There is a polarity in the organization of knowledge, between analysis and synthesis, between the sifting and organization of the hard data that were collected long ago for different purposes, and the models that are often based on intuition but serve to organize the facts as they are gathered. These models interpret historical events into intellectually satisfying constructs that are not necessarily grounded in sound evidence when they are proposed but that provide hypotheses to be tested in subsequent research. The progress of science is then determined by the success of efforts to either confirm the model or to bring the orthodoxy down and replace it with another one, more consistent with the facts. Whereas historical demography in its early years was dominated by the exploration of sources and the development of measurements, the emphasis has now shifted toward the investigation of whole social structures or demographic systems. The previous section of this chapter organized the discussion ''horizontally'' as it were, by looking at sources that may provide information on a variety of topics. This section and the next proceed ''vertically,'' by considering large topics that have retained the attention of historical demographers and historians of population. This section considers the more properly demographic topics: fertility, mortality, and marriage. The next section describes some exemplars of more interdisciplinary research in the demographic structure of societies of the past, where historians of population have played a primary role.
Louis Henry was led to the study of historical populations by his interest in the proximate determinants of natural fertility, i.e., the fertility of married couples that did not attempt to limit the number of their children by resorting to contraception or abortion. He did not invent the term natural fertility, which was commonly used before him to refer to the reproductive regime of high-fertility populations of the past. In early writings, Henry had called them ''non-Malthusian populations'' and qualified their fertility as biological; the terms were clearly inappropriate. He took an important step by providing a way of identifying natural fertility in statistical series, i.e., when the fertility of married couples was not a function of the number of children they already had. By comparing successive marriage cohorts of individuals in family reconstitutions, it was possible to distinguish clearly between the behavior of early cohorts where fertility was natural and that of later cohorts who practiced family limitation. The publication of Henry's (1961) result on natural fertility exerted an influence well beyond historical studies and laid the foundation for biometric models of the reproductive process and for the proximate determinant framework of fertility.
The dominant model in demography until then had been the theory of the demographic transition, a broad pattern of connected demographic changes in the course of modernization resulting from socioeconomic influences. One of the findings of the Princeton European project was that the decline of fertility took place under a wide variety of circumstances, at different levels of mortality, and in countries that had reached very different levels of socioeconomic development (Knodel and van de Walle 1979). This cast doubt on the theory of the demographic transition. The focus shifted to the timing of the fertility transition, the change from a regime of natural fertility to one of family limitation. A series of methodologies to measure the time of transition were proposed, based on the shape of the pattern of marital fertility (Coale and Trussell 1974) or on the average age of mothers at the time of their last birth (Knodel 1988; Wilson 1984). The Princeton European project proposed a dating of the fertility transition by computing the time when the index of marital fertility had declined by 10% from a predecline plateau (Coale and Watkins 1986). Such a predecline plateau of high fertility has been identified in most countries. For rural France, where the decline started before the existence of vital registration and censuses, David Weir (1994) connected the results of the family reconstitutions made by Louis Henry and his team with the 19th-century statistics and identified the end of the natural fertility regime at a time close to the
French Revolution. If France is excluded, close to 60% of the province-sized administrative areas of Europe experienced their marital fertility decline starting in the 30-year period from 1890 to 1920 (Coale and Treadway 1986:37). The concentration in time, in countries with vastly different economic and social conditions, suggested a common mechanism of ideational transmission rather than an adjustment to socioeconomic pressure.
A number of recent studies have attempted to rehabilitate the notion that marital fertility declined because of an adjustment to economic change, by using more sophisticated econometric models and a lower level of aggregation than the province (Galloway, Hammel, and Lee 1994; Brown and Guinnane 2002). The concept of natural fertility and the methods used for its measurement have been criticized for various reasons, for example, their lack of sensitivity that makes it very difficult to date the transition point with any precision. Low values of m, Coale and Trussell's indicator of the prevalence of parity-dependent family limitation, could be compatible with a significant proportion of the population practicing it (Okun 1994). Others have objected to the inability of the model to factor in the spacing of births (rather than the limitation of their number) that appears to characterize many reproductive regimes (Bean, Mineau, and Anderton 1990).
