddy, S.N. (2003) Discerning Palates of the Past: an
Tungabhadra Valley. Nagaraja Rao, Dharwad.
Osada, T. (2006) “How many Proto-Munda words in
ethnoarchaeological study of crop cultivation and
Sanskrit? - with special reference to aricultural
plant usage in India. Ethnoarchaeological Series 5,
vocabulary,” in T. Osada (ed.) Proceedings of the Pre-
International Monographs in Prehistory. Prehistory
Press, Ann Arbor.
Symposium of RIHN and 7th ESCA Harvard-Kyoto
Robbins, G. (1931) The Botany of Crop Plants, third edition.
Roundtable. Research Institute for Humanity and
Blakiston and Son, Philadelphia.
Nature, Kyoto. pp.151-174.
Paddayya, K. (1993) Further field investigations at Budihal.
Rowley-Conwy, P. (1989) Nubia AD 0-550 and the “Islamic”
Bulletin of the Deccan College Post-Graduate and
Agricultural Revoltuion: Preliminary Botanical
Research Institute 53: 277-322.
Evidence from Qa sr Ibrim, Eg yptian Nubia .
Archeologie du Nil Moyen 3: 131-138.
Paddayya, K. (2001) The problem of ashmounds of Southern
Deccan in the light of the Budihal excavations,
Runnels, C.N. and T.H. Van Andel (1988) Trade and the
Karnataka. Bulletin of the Deccan College Post-
origins of agriculture in the eastern Mediterranean.
Graduate and Research Institute 60-61: 189-225.
Journal of Mediterranean Archaeology 1: 83-109.
- 24 -
The spread of texitile production and textile crops in India beyond the Harappan zone
and agricultural significance. Centre for Agricultural
Ryder, M.L. (1965) Report of textiles from Catal Huyok.
Publishing and Documentation, Wageningen.
Anatolian Studies 15: 175-176.
Sherratt, A. (1981) “Plough and pastorlism: aspects of the
Ryder, M.L. (1984) “Sheep,” in I.L. Mason (ed.) Evolution of
secondary products revolution,” in I. Hodder, G. Isaac,
domesticated animals. Longman, London. pp.63-86.
Sanathnam, V. and J.B. Hutchinson (1974) “Cotton,” in J.B.
and N. Hammond (eds.) Pattern of the Past: Studies in
Hutchinson (ed.) Evolutionary Studies in World Crops.
Honour of David Clarke. Cambridge University Press,
Cambridge University Press, Cambridge. pp. 89-100.
Cambridge. pp.261-305.
Sankalia, H.D., S.B. Deo, Z.D. Ansari and S. Ehrhardt (1960)
Sherratt, A. (1995) Reviving the grand narrative: archaeology
From History to Prehistory at Nevasa (1954-56).
and long-term change. Journal of European Archaeology
Deccan College, Pune.
3: 1-32.
Sara swat, K .S. (1986) Ancient crop rema ins from
Sherratt, A. (1999) “Cash-crops before cash: organic
Sringeverapura, Allahabad, U.P. (ca. 1,050-700 B.C.).
consumables and trade,” in C. Gosden and J. Hather
Geophytology 16(1): 97-106.
(eds.) The Prehistory of Food. Appetites for Change.
Routledge, London. pp.13-34.
Saraswat, K.S. (1993) Plant economy of Late Harappans at
Sherratt, A. (2007) “Diverse origins: regional contributions to
Hulas. Purattatva 23: 1-12
Saraswat, K.S. (1997) Plant Economy of Barans at Ancient
the genesis of farming,” in S. Colledge and J. Conolly
Sanghol (Ca. 1900-1400 B.C.), Punjab. Pragdhara 7:
(eds.) The Origins and Spread of Agriculture in Europe.
97-114.
Archaeobotanical Investigations of Neolithic Plant
Economies. Left Coast Press, Walnut Creek, CA.
Saraswat, K.S. (2002) Banawali (29°37’5”N; 75°23’6”E),
pp.1-20.
District Hissar. Indian Archaeolog y 1996-97- A
Sidebotham, S.E. (1991) “Ports of the Red Sea and the Arabia-
Review: 203.
Saraswat, K.S. (2004) “Plant economy of early farming
India Trade,” in V. Begley and R.D. de Puma (eds.)
communities at Senuwar, Bihar,” in B. P. Singh (ed.)
