Cultural Determinants of Population Stability in the Havasupai Indians ANITA L. ALVARADO University of Arizona, Tucson, Arizona ABSTRACT The Havasupai are a small tribe who have traditionally practiced very productive agriculture i n a highly fertile, well-watered canyon. The agricultural subsistence base was formerly supplemented by hunting and gathering in adjacent areas which before significant American encroachment were plentifully supplied with desired game and wild vegetal products. Under these circumstances it would be expected that the tribe would be characterized by a n expanding population. However, the population has been stable throughout historic times, approximately 200 years. In order to determine the significant factors responsible for the maintenance of population stability, all available information regarding Havasupai fertility, morbidity, mortality, and emigration was examined. An analysis of these data indicates that the population balance has been maintained primarily through cultural practices which limit fertility. Among the most important of these practices are the age at mamage, the age span differential between spouses, the high number of unmarried adults to be found in the population at any given time, the spermatocidal effect of the steam lodge, and perhaps the practice of infanticide and induced abortion. J U A N COMAS AWARD The Havasupai are a small tribe who have traditionally practiced highly productive agriculture in a very fertile wellwatered side canyon of the Grand Canyon in Arizona. The agricultural base was formerly supplemented by hunting and gathering. Food had always been plentiful; hunger and crop failure were unknown (Iliff, '54). Under these circumstances we would expect to find that the tribe has been characterized by a rapidly expanding population. However, the population has been stable throughout historic times or approximately 200 years and has characteristically ranged between 166 and 300 (Spier, '28; Smithson, '59; Martin, '66). The population low of 166 occurred in 1906 after two years of measles epidemics and coincided with the initial occurrence of continual white contact which began in the first decade of the twentieth century. Before significant white contact, the population had peaked at 300 in 1869 and 281 in 1899 (Spier, '28). Since the advent of modern medical and obstetrical care in 1930, the population began to rise slowly to the present unprecedented peak of 350. The availability of wage labor outside the canyon has permitted the absorption of the recent increase in populatian, thus having an imAM. J. PHYS.ANTHROI S14. 1970 portant effect on demographic as well as other aspects of Havasupai life (Martin, '66). However, this paper is concerned with the population stability exhibited between the time of the first census of the Havasupai in 1776 and the 1940's, when modern medical and economic factors permitted the present limited expansion. In order to determine the factors responsible for the maintenance of that population stability, all available information regarding fertility, morbidity, mortality, and emigration has been examined. The information regarding disease and mortality are rather poor since no satisfactory records exist. Nevertheless I feel that the data obtained are sufficient to pinpoint significant factors in the maintenance of population stability. These data indicate that the population balance has been maintained primarily through cultural factors which limit fertility. The average Havasupai household consists of 4.2 persons. Relatives are regarded as those to whom one is consanguineally related through the grandparental generation. Marriage is prohibited within this range (Spier, '28; Smithson, '59). Marriage is not marked by any ceremony. It is herein defined as the relationship 9 10 ANITA L. ALVARADO wherein a person enters into a regular sexual union characterized by open cohabitation. Secret liaisons before this time are termed premarital sexual relations. Menarche is expected at about 12-13 years of age (Nag, '62). As may be seen from table 1, however, age at entry into regular sexual union is considerably delayed. The ideal age for first marriage is stated by the Havasupai to be 25-30 for women and 35-40 for men (Smithson, '59). While it is unlikely that this ideal was ever approximated, there is some evidence that the age at first marriage before the turn of the century did approximate those found by Smithson ('59) and Nag ('62) in the mid-twentieth century and that the age for women dropped at least two years in the 1910-1914 period in response to a severe shortage of women during this postepidemic period (Spier, '28), then rose