More importantly, controversy continues over the reality of the fertility transition. It has been asserted that there was not one fertility transition, i.e., a single historical movement, but many different transitions, each with its own motivations and characteristics, or even simply a continuous adaptation of fertility levels to local economic opportunities and the level of mortality. Simon Szreter (1996) hypothesized that the decline of fertility in England was the result of abstinence and spacing. There have been efforts to revive the theory of the demographic transition, by linking the secular decline of fertility to a decline of mortality whose effect was only appreciated with a very long lag by married couples (Chesnais 1992).
The concept of natural fertility started as a purely demographic concept, but it has become linked to the history of birth control. Two recent histories of contraception have deemphasized the innovation aspects of the fertility transition by claiming that women at all times had access to methods adequate for their needs (McLaren 1990; Riddle 1992). Janet Farell Brodie (1994), in an exemplary study of contraception and abortion in 19th-century America, has emphasized the extent to which the ideas and the paraphernalia of birth control were diffused in the public by pamphlets, commercial catalogs, advertising in newspapers, conferences, and word of mouth and were eventually stymied at the end of the century by legal interventions and a reversal of public opinion.
Crises of Mortality and the Epidemiologic Transition
Historical demography helped push back the date of reliable mortality data, particularly for infant and child mortality (Schofield, Reher, and Bideau 1991). Population reconstruction provided reliable life tables for the rural French and English population in the 18th century; official statistics for Sweden go back to 1750. These series show a steady decline of mortality from the high and fluctuating levels that prevailed at the earliest dates. In contrast, the decline of infant mortality is less obvious and does not start in earnest before the last decade of the 19th century (Vallin 1991). The prevailing general descriptive model of secular changes in mortality is that of the epidemiological transition from a stage of dominance by infectious and parasitic diseases to a situation where people die mostly in old age from chronic and degenerative diseases.
The statistical record before 1800 is dominated by peaks of exceptional mortality from crises caused by a mix of food shortages and epidemics (Livi-Bacci 1991). The history of such diseases as the bubonic plague (Biraben 1975), smallpox (Hopkins 1983), syphilis (Arrizabalaga, Henderson, and French 1997; Quetel 1990), or tuberculosis— even in the absence of precise quantitative information on causes of deaths—sheds light on the mechanisms of infection and on attempts at treatment in the past. In the 19th century, the frequency and intensity of crises diminish (Perrenoud 1991), although there are still occasional peaks of mortality in various countries from identifiable causes, such as the cholera epidemics of 1832 and 1856, the potato famine of 1846, and the 1870 smallpox epidemic. The Swedish data show a sharp decline in the frequency and mortality of smallpox epidemics in the 19th century as a result of the adoption of vaccination, the first successful use of preventive health measures against an infectious disease (Skold 1996).
Two explanations have been given for the importance of infectious diseases in the past and for their decline. In a first explanation, the periodic visitation of diseases, and their establishment in an endemic form, was essentially an exogenous influence, an accident of epidemiology under favorable environmental circumstances. William McNeil (1976) illustrates this position. Disease pools developed in isolated parts of the world and were diffused with a devastating impact when interregional contacts increased, eventually to form a ''common market of diseases.'' Infections such as smallpox and measles decimated the population at the time of first contact, but conferred immunity to their surviving victims and became childhood diseases. The epidemi-ological transition was the result of an increasing ability of preventive measures (e.g., smallpox vaccination) and the development of hygiene and public health (e.g., sewerage and clean water supply) to combat the infections.
The opposite position has represented the high levels of historical mortality as endogenous, the result of low standards of living and population pressure. The idea of a Malthusian equilibrium where population growth is kept within the bounds of subsistence by the positive check of mortality finds some support in the analysis of the record for Elizabethan England. By the 17th century, however, adjustments seem to occur through the preventive check on marriages (Wrigley and Schofield 1981: chapter 11). Thomas McKeown (1976) has argued that the decline of mortality in England during the 19th century, and hence the modern rise of population, was caused neither by medical innovations nor by an attenuation of the virulence of disease, and therefore must have been the result of better nutrition and an improvement in the standard of living. McKeown downplayed, and probably underestimated, the impact of smallpox inoculation and vaccination and of public health measures in the area of sanitation and water supply.