Rome and India: The Ancient Sea Trade. University of
Senuwar Excavations. Banares Hindu University,
Wisconsin Press, Madison. pp.12-38.
Singh, B.P. (2004) Senuwar Excavations. Banares Hindu
Varanasi.
University, Varanasi.
Saraswat, K.S. (2005) Agricultural background of the early
Smith, B.D. (1992) Rivers of Change. Essays on Early
farming communities in the Middle Ganga Plain.
Agriculture in Eastern North America. Smithsonian,
Pragdhara 15: 145-177.
Washington, D.C.
Saraswat, K.S. and A.K. Pokharia (2003) Palaeoethnobotanical
Southworth, F. (1988) “Ancient economic plants of South
investigations at Early Harappan Kunal. Pragdhara 13:
Asia: linguistic archaeology and early agriculture,” in
105-140.
Saraswat, K.S. N.K. Sharma and D.C. Saini (1994) “Plant
M.A. Jazayery and W. Winter (eds.) Languages and
Economy ay Ancient Narhan (Ca. 1,300 B.C. -
Cultures. Studies in Honor of Edgar C. Polome. Mouton
300/400 A.D.),” in P. Singh (ed.) Excavations at
de Gruyter, Amsterdam. pp.649-688.
Southworth, F. (2005) The Linguistic Archaeology of South
Narhan (1984-1989). Banaras Hindu University,
Asia. Routledge, London.
Varanasi. pp.255-346.
Sayre, L.E. (1917) A Manual of Oragnic Materia Medica and
Southworth, F. (2006) “Proto-Dravidian Agriculture,” in
Pharacognisy, 4th edition. P. Blakiston’s Son & Co.,
T. Osada (ed.) Proceedings of the Pre-Symposium of
Philadelphia.
RIHN and 7th ESCA Harvard-Kyoto Roundtable.
Research Institute for Humanity and Nature, Kyoto.
Schilling , C. (1993) The Body and Social Theory. Sage,
pp.121-150.
London.
Spring, C. and J. Hudson (1995) North African Textiles.
Schlingloff, D. (1974) Cotton-manufacture in ancient India.
British Museum Press, London.
Journal of the Economic and Social History of the Orient
Tengberg, M. (1999) Crop husbandry at Miri Qalat, Makran,
17: 81-90.
SW Pakistan (4000-2000 B.C.). Vegetation History
Seegler, C. J. P. (1983) Oil plants in Ethiopia, their taxonomy
- 25 -
Dorian Q Fuller
Indus Ethnobiology. New Perspectives from the Field.
and Archaeobotany 8: 3-12.
Lexington Books, Lanham. pp.175-198.
Tewari, R., R.K. Srivastava, K.K. Singh, K.S. Saraswat, I.B.
Singh, M.S. Chauhan, A.K. Pokharia, A. Saxena, V.
Wendel, J.F. (1995) “Cotton,” in J. Smartt and N.W.
Prasad and M. Sharma (2006) Second preliminary
Simmonds (eds.) Evolution of Crop Plants, second
report of excavations at Lahuradewa District Sant
edition. Longman, London. pp.358-366.
Wheeler, R.E.M. (1948) Brahmagiri and Chandravalli 1947:
Kabir Naga, U.P.: 2002-2003-2004 & 2005-06.
Megalithic and Other Cultures in Mysore State.
Pragdhara 16: 35-68.
Ancient India 4: 180-230.
Thanheiser, U. (1999) “Plant Remains from Kellis: First
Results,” in C. A. Hope and A. J. Mills (eds.) Dakhleh
Wild, J.P. and F.C. Wild (1998) “The Textiles,” in S. E.
Oasis Project: Preliminary Report on the 1992–1993
Sidebotham and W.Z. Wendrich (eds.) Berenike 96:
and 1993–1994 Field Seasons. Oxbow Books, Oxford.
Report of the Excavations at Berenike (Egyptian Red Sea
pp.89-93.
Coast) and the Survey of the Eastern Desert. Research
School CNWS, Leiden. pp.221-236.
Thomas, P.K . (1988) “Faunal Assemblag e,” in M.K .
Dhavalikar, H.D. Sankalia, and Z.D. Ansari (eds.)
