Social Structure and Fertility: An Analytic Framework Author(s): Kingsley Davis and Judith Blake Source: Economic Development and Cultural Change, Vol. 4, No. 3 (Apr., 1956), pp. 211-235 Published by: The University of Chicago Press Stable URL: http://www.jstor.org/stable/1151774 . Accessed: 09/10/2013 18:16 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact
[email protected].
.
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to Economic Development and Cultural Change.
http://www.jstor.org
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
AND FERTILITY: AN ANALYTICFRAMEWORK SOCIAL STRUCTURE feature of underdeveloped areas is that virtually all of them A striking than do urban-industrial exhibit a much higher fertility This wellsocieties. documented but insufficiently analyzed fact is known to be connected with proin social as between the two types of society, found differences organization for the comparative sociology and is therefore of reproduction. The significant and importance of the contrast, clarity however, should not be allowed to obscure the equally important fact that underdeveloped areas themselves differ markedly in social and that these differences organization, appear to bring about variations in fertility. of backward regions have Though the demographic statistics been so poor as to place in doubt the validity of reported differgenerally (e. g., as between ences, there are cases in which the evidence is reliable Of equal interest Puerto Rico and Jamaica, or Arab Palestine and Ceylon). are with differing the cases in which societies social have the same organization for they may reach this common result by quite different level of fertility, All told, ample opportunity exists for the comparamechanisms. institutional of social structure as it affects In view of the tive analysis fertility. trends on economic development, the pursuit of such bearing of future population has a practical as well as a theoretical analysis significance. an attempt to set forth and utilize The present paper represents an anaframework for the comparative sociology of fertility. It first lytical presents a classification of the intermediate variables factors through which any social the level of fertility must operate. It next tries to show, in influencing how some types and elements of social broad outline, organization, acting through these variables, Our hope is appear to enhance or depress societal fertility. that as more sociological and demographic information becomes available, the theories advanced can be refined further and tested empirically. The Intermediate
Variables
three necessary steps sufficiently The process of reproduction involves in human culture: obvious to be generally (1) intercourse, (2) conrecognized In analyzing cultural and (3) gestation and parturition.1 influences ception, on fertility, one may well start with the factors connected with these directly Such factors would be those through which, and only through which, three steps. For this reason, by way of convenicultural conditions can affect fertility. and can be presented variables" ence, they can be called the "intermediate as follows: schematically
(1)
these can all sees more steps in the process, Although the physiologist We are concerned only be subsumed under the three headings given here. and as they may be socially with the steps in reproduction recognized utilized.
-211-
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
AND FERTILITY SOCIAL STRUCTURE
212 I.
Factors A.
Affecting
Those governing ductive period.2
II.
sexual
into
("Intercourse
and dissolution
the formation
Variables").
of unions
in the repro-
unions.
1.
Age of entry
2.
Permanent celibacy: unions.
3.
Amount of reproductive period spent after or between unions. or desertion. a. When unions are broken by divorce, separation, b.
B.
to Intercourse
Exposure
Those governing
are broken by death of husband. within unions. the exposure to intercourse
abstinence.
Voluntary
5.
abstinence Involuntary temporary separations).
6.
Coital
(from impotence,
unavoidable
illness,
but
periods of abstinence). to Conception ("Conception Variables").
(excluding
frequency
Affecting
Exposure
7.
Fecundity
8.
Use or non-use
as affected or infecundity, of contraception.
a.
sexual
When unions
4.
Factors
of women never entering
proportion
and chemical
By mechanical
by involuntary
causes.
means.
By other means.3 as affected by voluntary Fecundity or infecundity, medical treatment, etc.). subincision, lization, b.
9. III.
Factors ables").
Affecting
Gestation
and Successful
10.
Foetal
mortality
from involuntary
11.
Foetal
mortality
from voluntary
Parturition
causes
(steri-
("Gestation
Vari-
causes. causes.
(2)
is not confined to wedlock, the term "sexual Since sexual intercourse A union is here defined as any to "marriage". union" seems preferable in which either actual intercourse occurs or heterosexual relationship Every society has a orgasm is produced for at least the male partner. is expected, approved, and type of union (marriage) in which reproduction runs the risk of unions in even enjoined. At the same time every society is condemned, either because they lack the legal form which reproduction one or more institutional taboos of marriage or because they violate K. Davis, "The or class enuogamy, etc.--see caste, (adultery, incest, Forms of Illegitimacy", Social Forces, Vol. 18, October 1939, pp. 77-89). Between the fully approved and the strongly proscribed unions, there may be other types which have a lesser grade than marriage but in which rein some cases Such unions may be frequent, normally occurs. production sothe majority of reproductive unions. Any satisfactory representing of reproduction must keep straight the different ciological analysis types of unions.
(3)
Means of contraception other than mechanical and chemical include the with"rhythm" method (which can also be classed as voluntary abstinence), without penetration, various '"perversions", drawal, simulated intercourse etc.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
ECONOMIC DEVELOPMENT AND CULTURAL CHANGE
213
It is clear that any cultural factor that affects must do so fertility in some way classifiable under one or another of our eleven intermediate variHence the latter ables.4 provide a framework in terms of which the relevance of cultural factors to fertility can be judged. In fact, attempts to explain causal relationships' between institutions and fertility without such a framework have led to inconclusive and confused writing on the subject.5 The cultural factors, or "conditioning are presumably many, and no effort variables", is made here to classify variables" offer a means them; but the "intermediate of approach to selecting and analyzing these factors. It is also clear that each of the eleven variables may have a negative on fertility. If by examining all socie(minus) or a positive (plus) effect ties we could find the range of influence of a given variable, any effect more than the midpoint of this range would be on the minus side, and any negative influence more positive would be on the plus side. If, for example, a society uses contraception it has a minus value with respect to variable successfully, number 8; if it uses no contraception, it has a plus value on this variable. The value of each variable refers to how it affects in each case; so fertility a positive use of something (e. g., contraception, abstinence) abortion, may mean that it has a "minus" fertility-value. One cannot say, as is frequently that some of implied in the literature, these variables are affecting in one society but not in another. All fertility of the variables are present in every society. This is because, as mentioned each one is a variable--it can operate either to reduce or to enhance before, If abortion is not practiced, the fertility-value of variable number fertility. 11 is "plus". In other words, the absence of a specific does not imply practice "no influence" on fertility, because this very absence is a form of influence. It follows that the position of any society, if stated at all, must be stated on all eleven variables. Societies in their social do not necessarily have differing organization different with respect to all the variables. On some of the fertility-values variables A nomadic tribe may have the they may exhibit quite similar values. same age at marriage as a settled a primitive agrarian village; group may practice the same rate of abortion as an industrial Two contrasting socisociety. eties are not likely, however, to manifest similar values for all the variables; to do this even when their general fertility level is practhey are not likely the same. The actual birth rate depends on the net balance of the values tically of all the variables. which generate a high fertility tend to Though societies be predominantly on the plus side, no society has the highest plus value on all eleven variables; and societies with low fertility turn out to be amazingly on a number of them. positive (4)
The reader will note that our list of variables does not include infanticide or child care. The reason for this omission is that our analysis is focused on factors defined. Infanticide affecting fertility strictly and may serve as does, of course, affect family size and natural increase an alternative to factors It is therefore discussed affecting fertility. at a later point. briefly
(5)
For instance, Frank Lorimer, Culture and Human Fertility, 1954, Paris, to make clear the ways in which fertility can be affected, by failing of how it is affected. The reader gives in some ways a confused picture outline of direct and may wish to compare our framework with a half-page indirect factors affecting given by Raymond Pearl at the end fertility of an article on "Biological Factors in Fertility", Annals of the American Academy of Political and Social Science, Vol. 188, November 1936, p. 24.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
214
SOCIAL STRUCTURE AND FERTILITY
It should, of course, be mentioned that cultural influences affecting do not necessarily our eleven variables rational represent attempts to govern conditions fertility. stemming from socio-cultural Many fertility consequences in underdeveloped are by-products, regions) being unanticipated (especially know scientists and unrealized by members of the society. Surely by now social actions or treat nonthat they cannot confine their attention only to rational The requirements of rational actions as somehow defying systematic analysis. a given society can be met just as well, and just as ill, by an unintentional as by an intentional level of fertility one. Institutional
Patterns and the Intermediate A Preliminary Analysis
Variables:
From the standpoint of comparative sociology, an important question is in of our intermediate distribute how the fertility-values variables themselves is that underdeveldifferent kinds of societies. A preliminary generalization for numbers 1, 2, 8, and 9 on tend to have high fertility-values oped societies the list; they may have high values for 3a, 3b, and 10; and they often have low is hard values for 4 and 11. As for the remaining variables--5, 6, and 7--it and to prove that there are any consistent between pre-industrial differences If this generalization then it beindustrial is roughly accurate, societies. as follows: comes meaningful to re-group the eleven variables The Intermediate Variables According to Their Values in Pre-Industrial Societies Usually 1. 2. 8. 9.
High Values
Usually
Age of entry into unions. Permanent celibacy. Contraception. etc. Sterilization,
4. 10.
