1It is largely for accidental reasons of historical development that sociology and demography exhibit the mutual isolation and indifference so often observed between the two sciences. The Greek word δημος described the land inhabited by a people. And then the people itself, ethnically defined. And then the political group, the set of all nationals, as well as a defined subset of that group, as in ancient Athens. The notion of society underpins the various Greek meanings of δημος. Nothing in the very word “statistics” predestined this science to be anything other than the political branch of sociological studies. It is only an accident, too, that “demography”, did not become by etymology the properly formed equivalent of the Greek-Latin hybrid “sociology”.

2The precarious nature of etymological discussions and conclusions is of little import. The difficulties of separating out the semantics of words concerning the study of Man provide a more important lesson. It is that the various branches of the science of Man are closely interdependent, although they were born separately, at a time when the desire to synthesize and the opportunity for overview were less than now, contrary to received wisdom. Examining the relations between sociology and demography lends support to this affirmation.

3That masterly compilation, Recent Social Trends [1], which reports on the state of society in the United States in the years around 1930, and whose contributors included many of the best American sociologists, begins with the following words written by the demographers Thompson and Whelpton:

4

“Human beings are the primary agency of social change. The rates at which the population grows, its geographical distribution and the proportions in which it is divided between farms and cities, the racial and national stocks from which it comes, its age trends, sex ratios and marital condition – all of these help to determine the rapidity and the direction of past and future changes. In surveying recent social changes in the United States it is appropriate to begin with these basic factors of births, deaths and the numbers living. With this definite knowledge in mind, we can better understand the changes in the ways that Americans make their livings, the values which appeal to them, their criticisms of themselves, the fears and hopes they entertain about the future” [2].

5It is always no doubt a good idea to remind sociologists of the importance of demographic data. The very earliest forerunners of sociology, Plato and Aristotle, showed by their example that an alliance must be made between the science of society and the science of population. In their plan for the ideal city, they regulated in the smallest detail the quantity and quality of the population (Republic V, Laws V and VI, Politics II). The great task was to determine the optimum size of the population and then take all necessary measures to prevent that number from varying:

6ιν∍ ως μαλιστα διασωζωσι τον αυτον αριθμον τϖν ανδρϖν

7(Republic V, 460a [*]); measures that did not exclude contraception, infanticide and emigration; and Aristotle tells us that Hippodamus of Miletus and Phido of Corinth before him had designed similar regulations.

8The demographic concerns of the Ancients in social matters related mainly to their economic and military effects. Nor are these the only effects. Sorokin has shown this in his attempt to bring some sort of order to the theories of modern sociologists [3]. The size of the population will influence national power and the progress of civilization according to Adolphe Coste; the division of labour according to Durkheim and Kovalevsky; economic prosperity according to D’Avenel, Gini and the optimum population theorists; language change according to Carli; political ideology according to Bouglé; and Sorokin cites other examples.

9Among contemporary French sociologists, the one who understood most clearly the sort of contribution that basic demographic data could make to sociological research was probably the late Maurice Halbwachs. Trained in the Durkheim school, which sought to study social realities as “things”, and consequently gave “particular importance to what in society takes on the character of the physical: area, number, density, movement, quantitative aspects, everything that can be measured and counted” [4], Halbwachs was a friend of the econometrician and sociologist Simiand, and himself a specialist in statistics and the calculation of probabilities [5]; he no doubt did more than any other in this country to create closer ties and coordination between sociology and demography.

10In his short Morphologie sociale [6], one of the last books he published, Halbwachs indeed examines the points of contact between the two sciences. Whether in terms of religious, political or economic morphology, social situations and their changes are closely linked to the demographic state of society and indeed largely determined by it from the outset. With respect to religion, for example, the conceptual schemes of demography apply directly to the description of religious life. One can design censuses of believers, maps of higher or lower religious practice, demographic distributions by age and sex. Within a given town, one can establish the ecology of religious practice as of population factors. Movements of the population of believers, from birth in the faith (christening) to death (at least provisional), mirror their physical movements. Independently of these parallels, which are useful in themselves, it can also be seen that demography affects religious observance. Isolation enables a group to gather round its church. The disintegration of the social group in large cities produces an immediately observable opposite effect. Among the Eskimos, who live close together in winter and apart in summer, the intensity of religious practice is much greater in the former state than in the latter. The success of Christianity at its origins was made easier by the volume and cohesion of the poor populations in large cities who formed its audience. Its growth depended on the expansion of the population by conquest and procreation. Similarly, political systems are shaped by population features. Halbwachs rightly recalls one of Durkheim’s remarks: the reason that civil justice in civilized states has taken over from the criminal justice that prevails in less polite societies is that the increase in population differentiates interests, infrastructures, services and situations, and makes civil relations more intense and more frequent. And the many aspects of economic life that are studied in sociology are also linked to demography. The size of the population regulates the supply of labour, and the demand for products. The unequal development of the different classes arises out of its dependence upon and the relative size of various industries, such as food and luxury goods, and the prices of these various goods. Halbwachs’ book contains a nuanced and convincing account of how social morphology in the widest sense is dependent upon demographic phenomena.