Wild, J.P. and F.C. Wild (2001) Sails from the Roman port at
Excavations at Inamgaon. Deccan College Post-
Berenike, Egypt,. The International Journal of Nautical
Graduate and Research Institute, Pune. pp.823-961.
Archaeology 30:211-220.
Traherne, P. (1995) The warrior’s beauty: the masculine body
Wild, J.P. and F.C. Wild (2005) “Rome and India: early
and self-identity in Bronze Age Europe. Journal of
Indian cotton textiles from Berenike, Red Sea coast
European Archaeology 3: 105-144.
of Egypt,” in R. Barnes (ed.) Textiles in Indian Ocean
Studies. Routledge, London. pp.11-16.
Tuck, A. (2004) Singing the Rug: Patterned Textiles and the
Wild, J.P., F.C. Wild and A.J. Clapham (2007) Irrigation
Origins of Indo-European Metrical Poetry. American
and the spread of cotton growing in Roman times.
Journal of Archaeology 110: 539-550.
Archaeological Textiles Newsletter 44: 16-18.
Turner, R.L. (1966) A comparative Dictionary of the Indo-
Willcox, G. (1992) “Some differences between crops of Near
Aryan Languages. London.
Vavilov, N. (1950 [1992]) The origin, variation, immunity,
Eastern origin and those from the tropics,” in C. Jarrige
and breeding of cultivated plants, 1992 reprint edition.
(ed.) South Asian Archaeology 1989. Prehistory Press,
Cambridge University Press, Cambridge.
Madison. pp.291-299.
Vishnu-Mittre (1961) “Plant economy in ancient Navdatoli-
Winter, J.G. and H.C. Youtie (1944) Cotton in Graeco-
Maheshwar,” in Technical report on Archaeological
Roman Egypt. The American Journal of Philology 65:
Remains. Department of Archaeology and Ancient
249-258.
Witzel, M. (1999) Early Sources for South Asian Substrate
Indian History, Deccan College, Pune. pp.13-52.
Languages. Mother Tongue Special Issue: 1-76.
V i s hnu-Mi t tre , A . S ha rma a n d Cha n c ha la ( 1 9 8 4 )
Witzel, M. (2005) “Central Asian roots and acculturation
Palaeobotanical and pollen analytical investigations.
Indian Archaeology 1981-82 - A Review: 105-106.
in South Asia: Liguistic and archaeological evidence
Weber, S.A. (1991) Plants and Harappan Subsistence. An
from Western Central Asia, the Hindukush and
Example of Stability and Change from Rojdi. Oxford
Northwestern South Asia for Early Indo-Aryan
and IBH, New Delhi.
language and religion,” in T. Osada (ed.) Occasional
Weber, S.A . (1992) “South Asian Archaeobotanical
Paper 1: Liguistics, Archaeology and the Human Past.
Variability,” in C. Jarrige (ed.) South Asian Archaeology
Research Institute for Humanity and Nature, Kyoto.
1989. Wisconsin : Prehistor y Press, Madison.
pp.87-211.
Zohary, D. and M. Hopf. (2000) Domestication of Plants in
pp.283-290.
the Old World, third edition. Oxford University Press,
Weber, S.A. (1999) Seeds of urbanism: paleoethnobotany and
Oxford.
the Indus civilization. Antiquity 73: 813-826.
Weber, S.A. (2003) “Archaeobotany at Harappa: Indications
for Change,” in S.A. Weber and W.R. Belcher (eds.)
- 26 -
Recent archaeological discoveries in Sindh, Pakistan
Recent archaeological discoveries in Sindh, Pakistan
Qasid H. Mallah
Department of Archaeology
Shah Abdul Latif University, Khairpur
Sindh, Pakistan
ABSTRACT
The paper is a compilation of archaeological sites documented up to this date within the lower portions of the Indus and Hakra
River system flowing through present geographical region of the Sindh Province of Pakistan. Sindh comprises several major
geomorphologic units such as western hilly sequence also called as the Sindh Kohistan , central alluvial plains, eastern desert and
southern coastal region all differ in their ecosystem as well. The Indus River occupies central plains in zigzag manner and is still
flowing actively. The Hakra River flowed in eastern side along the fringes of the Thar Desert and has seized its flow in the past.