High or Low Values
In attempting to analyze in a preliminary the variables, terns affect we shall find just given. Age of Entry into
Voluntary abstinence. Foetal mortality--involuntary
Indeterminate
3a. Time between unstable unions. 3b. Post-widowhood celibacy. 11. Foetal mortality--voluntary.
Number 1.
Low Values
5. 6. 7.
abstinence. Involuntary Frequency of coitus. Involuntary sterility.
institutional way how different patit convenient to follow the order
Unions
In beginning with age of entry into unions, we are dealing with one of the variables It should be noted that these governing exposure to intercourse. in themselves, however favorable particular variables, they may be to fertility in practice and gesmay be counteracted by other factors governing conception For example, even though sexual unions begin early, pregnancy or tation. childbirth This is often the case when the sexual union is may be prevented. not a marriage. intereven though they permit premarital Many societies, forbid illegitimate With respect to marital course, strongly pregnancy.6 (6)
for which he had information, Murdock found that, Among the 250 societies relations are fully permitted in apart from incest taboos, "premarital 65 instances, and are conditionally approved in 43 and only mildly disIn other words, approved in 6, whereas they are forbidden in only 44.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
ECONOMIC DEVELOPENTAND CULTURAL CHANGE
215
indeed expected. is specifically As sanctioned, unions, however, reproduction non-marital there may be, in addition, unions in which realready mentioned, in dealing with age of entry also normally occurs. Consequently, production into unions, we shall separate those unions in which offspring normally appear both marital and non-marital (including types) from those in which reproduction We shall now deal with the condemned that it is infrequent. is so strongly first leaving until mostly to marriage itself), general class (paying attention of non-reproductive sexual unions. later the discussion the age of entry into reproductive unsocieties Since in pre-industrial ions is generally young, the question must be raised as to why the fertilityit is usually positive value of this variable when on certain other variables is often negative. From a broad functional the explanation stems standpoint, from high mortality. Not only does a high death rate normally prevail in from year to year, but there is always the danger of societies underdeveloped rise in mortality. a sudden catastrophic Early marriage therefore represents the maximum possible in population rehedge against the threat of failure Entering a union at a young age does not commit one irretrievably placement. to a large family, because all other means of reducing fertility come after If a particular this point. union is resulting in progeny that are too numerous under current circumstances, this eventuality can be obviated by abstior infanticide. These means, precisely because nence, contraception, abortion, can be utilized at a time closer to the actual impingement of they come later, new individuals on the resources of those responsible. If, on the other hand, the age of entry into unions is late, the potential that is lost can fertility never be recovered. The threat of mortality, from a societal has standpoint, not only to the potential reference but also to the parents themoffspring selves. Early formation of unions helps to guarantee that the young adults will achieve at least some reproduction before they die. This broad functional does not, however, enlighten us conexplanation institutional mechanisms by which early marriage is insured. cerning the specific These can best be understood in terms of family and kinship organization (inand rules of descent) and the control of property. volving rules of residence Such mechanisms apply most clearly to formal marriage, although they may apply as well, though in lesser unions. degree, to informal reproductive is From the standpoint of kinship organization, distinction an essential that between a joint household and/or clan system, on the one hand, and an indeon the other. When the clan is the unit pendent nuclear family organization the property (whether the latter the in herds or land), consists controlling does not normally arise, because the clan is immortal. question of inheritance When the joint family is the controlling unit, the question arises only when the joint family divides; the joint family, however, does not divide when the at the earliest, when the father dies. offspring marry, but rather, Thus, in societies in those having a strong clan having a joint household (and a fortiori on the possession of sepaorganization), marriage is in no way made contingent rate property by the newly married pair. Furthermore, riages are usually arrangements early
with strong clan or joint-household control (or both), marwho are often motivated to make the arranged by the elders, in the lifetime of the prospective mates, i. e., before
in 70 per cent of our cases. In the rest, license premarital prevails the taboo falls a preprimarily upon females and appears to be largely caution against childbearing out of wedlock rather than a moral requireOn ment." New York, 1949, p. 265. George P. Murdock, Social Structure, different but the majority of figures, p. 5 the author gives slightly his societies sexual relations. still permit premarital
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
216
SOCIAL STRUCIUREAND FERTILITY
and the economic Religious may require this result, prescription puberty. in such a way as to yield an exchanges involved in betrothal may be structured If the system is one advantage to the parents who marry their daughter early. of patrilocal for example, a grown daughter remaining in her parenresidence, tal home is an anomaly. Not only does her presence run counter to the normal of labor by sex, which assumes the complementarity division of husband and wife, but she must adjust to the wives of her brothers coming into the household. Add to this fact that the daughter, as a prospective spouse, is most in demand when she is young, first because she then has a greater poby other families tential ahead of her, and, second, because she is more attractive fertility and fits more easily into a subordinate status in her husband's parensexually tal home. If, then, there is a substantial or groomprice at marribrideprice kin stand a better chance of a favorable age, the girl's bargain if they marry her off early. This may help them in procuring wives for their sons. In societies the forces having neither a strong clan nor a joint family, The Irish family, for leading to early marriage may be overbalanced by others. has long been organized in terms of neolocal residence and hence mariinstance, tal rather than filial This being true, land had to be obtainable solidarity. or marriage postponed. During the greater part of the eighteenth century land was scarce and could not be subdivided because the economy was predominantly of an obstacle to early marriage "was the difficulty Consequently, pastoral. acquiring a settlement upon which a new family might depend."7 Later, during the sixty years before the Famine, when the potato became the staple food and the economy shifted from pastoralism to cultivation, couples could get property the main obat marriage by subdivision of the land, thus removing temporarily to early marriage. stacle But with the crisis of of the Famine, the futility led to the Land Purchase Acts stipulating subdivision that the loans progressive which transformed tenants into owners were granted only on condition that no would take place. subdivision Since the annuities ran for 35 years, this repreon subdivision.8 sented some restraint was the fact A more powerful restraint to subdivide in bethat, once the tenants became owners, they grew unwilling half of their sons. The tendency was to retain only one son on the paternal land, the remainder of the children being dispersed, partly Through migration The independent nuclear family was maintained, abroad. but the son who remained at home could not establish such a family until the father was willing to resign both authority and property. the average age at marriage in Ireland As a result became extremely advanced, reaching 29.1 for women by 1926.9 Lest our characterization of Irish family organization as neolocal appear it should be noted that although the Irish have been interpreted surprising, as having a joint household and patrilocal the opposite seems to residence,10 be true. Even if one or two sons remain at home, the resulting menage is not what is ordinarily called a joint household; because in Ireland marriage imWhen the son brings a bride into what was plies the independence of the son. (7)
K. H. Connell, The Population ours]. [underscoring
(8)
See Elizabeth Chapel Hill,
(9)
A. M. Carr-Saunders, World Population, Meenan, "Some Causes and Consequences Journal of the Statistical and Social session, 1932-33, pp. 19-27.
(10)
E. g., Conrad M. Arensberg and Solon Ireland, Cambridge, 1938, p. 80.
of Ireland,
1750-1845,
R. Hooker, Readjustments of Agricultural 1938, esp. pp. 55-57, 106, 151, 208.
Oxford,
1950,
p.
89
Tenure in Ireland,
Oxford, 1936, p. 91. Cf. James of the Low Irish Marriage Rate", 86th Inquiry Society of Ireland, T. Kimball,
Family and Community in
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
217
CHANGE AND CULTURAL ECONOMIC DEVELOPMENT
as the paternal homestead, he brings her into a home that has been redefined both ownership of the The father has relinquished his, no longer his father's. to own the As long as the father continues over the son. farm and authority land, the son who remains at home cannot marry because the land is necessary for the fact that the parents are If marriage occurs, therefore, the "match".11 have entered "the age grade of in the home is merely adventitious--they still if irreconcilable conflict the dying".12 develops in the shared Significantly, "The not the son and his wife, who must leave. it is the parents, household, bond between them [husband and wife] is stronger than that between son and parThus in Ireland the fact of sharing a house with the parents is not a ent."13 The of the joint family ideal but of the force of circumstances. reflection defined in such a way as to comply with the fact of a common menage is socially and independent nuclear family. ideal of a neolocal is neither unique to Ireland This independent nuclear family organization inIn Northwestern Europe the custom of impartible nor modern in development. was found in many areas or ultimogeniture) heritance (e. g., by promigeniture it was apparently customary for the In some sections during the Middle Ages. their old people to give their land to the heir before they died. Surrendering The heir's marriage was they expected only their keep off the land. authority, and brothers on the land being turned over to him; if his sisters contingent The of marriage.14 stayed on, they could claim their keep but not the privilege no marriage,15 of no holding, operated to advance the average age beprinciple Furthermore, the notion of the indeyond what it otherwise would have been. in the master-apprentice itself pendence of the nuclear family also manifested for marriage often did not occur until within the medieval guilds; relationship or dower.16 an adequate guild status had been acquired by inheritance, purchase, There is thus evidence that European society has long emphasized the marital with a consebond as the basis of family organization, rather than the filial quent tendency to delay marriage.17 rather on neolocal The emphasis on marital rather than filial solidarity, which appears to have delayed marriage in Ireland than patrilocal residence, and Northwestern Europe contrasts sharply with the forces operating to preciIn a truly joint household the marriage in an extended family system. pitate of the elders continues after marriage; the marital bond is therefore authority bond and does not require economic independence on to the filial subordinate Such a family pattern is well known as the the part of those getting married. and Kimball,
op. cit.,
pp. 107-122.
