About the Book

The history of human population is of interest when human beings are aware of the burden they place on the ecosystem. Asia has long contained a major fraction or world population, and East and South Asia have accounted for most of that fraction. This book focuses on various aspects of the population of South Asia over the past twenty-five centuries.

Sumit Guha’s introduction highlights debates in the population history of Asia, Europe and the Americas. This leads into a chapter on the population of South Asia from 200BC to 1900AD. Later chapters focus on specific aspects of the interaction between demography, climate, health, medicine and culture.

A rare document on vaccination is translated for the first time and used to illustrate the interaction of cultural codes and medical techniques. Immensely detailed data on military populations is used to generate important conclusions regarding the efficacy of knowledge and hygiene in improving health. The book also includes a compact survey of the evolution of environmental hygiene in India through the twentieth century.

About the Author

SUMIT GUHA is professor of History at Rutgers University. He is the recipient of a John Simon Guggenheim Fellowship and his books include The Agrarian Economy of the Bombay Deccan and Environment and Ethnicity in India, c, 1200-1981.

Introduction

This planet has existed for some four billion years; our species for a comparatively insignificant fraction-perhaps 200,000 years or 0.005 per cent of that period. In that time, our numbers have changed dramatically. We started as a rare large-bodied omnivorous African primate-so rare that it seems likely that we are all descended from a single female ancestress who lived around 210,000 years ago (give or take 70,000 years). This large animal then proceeded to increase in number and spread over most of the earth in the space of a hundred millennia or so. In modern times, which is to say about seven or eight thousand years ago, the simultaneous domestication of plants and animals during the Neolithic Revolution allowed a massive further increase in its numbers, as well as a simultaneous drastic alteration in its life-ways. A recent review of the anthropometric data of this early population transition concluded that the ‘shift from foraging to farming led to a reduction in health status and well-being, an increase in physiological stress, a decline in nutrition, an increase in birth rate and population growth…’

The larger and more sedentary populations that developed under this new demographic regime allowed the appearance of unprecedented population concentrations possessing specialized features that allow us to classify them as cities. The societies in which these were embedded are what we term civilizations-in the Middle East, the Mediterranean world and China-generated systems of governance that required that required the enumeration of populations for various (usually exploitative) purposes. This, in turn, sometimes left records usable by later historians, though the difficulties of interpretation could result in very diverse estimates on the basis of the same sources. So, for example, Be loch and Frank, both specialists in the area, could estimate the population of Italy in 14 CE at 6 and 14million respectively. Perhaps the longest time-series is that available for East Asia where, by ‘the late second century AD, China’s total population was probably between 60 and 70 million people… It did not fundamentally change in size until the eighth century when population slowly rose to 100 million by the eleventh century….’

The upheavals consequent on the Mongol conquest and their overthrow by the Ming brought China’s population down to something over 65 million in 1393; it reached 150 million by 1700, being sustained by the development and diffusion of quick-maturing strains of rice. New World crops like maize then enabled a second tremendous expansion of the population to 430 million by 1850 but political and economic catastrophes checked growth thereafter, and it was 583 million at the 1953 census, having grown at only 0.3 per cent a year over a century. For South Asia, the deficiency of even moderately reliable data is such that, as Hollingsworth pointed out, it was even difficult to make a firm estimate for 1850, twenty-two years before the first regular census. Biraben, quite conjecturally, placed the subcontinent’s population at 45 million in 200CE but only 40 million in 1000 CE. Sanjay Lal has yet more speculatively proposed that it reached a carrying-capacity ceiling of 100million by 320 BCE, but was then periodically cut back by calamities before returning to that equilibrium point a hypothesis already suggested with little empirical support by Alit Das Gupta. In 1974, Durand’s judicious discussion of all population estimates for South Asia concluded that ‘[a] estimates of population in this region before the nineteenth century have a more or less flimsy basis…’

Wherever more reliable records are available, they clearly indicate that the sizeable populations sustained by early agriculture in the past few millennia typically led a precarious existence, and underwent massive fluctuations. The earliest record of this was unearthed in Adams’ marvelous study of the Tigris valley in Iraq. About 100,000 humans lived in the lower Diyala region in 2000 BCE, but only a tenth of that number was present in 1000BCE. The population peaked again at 600,000 in the seventh century CE but collapsed to some 50,000 in the twelfth century, a level at which it stagnated into the nineteenth century to take a more recent example: the lower Yangtze provinces of China in 1850 contained 137 million people more than most nation-states then or now; civil war, foreign invasion and famine ensured that the population of 1953 was actually smaller by 19 million than that in 1850! To take a European instance, it has also been credibly suggested that the population of the British Isles touched, and fell back from, a maximum of 5 million three times before the Industrial Revolution-once under the Romans in 200 CE, once in 1300 and once again in 1650. (Despite sending out a flood of emigrants that swamped much of the rest of the world, the population then exploded to 53 million by 1950.)

