Empire of Cotton

by Sven Beckert

  Worlds of Cotton: The First 5,000 Years

  Indian cotton technology also spread into Southeast Asia. As production skills advanced, cotton cloth emerged as the region’s most valuable manufactured product after foodstuffs. Buddhist monks brought it to Java sometime between the third and fifth centuries CE. Much later, between 1525 and 1550, cotton cultivation expanded into Japan. By the seventeenth century it had become an important commercial crop there, as small farmers grew cotton to earn extra income for tax payments, often in rotation with rice.21 With cotton’s arrival in Japan, the original Indian cotton culture had now spread over most of Asia.

  Fashioned by African, American, and Asian peasants, spinners, weavers, and merchants over at least five millennia, this cotton world was vibrant and expanding. Despite its diversity across three continents, the centers of this huge manufacturing industry had many things in common. Most important, cotton growing and manufacturing almost always remained small-scale and focused on households. While some growers sold their raw cotton into markets, including long-distance markets, and many rulers forced cultivators to part with some of their crop as tribute, no growers depended on their cotton crops alone; instead they diversified their economic opportunities, hoping to lessen risk to the best of their ability. In a large swath of Africa, and parts of South Asia and Central America, such patterns persisted until well into the twentieth century.

  For millennia, then, households planted cotton in a delicate equilibrium with other crops. Families grew cotton alongside their food crops, balancing their own and their community’s need for food and fiber with their rulers’ demands for tribute. In Veracruz, for example, double cropping of food grains and cotton was common, providing subsistence both for those who grew cotton and those who spun and wove. In the Yucatán, Mayan peasants grew cotton in fields that also produced maize and beans. In West Africa, cotton was “interplanted with food crops,” such as sorghum, in present-day Ivory Coast, or with yams, as in the area that is now Togo. In Gujarat, “the [cotton] shrubs are planted between the rows of rice.” In the cotton-growing areas of Central Asia, peasants grew the fiber alongside not only rice, but also wheat and millet, and in Korea alongside beans. No significant cotton monoculture emerged before the eighteenth century, and yet when that monoculture appeared so too did the hunger for ever more land and labor.22

  Like cotton growing, cotton manufacturing throughout the world began in households, and, with few exceptions, stayed there until the nineteenth century. In areas controlled by the Aztecs, for example, all cotton manufacturing was organized within households. In Africa as well, “in many cases the production of cotton goods was purely a family industry, each social unit being entirely self-sufficient.” We have similar testimony for India, China, Southeast Asia, Central Asia, and the Ottoman Empire. Household production enabled a family to produce the cloth they needed, but it also allowed production for markets. Since labor needs in most agricultural societies varied enormously by the season, and since picked cotton could be stored for months, peasants could focus on textile production intermittently and seasonally, during slack times. This was especially the case for women, whose activities focused on the house, with some of their labor available for the homebound production of yarn and cloth.23

  In every society a definite gender division of labor emerged, with a particularly strong association between women and textile production. Indeed, there was a premodern saying in China that “men till the soil and women weave.” Except among the Navajo, Hopi, and some peoples in Southeast Asia, women throughout the world have had a virtual monopoly on spinning. Because spinning can be done intermittently, and enables a simultaneous commitment to other activities, such as watching young children and cooking, women’s roles within households usually led them to be in charge of spinning as well. So close was the association of women and cloth production that in some cultures women were buried with their spinning tools. With weaving, on the other hand, no such stark gender divisions emerged. While men tended to dominate the weaving industry in places such as India and southeast Africa, there were many cultures in which women wove as well, such as in Southeast Asia, China, and North and West Africa. Yet even in societies in which both women and men wove, they usually specialized in different designs, produced distinct qualities, and worked on different types of looms. This gendered division of labor was reproduced in the emerging factory system as well, making gender relations in the household an important factor in the emergence of factory production.24

  Embedded within households and their particular strategies for survival, this premodern cotton industry was also characterized by slow technological change in ginning, spinning, or weaving. As late as the eighteenth century, a woman in Southeast Asia, for example, needed a month to spin a pound of cotton and another month to weave a piece of cloth ten yards long.25 This enormous time requirement was partly the result of what economists call “low opportunity costs” for the labor that went into spinning and weaving, and partly of a world in which rulers taxed their subjects’ production to the maximum extent possible. Moreover, since many households were self-sufficient in textiles, markets were of a limited scale, again reducing incentives to improve production techniques.