Archaeological information is collected from these geomorphologic units throughout the Sindh. The present data includes several
newly discovered sites; for the best understanding; this report illustrates some essential parts such as (i) major geographical features
and their environment, (ii) the major contributing settlements with inter-site and intra-site interaction network system, (iii) other
cultural aspects and basis of economy like availability and exploitation of resources; major products and production centres, trade
routes and trade mechanism, (iv) the possible deteriorating factors which caused the weakening of this strongly based civilization,
and (v) vital glimpses on the nature of settlements of last phase of the Indus period. The final part as an Appendix of the paper
describes the locational coordinates and the chronology of archaeological settlements. As the research work continues, so there will
always be additional information on many aspects and thus the explanations are subject to modification and revisions.
INTRODUCTION
structure starting right from the Acheulean tradition
of the Lower Palaeolithic onwards. The best data
The Indus civilization is one of the largest civilizations
set on the Palaeolithic Era in Pakistan has occurred
of the world that flourished along the major rivers
from two regions, i.e. the Potwar region and the
within South Asia with its core occupation on the
Rohri Hills both of these areas are further supported
Indus, Hakra and Ganga Rivers. It is estimated that
through several other discoveries from numerous sites
the Indus civilization was spread over more than a
in India and other parts of South Asia. The artifactual
million square kms; simultaneously it should be noted
repertoire has significantly shown the appropriate
that it is based on more than a million years of human
contact and adjustment of humankind within the
history which is deposited in the layers of different
surrounding environment.
type of landforms of Indus civilization territory.
Therefore, broadly speaking the series of discoveries
Ongoing research continuously adds the information
enabled researchers to describe the Palaeolithic
on various aspects and features of human culture;
human as first tool maker and hunter who lived in
their adoptability and change within the given
hilly environments; during the Mesolithic period,
environmental conditions. The South Asianists are
people got out of hills and became first nomad
successful in reconstruction of complete chronological
hunter and gatherer; the Neolithic period appeared
- 27 -
Qasid H. Mallah
with main factors of domestication and sedentism
after which the same pattern can be marked until the
which finally led people to establish their permanent
recent times, for example, Islamabad appeared after
residences with facilities of storage. This was early
the creation of Pakistan and Gawader port city is in
Indus period in which new technolog ies and
process of development.
reproduction system through domestication of plants
In this paper the major focus remains over Indus
and animals was achieved. The mature phase of the
period which combine its all three stages such as the
Indus period dated as 2600 to 1900 BCE became
Early Indus, Mature Indus and Late Indus. The stage
fully urbanised. Many cities existed on the surface
scheme has taken into consideration just to see the
of earth as Mohenjodaro, Harappa, Ganweriwala,
new and latest developments on research work within
Lakhanjodaro, Chanhudaro, Dholavira, Lothal are
the lower area of the Indus/Hakra River basins and
just a few to name. The urban people of mature Indus
around the Mohenjodaro, the mega city of the Indus
Phase were the builder of gigantic structure as in
Valley civilization. The Early Indus stage contains
Mohenjodaro and the manufacturer of long carnelian
three phases like Hakra/Ravi, Amri, and Kot Diji in
beads and very small/ micro beads measuring up to
which the Indus civilization holds the roots of growth
one mm from white paste of steatite; the stoneware
for the socio-economic complexity profoundly
bangles were manufactured with high degree of
fulfilling the preconditions of early urbanization.
precision and craftsmanship. The copper was quarried,
The Mature Indus period is an account of bringing
purified and alloyed to produce bronze. The society
up the socio-economic infrastructure in which inter
appeared with complex socioeconomic infrastructure;
-site and intra-site interaction is highlighted. The
with accuracy, sophistication, and beauty in material
Late Indus is period of sadness, an unhappy change in
culture and with maintenance of elegance and
which society experience weakening of their system at
hygienic conditions as can be viewed from the
large scale - some reviews from this unhappy change
covered drains, sealed bath rooms, toilets, chutes,
are also part of this paper. The data set is based on
garbage bins, and double storied buildings with built-
total of 148 archaeological sites documented till this
in staircases; all speak their living standards. They
date; some settlements were occupied repeatedly into
enjoyed these living arrangements within urban
later periods; for example, Amri and Chanhudaro.
environment for at least six-seven hundred years
Nevertheless, the registration of each period separately
and after which the civilization declined for next
shows that 75 settlements appeared during early Indus
millennium and reappeared having with mega city-
period, 91 sites are documented for Mature Indus
states within fort enclaves in the Ganga- Jamuna
Period and only 19 sites are listed for Late Indus
(Ganges-Yamuna) River plains (Allchin 1995: 12).