again. Thus, age at entry into regular sexual union is a highly significant factor in depressing Havasupai fertility. A problem in Havasupai demography is that of the previously imbalanced sex ratio. See table 2. To see what this means in terms of absolute numbers of individuals, see the figures for 1919. Information from the last 40 years indicates that the sex ratio has been balanced (Smithson, '59; Martin, '66). Prior to 1930, however, there was some loss of women in childbirth. In addition, it is possible that female infanticide may have been responsible for the imbalance. Cases of female infanticide in the case of twins is known but the actual extent of occurrence is in doubt since today the Havasupai deny its practice ever occurred despite the documentation of spe- cific cases in the past (Iliff, '54; Smithson, '59; Spier, '28). Thus, a combination of losses in childbirth and female infanticide may have been responsible. Whatever the causes of the previously imbalanced sex ratio, the important problem in fertility is not so much the absolute sex ratio as the number of women in the childbearing age brackets and how these are distributed in sexual unions. In the case of the Havasupai we see that, despite the balanced sex ratio of recent years, a combination of biological and cultural factors has resulted in a culturally created shortage of marriageable women. We have seen that there was a greater absolute number of males in the earlier decades (18811924). In addition, there is a greater longevity of males in the age brackets above age 62. For example, the 1964 population contained a total of 16 men and only three women in this age group (Martin, '66). This phenomenon has resulted in a relative shortage of marriageable women because the older men not only tend to remarry as they are widowed, but often marry women much younger than themselves. In 1950 there were six marriages in which the men were older than their wives by 15-41 years. In each case the men had all been married at least once before, but it was the first marriage for five of the six wives (Smithson, '59). The result of this practice is an average age span differential of 11-12 years for the married population. Again see table 1. This in itself might not be a significant factor in a larger population. But Lasker ('54) has calculated for 1919 a breeding population of 39. Thus with a very small breeding population I feel that TABLE 1 Mean age at marriage Years Men Women Mean age at marriage with both spouses previously unmarried: 1910-1914 22.4 15 1950-1954 24.6 17 Mean age at marriage regardless of whether either party previously married: 1910-1914 26.8 15 1950-1954 30.5 19.1 Average age difference 6.6 7.6 11.8 11.0 Mean age at first marriage: 2 24.6 1 2 Smithson, '59. Nag, '62. 18.9 5.7 HAVASUPAI POPULATION STABILITY such a wide age span differential between spouses is also a significant factor in the depression of fertility. Table 3 shows the average age at the births of the f i s t and the last child among the Havasupai. The average age at the birth of the first child is almost 20 while that at the birth of the last is 38 (Nag, '62). The delay in onset of reproduction activity is clearly related to the delay in the age at marriage. I feel that the early termination of the reproductive period may be due, at least in part, to the 11-12 year age span differential between the spouses. Table 4 shows the average number of live births for all married women aged 1554 in 1960. Thus married couples were producing an average of 4.4 children, an extremely low figure. Induced abortion is acknowledged by informants for both the past and the present and various methods of inducing abortions are widely known. However, no one admits to having had any and no one will discuss specific cases (Nag, TABLE 2 Sex ratio Ratio of males to females Year 113/100 145/100 151/100 125/100 122/100 (= 98 males, 79 females) 12A/100 1881 1905 1906 1918 1919 1924 1964 101/100 Adapted from Dobyns and Euler, '67; Spier, '28; Martin, '66 TABLE 3 Childbearing period Woman's general childbearing period: 1 ages 15-49 Havasupai woman's childbearing period: average age at birth of first child 19.7 average age at birth of last child 38.0 '62; Smithson, '59). Thus the role of abortion in reducing fertility levels is difficult to ascertain. However, I think that the open admission of its practice in the face of a pronounced reticence to discuss specific cases is an indication that it is not an uncommon practice. Nag ('62) found that five out of 46 or 10.6% of married women over age 19 were childless, a high incidence in such a small population. Of course some of these women are quite young and may