CKNESS

AND
P

Perspective

ROBERT A . HAHN

THE ROLE OF SOCIETY AND CULTURE IN SICKNESS AND HEALING

la view of the history of infectious disease in the last IS0 years, it is at least as sensible to say that unregulated capitalism causes bronchitis as it is to blame bacilli.

Humans are born into societies that inform them how the world is and how to behave in it. People born into one society are iikilyto have views.' of the world and of proper behavior very different from rhe views of pkpie' ........... born into another society. In p a t because they see the world differently, . . . - ............ ~ in part because what they see i~ different, they live in different worlds. . In this chapter, I analyze the ways the society in which people gtow up &ects their sickness and healing. The organization o f societies affects not on1y what conditions people "get"and "have, but who gets which ones, how sick persons and their condi rims are perceived, and what is done in response to sickness. I argue that the effects of societal organization and culture on sickness and healing are powerful and pervasive. Sociocultural effects are causal in the same way that environmental carcinogens, toxins, and bacterial and viral pathogens are. Sociocuftural effects do not preclude but rather complement biological modes of causation in sickness and healing. They present a profound challenge to the underlying theory of Biomedicine, because in Biomedicine it is commonly assumed that sickness and healing are essentially biological events in which sociocultural phenomena play at most a secondary role. In the sections that follow, I describe three distinctive bur relared forms of sociocultural influence on sickness and healing; I refer to these forms of
.

"

76

: . O F SOCIETY A N D CULTURE
,I

77
"

4Econsrruction, " "mediation, and "production. I give examples, ith the effects of nationaI and societal organization on the health ions and ending with the ef6ec.t~ of immediate social environments alth of individuals. Finally, 1 review the place o f soctocuIrural within Biomedicine and formulate an alternative, interdisciplinch in which sociocultur.al phenomena are central.

THREE MODES OF SOCIOCULTURAL INFLUENCE ON SICKNESS AND HEALING The sickness and healing we experience are constrained bur not ined by our biological makeup and the physical environment we . Within biological and physical limits, and perhaps stretching these human societies and their ctritul-es affect events of sickness and healhree basic but related ways. Human biology and the physical environin turn, may also be modified by society and culture. s t , borrowing a concept from the phenomenological school o f sociology ample, Berger and Luckmann 19661, the culture of a society constrzkctj y societal members think and feel about sickness and healing. That is the members of a society are taught by others about di&rent sicknesses eir names, their characteristic symptoms and courses, their causes and ating circumstances, their cosrnologicd and moral significance, and opriate responses. What counts as sickness may differ from society to ty, and given conditions of sickness are understood in very different Anthropologjsts refer ro the part of a society's cultural reality coned with sickness and healing as its "echmmedicine."The metaphor of truction suggests that reality is a structure of ideas built by society ugh social interaction that may include informal as well as formal educan. The reality constructed by society makes sense of the experience of kness and healing to its members. rriculady in complex societies, difkring medical realities may be conted in different social sectors; professional healers often mainrain one lity (or several), folk healers others, and laypersons still others (Meinman 980). All are first introduced to avshared, popular reality of sickness and ealing; this orientation shapes the subsequent training of those who become rofessional healers. In turn, popular media may carry professional medical ealities back to folk healers and the h e l a y communiey. The members of a ociety may encounter multiple medical realities in rhe course of life. When sick, patients may resort to several different societal sectors for treatment. The reality constructed by society also affects events o f sickness and healing in two additional ways. Medzdon, a second mode of sociocultural influence on sickness and healing, is perhaps the best recognized o f sociocuItural effects. The concepts,
+