Period. As the documentation focus has remained
For example at Ujjain, the city state which is dated as
only on two portions, for example the Sindh-
398 cal. BCE has fortification ‘measures 75 m at the
Kohistan and the Thar Desert, therefore, the majority
base, soars to a height of 14 m and stretches for over 5
of sites are listed in connection of those areas. This
km’ (Erdosy 1995: 111). Nevertheless, after the Indus
does not, however, mean that there are no sites
period, Sindh never had such an excellent city system
within remaining areas like alluvial plains and coastal
thus only smaller trading posts and/or the capital
region but it simply requires a shift of the focus of
cities of the ruling families appeared and deserted
the research towards those areas in order to refresh
such as the Aror and Bhambhore of Hindu/ Arab
our understanding and knowledge about the Indus
period; Brihmanabad/Mansurah of the Arab period;
Civilization.
Khuda Abad and Shikarpur of Kalhora period and
- 28 -
Recent archaeological discoveries in Sindh, Pakistan
Judeirjodaro
R
us
Ind
r
ive
Thariri Iban Odho
Bhando Qubo
Loal Mari
Lakhanjodaro
Jhukar
Peer Sarihiyo Bhir
Khipro II
Khosa Daro
Ghob Bhir
Mohenjodaro
Kot Diji
Bamba
Garhar
Angiaro
Bakri Waro East
Dubi 4
Poonger Bhanbhro
Choondiko
Bakri Waro West
Kathgarh/Thikrao
Lohumjodaro
Bhankio Veero I
Hadi Bux-ji Wandh
Ganero 8
Chhuti jo Kund
Hingorja
Kandharki
Kathore Deh
Taloor-ji Bhit
Ali Murad
Lak Sharief
Thar Desert
Ghazi Shah
Sindh Kohistan
Kander Bhit
Amri
Bandhani Flint site
Chanhudaro
Maliri Landi
Bibi-ji Bhit
Kohtrash
Hothiano Flint site
Ahmed Shah
Dhillani jo Kot
Desoi (Bedi Kotiro)
Allahdino (Nel Bazar)
Figure 1 Main archaeological sites in Sindh
GEOGRAPHY OF SINDH
paper. The environment in these units varies when
climatic factors such as temperature, humidity, rain-
The province of Sindh contains diverse geography
fall, cloud cover, fog, solar insulation, winds, and
in its overall layout and thus this diversity can be
total amount of accumulated heat are taken into
comparatively divided into four major units i.e.
consideration. M. H. Panhwar has divided Sindh into
western foot hill region known as the Sindh Kohistan,
several micro-zones based on the variation on above
central alluvial plains, eastern desert and southern
mentioned climatic factors (www.panhwar.com). In
coastal belt. In the northeast corner of Sindh a small
this scheme the Sindh Kohistan area is not included.
hilly sequence named is the Rohri Hills contains
However, if this situation within these micro-zones
specific geographic character and is included in this
prevailed in prehistory it might have grater effects on
- 29 -
Qasid H. Mallah
the socio-economic organization of the Indus period.
thick walls that can sustain not only the high volume
of water but also the force of water coming from
SINDH KOHISTAN
high mountains; if necessary the height of walls is
This region is located in between the main hilly ranges
increased. This is archaeologically important because
bordering Balochistan/Sindh, i.e. the Kirthar Range
it can provide a detail of dam construction and usage
and the Indus alluvial plains and contain scattered
if scientifically investigated; until now, nobody has
low-lying hilly tracts/ranges and gravely soils some
done such type investigation. The present author
time covered with alluvial soil suitable for cultivation
has observed some portions of the Baran Nai, a rain
purposes. This alluvial soil is present in the valleys.
feed river and have recorded several such dams along
The main hilly sequences of the Sindh Kohistan
its length and have concluded that these dams were
are Lakhi, Kambhu, Badhar, Bhit and Dumbar
perhaps for the storage of waters used for cultivation
(Quddus 1992: 197). Within these ranges, there are
in alluvial valleys, for example Taung Valley where
several freshwater springs; rain feed rivulets (e.g. Gaj
majority of settlements (both prehistoric and historic)
and Baran). There are numerous passes wherefrom
are concentrated. Beside the Dams/‘Gabar Bund’;
people in retrospect have communicated. The
the water was also stored in wells and ponds like
concentration of ancient settlement is documented
ditches for consumption of herds during no-rain
around Phusi pass in north and Darwat pass in south
years. Sometimes the springs also add the water supply
Sindh (Fairsevis 1975: 211). The climate of the area
(Personal observation 2004). Along this Baran Nai
is categorized arid with scanty and unpredictable
numerous ancient settlements have been recorded.