eventually bear children. Yet, in the 1950s, two older married women had never borne children. In addition, three brides of five years were childless (Smithson, '59). Thus at least 14% of the women in the breeding pool were not producing for at least five years, a sizeable proportion in so small a pool. There is no social stigma attached to barrenness (Nag, '62; Spier, '28). Among the most common causes of sterility in a population are venereal diseases and vitamin deficiency. Yet venereal disease among the Havasupai is said to be rare (Nag, '62). Vitamins A and E, believed to be essential to normal spermatogenic activity, are adequately supplied by the diet, which includes both green and yellow vegetal materials. Perhaps a major contributing factor to the low fertility of the Havasupai may be traced to the ubiquitous use of the steam bath by the men. The human testes function at a temperature which is lower than that of the abdominal viscera and it appears that the human male may be temporarily sterilized by raising the temperature of the scrotum. Sperm production is affected by only slight changes in the temperature of the circulation in the scrotum and heat also affects normal mature sperms, rendering them infertile. The application of cold to the scrotum also reduces fertility (Glover and Young, '66). The Havasupai men customarily enter their sweat lodges four times an afternoon for periods of about ten minutes each. Then each entry is usually followed by a TABLE 4 Maternity ratio Women, aged 15-54,having borne live children: 5.5 Women, aged 15-54,including all married childless women: 4.4 1 Nag, '62. 11 12 ANITA L. ALVARADO plunge in an icy stream. Investigators have widows were some who were widowed after found the temperature of the sweat lodge their childbearing period and therefore to range from 118-157" Fahrenheit (Spier, have no significance for the fertility level. '28; Smithson and Euler, '64). Steam and But Martin ('64) found five unmarried other forms of heat being effective sper- women between the ages of 18-25 and matocidal agents, as are also local applica- eight unmarried men in the same age tions of cold, it may be readily seen that group. Thus the high number of unmarried this practice may be the single most sig- persons at any given time would appear to nificant factor in the depression of Hava- be another significant factor in the depressupai fertility. sion of fertility. Premarital sexual activities are quite Among other possible contributing faccommon, but pregnancies resulting from tors are the use of native contraceptives them are said to have been rare until re- whose effectiveness is unknown and the cently. Abandonment of such children by formerly rather lengthy lactation period of the mother is not uncommon, so that il- approximately two years. legitimate children are often raised by the Emigration as a measure of population maternal grandparents ( Smithson, '59). control seems to have been inoperative The rarity of illegitimate children during until the availability of wage work in the the first part of the twentieth century may 1940's. There being virtually no water on have been due to the reduced age of wom- the plateau above the canyon, an agriculen's first marriage. However, Smithson tural subsistence base could not be estab('59) feels that infanticide may have had lished there. Game and wild vegetal proda role. I agree since the illegitimacy rate ucts were formerly plentiful on the plateau seems to have risen at about the same time and were at one time regularly exploited when hospital deliveries became almost during the winter months. If the Havasuuniversal. Thus the birth of the illegitimate pai had been willing to give up their agrichild was legally established and i t could cultural base for a poorer hunting and not be destroyed. gathering economy, they could theoretically Another important factor affecting the have expanded out of the canyon. Through fertility level is the fact that, while few historic times, however, their plateau range people remain permanently unmarried, the actually shrank as the Walapai and Paiutes number of unmarried persons at any given steadily crowded them out (Spier, '28). time may be quite high. Both widows and Thus, pressure from other plateau dwellers widowers are expected to delay remarriage prevented expansion out of the canyon for one to three years although this waiting the Havasupai as a group. period is not always observed (Spier, '28; Yet, if emigration as a group was not Smithson, '59; Nag, '62). Divorced persons possible, surplus population might still be may remarry with no delay. Yet despite the drained off by the absorption of individuals tendency of the widowed and divorced to into neighboring tribes through intermarremarry, there was a total of 27 unmarried riage. However, according to Spier ( ' 2 8 ) , persons over age 30 in 1955. See table 5. intertribal marriages were relatively few Notice that five men and two women had and of short duration, occurring primarily never married (Smithson, '59). Among the when foreign women came into the canyon TABLE 5 Unmarried adults Year Age 1955 30 or above 1964 3 0 4 3 only 18-25 Men 6 widowers 7 divorced 5 never married Total 1 2 Smithson, '59. Martin, '66. Women 5 widows 2 divorced 2 never married 18 9 7 ? 