(11)
Arensberg
(12)
Ibid.,
p. 123.
(13)
Ibid.,
p. 128.
(14)
George C. Homans, English 1942, Chs. 9-10.
(15)
Josiah C. Russell, "Demographic Values in the Middle Ages", 1949, p. 104. George F. Mair, ed., Princeton, Population,
(16)
Josiah C. Russell, 163-164.
(17)
shows a retarded age with neolocal residence Of course, not every society where no formal In a primitive at marriage. economy with high mortality, to adult status must be hurdled, and where or other obstacles training of land is the felt need, indeof persons rather than scarcity scarcity be formed families nuclear early marriage, e. g., the Netby may pendent Ruthenians. Andaman Islanders, silik Eskimos, Fox Indians,
British
Villagers
Medieval
of the Thirteenth
Population,
Century,
Albuquerque,
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
Cambridge,
Studies
in
1948,
pp.
218
AND FERTILITY SOCIAL STRUCTURE
ideal one in traditional and many other peasant or China, India, Bantu Africa, In the Chinese case, the father maintains his tutelage cultures. primitive over the married son and his control over the familial property until death. He consequently need not fear the marriage of his son as a threat to his authand therefore, unlike the Irish father, has no motive (at least in this ority, such marriage. On the contrary, to the extent that his regard) for postponing son brings a wife into the house and has children, the old man's authority is extended. Indeed, it is only by the marriage of his son that the patriarch can fulfill his filial to his father.18 obligation Number 2.
Extent
of Permanent Celibacy
If late marriage can have a minus effect on fertility, so can permanent In both cases, if this effect is to be produced, there must be non-marriage. either continence outside of marriage, or the use of means to prevent intercourse from resulting in childbirth. In practice, does produce a non-marriage usually low rate of reproduction as mentioned already, among the unmarried, because, is the preferred institutional marriage in all societies arrangement for having It seems wise, therefore, children. to discuss in terms "celibacy" primarily of non-marriage, and to consider sexual continence only in so far as it illuminates that factor. a more potent factor than Although permanent non-marriage is obviously mere postponement of marriage, it actually occurs less frequently and hence has less negative influence on fertility. be found Only rarely can a population where more than 20 per cent of the women complete the reproductive period without ever having married. Ireland is an extreme case, with 26.3 of its women If we assume that these women, had they aged 45-49 in 1946 still single.19 as those who did, then married, would have had the same completed fertility their proportion an estimate of the loss of fertility due to nonrepresents Thus the loss due to permanent births).20 marriage (excluding illegitimate to exceed one-fourth. non-marriage seems, even in the extreme case, scarcely Such a loss in fertility is greatly exceeded by that due to late age at marriFor example, in Switzerland if (where the data are readily age. available), all women in 1941 who had ever married by ages 40-44, had married at ages 1519 and had subsequently manifested the same age-specific as those who fertility had actually married then or did marry at some point prior to age 40, the reIn other words, production would have been 75% greater than it actually was.21 (18)
Marion J. Levy, Jr., The Family Revolution in Modern China, Cambridge, When the family head dies there is the problem of one 1949, pp. 168-170. of the sons assuming authority over the others. It is precisely at this but if it survives this point that the joint household often dissolves; as it may, it does so because of the past institutionalization of crisis, relative age as a factor in authority.
(19)
Other cases Switzerland 13.3%.
(20)
in mortality Differences and possible unmarried women may introduce a small this estimate.
(21)
This calculation excludes non-marriage as a factor, because the women who had never married by age 40-44 were subtracted from the women under conin each age group. sideration In other words, 21.4% of Swiss women at But the remaining 78.6% had married at ages 40-44 had never married.
of high proportions never married are Sweden (1945) 20.9%, (1941) 20.1%, England and Wales (1931) 16.86, Belgium (1930) as between married and fecundity but probably not serious error into
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
CHANGE DEVELOPMENT AND CULTURAL ECONOMIC
219
would have been apif late marriage had been eliminated the gain in fertility three times the gain (25%) if permanent non-marriage had been elimiproximately nated. societies that the proportion of women It is mainly in urban-industrial never marrying by the end of the reproductive span exceeds 10%. In India in 1931 it was only 0.8%; in Ceylon in 1946, 3.4%; and in Malaya in 1947, 3.3%. areas generally show a very high plus value for fertility Thus the underdeveloped number 1 (age at marriage) and variable number 2 with respect both to variable societies often show rather low ever married), whereas industrial (proportion on these. fertility-values We thus have to answer two less use of non-marriage than of underdeveloped peoples make less Let us attempt trial societies?
make Why do all societies generally questions: late marriage in depressing fertility? Why do use of both of these mechanisms than do industo answer these two questions in order.
of the human species, can hope to reGiven the low fecundity no society unless either a majority of its women participate in reproduction place itself or its mortality is rigorously has ocSince most of man's history controlled. of heavy mortality--conditions for curred under conditions which still prevail viable societies mechanisms have evolved social many of the world's peoples--all in reproduction. Their particithat lead the majority of women to participate of marriage, which links sex and pation is organized through the institution to the care and socialization of children. This institution is in reproduction The mariturn supported by its articulation order. with the rest of the social tal relation thus becomes a general norm in terms of which the hopes and expectations of virtually all individuals are channelized. If for some reason the is relaxed, the norm still in effect. Not only continues pressure of mortality do normative systems change slowly, but there still for remains the necessity in terms of which reproduction a family organization and child-rearing are provided for. Thus individuals continue to anticipate marriage as a normal and an event more easily postponed than foregone altogether. important part of life, In any case, an increase in non-marriage would not reduce fertility unless either coitus outside of wedlock were successfully banned or contraception and If the latter were readily abortion were freely used. they could be available, used within marriage, and the consequent reduction in marital fertility would of denying marriage to a substantial obviate the necessity portion of the popuIf contraception and abortion were not readily lation. available, non-marriage brake on fertility would be an effective only at the price of permanent sexual that this price is indicates celibacy. Everything we know about human society so high that no population to pay it. is willing Since no society has ever attempted to incorporate as permanent celibacy a widespread custom, we have no conclusive evidence as to what it would do to a social We can, however, obtain some clues by examining countries in system. has appeared to an unusual extent and by examining which permanent celibacy We can also say something on which have enjoined it as a rule. organizations theoretical as the chief purely grounds concerning what it might do if utilized means of reducing fertility to a modern level. of space prevent Limitations various ages. If this 78.6% had all married at ages 15-19 and had from the same age-specific that age experienced as those ever marfertility ried at each age, their total fertility would have been 76% greater. Stated in terms of the potential lost by late marriage, the fertility is rough, because the data 64%. The calculation figure is approximately refer to 1941 and thus do not represent a true cohort analysis; but a refined calculation on a cohort basis should yield rather similar results.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
SOCIAL STRUCTURE AND FERTILITY
220 our giving a complete about each of them.
treatment
along
these
lines,
but something
can be said
late age at marriage and a high proporBecause Ireland has an unusually tion who never marry, together with a strong prejudice against coitus outside of it provides the main example of a rather extensive of marriage, practice Has this adjustment exacted a price? To answer such a question celibacy.22 as a consetoward sex cannot be listed attitude is difficult. A puritanical itself. That the Irish avoid requence, because this is part of the celibacy rate--2.8 outside of marriage is shown by their low illegitimacy production of all live births in 1921-1930 and 3.3% in 1931-1940.23 However, such deas we have suggest that a great amount of attention, community efscriptions sexual expression. and personality conflict Having go into controlling fort, the a social system that emphasized the marital bond and the nuclear family, unmarried females, as is done in Moslem counIrish cannot completely segregate in courtship The young people must have some chance to participate and tries. mate selection. But, given this system, the Irish seem to make an unusually under a to control sexual behavior. For a country not living strong effort of literature and ideas is exceptionally the official censorship dictatorship, to and has as its main purpose the suppression of material pertaining rigid, sex and reproduction.24 the data on mental illness, which show Furthermore, indicate a possible a high rate for Ireland, consequence of such repression.25 (22)
to Malthus, New York, 1953, pp. 27-54, shrewdly David Glass, Introduction notes that Ireland is the only country which has come close to following In restraint" and no birth control. Malthus' rules of conduct--"moral of Northwestern Europe, such as Sweden and Norway, a late other countries not only because illeage at marriage does not imply sexual abstinence, but also because contraception is more freely gitimacy is more tolerated practiced.