  Yet slow technological change was also related to constraints on the supply of raw materials. In most regions of the world, raw cotton could not be transported efficiently very far. Beasts of burden or humans sometimes carried raw cotton over relatively short distances. In the Aztec Empire, raw cotton was transported into highland areas to be manufactured, at distances of perhaps a hundred miles. More efficient and common was the waterborne cotton trade. In the second millennium CE, for example, observers reported hundreds if not thousands of boats floating cotton down the Yangtze to the region of Jiangnan. Gujarati and central Indian cotton was similarly shipped on the Ganges and along the coast to South India and Bengal. Nonetheless, until the nineteenth century the overwhelming bulk of raw cotton was spun and woven within a few miles from where it was grown.26

  So many people in so many parts of the world grew cotton, spun it, and wove it into fabrics that it was very likely the world’s most important manufacturing industry. And while household production for household consumption would remain until the nineteenth century its most important sector, there was significant change before the Industrial Revolution of the 1780s. Most importantly, cotton goods—partly because they were so labor-intensive to produce—became an important store of value and a medium of exchange. Rulers everywhere demanded cotton cloth as tribute or taxes, and indeed it might be said that cotton was present at the birth of political economy as such. Among the Aztecs, for example, it was the most important medium for tribute payments. In China, beginning in the fifteenth century, households were required to pay some of their taxes in cotton cloth. And in Africa the payment of tribute in cloth was common. Practical as a means to pay taxes, cotton cloth was also used as currency in China, throughout Africa, in Southeast Asia, and in Mesoamerica. Cloth was an ideal medium of exchange because unlike raw cotton it could be easily transported over long distances, was not perishable, and was valuable. Nearly everywhere in the premodern world, a piece of cotton cloth could buy needed things: food, manufactured goods, even protection.27

  Cotton’s use as proto-money illuminates the fact that not all cotton textiles, with their favorable ratio of value to weight, were used in the immediate vicinity of their production. Indeed, the cotton centers that had emerged separately in the Americas, Africa, and Asia all developed increasingly sophisticated networks of trade, connecting growers, manufacturers, and consumers over long, eventually even transcontinental distances. In Iran, the ninth- and tenth-century cotton industry led to significant urbanization, drawing raw cotton from the surrounding countryside, spinning, weaving, and tailoring it to sell into long-distance markets, especially in what is today Iraq. In precolonial Burkina Faso one author finds that “cotton was at the center of trade.” Gujarati cotton cloth, as early as the fourth cent
ury BCE, came to play a very significant role in the trade between the various lands bordered by the Indian Ocean, and large quantities were sold along the East African coast, to be traded far into the African hinterland. In all of these exchanges, traders, especially if far removed from the polities they originated from, had to adjust to local tastes, and had to offer their products at prices attractive to local consumers.28

  In Mesoamerica, cloth was traded over many hundreds of miles, including to neighboring states, as, for example, when merchants brought cloth from Teotitlán (in modern-day Oaxaca) to Guatemala. In the Southwest of what is now the United States, yarn and cloth were also important trade items. Cotton goods have been found in excavations far from regions in which cotton could grow. Since the thirteenth century, Chinese merchants imported cotton yarn and cloth to supplement domestic production from as far away as Vietnam, Luzon, and Java. In similar ways, African merchants traded cotton textiles over long distances, as for example when they exchanged Malinese cotton cloth for salt brought in by desert nomads. Ottoman cotton textiles found their way to places as distant as western Europe, while cotton goods were already being imported into Japan in the thirteenth century.29