5 8 13 HAVASUPAI POPULATION STABILITY on trading expeditions, remained for a time, and then usually returned to their own people. At the time of his fieldwork, in 1918, he found that a total of 11 intertribal marriages had resulted in the immigration of eight persons into the canyon and the emigration of only three Havasupai. Thus, immigration exceeded emigration at this time. If, however, most intertribal marriages were characteristically of relatively short duration and both immigrants and emigrants generally returned to their own tribes, then this practice would appear to have little effect on the population structure. In any case, emigration as a factor in the maintenance of population stability seems to have been inoperative. While there is little good data on morbidity and mortality, we find that the Havasupai historically lived under generally sanitary conditions, had well-balanced diets, were and are vigorous outdoorsmen who are generally healthy (lliff, '54; Spier, '28). They take meticulous care of babies and even had the original disposable diapers. Babies are placed on pre-masticated foods by four months of age (Smithson, '59) so that their diets conform to the general level. Although many new diseases and consequent epidemics assailed the Havasupai sporadically following white contact, the problem seems to have been of short duration. In 1965, the Arizona Commission of Indian Affairs ('65) reported the health problems in the following order of importance: ( 1 ) diabetes; (2) dental care; ( 3 ) pneumonia; ( 4 ) children's diseases. It is noteworthy that dental care precedes pneumonia and children's diseases. Nag ('62) has calculated a modern infant mortality rate of 6.4% and a child mortality rate of 2 5 . 1 % . This would seem to confirm the relative unimportance of children's diseases. It is known that the general mortality rate rose sharply with a population loss of up to 6% a year (Martin, '66) after the turn of the century, following white contact and the establishment of the reservation. It is further known that the infant mortality rate dropped sharply in the 1950's (Nag, '62). Yet it is difficult to say with any accuracy what the long-term mortalitv pattern was. Since the mortalitv rate rise at the turn of the century was the result of white contact and the introduction of new diseases, it may be that the present mortality rate approximates more closely the pre-intensive contact situation. The elimination of deaths in childbirth is the one certain effect o€ modern medicine on the mortality pattern. Mortality by causes other than disease is evidently unusual. Murder and suicide are said to be uncommon (Smithson, '59; Spier, '28). Warfare was in aboriginal times avoided to the extent possible. Stating that their population was too small to be risked in warfare, the Havasupai claimed that in the event of an attack they simply hid in the vastnesses of the Grand Canyon until all danger was past (Spier, '28). While the data, then, on mortality are too poor to make sound inferences on its specific role in the maintenance of Havasupai population stability, a look at table 6 will quickly reveal that it is the depression of fertility which is primarily responsible. With a live birth rate of 4.4 children per married woman and a child mortality rate of 25.1 % , which equals 1.1 child mortality rate per woman, we see that each woman produces an average of 3 . 