Whatever the reasons for the increase in expectation of life at birth during the 19th century in Europe and the United States, they do not seem to have operated on infant mortality, which remained practically unchanged before the last decade of the 19th century and then underwent a rapid decline. Although there is no consensus on the reason for this late decline of infant mortality, the most convincing analyses for England (Woods 2000), the United States (Preston and Haines 1991), and France (Rollet-Echalier 1990) suggest that the crucial element was a revolution in public and private attitudes toward the child, combined with new principles of hygiene in the home, including boiling milk and sterilizing bottles, which represented a breakthrough in the practice of artificial feeding.
The Western European Marriage Pattern and the Structure of Households
In 1965, John Hajnal published a pathbreaking study of European marriage patterns, in which he pointed out the existence of two distinct types of marriage separated roughly by an imaginary line going from Leningrad (as it was then called) to Trieste. The western European countries were characterized, around 1900, by late female marriage (typically above 25 years of age) and extensive celibacy. Eastern Europe, in contrast, had early marriages and high proportions marrying, resulting in higher levels of overall fertility. Hajnal attempted to trace back the temporal origins of the pattern and concluded that it had existed in many countries of western Europe at least since the 17th century. He associated it with an original pattern of household formation: ''In Europe it has been necessary for a man to defer marriage until he could establish an independent livelihood adequate to support a family; in other societies the young couple could be incorporated in a larger economic unit, such as a joint family'' (Hajnal 1965: 133). Alan Mcfarlane (1986) linked the European pattern of marriage with early capitalism, the dominance of market forces (particularly for the sale of land in private ownership), and the prevalence of wage labor. He believed that these conditions were attained in England by the 13th century and later in other European countries. Research has confirmed that the western European marriage pattern is distinctive, although there are other systems throughout the world where marriage is relatively late and the young couple sets up an independent home. Tokugawa Japan, offered such an example, although permanent celibacy was rare.
In theory, a marriage could only take place when a new couple had the economic resources to allow its independent establishment. The reconstruction of the English population by Wrigley and Schofield (1981) clearly demonstrated the role of nuptiality as the feedback mechanism regulating the growth of population in the 18th century so that it would not exceed the resources available for its support. Favorable real prices and agricultural wages would lead to earlier and more frequent marriages; times of scarcity would lead to the opposite result(for a recent review of the English nuptiality story in a long-term perspective, see Smith 1999). It has been argued that the development of cottage industry, and later the industrial revolution, by providing new opportunities for both male and female wage labor, must have led to younger and more universal marriage (Levine 1977). The evidence from the study of microdata from the 19th century is mixed on this, and the most recent research suggests that many factors regulated the contracting of unions during the 18th and 19th centuries in somewhat conflicting directions, depending on country, region, and the local economic organization (Devos and Kennedy 1999). These factors included the dominant forms of wage labor (e.g., day labor versus living-in servants), employer's control over housing, the intensity of human labor in the prevailing technology, the substitution of cheaper female workers for male workers in some types of production, and the elasticity of the marriage market with respect to increases in the standard of living. In the aggregate, and with some exceptions, the vast expansion of the industrial labor force and the steady increase in wages during the industrial revolution did not produce a commensurate increase in the proportions married before the 20th century in western countries (Coale and Treadway 1986).
The study of household structure from population listings, although at face value outside of the scope of demographic studies, provided one of the missing links in the reconstruction of past societies. It was initiated by Peter Laslett and his colleagues at the Cambridge Group for the History of Population and Social Structure. Household structure is a topic that had received little attention in traditional demography; the household was usually treated simply as a unit of enumeration in the census, without theoretical interest. The common misperception was that families of the past were very large and extended.
A large number of manuscript census listings from the past have been examined, and they show conclusively that with rare exceptions, the conjugal family unit (a man, his wife, and their children) constitutes the basic residential pattern of western populations and that this has been the dominant household structure among the common people of western Europe as far back as the evidence from enumerations will carry (Wall, Robin, and Laslett 1983). Households, as today, were typically limited to the conjugal pair and their children and were therefore small, and the marriage of the children entailed the founding of new households.
Was this article helpful?
Are You Expecting? Find Out Everything You Need to Know About Pregnancy and Nutrition Without Having to Buy a Dictionary. This book is among the first books to be written with the expertise of a medical expert and from the viewpoint of the average, everyday, ordinary,