11But the number, distribution and composition of a country’s population do not merely give shape to its sociological structure. Problems of social action may be an effect of demography; there are social inadequacies whose causes are demographic. Over-population in the East, like the falling population in most Western nations, affects ways of life both cultural and material. Urban concentration, in appearance a purely demographic question, has justified the development of an almost independent branch of sociology, urban sociology. The specific age structure of the population is just as important. Population ageing will affect not only the economy and employment, but also the arts and politics. It is no exaggeration to say that demography largely commands social life and that every sociologist should also be a demographer.

12Halbwachs undertook to examine how demographic phenomena exert this social influence. It is not, he found, through external determination, like some sort of physical force, but through the understanding – however imperfect – of these phenomena by the members of the social body: “The material shape of society acts on that society, not as a physical constraint, as one body might act on another, but by the understanding we have of it as members of a group that perceive its volume, physical structure and movement in space. This is a type of collective thinking or perception that one may call an immediate datum of social awareness, quite different from all the others, and which has not yet been sufficiently examined by sociologists” [7].

13If the sociological action of population phenomena owes its efficacy to the perception of these phenomena by group awareness, it is because they are in themselves human phenomena; over and above their material sense, they are charged with meaning. When a body falls freely in a vacuum, there is no thought or intention involved. The event is sufficiently described once it is explained, once the details of the reasons for its fall have been given. It will be said to cover a distance proportional to the square of the time elapsed, and this law of free fall can be attached to a general theory, such as the basic equation of mechanics. This is not the case with a population event. Nuptiality, for example, does not simply depend on the biological fact of sexual maturity. From one people to another, one civilization to another, the legal or usual age of marriage may precede or follow the age of puberty, sometimes by years. Marriage has a social significance, and nuptiality is a human phenomenon that needs to be understood.

14Even at the basic methodological level of conceptualization, therefore, demographic phenomena involve sociology. In practice, there is no need to make this point to population specialists. However material the facts with which they concern themselves, these specialists do implicitly admit that births and deaths, marriages, age and sex distributions, internal and external migration not only have social effects, but are by their very nature charged with sociological meaning. However, this admission is all too often only half-conscious. The reason is partly that demographic findings are collected in a national framework within which this meaning rarely changes. Halbwachs is right to remind us, however, that the balance of the sexes, for example, is not solely due to organic causes. It can be upset in societies that do not fully accept it, and that practise selective infanticide. Similarly, the age distribution is regulated by social intervention. For our current balance to be maintained, society must exercise a moderating influence on the competition between the stronger and the weaker, between children and old people and workers in the prime of life. The extension of average life expectancy does not merely depend on the progress of public health and medicine, and even less does it indicate some change in the biological nature of humankind. It is conditioned to some extent by an increasing respect for the value of the individual life, which manifests itself to varying degrees in different national communities and socioeconomic groups, and is expressed by greater or lesser endeavours by individuals and groups to cure or prevent illness. For fertility, the psychosocial aspect is even more obvious. We shall return to this below. Meanwhile, let us note that demographic phenomena, which may seem at first sight to be as material as, say, geological data, are laden with social meaning and belong in their definition and interpretation to the field of sociology.

15This is not merely an academic point. In his article on “Demography” in the Encyclopaedia of the Social Sciences, Wolf defines demography as “the numerical analysis of the state and movement of the human population, including census enumerations and vital records, and any general quantitative statistical analysis of the state and movement of the population that may be made on the basis of fundamental census and registration data”[retranslated]. By this definition, demography is closely related to censuses and the organization of vital records, its scope seems to be restricted to such data, and its task is merely to examine them quantitatively. However, these enumerations are administrative acts carried out in administrative districts. They thus impose the same frameworks for all demographic analyses. These frameworks, often highly artificial, conceal the real shape of population phenomena wherever these phenomena are linked to sociological or cultural conditions independent of administrative divisions.

16For example, Ravenstein’s hypothesis on the forms of internal migration contains, among other points, the assumption that an individual moves by small steps, from village to nearest small town, from there to a larger town, and only reaches a major conurbation at the end of the process [8]. In the current state of the data, it is almost impossible to verify this theory. The best-designed censuses generally only take account of individuals’ district of birth and ignore any intermediate moves. Similarly, it appears in most countries where the question has been studied that the rural birth rate is higher than the urban rate. In the current state of French documentation, presented by départements only, it is impossible to verify this fact. In general, the sociologically important distinction between rural and urban areas of residence, which has such a major impact on demographic phenomena such as fertility, mortality, nuptiality, divorce, sex and age distribution, can virtually never be measured because the facts are presented by département, and are not broken down by type of living environment.