(23)
Ibid.,
(24)
For attitudes toward sexual behavior see Arensberg and Kimball, op. cit., and popular sources as Frank O'Connor, Ch. 11; and also such literary "Love Holiday, Vol. 6, December 1949, p. 40; Sean O'Faolain, "Ireland", Life Magazine, Vol. 34, March 16, 1953, pp. 140-157. Among the Irish", the following Regarding censorship, passage from O'Faolain is pertinent: of books and publications, "...Our censorship instigated by the clergy is a symbol of this fear of and submitted to, willy-nilly, by everybody, of books and the 150 close-packed sex...In register pages of the official banned by the Irish Censorship Board we find the names of periodicals Irish writer of note, some for one book, some for almost every single The banning is done in secret. There is no appeal to the courts several. of law..." See also an article, "Irish Challenge Censors' Methods", The New York Times, August 14, 1955, where it is pointed out that the Irish Censorship Board "has banned books by the most reputable Irish authors, and Ireland's Sean O'Casey, Liam O'Flaherty, Sean O'Faolain, including most brilliant short story writer, Frank O'Connor. Nobel prize winners have even come under the interdict...many works of worth are condemned on a few isolated marked passages, while the general tenor of the book is the works of Roman Catholic authors approved by the church ignored...Even in Britain have not escaped the five Irish Roman Catholic authorities Censors."
(25)
In 1949 the proportion beds devoted to mental cases was 57% of hospital in Ireland, The rate of whereas it was only 49 in the United States. mental patients in 1948 was 603 in Ireland as conper 100,000 population seem not trasted to 382 in the United States. Adventitious circumstances
p. 37.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
ECONOMIC DEVELOPMENT AND CULTURAL CHANGE There appear lost through any nation in that pothesis
to be few features in Irish Ireland has, for celibacy. Northwestern Europe. All Ireland is paying a price
221
life that compensate for whatever is of example, the lowest level of living told, there is some ground for the hyfor its unusual degree of celibacy.
rule has been almost solely Celibacy as an organizational applied to rewhich have adopted such a rule for Among those few religions ligious personnel. available for the Roman Catholic their clergy, our evidence is most readily of the rule in this case encounAs is well known, the application priesthood. It required nearly nine centuries before the edict of tered great difficulties. Priests were first could be enforced with relative success. non-marriage itself in 385 A. D. After commanded to separate from their wives and remain continent that date there were periods when the ban against marriage could be safely igfollowed by periods when the Church was militantly nored by priests, purging its in enformarried clergy. encountered such obstacles Pope Gregory (Hildebrand) to withdraw their obecing the rule of non-marriage that he ordered the laity dience from all members of the clergy who disregarded the papal canons on simony and incontinence. of the Church-By so doing, he undermined a basic principle laid one of the foundaclerical immunity--and thus as early as 1074 directly tions of the Reformation. Only by placing the sacrament of marriage in a lower than that of the religious vow (Lateran Council of 1123) did the Church position the issue of clerical settle such marrifinally although in practice marriage, late as the nineteenth ages occurred with some frequency after that--as century in some parts of Latin America, for example. In periods when the ban against the Church still had to deal with sexual incontimarriage was being enforced, nence among its priests of female peni"Solicitation" and nuns. (the seduction and other violations were so common as to cause chronic tents), concubinage, In some areas priestly a concubinage became, for long periods, public scandal. and the sons of priests received preferment.26 We can thus customary practice, of the popusee that the enforcement of celibacy even for that small fraction lation represented by the clergy was anything but easy.
to account for this result. of Though Ireland has a larger percentage (24.7% at ages persons in the advanced ages than does the United States 50 and over as against 22.4% at these ages in the United States), she has a higher proportion under age 30. The fact that Irish medical services are less developed than in this country suggests that the comparison the difference understates in mental illness. In 1949 Ireland had only one hospital bed per 1,000 inhabitants, whereas the United States has of mental cases in Ireland may never 9.6, so that a higher proportion appear in the statistics. (26)
of clerical For the history in Europe, see Henry C. Lea, Hiscelibacy tory of Sacerdotal Church, London, 1932, and Celibacy in the Christian A History of the Inquisition of the Middle Ages, Vol. 1, New York, 1888, Alexander C. Flick, The Decline of the Medieval Church, New pp. 31-32; J. R. Tanner et al. (eds.), Contest of York, 1930, Vols. 1-2, passim.; New York, 1926, Empire and Papacy, Vol. 5 of Cambridge Medieval History, esp. pp. 11-14, 40, 61-62, 73, 695; Eileen Power, Medieval English Nunneries, Cambridge, 1922, Ch. 11; Geoffrey Baskerville, English Monks and the Suppression of the Monasteries, New Haven, 1937, pp. 261-266; Joseph St. Louis, 1944, McSorley, An Outline History of the Church by Centuries, Decrees of the pp. 83, 154, 206-207, 237; H. J. Schroeder, Disciplinary General Councils, For Latin America, see J. St. Louis, 1937, p. 193. Lloyd Mecham, Church and State in Latin America, Chapel Hill, 1934, p. 48; Mary Watters, A History of the Church in Venezuela, 1810-1930, Chapel New 1933, p. 211; Gilberto Freyre, The Masters and the Slaves, Hill, York, 1946, pp. 446-452.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
AND FERTILITY SOCIAL STRUCTURE
222
is institutionalized and becomes in which celibacy If we imagine a society and a norm rivaling marriage, we can see that the result would be paradoxical Should the celibate class be large enough to reduce the birth rate impossible. to a modern level without other means, it would have to contain at least half on such a scale to be induced to make the the population. For individuals of celibacy, sacrifice (perthey would not only have to be firmly controlled from the rest of the community and thus divorced from the temphaps segregated of everyday life), but would also have to be ideologically tations indoctrinated, If the rewards were great enough to recruit rewarded. socially and, above all, this class would portion of the population, people for the numerous celibate But the celibate class would ladder. inevitably occupy the top of the social would not be too big to be an elite. Furthermore, the sheer fact of celibacy of the society. in itself a contribution to the productive capacity represent of If the celibate were given useful tasks to perform, the variety population an indiscrimibe great; and if all these received would necessarily functions this return not because would be receiving nately high reward, some celibates but because of their celibacy. In this way, contribution of their productive advantages that at best only a seeking to give half or more of its population of productive the society would few can be given (and doing so regardless merit), economic and social suffer an intolerable burden.27 in minor role of permanent celibacy of the relatively After this analysis we are now ready for our second question: Why are late fertility limitation, than in pre-industrial marriage and non-marriage more frequent in industrial societies? because societies Perhaps non-marriage occurs more often in industrial orthese societies depend less upon kinship and the family as bases of social less the individual's The fact of being or not being married affects ganization. In pre-industrial where the family is a produceconomic chances. societies, tive unit, marriage has a high value for the individual. Also, where the,partas in ners to marriage are self-selected by a competitive process of courtship, who are not sucthere tends to be a substantial modern countries, proportion a suitable in attracting mate. cessful nations can be The greater postponement of marriage in urban-industrial in for skilled The necessity of lengthy training positions similarly explained. an industrial economy, the often lengthy trial-and-error process of courtship, on the part of the newly married the necessity of economic self-sufficiency are conducive to marital postponement. couple--all is non-marriage likely to be as important But in neither type of society of fertility as late marriage, because marriage remains the instia depressant but norm in both cases. Wedlock may be postponed with some equanimity, tutional who actually individuals never marry have, in most cases, hoped that this would is certainly In Ireland, for example, clerical not be their fate. celibacy among laymen.28 valued, but not permanent celibacy ration
Once again let us note of marriage necessarily
the postponement nor the total abjuthat neither Hence no industrial implies sexual celibacy.
(27)
Of course, a society could be imagined in which half or more of the women in polyandrous were forced to be celibate, the rest of the people living But such a speculation would evoke more paradoxes than that marriage. A society could organization already sketched. capable of such deliberate be expected to use celibacy alone as its means of controlling scarcely With other less drastic means available, the end would hardly fertility. the means. justify
(28)
Arensberg
and Kimball,
op. cit.,
p. 69.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
ECONOMIC DEVELOPMENT AND CULTURAL CHANGE
223
today is required to use either method as a dominant means of controlsociety methods are less sacrificial, because other less drastic, ling fertility, and abstinence It is clear that marital postponement, available. non-marriage, in limiting all have a common if they are effective within marriage, fertility, that this entails. and all share the difficulties feature--sexual denial; Number 8.