78

AN ANTHROPOLOGICAL PERSPECTIVE

ideas, and values of a society's culture guide the behavior of societal members, distributing them in time, space, and activity. The members of a society live in certain kinds of dwellings and cerrain kinds of communities; they pursue diverse occupations and social and recreational activities. By such guided movement, they are brought into greater or lesser contact with pathogenic sources; thus sickness may be fosrered or prevented. In Western society, activity-specific types of medical practice-for example, occupational medicine and spom medicine-attest a partial recognition of the sacioculturaf mediation of s ickness . Sociocultural mediation brings the members of society into greater or lesser contact not only with causes of sickness but with therapeutic resources as well. Resources may be located in different places and institutions, and access to them may be determined by the patient's condition, characteristics, o r circumstances. A society's sick roles also guide the activities of members by telling them what to do when sick-when to take to bed (or hammock, sleeping mat), to consult with family members or friends, to seek medical attention. The sociocultural mediation of sickness and healing may be regarded as a form of cultural transportation of persons, pathogens, and thetapeuric agents. Proalucta'on is a third sociocultural effect on sickness and healing. A society's beliefs and patterned relationships produce events of sickness and healing not only by transporting persons, pathogens, or therapies-as in sociocultural mediation- but by more direct causation as well. Relationships and cultural beliefs may themselves be pathogenic or therapeutic, regardless of what they lead the members of society to do. Sickness or healing may result from beliefs and interpersonal relationships with regularity just as they result from better recognized biological pathogens and medicines. Perhaps the most striking example of the production of pathological and therapeutic events is the placebo phenomenon, in which beliefs and expectations produce the reality to which they refer-belief in a pill fosters its efficacy. But beliefs may also have other effects thac are seldom noted and . perhaps not as spectacular: it is likely thac beliefs o f certain kinds &ct not only the conditions they refer to but other, tangential conditions as well. For example, it. may be that fear o f cancer does not produce cancer but does produce some other condition, perhaps elevated blood pressure or cornpromised immune function. Such side effects of beliefs di&r from the placebo phenomenon i rsel f. There is evidence that social relationships, as well as beliefs, produce sickness and healing not only because they guide people to sources ofsickness and healing but because they are themselves more directly pathogenic and therapeutic, for example, in the death of one spouse shortly after the death of the other. These effects may seem puzzling if we think of relationships as being "outthere,"disconnected from physiological functioning. If, however, we acknowledge that relationships, and beliefs as well, are embodied within

ROLE OF SOCjETY A N D CULTURE

79

well as among us, rhen the sociocultural production o f sickness and ng becomes more reasonable. &aracterizing the role o f society and culture in the causal production of ss and healing, it is important to recall several basic features of causal~ r s t evencs , of sickness and healing are rarely, if ever, caused by a sinvent; more often, causation requires the concurrence of many circumes. For example, exposure to M y r o b ~ c t t ~ - tubwcuIasis iw~ may be necessary not sufficient to cause tuberculwis; no$ a11 exposure leads to infection or c (Dubos and Dubos 1952; Harris and McClemenr 1983). The amounr means of exposure to the organism affects the occurrence of infection; and acteristics o f the host (for example, age, nutritional and immune status, the presence of other diseases such as diabetes o r alcoholism) affect the ihood of contracting disease once a person is infected. Such conditions not be necessary for the occurrence of tuberculosis, but in the presence of cob,mterizkrn agent, chey may be sufficient causes. Poverty and crowding known to increase the likelihood o f tuberculosis cxansmission; and ical treatment for tuberculosis decreases the likelihood of transmission ducing the number of infective transmitters. n fact, &fycob~nwi&m tba'bercaloszs is not a necessary cause of tuberculosis; er mycobacteria also cause tuberculosis. The causarion of tuberculosis by er mycobacteria points to an important characteristic of the causal pro. If contact wich one of several different organisms results in a common come, then it is likely char some feature shared by these organisms, rather n each organism as a whole, is causal. Experimental evidence indicates t several fragments of mycobacteria can cause tuberculosislike reactions in imals. indeed, similar fragments have even been synthesized in rhe absence rnycobacterial organisms, so that mycobacteria themselves are no longer cessluy causes, but rather the vehicles o f causal fragments. Mycobacteria e regarded as "the cause" of tuberculosis because, in human environments, e causal fragment occurs most often in their company. Were the causal ments more commonly attached to another entity in the human environt , this enrity would more iikely be regarded as the cause of tuberculosis. The notion of causation I am presenting here might be couched in terms of I and pmximd causes, causes more or fewer steps removed from the tcome. For example, the My~ob~c~w2'urn i s a distal cause of tuberculosis in rnparison with the fragments that are believed to be the more direct, real auses. It should be noted that distal and proximal causes are assessed relative other causes. The fagrnents themselves may be shown to be distal to maller bits that cause tuberculosis, whereas the Mycubdcwium is proximal to he social conditions that promote its spread, subsequent infection, and