rainfall. Archaeological investigation shows that
There are several passes connecting settlements of
human beings living there have learnt to utilize the
Balochistan and the Indus plains, for example the
rainwater efficiently and have established the excellent
Darwat Pass in south and the Phusi Pass in north
water storage system known as ‘Gabar Bund’; where
(Fairservis 1975: 211, map16).
rainwater was collected artificially for agriculture
Another major rain feed rivulet is the Gaj Nai
and other usage. These Gabar Bunds are efficiently
which originates in highland of Balochistan and
established in such a way that even small rainwater
ends into the Lake Manchar. The Lake Manchar
can be collected. The system has two portions (a)
is a huge natural lake that also receives water from
diverting of water from hilly slopes towards storage
Indus River and was major source of the food supply
facility and (b) the storage tank known as ‘Gabar
to many prehistoric as well as historic settlements
Bund’. At the slopes several smaller walls usually
located around it. The fisher folk dwelled within
with height of one to two feet high from the ground
the lake on the boats. Richard F. Burton writes ‘they
level are setup for the purpose of channelling water
eat; drink, smoke and sleep on board their vessels’.
in one direction and retaining enough water to soil
Burton further explains that they dry fish on the bank
for cultivation. If the land is not cultivated for any
and sell it in nearest village market for the daily-use
reason then another benefit of retaining water is to
items (Burton 1851, reprinted in 1999: 235). The
grow the wild grasses for pasturage. The cultivation
under water archaeological investigation is suggested
in Kohistan is called ‘Khushkaba’ or ‘Barani’ means
for scientifically explanation about the lake and its
‘the rain crop’ (Fairservis 1975: 171). The diverted
utilization.
water flows towards the main dam. The dam is built
in the flat bed of main rivulet at the place where good
ALLUVIAL PLAINS OF SINDH
amount of water can be stored, supported with high
The a l luvia l pla ins of Sind h are ta ken into
- 30 -
Recent archaeological discoveries in Sindh, Pakistan
Figure 2 Desert conjoining with the Indus Plains
consideration within its present administrative
tamarisk for mates, baskets and material for the house
and political boundaries. The Indus River flows
roof cover and ropes called as ‘Wann’ for netting the
within these plains and merges into sea in the south
cot known as ‘Khat’ or ‘Charpaie’.
direction. In its present location where it flows from
In past, we do not have direct evidences on the
northeast to southwest and hits the Kohistan area
irrigation system or the water works except the
below Sehwan/Amri and turn towards Southeast and
artificially dug well for drinking water. The discovery
than south via Hyderabad and merges into the sea at
of the Gabar Bunds is specific to the highland areas.
Gharo Creek, where Bhanbhore site is located.
The best assumption for irrigation system would be as
The Indus has been freely wandering in the
lift system, and ‘Bosi’ or ‘Selabi’, for which the flood
vast alluvial land until the protective bunds in
channels may have been utilized as canals. During
1860-1960CE were established limiting the river
historic time, for instance, the Kalhora period of
to flow within 6 to 8 km in width (Panhwar n.d.).
Sindh from 1701 to 1758CE, an extensive irrigation
The land between bunds is called ‘Kacho’ which is
system was established. M.H. Panhwar explains that
frequently flooded each year and after the flood the
the Kalhora were the master builder of canals who
crop is grown on preserved moisture in soil called ‘Bosi’
built hundreds canals having with six yards of width
or ‘Selabi’. Prior to the construction of the Gabar
thorough Sindh. In their schemes, some of the Dhoros
Bunds, the Indus River moved freely and made several
might have converted into canals. In 1755/56CE,
lakes and courses varying in size and length, now filled
due to increase of water in lower portion of the Indus
with saline water; the Manchhar, and Kenjahr are very
River it changed the course and at least five hundred
famous lakes. The ancient channels which are known
canals in southern Sindh were deserted which resulted
as ‘Dhoros’ and the low-lying area turned as lakes are
huge socioeconomic effect on the population as they
called ‘Sim’; both Dhoros and Sim (lake) provide
were turned as nomads and pastoralists. The successors
pasturage grounds supplement economy by being
of Kalhora were Talpur Mirs who could not maintain
source of migratory birds, and fish for food; Typha
the extensive water system due to fluctuations in the
domingensis (reed), Saccharum bengalensis (Boro), and
Indus River flow and lack of expertise in canal system
- 31 -
Qasid H. Mallah
n.d.). Nowaday, the forest area is being reduced and
Home
archaeological sites are being levelled for agricultural
purposes.