  India, at the center of this increasingly global reach, traded with the Roman Empire, Southeast Asia, China, the Arab world, North Africa, and East Africa. Indian cottons crisscrossed South Asia on the backs of people and bullocks. They crossed the seas in Arab dhows, traversed the great Arabian Desert to Aleppo on the backs of camels, moved down the Nile to the great cotton mart of Cairo, and filled the bottoms of junks on their way to Java. Already in the sixth century BCE Indian cotton was traded to Egypt, as merchants brought Indian cotton to Red Sea and Persian Gulf ports. Greek merchants then took it from Egypt and also Persia to Europe. Roman merchants eventually participated in this trade as well, making cotton a coveted luxury good among the imperial elites. Throughout eastern Africa, Indian cottons were an important presence as well. And throughout the Arab world and Europe, India remained a major supplier until the nineteenth century, with Gujarati merchants, among others, unloading huge quantities of cloth. As an Ottoman official complained in 1647, “So much cash treasury goes for Indian merchandise that…the world’s wealth accumulates in India.”30

  Indian cloth also traded eastwards into other parts of Asia. Merchants sold it in the marts of China in very ancient times. Huge quantities of Indian cloth also found their way to Southeast Asia to clothe the local elite: Imports to Malacca in the early sixteenth century, it has been estimated, filled the holds of fifteen ships that arrived annually from Gujarat, Coromandel, and Bengal. So dominant was Indian cloth on world markets that around 1503 the Italian merchant Lodovico de Varthema observed about the Gujarati port town of Cambay, “This city supplies all Persia, Tartary, Turkey, Syria, Barbary, i.e., Arabia Felix, Africa, Ethiopia, India and a multitude of inhabited islands, with silk and cotton stuff.” The Sanskrit word for cotton goods (karpasi) entered into Hebrew, Greek, Latin, Persian, Arabic, Armenian, Malay, Uigur, Mongolian, and Chinese. Even the names of particular fabrics became global brand names—chintz and jackonet, for example, are corruptions of terms in Indian languages that eventually came to describe a particular style throughout the world. Beginning in the seventeenth century, Indian cottons, in fact, were what historian Beverly Lemire has called the “first global consumer commodity.”31

  As demand grew, cotton took its first tentative steps out of the home. During the second millennium CE, production in cotton workshops became more common, especially in Asia. Professional weavers emerged in India; they focused on supplying the long-distance trade, providing rulers and wealthy merchants both at home and abroad with cotton cloth. In Dhaka, weavers labored under tight supervision to produce muslins for the Mughal court, “forced to work only for the Government which paid them ill and kept them in a sort of captivity.” Workshops containing more than one loom are also reported to have been located in Alamkonda, in modern-day Andhra Pradesh, as early as the fifteenth century. In contrast to the subsistence weavers, the long-distance tradesmen were geographically concentrated: Bengal was known for its fine muslins, the Coromandel coast for its chintzes and calicoes, and Surat for its strong but inexpensive fabrics of every kind. Though weavers could occupy very different positions within India’s caste system, in some parts of the subcontinent they found themselves in the upper reaches of social hierarchies, prosperous enough to be among the leading donors to local temples. Groups of full-time cotton manufacturers emerged in other parts of the world as well: In fourteenth-century Ming China, for example, higher-quality textiles were worked up in “urban loom houses,” which collectively employed many thousands of workers. In the Ottoman city of Tokat, highly skilled weavers produced significant quantities of cotton textiles. Baghdad, Mosul, and Basra, among other cities in the Islamic world, had large cotton workshops, and indeed the word muslin for fine cottons derives from Musil, the Kurdish name for Mosul. In Bamako, the capital of present-day Mali, up to six hundred weavers plied their trade, while in Kano, the “Manchester of West Africa,” a large weaving industry arose, supplying the people of the Sahara with cloth. In Timbuktu, already in the 1590s twenty-six cotton-producing workshops plied their trade, each with fifty or more workers. In Osaka as well, thousands of workers wove cotton textiles; workshops spread throughout the region employing thirty to forty thousand people by the early eighteenth century.32