When I claim that sociocultuml conditions produce or cause sickness and ealing, the sense of causation is the same as chat in claims that Mycobactwiabm twbwc~losis i s the cause of tuberculosis. The difference may be one of prox-

80

AN ANTHROPOLOGICAL PERSPECTIVE

imity to the outcome in a causal chain. In sum, the presence of a pathogen alone is not sueicient to cause a pathological outcome; ocher conditions must also be met. Even given known pathogens, it may be some facet of the pathogen rather than the whole thing that, in conjunction with ocher causes, results in the outcome.

SICKNESS, HEALING, A N D THE LXf E OF SOCIETY: BODY POLITIC A N D BODY SOCIAL In this section I illustrate the sociocultural effects of construction, mediation, and production at different levels of social life, from nations as wholes co the more intimate social environments of persons. I use examples from the developing as well as the developed world.

E x m p h I :Ndtkonal Economy and lndi~zdml Health

A society's economy is an index of the way in which it organizes its activity and exphi ts and distribures its resources. Thus, one way to assess the effectsof societal organization on the health o f a population i s to examine the association of national economic conditions and trends and correspondiig health rates. f n general, we would expect chat a lack of resources would lead borh to increased exposure to the causes of disease and to a diminished ability to treat persons for disease once it has occurred. An uneven distribution of resources in society would lead to a parallel uneven distribution of sickness and healing. M. Harvey Brenner has conducted studies to assess the association between economic events and mortality from different causes in nine industrialized nations: Australia, Canada, England and Wales, Denmark, rhe Federal Republic of Germany, Finland, Prance, Sweden, and the United States (Brenner 1973, 198 I, 1383, 1987). Brenner draws two basic conclrrsions. First, he finds that long-term trends in per capita income are inversdy associated with rates of mortality: as per capita income in a nation rises, the nation's mortality rates fdl. Long-term increases in per capita income are associated with declines in infant mortality, mortality from coronary hear disease, and several types of cancer. This association is found in at1 ofthe ni nations except France (for reasons not clear). Second, Brenner has also examined associations between mortality rare and shorr-term economic events- recession (as measured by unemploy ment), rates of business failure, and national welfare expenditures. He argue that, while unemployment rates are likely to affect mostly the lower soci economic classes of a population, business failures are more likely to &e orher classes as well. He finds that short-term fluctuations in econom

OF

SOCIETY A N D CULTURE

81

associated wich increasing rates of death from the same causesrtality, coronary heart disease, and several types of cancer. The event and changes in mortality sates varies by cts appear to be rapid-within a year-for infant mortality, e , and suicide, and to peak after two to three years for coronary hearr