The effect of organized intra-species violence was, directly or indirectly, one of the most important causes of past population declines, but exploited peasant populations could also be ravaged by harvest failures resulting from climatic variability or land degradation. The first western thinkers on this subject attributed peacetime population decline to vice and misery, and associated it with misgovernment. The prominent French savant Holbach even proposed a cyclical moral economy of reproduction, in which luxury led to lapses into tyranny which were followed by depopulation and the consequent fall of empires, thus allowing a return to the simple virus and high fecundity of savagery to be once again succeeded by repopulation and a resumption of the cycle. As a corollary to such ideas, it was widely held that the population of the West had been far larger in ancient times than in the seventeenth century.

A major attack on this doctrine came from the celebrated Thomas Malthus who sought to use demography to check popular expectations of amelioration in the era of the French Revolution, and developed the first integrated demographic model to do so. He argued for a continuous tendency for human populations to outgrow their sources of subsistence, something which resulted from ‘the constant tendency in all animated life to increase beyond the nourishment prepared for it.’ As regards human beings, they could either surrender to their instincts, or they could resist them. The latter course usually led to what Malthus termed vice; on the other hand, the commoner course of marriage and reproduction meant that ‘the human race will constantly be attempting to increase beyond the means of subsistence.’ This actual or impending deficit in turn meant that ‘difficulty must fall somewhere, and must necessarily be felt in some or the other of the various forms of misery, or the fear of misery, by a large portion of mankind.’

The crude arithmetical and geometrical progressions of the early Malthusian model were soon discarded, but the model itself has proved highly influential. Its fundamental pessimism about progress, and its convenient explanation of lower-class (and later, Third World) misery as due to the feckless reproduction of the poor have made it attractive to many down to the present. This ideological load also attracted many critiques.

As data accumulated, however, it soon became evident that the simple model of human populations growing limitlessly unless checked by superhuman continence, the practice of ’vice’, or the onset of calamity, was quite inadequate to explain the observed facts. Demographers and population historians soon came to recognize that many forces acted to shape population change a recognition necessitated (among other things ) by the rapid fall of first French and then western birth rates from the nineteenth century onward. Indeed, as Caldwell recently pointed out, the ‘fertility transition has a dominance in demography possessed by no other theme. Its study is largely responsible for the growth of demography as a discipline over the past half century.’

This exclusive focus is not perhaps as appropriate for historians; in the long centuries since the Neolithic revolution, many populations have been dominated by variations in mortality rather than fertility. The emerging science could also err. At the end of the (justly) pessimistic and depressed 1930s, for instance, expert population projections substantially overestimated the impending fall in fertility, and thus underestimated the size of European populations in 1970. This led Bourgeois Pichat to propose the inclusion of periodic waves in models of population transition. Similarly, the capacity of Asian societies to change and adapt was also underestimated in the 1950s, with as serious a thinker as Julian Huxley suggesting in 1956that Japan’s overpopulation problem was so serious as to make its situation even more desperate than that of India unless a drastic population policy took quick effect!

Some of the mechanisms by which culture and society had impinged on fertility were known in the eighteenth century, and discussed by Malthus and others. Folkways of checking fertility were present even before the spread of modern contraception. Furthermore, apart from pre-natal intervention, family size and composition could be (and had been) adjusted by the post-natal practices of adoption and infanticide. In Western Europe, an important breakthrough in the study of the social regulation of fertility was the identification by John Hajnal of the ‘European marriage pattern’- one in which a significant number of women remained unmarried, and therefore excluded from legitimate reproduction. Generally speaking, this proportion varied with economic conditions-prosperity encouraged earlier marriage and higher fertility, while hard time resulted in a postponement of marriage and reproduction with some persons remaining celibate for life. This pattern has been extensively investigated in the English case, where Wrigley and Schofield found it to hold until the middle of the eighteenth century, with the proportions never married fluctuating from 5 to 22 per cent of the female population. With the onset of the Industrial Revolution and the boom in the demand for labour, however, ‘there was little change in this aspect of nuptiality but a rapid and substantial fall in age at marriage, followed by a compensating rise in the nineteenth century.’ Throughout the period it was fertility rather than mortality that determined the course of population change.