17Indeed, identifying urban areas is a difficult problem that is at least partly sociological in nature. In its preparations for the 1930 census, the US Bureau of Census asked cities of more than 50,000 inhabitants to produce maps of their territories including places that met the following conditions, among others:

18

“suburb where no fewer than 10% of the working population -travels daily to the central city; …zone receiving daily newspapers from the central city delivered by these newspapers’ own deliverymen;… residential areas of people belonging to social and sports clubs headquartered in the central city;… business area of real estate companies of the surrounding area…” [9].

19Clearly these criteria are economic and sociological in nature.

20Beyond the limits of the metropolitan city there is also the zone of influence of the city which affects demographic phenomena as much as any other social phenomena. Thompson established from a study of the environs of sixteen conurbations that for nine of them the fertility rate in rural localities increased in inverse proportion to the distance from the main city [10]. The Americans, well-aware of such realities, have been trying for more than fifteen years to define a regional classification of their country, independent of administrative and political divisions, that presents the facts in the appropriate sociological and economic framework. One of the first attempts was the Atlas of Wholesale Grocery Territories, produced in 1927 by the Bureau of Foreign and Domestic Commerce on purely economic bases. Later exercises took the circulation of metropolitan newspapers [11] as their main criterion, producing a clearly cultural division.

21However, the framework for demographic phenomena is not solely regional, or geographical. US census data prove that fertility is not only a function of the place of residence. It also varies with national origin and profession. In every type of residential area, the fertility of foreign-born populations was higher than that of the native-born in the censuses of 1910, 1920 and 1930. In rural areas, whatever the origin of the group analysed, those who work in farming have the highest fertility [12]. In this way, sociology makes a major contribution to demography. By using its own methods to analyse the variables that affect demographic phenomena, it provides a better breakdown of the information that demographers have to analyse. Sociology prepares the frameworks into which the data of population science will be inserted.

22Demography would not need to turn to sociology for its frameworks if social phenomena did not, conversely, affect population phenomena. The former also help to explain the latter. A striking example is that of fertility. It is simplistic to see fertility, as some have done, as an essentially biological phenomenon. It is probably not even sufficient to see it as an economic one.

23An inverse relationship has been found between income level and net reproduction rate [13]; it has even been shown that the inverse relationship previously discovered between socio-occupational status and fertility [14] appears to correlate much less than fertility and resources, except at the lowest economic levels [15]. But even here there are exceptions, particularly in Sweden [16]. This would tend to illustrate the role of the national factor, and requires at least some discussion. Above all, this inverse relationship, which seems to hold in most cases, does not establish the influence of economic factors on fertility. On the contrary, it indicates the presence of processes unconnected with economics. If birth control were economically determined by the desire to escape the cost of having children, we would expect to find that it is the poorest who make the greatest effort to avoid children. In the actual state of things, to say that the economic-demographic relationship is a causal one is simply nonsense.

24In reality, this relationship merely accompanies other processes, which are not economic but sociological. In the current state of research and knowledge, it is impossible to claim to give an exhaustive or even adequate account of them. It is not even possible to propose any partial but definitive interpretations. A few observations, however, which may seem impressive, give some provisional clues. A study by Burgess and Cottrell of a sample of middle-class families reveals that the couples best adapted to our civilization tend to have fewer children [17]. Furthermore, some clinical observations suggest that at the lower socioeconomic levels there is sometimes a sort of compensation effect, whereby having a large family makes up in the parents’ minds for their economic and social failure [18]. It is in the very structure of our cultural lifestyle, observed by sociologists in Western societies, that we may need to seek an explanation for the short- or long-term decline in fertility. The civilization we know, which finds its apogee in urban life, promotes the spirit of individual competition and the development and success of the person. The family spirit clearly belongs to a totally different system of values. It appears indeed to be the case that the classes that present the lowest birth rates are precisely those most engaged in this industrial, urbanized and individualistic civilization. A prime example is that of the ruling classes, who have very few children, and enjoy the fruits of this civilization to the full, both culturally and materially. Another, more tragic, example is that of the white-collar workers, the intelligentsia, who are increasingly proletarianized and fight desperately to maintain their level of civilization [19]. At all events, it is within this class that demographers can understand that the problem of fertility depends at least as much on sociology as on economics or biology.

25In general, demographic phenomena, like all human phenomena, must be interpreted as the result of the general twofold process of competition and cooperation that characterizes the struggle for life. Be it a question of growth, contraction, migration or composition, these are the processes that underpin all forms of population change. What appears chaotic, irrational, inevitable in population phenomena may only be so in the absence of any reference to these basic concepts. As such, of course, they are too general and too wide. It is the detail that counts. Ecological studies must be undertaken to examine the distribution of human phenomena within frameworks that are not only geographical but also social. Human ecology, the science of distribution, half-way between that of population and that of social phenomena, is the obvious bridge between the two branches of knowledge. The Chicago school, in particular, has rendered an inestimable service here. It is, no doubt, time that we caught up by combining the endeavours of both demographers and sociologists.