he associarions of long- and short-term economic condirates by the distribution of recognized risk factorssmoking, alcohol, and diets high in fat; these are examples of have referred to as "mediation."Bur Brenlaer also notes ion of social relationships chat accompany economic , in many nations, rapid economic fluctuations are ed with increased mortalicy rates, independent of changes in riskrevalences. Thus it seems that psychological reactions and social ships may be more directly associated with mortality outcomes; these oculcural effects 1have referred to as "production. ough smdies of this sort suggest a powerhl connecrion between the th of nations, it is important to note crucial rnethodal limitations. Brenner's studies are based on separate information on onomic conditions of nations and on their mortality rates; thus, we t know that the individuals affected by economic fluctuations are the nes who suffer adverse health events. in addition, tong-term declines rtality have coincided with many changes in addition to increases in per a income-for example, improvements in sanitation and public health. e changes may also explain changing rates of mortality. formation an the ecoaomic and beaith conditions of nations has atso interpreted i n a different way. For example, Marxist researcher Joseph believes that capitalist societies exploit workers for the benefit of enrreeurs (Eyer 1977). Eyer has claimed, to the contrary of Brenner's analysis, economic prosperity is associated wirh incrwsking rates of mortality. society prospers, another suffers. Eyer believes that, r than unemployment, it is employment in capitalist society that is sful for workers because they must be more mobile, thus disrupting their stems of social suppoa. urther:evidence for Brenner's claim has been found in other srudies, hower (1973) have shown that low levels of education d income are both, independently, associated wirh high rates of infant morlity and overall mortality in individuals in the United States. Syme and e r h a n ( 1976) and Kaplan and colleagues ( 1987) have shown that, for indiortality levels are inversely associated with income r a wide range of conditions; even when known risk facton-smoking and besity, and lack of access to and utilization of medical care-are taken into ccount, differences in h d t h status by income remain. Colleagues and I have recently esrimated thar poverty accounts for 6 percent of mortaity among
II

82

A N ANTHROPOLOGICAL PE RSPECTlVE

black and white adults in the United States (Hahn, Eaker, Barker, Teutsch, Sosniak, and Krieger in submission). Researchers hypothesize that poverty compromises resistance to a broad range ofdiseases by placing greater burdens on people and by taxing their capacities to respond (Kaplan et al. 1987).

Example 2: The Health o f Blacks in South Africa and the United Strjlte~
A comparison of the health conditions of blacks in South Africa and the United States provides another striking illustration of the effects of culture and societal organization on the health of populations. In South Africa, at least until recently, the dominant white regime succeeded in enforcing a comprehensive system ofthe oppression of blacks and other non-white p o p lations. In the United States, a system of black slavery and formal segregation has been replaced by one more egalitarian in principle, but nevertheless one in which discrimination persists. (Myevidence on South Africa is provided in reports by Andenon and Marks 1988; Jinabhai, Coovadia, and AbdoolKarirn 1986; Nightingale, Hannihl, Geiger et al. 1990a, 1990b; and World Health Organization 1983. My evidence on the United States is provided by the Centers for Disease Control 1993b; U. S. Department of Commerce 1990; Thernscrom, ed., The Haward E;rzcycIo#edia o f Am&icd Ethnic Groupr; and ocher individual researchers.) In South Africa, race has been the most prominent theme of societa organization i n recent decades. The white cultural system of apartheid con structed a rationale for white domination and maintained domination by 1 and by force. The 14 percent minority white population has controlled percent of the disposable income and 82 percent of the land (and a greate proportion of the high-quality land). While Asian Indians and mixed-rac "coloreds"had separate and proportionately small legislative representatio compared with that for whites, until 1994, black South Africans had representation at all. Yet, it was the white government that dictated societal distribution of resources. (South Africa provides better evidence Eyer's theory of national prosperity as a cause of death than the democrat industrialized nations studied by Brenner. ) Places of residence for blacks were determined by the apartheid g ment. Black communities were forcibly relocated; 7 4 percent of the b population was moved to "homelands" where there was little or no emp menc. Housing shortages were estimated to be thirteen times greater blacks than for whites. Access KO basic sanitation, water, and food ha usually been inadequate in black settlements. The movement of black individuals has also been strictly regulated. La provided for the legal detention o f blacks, including children, wicho charge; torture of children as well as adults was reported ro be comrno Because they have owned most narural resources-principally gold and di