COASTAL AREAS
The coastal area of Sindh stretches from Karachi
to Rann of Kachchh. The region beyond Thatta is
generally the delta area. At least six administrative
Bullock Cart
Farm
Taluka named as Sakro, Ghora Barri, Kharo Chhan,
Market
Jatti, Keti Bander, and Shah Bander; all Talukas cover
Figure 3 Modes of transportation
total of 350 km strip. Syed Abdul Quddus defines the
(Panhwar n.d.). The British government perhaps
delta area as ‘uncultivable and unstable, full of silted
rejuvenated the canal system which was originated
creeks…[where] the tide submerges the shore up to 3
during the Kalhora period and added some more
to 4 miles. During inundation it is flooded up to 20
canals on the Indus through establishing the dams/
miles inland (Quddus 1992: 197). Theses coastal area
barrages.
has been very important source of communication,
On the Indus River three major barrages named
subsistence and other things including the source
as Guddu, Sukkur and Kotri were constructed for
of shell and fish. Several archaeological settlements
irrigation through which the water is distributed all
have been recorded associated with Mesolithic
over Sindh and parts of Balochistan. This irrigation
to Indus period. In the coastal area the historic
system is major source of agriculture for both the
periods settlements are also present; Bhanbhor and
rabi (winter) and kharif (summer) crops. In rabi crop
Chukhundi are famous sites of historic period.
season, the wheat, pulses, oil seeds are grown and
in kharif, cotton, sorghum, vegetables and fodder
DESERT AREAS
are grown. Bullock cart has played major role in the
The sandy desert occupies the entire eastern side of
agrarian communities. Many terracotta frame have
Sindh which is part of the Thar Desert. The Hakra
been found from the Indus period sites and the wide
River makes its flow along the western fringes as it
streets of Mohenjodaro suggest the accommodation
enters into the administrative boundary of Sindh,
of two-way traffic of bullock carts. That is an easy
flowing within sand dunes towards south. Near the
tool from home to farm and for nearby markets as
Salehpat and/or Pharhiyaro, it enters into Nara valley
well. The long distance transport would be only
and takes the name as the Nara River (generally
possible in the dry plain areas; any muddy, sandy, and
known as the eastern Nara) until Jamrao head and
elevated areas are never preferable for the bullock cart
then takes the western fringes of the desert until
transportation, therefore, it is very effective and useful
it ends in the sea. This alluvial valley is four to 10
for short distances.
km wide where the Nara flows mainly along the
The climate in Sindh also varies, divided into three
eastern side. The desert continues on both sides ip
major parts i.e. (a) Sarro the northern part extremely
to Jamrohead, joining with the Indus plains on the
hot and arid with less than 125 mm rainfall; (b)
western side.
Vichollo, the central part hot/arid receiving 125-
In general setup, the southern portion of the Thar
255 mm rainfall, and Lar the lower portion warm/
Desert has higher dunes in a cardinal orientation,
semi-arid with 255-355 mm rainfall (Panhwar
measuring more than 500 feet in height with greater
- 32 -
Recent archaeological discoveries in Sindh, Pakistan
rainfall (Pithawalla 1959). The height of sand dune
pastoralists who repeatedly visit the area even if there
diminishes as moving towards the north and the very
are scanty monsoon showers and fewer grasses grown.
low sand dunes can be seen in Cholistan (Mallah
The pattern of their mobility has been observed
2000). The entire desert is covered with sparse
through archaeological as well as ethno-archaeological
vegetation mainly with Khabar Salvadora oleoides,
perspective.