Use or Non-Use of Contraception
on fertility have a negative effect Whereas the "intercourse variables" revariables neither the conception nor the gestation only through abstinence, necesor the institutionalization behavior by the individual quire this drastic variables" With the "conception (of which the use sary to insure such behavior. is not foreof intercourse or non-use of contraception is one), the pleasure from paying a heavy appetitive thus released The individual, gone. penalty for is much freer to decide this issue in terms not to have children, the decision of his economic and social interests alone. its apparent efficiency With reference to contraception in particular, of fertility. might lead one to expect a widespread use of it as a depressant Yet we have already stated that this is one of the three variables which almost in pre-industrial have a strong plus fertility-value societies. universally Why, so widely exhibit the non-use of contraception? To then, do these societies the two types of contraception. answer this question, we must consider separately In many primitive and 8a. by chemical or mechanical means. Contraception the idea of chemical and mechanical contraception is known and peasant cultures the indiviattempts are made to apply it. motivating Yet, even in situations dual to limit his fertility, this is not usually the means adopted, simply because the technology of underdeveloped methods. societies cannot supply effective In the absence of a knowledge of reproductive physiology, people in these soto look for. cieties sense of even the kind of instrumentalities have little there is not enough knowledge of chemistry to give command over maSimilarly, The methods, therefore, tend to be hit or miss, with magic rather than terials. science playing a prominant role. Lack of experimental technique leads one method to be valued as highly as another.29 Even the methods that would actually are apt to be clumsy, sexually unsatisaccomplish the purpose of contraception and unhealthful, e. g., insertion of an okra-like seed pod in the vafactory, of rags or finely gina (Bush Negroes of British Guiana); insertion chopped insertion of dung (Egypt and grass (Bapindas and Bambundas in Central Africa); other societies).30 method is Furthermore, granted that a really satisfactory hit upon, such as possibly the use of a douche containing lemon juice or a dethe materials coction of the husks of mahogany nut (Martinique or Guiana),31 to be available only in one locale are likely or in certain seasons of the year. Thus the technology and economy of pre-industrial societies have not been equal to the task of providing a chemico-mechanical that would be at contraceptive once cheap, satisfactory, and readily available. effective, 8b. without chemical or mechanical Contraception methods as withdrawal, intercourse without penetration,
means. Clearly such and various heterosexual
(29)
Norman E. Himes, Medical History of Contraception, Baltimore, 1936, pp. 53-54, 99. See also Clellan S. Ford, A Comparative Study of Human ReproNew Haven, 1945, pp. 40-42. duction,
(30)
Himes,
(31)
Also see M. Soors, Ibid., p. 17. Vol. 4, May 1950, pp. 525-532.
op. cit.,
pp. 10,
18-19,
63. "La denatalite
chez les
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
Mongo", Zaire,
224
AND FERTILITY SOCIAL STRUCTURE
and technological do not depend on scientific They progress. "perversions" Yet in one form or another in nearly all societies.32 are known and practiced a major control over feremployed to represent they seem to be insufficiently but apparently societies, They may be so employed in a few primitive tility. such as that of China, India, and the Near East where not in the civilizations For the most part, it seems, they are are found. aggregates huge population interor in those cases where premarital relations employed in extra-marital But it is doubtful that course is permitted but premarital pregnancy forbidden. to fertility control in whole an important contribution such practices represent with a good share of the world's people-Numerous societies--some societies. or either do not permit the ordinary female to engage in premarital intercourse, would play a small role have such a young age at marriage that such intercourse those societies which permit them in any case. As for extra-marital relations, concerned about the woman's are not particularly under certain circumstances is not stressed. because biological Only those paternity becoming pregnant, would condemn the maras illegitimate children societies branding adulterous ried woman's pregnancy by another man than the husband, and these would be socito have an For these reasons, intercourse. extra-marital eties which restrict non-mechanical on fertility, effect and significant contraceptive independent forced to ask We are therefore methods would have to be used within marriage. sowhy such methods are not more widely used within wedlock in pre-industrial cieties. must in with a high mortality that any society The reader should recall Under favorably. reproduction general motivate its members to view legitimate in question, as already pointed out, are so organized the cultures this pressure values in the early stages of the reproductive as to maximize fertility process it is still is one step later, --e. g., by early marriage. Although intercourse If conditions so early as to involve a risk of inadequate fertility. subsequently be taken after conception. measures can still make children undesirable, is consideration An additional and the responsibility childbearing, (32)
burden and danger of that the physical fall and rearing the child, for nourishing
the is doubtless Himes, speaking of Europe, says that "coitus interruptus has been for method of contraception...and most popular, widely diffused of man." is probably nearly as old as the group life centuries...[It] in which He also cites numerous primitive tribes 0p. cit., pp. 183-184. is practiced. I. Schapera, writing of the Kgatla of coitus interruptus locally pracsays: "The commonest method of contraception Bechuanaland, is widely employed not only by married ticed is coitus interruptus...It Sometimes the woman, by moving but also by unmarried lovers." people, her hips so as to extrude the penis just before ejaculation, accomplishes Married Life in an without the male's cooperation. coitus interruptus Coitus inter femora is pracAfrican Tribe, New York, 1941, pp. 222-223. Girls may ticed in many societies, by the Bantus in Africa. particularly C. Daryll Forde, wear special designed to avoid penetration. girdles London, Nigeria, Marriage and the Family among the Yako of South-Eastern but not pregnancy sexual relations Bantu tribes, 1941, p. 14. permitting before marriage, teach (or did teach) their young people how to have the unbroken hymen in some tribes being intercourse without penetration, on at marriage. insisted regarded as an important index of virginity, to climax" to have been pracAlfred C. Kinsey et al. found "petting the U. S. male ticed by 24% of the male sample (blown up to represent The cumulamales. by age 21, and by 50% of college-educated population) tive incidence being 24% substantial, among females was less but still at age 20. Sexual Behavior in the Human Male, for the college-educated and ...in the Human Female; 1953, p. 270. 1948, pp. 531-542, Philadelphia,
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
ECONOMIC DEVELOPMENT AND CULTURAL CHANGE mainly on the mother. wish is apt to be hers non-chemico-mechanical operation and partial his sures that affect avoiding pregnancy.
225
If therefore there is a wish to avoid childbirth, this rather than her husband's. It happens, however, that the methods of contraception are the ones requiring the coof the male. Since he is not under the presfrustration wife in this matter, he may be reluctant to aid her in
of the two sexes is often carried so far that comThe social insulation This insulation munication between them is difficult. is particularly observable in regard to sexual behavior, which tends to be surrounded by taboos and rituAs between husband and wife, sexual intercourse, als. by virtue of being the the focus of anxiety and conflict bond and therefore between them, may special be the topic they discuss with least freedom. Thus the cooperation necessary for contraception is made difficult. In such terms we can understand why the available methods of contraception receive scant use in underdeveloped societies. Which of the considerations mentioned plays the greatest role is hard to say, but the fact should be emphasized that not all the reasons for limiting births are predictable at the time of intercourse--particularly in simple societies that live close to the environment and are threatened The individual by quick catastrophe. couple may, as we shall see later, limit fertility after rather than at the time therefore, of intercourse. Number 9.
Voluntary
Control
over Fecundity
Like chemical and mechanical contraception, control of fesatisfactory of pre-industrial societies. Neither cundity is beyond the technical capacity the reduction nor the enhancement of fecundity by harmless medical measures in such cultures. appears possible genitalia Operations on the male external can be performed, such as subincision and castration, but these are either too drastic to be harmless or have little effect on fecundity.33 We may conclude, societies are plus on variable then, that pre-industrial number 9. But so are industrial The latter societies. may have even more of a plus fertility-value on this variable than simple societies because they can, and usually do, forbid sterilization and, at the same time, foster medical treatment for sterility, thus enhancing the fecundity of partially sterile couples. Although yet been used,
modern science makes harmless sterilization it has not possible, except in Puerto Rico, as a popular method of avoiding children.34
(33)
Castration is so drastic that it is apparently never used with enough the splitting of the frequency to affect Subincision, group fertility. from the lower part rather penis in such a way that the semen is expelled than through the glans, seemingly has little effect on fecundity, depenassumed during intercourse. Also the pracding in part on the position tice has a very limited distribution even in primitive and seems society unknown in more advanced pre-industrial societies. Among the Australian where it is found, opinion differs as to its effects. German aborigines, held that the operation theorists, according to Himes, have generally lowers fertility and is so intended. Modern anthropologists, on the other Himes himself believes it may hand, have denied both these contentions. have some negative effect of this kind. Op. cit., pp. 41-51.
(34)
See J. M. Stycos, "Female Sterilization Vol. 1, June 1954, pp. 3-9. terly,
in Puerto
Rico",
Eugenics
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
Quar-
AND FERTILITY SOCIAL STRUCTURE
226
The Puerto Rican case suggests, however, that sterilization may in the future in underdeveloped If the operational techareas. become more widely diffused that it reversed--so nique were improved to the point where it could be easily the total number, of as well as for limiting could be used for the spacing, in backward means of reducing fertility children--it might become the principal areas. Number 3a.
Time Between Unstable
Unions.
of both 3a is a function on fertility from variable effect Any negative of unions and the time lost between them. If unions are the rate of dissolution will or if they are unstable but no time is lost between them, fertility stable, not be affected adversely. seem generally societies to marital unions, pre-industrial With reference to this to have a low rate of dissolution. True, there are certain exceptions Some of the Islamic peoples show a tendency toward marital instability, rule. the clan or joint household takes such precedence societies and in some primitive On the tends to be somewhat unstable.3 over the nuclear family that the latter of pre-industrial structure groups buttresses whole, however, the institutional stability. marriage in such ways as to give it considerable of informal unions which it has a significant When a society proportion is nevertheless to legal marriage but in which reproduction regards as inferior unions" in Latin America and "common law" unions in expected (e. g., "consensual of such unions is that they tend one of the features the British West Indies), a In such cases the woman may wait some time before entering to be unstable. For a small sample of lost may be substantial. new union, and the fertility the reduction women in Jamaica (where around 70% of the births are illegitimate) 37%.36 The indue to the instability of unions was approximately in fertility form from various historical as an institutional formal type of union arises that have been disorganized In societies causes. they may by Western contact, In other and legal marriage itself may become unstable.37 appear abundantly, out of a former slave class, order has grown largely where the social instances
(35)
See Ralph Linton, Study of Man, New York, 1936, Ch. 10. Murdock, op. cit., Linton for holding that in some societies organized on p. 3, criticizes a "consanguine" basis the nuclear family plays an insignificant role, but dismarital instability the fact is that in such cultures may have little See K. Davis, "Children of Divorced Parents", Law and effect. organizing Contemporary Problems, Vol. 10, Summer 1944, pp. 700-710.