83

ines--whites have also controlled labor opportunities. Many blacks n forced r o adopt a migratory life, disrupting family and other social ens. It is estimated that 80 percent of the male population has been in migratory labor. Unemployment among urban blacks was reto be 88 percent, compared with 5 percent for whites. produced severe health problems elative to Indians and coloreds as well as to whites). "Poverty rimary cause of the prevalence o f many diseases and widespread malnutrition among black South Africans" (Nightingale et al.

ms, the quality of public health dissemination of health infurmas than for whires, and least adelacks in remote, rural regions where the burdens of sickness are , it is likely that available statistics understate health conditions an blacks. But they provide at leasr an indication of the health expectancy at binh for white South Africans was seventy-one years in s was sixty-two years. Infant mortality i s another good 's overall health; whites have a rate of infant mortality of (comparable with U.S. whites); blacks have a rate of 0 (more than 6 percent, compared with less than 2 percent among lack South African infants die at a m e 6.8 times thsr of
ng children who survive, rnalnutririon i s common, leading to ed that 2.9 million South African childrenhed and that 15 ;000 to 30,000 die annually diseases also occur at far higher rates among ks than whites in South Africa; in 1979, for example, rates of tuberosis were fifty-nine times greater among blacks than among whices. Black ners, mostly relatively healthy adult males, are estimated to have died at a e of 390/ 100,000per year between 1979 and 1983-a rate slightly less an that for black men 25-34 years of age in the United States. Not only does the structure of South African society cause (that is, medits black population; it has also withheld the eid government had fourteen separate health s, biacks, coloreds, Indians, and d l races mbined, and one for each of the semi-autonornous homelands. Per capita ealth-care expenditures for blacks have been 25 percent and those for coloreds and Indians 5 3 percent of the expenditures for whites. Researchers timate chat there is one physician per 400-6,300 white patients and one ack physician per 90,000-326,000black patients-a ratio of 22511, ~nversely proportional to medical need. Hospitals have commonly been seg. .. S; . . regated, and hospitals for whites may have empty beds, while hospitals for

84

AN ANTHROPOLOGICAL PERSPECTIVE

blacks (with half the number of beds per capita) are greatly overcrowded. Physicians who resist the oppressions of apartheid have been harassed and

killed.
Finally, the society of apartheid also restricted access for blacks to general and medical education. Per capira expenditures on education for whites have been 10.2 times higher than expenditures on education for blacks, and though only I percent of whites cannot read, 33 percent of blacks ate reported to be illiterate. Only 4 percent of a recent medical school class were blacks. The socioeconomic and health conditions of blacks in the United Scares, though now generally better than conditions of blacks in South Africa, were perhaps worse (than chose of South Aftican blacks at the same time) until at least the end o f the Civil War (Savitt 1978). Under slavery, U. S. blacks were treated as less than human; they were sold and traded, breaking family and community relations;they were forcibly subjected to arduous labor and given minimal food and quarters (Savitt 1978). Although the figures may not be fully reliable, a report from the U .S. Bureau of the Census in 1850 indicates that, despite apparently higher rates of some diseases among whites (for example, malaria and tuberculosis), overall mortality in Virginia was 1.6 times greater among black slaves and 1.3 times greater among freed blacks than among whites (Savitt 1978 ) . Most deaths were ascribed to mberculosis, respiratory diseases, unspecified nervous system diseases, diarrhea, cholera, and "old age ." The Emancipation Proclamation and subsequent constitutional amendments and legislation formally ended slavery and gave b k k men the vote. Despite formal emancipation, however, the civil rights of blacks were obstructed in many regions of the country in an apartheid-like sysrern that predominated until the second half of the twentieth century, when the civil rights movement- and civil righcs legislation more firmly established legal access for blacks in many arenas of U. S. life. ~ o n f l i c t i n cultural ~ ideologies in the United States have disputed appropriate approaches to equality-by redressive affirmative action, by anridiscriminarion legislation, or by a more individualistic "freedom o f choice." With relatively greaxer civil rights in the United States, substantial inequities in well-being between blacks and much of the rest of society have been reduced but not eliminated. Socioeconomic opportunities and resources are unequally distributed in U.S. society, and eEorts to create a balance are resisted. In 1988, more than a quaner of black families lived on incomes below the poverty level, a rate almost four rimes that of white families; even among black families in which the householder had some college education, 1 1.5 percent lived below the poverty level, a rate still more than four times that of white families. Unemployment in 1988 was also 2.3 times greater among blacks than among whites, independent of educational level. Blacks have higher prevalences than whites of many risk factors far major