Kirar Capparis deciduas and Kandi Prosopis cineraria;
On the western side of the Nara valley, the desert
the shrubs are Phog Calligonium polygonoides, Ak
continues, covering some portion of the Rohri Hills
Calotropis procera, Khip Leptadenia pyrotechnica and
on the south and east sides. The Rohri Hills is a small
Booh Avera javanica, Lano Haloxylon stocksii; the
hilly sequence stretching some 73 km from north
herbs are Chhapri Neurada procumbens, Ghorawal
to south and 20 km from east to west. The hills are
Cassia italica, and the grasses are Katan Cymbopogon
surrounded by the Nara valley and Thar on east
jawarancusa, Lumb Stipagrostis plumosa, Boro
and south, while the Indus plains are on the west
Saccharum bengalensis and many others are most
side. These hills contain deposits of fuller’s earth,
frequently available. The monsoonal grasses also
limestone and chert/flint and banded chert. The
grow and supplement the pasture resource. This
chert nodules of various size and quality are littered
vegetation is best for the goat, sheep, cattle, and
on the surface throughout the hills; however the
camels; some animal herds still utilize the desert
banded chert deposits are only discovered in the
resources. Because of the intensive hunting , few
northern tip. A poor quality of banded chert is also
wild animal species prevail today. Some fox, rabbit,
observed in the Veesar valley that could not have
jackal along with reptiles like lizard and snakes can
been used for weight manufacture. In addition to
be sparsely encountered. There are several lakes and
raw material, these hills also provide several types of
alluvial valleys specifically in the desert region having
monsoonal grasses and supplement the subsistence
with above mentioned flora and were once watered
resources. Several archaeological sites are located
by the Hakra River. These lakes were the best source
around the hills, therefore it is quite possible that
for game, pasturage and other economic activities.
some of the pastoralists were engaged in chert tool
Many archaeological sites have been documented
production. As these hills are rich in raw material
around the lakes. This sandy desert is very suitable for
sources, they were heavily exploited for manufacturing
Figure 4 General view of the Thar Desert
- 33 -
Qasid H. Mallah
Figure 5 General view of the Rohri Hills
stone tools. A collaborative project of Italy and
his discoveries through several publications of
Shah Abdul Latif University, Khairpur, Pakistan
Archaeological Survey of India, explaining the Indus
(hereafter ‘SALU’), was framed from 1993-1999
civilization.
and after that continuously by the Department of
In 1927-31, N.G. Majumdar carrie d out a
Archaeology, SALU independently. Although a series
comprehensive exploration of Sindh and recorded
of investigations were carried out in these hills, the
numerous archaeological sites (Majumdar 1934).
research is still continued and thus the information
Majumdar put test trenches at several sites and
would be added in future.
reported them based on context and typolog y.
Meantime, Ghur yes (1936) reported two ver y
PREVIOUS WORK ON
ARCHAEOLOGY IN SINDH
important sites, i.e. Naru Waro Dhoro and Kot Diji
in the upper Sindh. H.T. Lambrick did some work
in Sindh and published a series of papers (Lambrick
The academic research in Sindh began when John
1941, 1942, 1944, 1946 as cited by Shaikh 1995). He
Evans (1866), William Blanford in (1880), Henry
also wrote two volumes ‘the History of Sindh’, the first
Cousens (1929), H. De Terra and T.T. Patterson
volume of which includes discussion on the Hakra
(1939) reported the presence of remains of ancient
River as well (Lambrick 1975). The excavation at Kot
settlements worth to be investigated further. R.D.
Diji (Khan 1965) and at Amri by J.M. Casal in 1959-
Banerji in 1922-23 surveyed some portions of Sindh
62 (Casal 1964) imposed significant impact on the
and found Mohanjo-Daro (Banerji 1923). These were
archaeology of the Indus valley.
the some of pioneering studies in Sindh in the early
In 1 9 7 2 , M . S ha ri e f f ro m D ep a r tm ent o f
periods of the research.
Archaeology and Museums explored the southern
J.H. Marshall started excavation at Mohenjodaro
Sindh and documented many historical sites. Dr. N.A.
in 1921, and E.J.H. Mackay excavated Mohenjodaro
Balouch in 1973 added Gharo Bhiro site located in
(Mackay 1927-34) and excavated Chanhudaro
the southeast corner of Sindh. Louis Flam adventured
in 1935-36 (Mackay 1943). Mackay reported
in Sindh by following the classic works by Majumdar,
- 34 -