(36)
and Reproductive Behavior in Jamaica", Judith Blake, "Family Instability Milbank Memorial Fund, New York, Current Research in Human Fertility, 1955, pp. 26-30.
(37)
Margaret Mead, Changing Culture of an Indian Tribe, New York, 1932, pp. Ch. 10; Migrant Labour and Tribal 14-15, Ch. 10. Schapera, op. cit., Life, London, 1947, pp. 183-189; and "Cultural Changes in Family Life", The Bantu-Speaking Tribes of South Africa, London, 1937, pp. 380-385. The literature covering the impact of Western culture on native peoples is the tendency of such conso enormous that one could document indefinitely in such unions and tact to produce illicit sexual unions and instability in marriage.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
ECONOMIC DEVELOPMENT AND CULTURAL CHANGE informal ages.38
unions
may be both more numerous and more unstable
than legal
227 marri-
to premarital With reference unions, there is every evidence that in the where these are permitted they are, as a rule, highly unstable, many societies amounting in many cases to adolescent promiscuity. However, there is ordinarily time lost between such liaisons; few societies in little permit reproduction them; and, given a young age at marriage, most such unions occur at an age when adolescent seemingly reduces the number of conceptions. sterility It follows that pre-industrial societies have a plus fertilitygenerally number 3a, buf the exceptions are more numerous value with respect to variable than was the case with the other variables so far considered. Number 3b.
Post-Widowhood
Celibacy
the high rate of widowhood found in pre-industrial What effect societies has on fertility of the widow. In many depends on the institutional position time from exposure to intercourse, she loses little such societies because she In other pre-industrial soon marries again. cultures, however, the widow either must wait for a protracted period or is subject to a distinct against prejudice An important problem in analyzing the institutional remarrying at all. impingeis the discovery of why some societies ments on fertility take one course in this regard and others take the opposite course. If we study those societies in which remarriage occurs universally and the widow to marry a kinsman of soon, we find that they are the ones requiring the deceased husband (levirate). Such societies are usually primitive, praca shifting or pastoral and are characterticing cultivation, hunting, pursuits, ized by strong clan or lineage substantial organization. Marriage involves economic exchanges and, if the system is patrilineal and patrilocal, these are The woman brought into weighted in favor of the bride's lineage (brideprice). as a wife is conceived as belonging the clan or lineage to this clan, which has her children, who are automatically members of the husband's paid the brideprice; her contribution in return for the cost of procuring her. lineage, represent When the woman is widowed, the lineage retains control over her, not only because a price has been paid for her but also because her children must remain with the If she still is fecund, the lineage feels it would be losing potential lineage. if she did not remarry. But remarriage to an outsider would be unsatchildren of that union would belong to another lineage. because the children isfactory, Hence the remarriage must be within the clan. Since in the exchanges cementing the husband's nearest relatives bore the main cost, it is the first marriage, natural that his close kin (notably his brothers) should have first claim on the If the deceased husband has no actual brothers, widow. one of his "classifiIn anticipation of her possibly catory brothers" can be substituted. entering a leviratic with her husband's actual and classifiunion, a woman's relation The term for "husband's catory brothers is often one of privileged familiarity. brother" may be the same as that for "husband". The social structure clearly that the clan is thinking of the widow in terms of her potential demonstrates of children. even if the widow should Among the Nuer, for instance, production take as a lover a person outside the clan (she cannot legally marry outside), (38)
T. S. Simey, Welfare and Planning in the West Indies, Oxford, 1946, passim. F. M. Henriques, G. W. Family and Colour in Jamaica, London, 1953, passim. in the West Indies", PopuRoberts, "Some Aspects of Mating and Fertility lation Studies, Vol. 8, March 1955, pp. 199-227. R. T. Smith, "Family in British Vol. I, FebOrganization Guiana", Social and Economic Studies, ruary 1953, pp. 87-111.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
228
SOCIAL STRUCTURE AND FERTILITY
of the dead husband and therefore the children are viewed as the descendants as members of his, not the lover's, clan.39 on the other hand, the widow is forbidden to marry a In many societies, close relative of the deceased husband. These seem to be cases in which the clan, however important it may once have been, has receded in economic and poadvance and greater litical seemingly as a result of technological significance, The economy is that of a more stable agriculture in class stratification. Under such cultivated which the same land is intensively year in and year out. the joint household acquires more independence and more significircumstances cance as an economic unit than it seems to have in most primitive societies. in different households thus takes precedence The distinction between relatives as members of the same lineage or clan. To be sure, the over their solidarity woman marrying into the joint household may do so in terms of some form of ecohouseholds rather than nomic exchange, but this exchange is between individual The widow and her children accordingly clans. belong to the deceased husband's household. Remarriage to one of her dead mate's brothers or other close male because the joint however, would be structurally relatives, inappropriate, and must be so organized as to minihousehold is always subject to dissolution of suuh dissolution. Unlike the clan or lineage, which mize the complications the household is a residential economic' is immortal and indefinitely expandable, unit which can easily With stable grow too large for its immediate resources. the household must be near the land it works. If its memberships agriculture, it must ultimately break up because the land required for sustenance increases, will be too distant. When the household does break up, usually at the death of the male head, it does so by the separation of its nuclear families.40 Aceven when the nuclear family forms part of a joint household, it is cordingly, not only as a separate unit but also as one that may in the future visualized A widow's remarriage to one of her husband's have its own independent residence. indewithin the household would conflict with this idea of potential relatives It would inextricably It would require merge two nuclear families. pendence. of the sibling rather polygyny and would emphasize the solidarity relationship than the father-son relation so central to the independent joint household. Stable agrarian societies not only forbid the widow to marry within the of her husband's kin but also often frown on her marrying anyone at all. circle This additional seems likewise to be explicable in structural terms. prejudice For the widow to marry outside would require that some agency make a match for are arranged by persons her, because marriages in traditional agrarian societies to the union. other than the parties is no However, her family of orientation for her. The family of her deceased husband is restrained longer responsible from taking the responsibility for several reasons. It would, in seeking a mate for the widow, have to treat her as a daughter, which might interfere with the rights of the actual daughters. Furthermore, since she is a widow and is older, she has become less valuable than upon her first marriage, so that it is difficult to get her married at a social level reflecting on the family's favorable If the widow has children, her marriage outside the immediate kin prestige. from them. It is thus understandable would require her separation why traditional agrarian societies, where the joint household is normally especially a prejudice should exhibit Such unions against widow remarriage. preferred, do occur, particularly in the lower classes which cannot carry out the certainly but the prejudice joint family ideal, may be strong enough to prevent a high
(39)
E. E. Evans-Pritchard, pp. 112-123.
(40)
For mention of the joint household's vulnerability sequent fissive tendency, see Murdock, op. cit.,
Kinship
and Marriage
among the Nuer, Oxford, to change and its p. 36.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
1951, con-
DEVELOPENTAND CULTURAL ECONOMIC CHANGE
229
In India the caste from remarrying.41 of widows in the upper classes proportion widow remarriage. those of the joint household in preventing reinforce controls and since marriSince such unions are thought to lower the caste's prestige, to a remarriage are condemned. age is endogamous within the caste, both parties due to widow agamy is probably greater For this reason the reduction of fertility because of the early age at marin India than in any other country, especially there. riage and the high mortality Number 11.