EE OF SOCIETY A N D CULTURE

85

rcent higher for black than white men, 83 hi te women), hypertension ( 13 percent higher k than white men, 48 percent higher for black than white women), rcent higher for black than white men, 6 t higher for black than white women) (Centers for Disease Control 1. At all educational levels, blacks also suffer higher rates of occupament to positions with greater exposure to

health care is also substantiaily lower ong whites in the United States. For example, 1.6 3 as many blacks as whites under sixty-five years of age were wirhouc any of health insurance in 1986. And for the same diagnoses, kidney disease heart disease, blacks were less likely ro receive medical procedures or asiske et: al. 1931; Ford er al, 1989). As an overall of health care, blacks in the United States are 4.5 whites to die of medically preventable conditions
ccess ro and/or utilization of

blacks persistently differs from that of whites, aldiminished substantially since the beginning of the e in 1900, life expectancy of white newborns was cy for nonwhite newborns (including populations 33 years-an excess o f 44 percent. In 1990, U.S. tes could expect to live 76.1 years and blacks 69.1 years-an excess of y 10 percent; however, life expectancy fur blacks has recently declined for first time this century. The proportion of blacks who assess their health as or poor is almost twice that of whites, and blacks are only 7 1percent as ir own health as excellent, Infanr mortality in 1987 was reported to be 2.1 times greater among acks (18.7) than among whites (9. I), a difference present regardless of che other's level of education. EmanueE ( 1986) has reviewed evidence char not be easily corrected from one generadon to the re small because of the small stature of their awn others are themselves likely to bear low-weight infants.' In a study of hospital deliveries in Seattle, the risk of abnormal births was 2.5 rimes reater for women who were themselves born with a low weight than among was higher (Hackman et al. 1983). Poor health m generation to generation, a cycle that may equire more than routine prenatal care to inrermpr . In 1987, the age-adjusted death rate from all causes was 1.5 rimes greater among blacks than among whites. Rates of homicide (for victims) were 7 times greater among black than among white men and 4.2 times greater among black than among white women. Many infectious diseases-for example, acquired immunodeficiency syndrome, ruberculosis , syphilis, and hepatitis B-also occur at subs~antiallyhigher rates among blacks than

86

AN ANTHROPOLOGICAL PERSPECTIVE

Table 4,3 Health Conditions of Blacks and Whites in South Afr'ricaandthe United States
South Africa

United States

Black
Life Expectancy (years)

White
KIA 71

Black 33.0

White

1900 1980s
1Inf;anc Mortality

NA 62

62
170
71

69.4 18-7
2 1.9

47.6 75.6 9.1

(deaths/ 1,000 livebirths)

TubercuIosis (cases/ 100,000 population) Measles (cases/ 100,000 population)

et al.