In other parts of the world, marriage had different functions, and age at marriage had less demographic significance. So, for example, in early-nineteenth-century Japan, the north-east had low age at marriage (between 13 and17 years for women) but also low completed fertility and early termination of child-bearing (at mean age 32.9 years). This was related to the4 lack of economic opportunity-a newly married woman was ‘supposed to have children as soon as possible so she can soon become part of the labour force herself.’ A study of central Japan also supported the hypothesis that age at marriage was unrelated to fertility, since marriage was seen as a labour recruitment device. The age of first marriage for women was high by Asian standards, being 19 in the late eighteenth century. Here, as in the western case, marriage may be regarded as commencing cohabitation; but in regions such as South Asia, ceremonial marriage might well precede consummation by several years. Data on the latter are hard to find, but in the 1930s, a large number of rural doctors were asked about common practices in their areas of work. They reported that in different provinces cohabitation began when girls were thirteen to fifteen years old, and the first child usually borne at the age of sixteen. This would imply that cohabitation usually began at menarche.

Things may have changed somewhat by 1972 when a fertility investigation formed part of the Sample Registration Scheme in India median age at ‘effective marriage’ for girls was 16.97 years in the rural areas, and 17.26 in urban. There is also a possibility that women’s ages in the 1930s were being understated by one or two years. This is suggested by the fact that the gap between commencement of cohabitation and birth of the first child was estimated at 18 months in Madras state, and 24 months in five other states. Only in the Central provinces (36 months) and Punjab (42 months) was it thought to be higher. The 1972 Survey reported, however, substantially longer intervals between ‘effective marriage’ and first birth for rural women the only exception being the north-eastern states, where adult marriage was common. Thus the gap was 38.07 months in Uttar Pradesh, 48.24 in Madhya Pradesh (Central Provinces), 44.10 in Bihar, 30.45 in West Bengal and 32.93 in Punjab. Tamil Nadu had a relatively small interval 27.53. Now, there seems to be no good reason to accept that the gap between effective marriage and conception increased between the 1930s and 1960s, and thus I suggest that there may have been underestimation of age in the earlier survey. This statement is supported by the fact that median age at first birth all-India was 21.0 in the countryside and 21.8 in the towns in 1971-2. Thus the apparent increase in age at effective marriage from early to late teens would have had little impact on fertility since it would appear that fecund ability was significantly lower in very young women. As result, the total marital fertility for rural women surveyed in 1972 was 6.9 for those married below 186.7 for 18-20, and only fell to 5.2 where age at marriage exceeded 22.

The completed fertility for marriages lasting twenty-five years in historic China has been estimated at 5to 8 while the Indian Sample Registration Scheme found it to be 5.5 in the rural areas for 1972. In both countries, it is likely that marriage was seen as important to reproduction of the family. Thus it becomes evident that the way families are organized and the cultural values that regulate marriage and reproduction had, and have, an important demographic impact. If the ideal of early and universal marriage blocked one of the mechanisms of Hajnal’s western marriage pattern, a similar reduction in the number of child bearing women could result from the practice of female infanticide a widespread phenomenon in India into the late nineteenth century. This was (and in the modern form of neglecting girl-children is) a consequence of the cost of marrying them suitably when grown. There is also qualitative evidence for the postponement of marriage on economic grounds in north India thus in 1852 upper-class Muslims in Punjab are described adjust as extravagant as Hindoos in their marriages, but then they need not marry their daughters at all if they cannot afford it; and it is notorious that on this account the daughters of the best Mahomedan families often remain unmarried to the age of thirty.

This practice does not seem to have been significant enough to show up in the Census records. Much demographic analysis was and is organized around the search for the fertility transition the shift from a regime of high fertility thought to be characteristic of ‘traditional societies’ towards the low, controlled fertility of the modern. The fragmentation of this binary dichotomy has been a major achievement of historical demography, and recently allowed Chris Wilson to point out how enormously both fertility and mortality varied between regions and over time in so-called ‘pre-transitional’ societies.34 Thus an important recent paper by Monica Dasgupta develops a model for north India (Punjab) in which marriage restriction limited the size of families in which several brothers stood to inherit a small landholding. The limitation was sometimes combined with informal fraternal polyandry. This automatically limited the number of male heirs in the next generation. Such calculations, however, were based on relatively high levels of background and crisis mortality, and were initially swamped by the accelerating decline in mortality that began from the 1920s. By the1940s, however, landholders had begun to control fertility to accommodate hi8gher expected survival rates, and this became common in rural society as a whole by the early 1980s. Points of major significance for historians emerge from this article. One of these is that the relatively high fertility observed in the twentieth century had in earlier periods been related to historically experienced levels of background and crisis mortality, and that the adaptation of fertility to the twentieth century decline in mortality had already begun. At the global level, as S.H. Preston pointed out in 1990s levels. Thus efforts at fertility control, earlier restricted to some social groups, may spread through the population when social and economic circumstances change. But while fertility may be the central problem for contemporary demography, historical studies must give equal or greater weight to mortality.

Contents