Voluntary
Control
over Foetal
Mortality
foetal mortality, have few means for lessening Underdeveloped societies such for increasing available but they do have readily means, through abortion, in pre-industrial In fact, abortion is widely practiced societies, mortality. Since medical means of limiting fertility.42 being the individual's principal do not, at least as yet, have as much influmeasures to avoid foetal mortality as voluntary abortion can and does, we can say that whether ence on fertility 11 dewith respect to variable has a plus or minus fertility-value a society some abortion. on the extent to which it practices Accordingly, pends primarily abortion and pracare on the "plus" side (forbidding societies pre-industrial abortion but many others are on the "minus" side (practicing it little) ticing is with conception to a considerable If we grant that interference extent). with pregnancy, an important questhan interference le3 s hazardous to health used in underdeveloped tion for us is why abortion is so much more frequently than contraception. sc ieties one can point to the following In answering this question, (a) as compared to mechanical and chemical means of contraception,
considerations: abortion is
(41)
Levy, op. cit., p. 46, points out that although the Chinese gentry have the peasants have usually practiced always frowned on widow remarriage, if a peasant widow was young and lacked grown sons, remarIn fact, it. As the peasantry is said to comprise as much as riage was inevitable. of is hardly characteristic 80% of the population (p. 44), widow celibacy set the ideals for the entire China as a whole, although gentry patterns between gentry and peasantry, society. Olga Lang, without distinguishing New says that remarriage is frowned on. Chinese Family and Society, She says (p. 126) that poor men often marry widows Haven, 1946, p. 53. because they are easier to get than virgins. Any divorcee or widow can With reto marry beneath her status. find a husband if she is willing in China, it is integard to the absence of anything like the levirate to note that Miss Lang says (p. 21) that "early in the feudal resting period, under the Chou dynasty (ca. 1027-256 B. C.), the clan began to divide into economic families." Today, even in the South where clans in family matters. are of some importance, they have no real authority of clan The strongest clans in Central and North China lack the essential a fair amount of common property (pp. 177-178). life,
(42)
Ford, op. cit., pp. 50-51, found that most of his tribes took cognizance and stated to be forbidden, In eleven it was specifically of abortion. in 21 it was permitted to to be forbidden; in eight it could be inferred and in 4 this could be inthe young girl who finds herself pregnant, ferred to be the case; and in 12 a married woman was allowed to practice that she had become pregnant through an adulterabortion if she believed societies Himes regards abortion as widespread in primitive ous intrigue. A recent study by George Devereaux, Abortion in Pri(op. cit., p. 52). New York, 1955, pp. 25-26, cites cases of tribes where mitive Society, abortion is quite frequent.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
SOCIAL STRJCTUREAND FERTILITY
230
to such non-chemico-mechanical methods (b) In contrast technically simple;43 or coitus inter femora, abortion is not applied at the as coitus interruptus and does not require cooperation time of intercourse between man and woman. It is a woman's method and can be practiced without the man's knowledge. (c) Unit is completely like contraception, effective. (d) Once an undesired pregnancy has occurred, the need for abortion is certain, whereas at the time of intercourse there is always the chance that pregnancy will not eventuate anyway. (e) Although a child may be desired at the time of intercourse, subsequent events is a at which time abortion rather than contraception may alter this attitude, remedy. is not dealt with as an inteA note on infanticide. Although infanticide one should note because it does not affect fertility, gral part of our analysis of abortion in controlling a functional that it is virtually family equivalent and that it too is practiced much widely in pre-industrial size, societies, The rationale more so than contraception. for its use is much the same as that in at least three respects. for abortion, but it does differ from the latter infanticide allows the progeny to be selected First, by sex, as shown by the The logic of this practice custom of female infanticide. is exemplified by the Netsilik Eskimos: The most glaring consequence of the struggle for existence is manifested in the way in which they try to breed the greatest number of boys possible and the fewest possible if they have are killed at birth, girls...girls not already been promised to a family where there is a son who some day is to have a wife...They hold the view that if a woman is to suckle a girl child it will be two or three years before she may expect her next conhunter must take into consideration that he can only subject finement...A himself and his constitution for comparatively few years to all the strain that hunting demands...Now if he has sons, they will as a rule be able to Thus it step in and help just when his own physique is beginning to fail. is life's own inexorability that has taught them the necessity of having as many sons as possible. Only by that means may they be certain that they will not need to put the rope around their own neck too early; for it is the common custom that old people, who can no longer keep themselves, prefer to put an end to their life by hanging...44 the persistence of the immemorial custom of female infanOlga Lang discusses ticide in China. The hospital records used for her study "contained matter-offact references to infanticide made by Chinese social and medical workers that it was taken for granted. Much more often, however, infant indicating What happens is that the small amount daughters have not been killed outright. of food available for the family is unequally distributed: the son gets the Hence the frequent starved. larger share and the daughters are practically than of boys."45 Much the same epidemics have taken a heavier toll of girls could be said of India. also allows Second, infanticide status, weeding out thosewith physical
the offspring to be selected according or unacceptable deformities, badhealth,
to
(43)
Premature labor can be induced by killing the foetus. This can be done or massaging the abdomen; by drinking poisons or by beating, pressing, the foetus or amniotic sac with strong emetics or laxatives; by piercing or by wearing a tight belt. See Ford, o. sharp reeds or instruments; p. 52; Devereaux, op. cit., cit., pp. 27-42.
(44)
Knud Rasmussen,
(45)
Lang,
op. cit.,
The Netsilik
Eskimos,
Copenhagen,
1931, pp. 139-140.
p. 150.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
ECONOMIC AND CULTURAL CHANGE DEVELOPMENT
231
when the ciror racial characteristics.46 Third, it can be practiced physical to be abnormal and ritualistically cumstances of birth are considered taboo. or with teeth, infants whose mothers died Twins, children born with feet first and offspring born on unlucky days are typical at their birth,47 victims.48 obFourth, whereas abortion may injure the health of the mother, infanticide does not. viously of infanticide A disadvantage may seem to be that since a child has already been born, a living person is being killed. However, the newborn child is until he has passed through some sort of often not viewed as a member of society of the child to the father ceremony (amphidromia in ancient Greece, presentation The destruction of the child is therefore in China) which defines him as such. in much the same light as abortion. viewed psychologically Number 4.
Voluntary
Abstinence
within
Unions
much more, on the average, in preAbstinence within unions is practiced on ferthan in industrial The effect of such abstinence industrial societies. for there are at least four types however, depends on the circumstances; tility, The first and menstrual. of restriction--post-partum, occasional, gestational, two types tend to limit fertility, while the last two, if they have any effect at all, tend to increase it. our own. occurs in nearly all societies, including Post-partum abstinence the way from one to The amount of time involved, however, varies greatly--all to two to three years in others. two weeks in some societies Many pre-industrial insist for an arbitrary societies upon abstinence period of time after birth, In a few instances the duration of abstiweeks or months. for several usually nence is fixed by some developmental stage of the child--e. g., when the baby In many cases the taboo on first crawls, sits up, walks, or cuts its teeth. coitus extends through the lactation period, which may last two to three years.49 of course, represents a loss of fertility, because Not all of the time involved, for a time after parturition. is often delayed or occurs sporadically ovulation It is only when the period of abstinence extends to two months or more that a loss of fertility can be assumed, although even then it may not be quite commensurate with the amount of time covered. These longer periods, though found in primitive and peasant societies,50 are not customary in industrifrequently al countries. taboos on intercourse Long post-partum obviously help to space out chilsuch dren, but this is not the reason usually given in communities that practice a violation of the taboo is viewed as being magically taboos. dangerInstead, ous to the child or the parents.51 Such notions probably lead to the observance Taboo: A Sociological
pp. 59-61.
(46)
Hutton Webster,
(47)
Ibid.,
(48)
to the Tanala Linton, Study of Man, op. cit., pp. 194-195, with reference to W. Lloyd Warner quoted by Himes, op. cit., In a letter of Madagascar. "I do not think that there was any idea of limiting p. 8, Linton says: In at least but the losses were severe. in it [infanticide], population one tribe all children born on three days in each week were killed."
(49)
Clellan S. Ford and Frank A. Beach, York, 1951, p. 219.
(50)
Webster,
(51)
Ford and Beach,
Study,
Stanford,
1942,
pp. 59-65.
op. cit.,
Patterns
of Sexual
Behavior,
pp. 67-71. ap. cit.,
p. 219.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
New
232
AND FERTILITY SOCIAL STRUCTURE
of the abstinence In addition, it should be noted that in many instances rules. the male has access to another wife (if he is polygynous) or to a concubine or other available woman. The social structure may encourage observance of the taboo in another way. lWhen, as in India, the wife customarily goes to her parents' home to bear each of her first two or three children and stays there for a few months after the confinement, the taboo is enforced with ease. Thus the in one study reported post-partum abstinence fact that 80~ of Indian villagers of six months or more indicates a significant loss of fertility from this cause. cause.52 Doubtless similar or greater losses occur in many other agrarian societies. The "occasional" restrictions on sexual intercourse are those occurring in connection with regular holidays and special tabooed days of the ceremonies, The etc.).53 week, and important communal tasks (war, economic undertakings, exact amount of time lost to reproduction in this way has seldom been calcubut the Indian field study just cited found that the average number of lated, reasons was 24 per year in a rural village, days of avoidance for religious while in a middle class housing project it was 19.54 If these days occur spomuch loss of fertility, because they are pracradically, they hardly represent but in many socitically comprised within the normal frequency of intercourse; the abstentions eties extend over substantial "The natives of the periods. a part of the Caroline group, proscribe Mortlock Islands, any sexual intercourse in time of war; a man who violated the rule would die a sudden death. During the fishing season, which lasts for six to eight weeks, every Yap fisherman is are very strictly tabooed to him..."55 subject to many restrictions...Women In contrast to post-partum and "occasional" taboos on coitus, gestational abstinence cannot diminish fertility. The only question is whether obviously it may slightly increase Most societies intercourse fertility. during proscribe some part, but seldom during all or even the major portion, of the gestation Only seven of the primitive period. groups in Ford's sample extended the taboo to the greater part of the period.56 Usually it is toward the end of the pregIf intercourse nancy that the prohibition during the later stages applies. induces miscarriage or causes puerperal infection, as is sometimes occasionally then the taboo may enhance fertility, but only slightly. claimed,57 the almost universal of coitus during menstruation Similarly, prohibition can have little or no negative effect on fertility. Such abstention, when fertilization is least likely, tends to concentrate sexual activity in the more fertile In some pre-industrial the taboo cultures part of the menstrual cycle. is extended for a few days after the menstrual flow has ceased (as among the ancient Hebrews), which has the effect of concentrating coital still activity more directly on the days when conception is most likely. (52)
C. Chandrasekaran, "Cultural Patterns in Relation to Family Planning in of the Third International Conference on Planned ParIndia", Proceedings enthood, 1952, Bombay, p. 78.