3 surviving children. Since i t takes the production of two children to replace the parents, there is a surplus production of only 1.3 children per couple. Sufficient information is available to compute only one completed pregnancy ratio for approximately the same period. This shows a surplus production of 1.7 children per couple, thereby corresponding quite closely. These figures are for the middle decade of the twentieth century. Under earlier conditions i t could be expected to have been even less. For example, there has been an increase in illegitimate births in the last few decades. Therefore, i t can be seen that the population stability exhibited by the Havasupai has been due predomiTABLE 6 Number of chiEdren per couple mid twentieth century Maternity ratio Child mortality (25.1%) Surviving children Replacement of parents Surplus production of children 1 2 Nag, '62. Nag, '62. - 4.4/woman l.l/woman 3.3/woman -2.0 1.3/couple 14 ANITA L. ALVARADO nantly to cultural practices which limit fertility. Among the most important of these are the age at marriage, the age span differential between spouses, the high number of unmarried adults to be found in the population at any given time, the spermatocidal effect of the steam bath, and perhaps the practice of infanticide and abortion. I would like to emphasize that this is a preliminary study and that many more problems still remain. For example, I SUSpect that the average Havasupai woman may not produce 4.4 living children, but that many produce more while others produce less and a few produce none at all. The questions, then, are who produces how many children and what are the factors which are directly involved. Are more children produced by those couples living outside the canyon where the use of the steam bath has declined? Do some men within the canyon use the steam bath more frequently than others? The effects of the steam bath still need further exploration. Sperm counts are necessary to determine its exact effects on male sterility and the length of time such effects last. The frequency of coitus, which I believe to be quite high, needs to be investigated. There then remains the problem of the effect of frequent coitus on fertility, a question which is still under dispute by experts. It is hoped that, because of changing attitudes toward abortion in the United States, the Havasupai will become less reticent to discuss this aspect of population control with investigators. On the other hand, the problem of the role of infanticide may never be resolved. In the 1950’s, Smithson (’59) found an illegitimacy rate of 20%. If the practice of infanticide was at one time a significant factor in the rarity of illegitimate births, then the present illegitimacy rate may be su%cient in itself to account for the recent increase in population. In any case, there is no question that the Havasupai would presently like to limit their population by legal means. In the mid-l960’s, when a United States Public Health physician interested in population control offered them such means, they were eagerly accepted and effectively utilized (Robert S . Drew, pers. comm.). Therefore, it is reasonable to assume that abortion and infanticide, acceptable within the aboriginal cultural context, were and may continue to be employed surreptitiously. It is hoped that detailed answers to some of these questions will be elicited by proposed fieldwork in the near future. ACKNOWLEDGMENTS I would like to take this opportunity to express my appreciation to the American Association of Physical Anthropologists for the recognition and encouragement which the receipt of the Juan Comas Award has brought to me. In addition, I would like to thank Dr. Hennann K. Bleibtreu whose seminar on Human Populations showed me a means by which to integrate the discipline of cultural anthropology with a longtime interest in human demography. The intellectual spark which initiated this study is his; any defects in the study are entirely my own. LITERATURE CITED Arizona Commission of Indian Affairs 1965 Reservation survey on health. Dobyns, H. F., and R. C. Euler 1967 The ghost dance of 1889 among the Pai Indians of Northwestern Arizona, Prescott College Press, Prescott, Arizona. Glover, T. D., and D. H. Young 1966 Temperature and the production of spermatozoa. In: Human Ecology. Jack B. Bressler, ed. Addison Wesley Publishing Co., Inc., Reading, Massachusetts. Iliff, F. 1954 People of the Blue Waters; My Adventures Among the Walapai and Havasupai Indians. Harper & Brothers, New York. Lasker, G. W. 1954 Human evolution in contemporary communities. Southwestern J. Anthropology, 10: 353-365. Martin, J. F. 1966 Continuity and change in Havasupai social and economic organization. Unpubl. Doctoral Dissertation, University of Chicago. Nag, M. 1962 Factors affecting human fertility in nonindustrial societies: a crosscultural survey. Yale University Publications in Anthropology No. 66. Smithson, C. L. 1959 The Havasupai woman. University of Utah Anthropological Papers No. 38. Smithson, C. L.,and R. C. Euler 1964 Havasupai religion and mythology. University of Utah Anthropological Papers No. 68. Spier, L. 1928 Havasupai ethnology. Anthropological Papers of the American Museum of Natural History 2 9 ( 3 ) . Wrong, D. H. 1961 Population and Society. Rev. ed. Random House, New York.