  As the workshop became more common, so too did a new type of weaver: an individual, usually male, who produced specifically for sale in a market. Yet even as workshops emerged, this specialized production for markets typically took place in the countryside, not towns, and in homes, not workshops. What set these rural market producers apart from those who produced for subsistence only was their reliance on an emerging force in global commerce: putting-out networks held together by merchant capital. In these networks, which would form the nuclei of nineteenth-century mechanized cotton production, spinners and weavers worked up cotton thread and cloth for urban merchants who would collect the products of the spinners and weavers and then sell them on distant markets. The particular ways merchant capitalists and producers related to one another varied widely. On the Indian subcontinent, for example, rural weavers relied on merchants for the capital needed to purchase sufficient yarn, and for the food they needed to subsist while weaving, yet these weavers generally owned their own tools, worked without supervision, and enjoyed some control over the disposal of their products. In other parts of the world, rural weavers enjoyed considerably less power. In the Ottoman Empire, for example, merchants advanced cotton and yarn to peasants, who spun and wove it, then returned the product to the merchants for a small profit. Unlike weavers in India, they did not have any control over disposal of the product. In China merchants also enjoyed great control over production. “They bought up raw cotton, put it out at local markets for peasant women to spin and to weave, had the cloth dyed and calendared in town or city workshops and then exported it all over China for sale.” Merchants, in fact, controlled every stage of production, foreshadowing their central role in the nineteenth-century construction of a globe-spanning empire of cotton.33

  With expanding markets, cotton technology changed as well. While the basic principles of cotton processing were quite similar throughout the world, and productivity was dramatically lower before the invention of the novel gins, spinning machines, and looms of the late eighteenth and early nineteenth centuries, there were some significant innovations. In Mesoamerica, for example, spinning was improved by the introduction of “specially formed ceramic spindle whorls.” After 1200 CE, Mesoamericans also used specifically designed spinning bowls, which increased the productivity of spinners, enabling them, among other things, to feed the voracious appetite for tribute of their rulers. The center of technological innovation, however, was Asia: The roller gin (to remove seeds), the bow (to clean and disentangle ginned cotton), the spinning wheel, and new kinds of looms, including the upright warper, a
ll originated in Asia. The spinning wheel, invented in the eleventh century, was an especially significant innovation as it allowed peasants to spin cotton much faster. Weavers in the same regions also invented a novel kind of loom—the treadle loom. While its exact origins are uncertain, it was introduced into India sometime between 500 BCE and 750 CE, and into China (where it was first used in silk manufacturing) in the third century CE.34

  The greatest innovations occurred in the domestication of the cotton plant itself, indeed so much so that the cotton picked by slaves in the nineteenth century would be nearly unrecognizable to Indian farmers of two thousand years earlier. Human selection made cotton compatible with highly varied environmental conditions and rendered its fiber ever more applicable to the production of textiles. Rural cultivators in China, Japan, Southeast Asia, North and South America, western Africa, and Anatolia brought cottonseeds from adjacent territories and added cotton to their crop mix. Through the centuries, this process of domestication drastically altered the physical properties of cotton, creating plants that produced longer and brighter fibers (later-day cotton experts would refer to the length of the fiber as “staple”), ever more plentiful and easier to remove from the filbertlike shell. Moreover, advances in irrigation techniques and agronomy allowed for the expansion of production into new regions. Through seed selection and improved technology, the cotton plant flourished in drier and colder parts of Africa, Asia, and the Americas, including the mostly arid soils of the Islamic world. In Iran, for example, investments in irrigation systems as early as the ninth century enabled a significant extension of cotton agriculture. Nonetheless, compared to eighteenth- and nineteenth-century changes, overall productivity increases in the two thousand years prior to the Industrial Revolution were small. For much of its history, the world’s cotton industry expanded primarily because ever-increasing numbers of people spent ever more time growing, spinning, and weaving cotton.35