(53)
Ford, Comparative Study of Human Reproduction, op. cit., ster, pp. 132-139.
(54)
Chandrasekaran,
(55)
Webster,
(56)
Ford,
op. cit.,
(57)
Ibid.,
p. 49.
o.
cit.,
op. cit.,
op. cit.,
pp. 28-29.
p. 78.
p. 134. p. 48.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
Web-
ECONOMIC CHANGE DEVELOPMENT AND CULTURAL
233
and peasant societies On the whole, primitive appear to have a greater abstinence loss through intra-marital number 4) than do (variable fertility more post-partum and "occasional" industrial societies. They have considerably of these in inhibiting and the effect is not fully abstinence, reproduction societies also occasionally counterbalanced by the fact that underdeveloped taboos (which may slightly have longer menstrual and gestational enhance fertility). The Other Intermediate
Variables
There remain four variables--number 10 (which usually has a low fertilityand numbers 5, 6, and 7 (which seem indetervalue in non-industrial societies) minate in their values). All four of these variables appear not to be clearly in different determined by institutional If there is any cultures. patterns in their fertility-values as between one type of society difference and another, the difference seems to be more a function of the general level of living than of the specific institutional structures. Perhaps one clue to this circumstance lies in the fact that three of the four variables (10, 5, and 7) are defined as in the sense of not being under control and hence not amenable to involuntary The other variable motivational determination. (number 6, frequency of coitus), is possibly too private and too linked control, though subject to individual to be culturally controlled. up with organic capacity With respect to number 10--foetal from involuntary causes--we mortality is generally low in pre-industrial have said that the fertility-value societies; that stillbirth because the data available indicate rates are greater in such is tentative, because adequate comparative societies. However, the conclusion does not exist for miscarriage information rates. Number 5--involuntary varies according to several abstinence--presumably In so far as health or sickness the nonfactors. may be involved, disparate a higher degree of such abstinence. industrial peoples would probably exhibit The same inference might be drawn with regard to impotency, except that this is often caused by psychological determinants condition which may be more prevalent in industrial Another cause of involuntary the cultures. abstinence, of couples due to migration, would seem to vary according to the separation of the society. circumstances historical particular Except under conditions of European contact, mobilindividual indigenous groups apparently have little these divergent influences abstinence can Clearly, ity. affecting involuntary It is therefore run counter to each other. to claim, for this varidifficult between societies. overall differences We are also handiable, any consistent has been capped by an almost total lack of data, for no comparative information with this issue in mind. collected of intercourse--possibly Variable number 6--frequency favors fertility more in underdeveloped than in industrial But at best the evidence societies. for this view is indirect, drawn solely from a few advanced societies where coital frequency appears greater among the manual than among the sedentary Such direct evidence as we have supports no view at all. classes. Average on "coital figures frequency" given in the literature, usually stated as so many times per week, are ambiguous, because it is unclear whether they mean because of sickevery week or only those weeks when coitus is not impossible or other taboos, etc. ness, absence, menstrual, Also, the comparative frequencited in the literature are fantastic from one showing variations cy figures to another that are wholly inexplicable.58 We have found no reliable society (58)
Thus Ford and Beach report as an apparent as often as three tralia have intercourse
fact that "the Aranda of Ausor five times nightly, sleeping
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
234
AND FERTILITY SOCIAL STRUCTURE
for comparable age groups evidence that the average frequency of intercourse none as between one society and another, and certainly varies significantly variations factor in inter-societal that this is a significant which indicates in fertility. we again have number 7 (involuntary to variable With respect sterility) of life in pre-industrial societies The hard conditions little evidence. may or absolute sterility--paramount of low fecundity give rise to a considerable in the latter ticularly part of the woman's reproductive span; and in given invenereal disease may have after contact with highly civilized stances, peoples, a pronounced effect On the other hand, the nervous tension and of this sort. tend to modes of life in urban-industrial artificial may possibly populations to some extent. lower fecundity of the four intermediate the comparative fertility-values variPatently, are unknown. Not only is evidence lacking, but there is ables just discussed no sound line of reasoning by which the behavior of these variables can be At most, there may be some institutional linked up with specific patterns. in each case with the general level of living. The evidence for connection this is best with respect to number 10, but the other three must be left for the time being as indeterminate. Conclusion:
The General
Pattern
must first of institutional factors in fertility Any analysis explain in general have a higher rate societies the well known fact that underdeveloped The explanation, in brief, is that than industrial of reproduction societies. in the face of high mortality, have had to develop the pre-industrial peoples, which would give them sufficient to an institutional organization reproduction at this level does not carry us very far. In order survive. However, analysis of institutional one needs to break down the reto study the effects factors, the various mechanisms so as to distinguish clearly process itself productive fertilthrough which, and only through which, any social factor can influence In trying to do this, we have found eleven "intermediate variables". ity. it can be seen that the generally is made along those lines, When analysis high of underdeveloped areas does not mean that these areas encourage fertility in every respect. As we have seen, they do not have high plus high fertility values on all the intermediate variables. Why, then, do they have low values in some respects and not in others? It is possible to discern a systematic difference between underdeveloped In general, and developed societies to the eleven variables. with reference for those variables the pre-industrial societies have high fertility-values farthest removed from the actual moment of parturition and which, therefore, to fertility. outlook favorable To a much greater degree imply an overall than industrial societies, they tend to encourage early exposure to intercourse a far younger age at marriage and a higher proportion married. --exhibiting
between each sex act", and that for Chagga men "intercourse ten times in a single night is not unusual". Nothing is said about how these bizarre are gathered, statistics or about what age groups in the population are The authors say simply, "it is reported that", or "it being considered. is not unusual that", etc. Such reports are all the more questionable since societies are said to apparently with a similar level of living have extremely different at "once a week" or "once or twice figures--some a week"--without of why they should be so low and others any explanation fifteen or twenty times as high. Op. cit., pp. 78-79.
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
ECONOMIC DEVELOPMENT ANDCULTURAL CHANGE
235
by delaying or avoiding the formation They thus lose little potential fertility tend to enjoin more of unions. After unions have been formed, these societies abstinence than industrial societies do (and therefore have lower values on variable number 4), but such "sexual fasting" arises from religious and magical motives rather than as a deliberate fertility control measure, and it does not appear to be great enough to have a substantial negative effect on fertility. also have high fertility-values for the concepUnderdeveloped societies tion variables. They practice little contraception and virtually no sterilization. Consequently, the tendency is to postpone the issue of controlling pregnancy until a later point in the reproductive process, which means that when a couple wishes to avoid children, those methods nearest the point of parturition-abortion and infanticide--are employed. These have the advantage, in societies living close to privation, of being nearer to the actual momentwhen the child must be supported. on the other hand, exhibit low fertility-values Industrial societies, for those variables involving the early stages of the reproductive process, especially age at marriage, proportion married, and contraception; and they manifest for the variables in the later stages, especially infanhigh fertility-values It follows that for many of the variables the two types of society ticide. exhibit opposite values. This is true for age of entry into unions, permanent celibacy, voluntary abstinence, contraception, and (if included as a variable) It is not necessarily true of the time spent between or after infanticide. or of abortion; and it, of course, is not true of unions, of sterilization, those variables characterized as "indeterminate"--involuntary abstinence, freBut the general contrast is quency of coitus, or involuntary infecundity. clear to require explanation. sufficiently A key to the position of the industrial societies lies in the fact that, as compared to pre-industrial cultures, they have achieved their lower reproducfor all the intermediate variables, tion, not by acquiring low fertility-values but by singling out particular ones as the means to that result. They took those means of reducing fertility which involved the least institutional organization and re-organization and which involved the least human cost. In the secular decline of the birth rate they relied more heavily on the mere postponement of marriage than on non-marriage. They relied less on abstinence, which makes heavy demands on the individual, and more on contraception and abortion, which do not. They dropped infanticide altogether and, in the later stages, tended to reduce In other words, they have undertaken to lower fertility, abortion. not primarily was by extending further the negative effect of the variables by which fertility lowered in the pre-industrial stage, but by using readily available institutional mechanisms with respect to marriage and by employing the possibilities of their advanced technology for conception control. Marital postponement was easily extended in the early and middle stages of industrialization because the basis for it already existed in Western society and because contraception and relatively safe abortion freed those who married late from the necessity of premarital celibacy. Gradually, in the late stages of industrial development, contraception has on the other varigained such predominance that it has made low fertility-values ables (including abortion and late marriage) unnecessary. Kingsley Davis Judith Blake University
of California
This content downloaded from 201.174.75.10 on Wed, 9 Oct 2013 18:16:17 PM All use subject to JSTOR Terms and Conditions