I5
24

6.1
0.5

0.4

Sw~ces: Dara from CDC 199Ua, Buehler er al. 1989, Anderson and Mark 1988, and Nightengale

1990b.

among whites (Buehler et al. 1989; Hahn et al. 1989; Selik et al. 1988; McQuiflan et al. 1989b A recent study by Otten and colleagues (1990) indicates that 3 1 percent of the black adult excess in mortality in the United States reflects differences i n the prevalence of known risk factors (for exampie, hypertension and alcohol consumption); another 3 8 percent is arcributable to differences in family income; the remaining 3 1 percent is unexplained. Similarly, Navarro has recently presented evidence (1930) that much of the differences between the health status of blacks and whites in the United States are due to diEexences i n socioeconomic status. In 1986 U.S. blacks were onjy 54 percent as likely to have completed college as whites. In 1987-88 blacks represented 6 percent of medical school classes and approximately 12.2 percent of the U .S . population as a whole. Comparison of South Africa and the United States (table 4.1) indicates that differences in the health conditions of blacks and whites are caused principally by the sociocultural and socioeconomic circumstances of these nations rather than by biological, "racial" differences between blacks and whites. Cultural systems have fostered forms of societal organization, which, in turn, through socioculrural mediation, have differentidly dist rib u d the c a w s of sickness and the resources for its remedy among blacks and whites. Cultud systems have also affected social relationships and expecrations, : which may in turn have affected the health conditions of blacks in the United States and South Africa. During the slavery years and the period immediately thereafter, the health '; of blacks in the United States was apparently far worse than that of blacks in South Africa today and was certainly worse than that of U.S. whites at.::! the rime. Sociocultural movements, following emancipation in the ~ n i r e d $ States, gradually reversed this difference in the status of South African and

{:.

..

QLE OF SOCIETY AND CULTURE

87

blacks. W e can hope that the emancipation of South African blacks will ve their health status and thar further efforts to create equalopportunity in the United States, following generations of oppression, educe the remaining gap between the health of blacks and whites.

Exampie 3 :Social Relationships and Health Status in the Locd E nviro~ment

Having considered examples of the effects o f national and societal nization, I now examine evidence for the effecrs o f people's immediate nts on the mediation and production of the sickness and
number of studies demonstrate thar a person's connectedness with rs is an important determinant of sickness and health (House, Landis, Umberson 1988; Berkman 1984). In a study in Alameda County, Catia, almost 7,000 adults were first quesrioned in 1965 about their health , behavior, and social conracrs, including their mariral status, relations with other family members and close friends, and church and other p membership (Berkrnan and Syme 1979). A Social Network Index was as an overall measure of connectedness. Over the next ten years, 632 of persons first interviewed in 1965 died. Men with the lowest Social 3 times as likely to have died as men with the highest ex; women with the lowest Index were 2.8 times as likely to have died as t Index. Among persons with a low Social Network ex, there was an excess mortality from heart disease, all types of cancer, ses of death combined, suggesting that the effect of nnecredness on mortality was associated not with the etiology of specific ore general h c tor of resistance or susceptibility. Since it is possible that the observed association between sociai connecteds and mortality might be explained by a lack of connectedness among ple with poor health (that is, poor health causing fewer connections), earchers anempted to confirm their own explanation (that is, lack of nnectedness as a cause of increased mortality) by examining the initial ' ealth status of the persons they studied. Even when factors such as health atus in 1965, risk behavior (for example, smoking, hypertension, overeight, and heavy alcohol consumption), socioeconomic status, and prevenve health-care practices were taken into account, persons least connected ith others were still more than twice as likely to have died in the ten-year were more connected with others. The persistence n association between social connectedness and mortality-having confor initial health status and risk factors-suggests both that social or rather than (or in addition to) a result of poor f connectedness is independent of other major risk

88

AN AMTHRQPOLOEICAL PERSPECTIVE

Researchers have also focused more narrowly on the health effects of marital status (Helsing, Szkto, and Cornstock 198 la; Koskenvuo et al. 1980). They have found that widows and widowers are more likely than comparable married persons to die from several causes, for example, coronary heart disease and all types of cancer. Most studies indicate that excess mortality is greater for men than for women. Several studies indicate that the excess peaks within several months following a spouse's death, but persists for years. A study comparing the prevalence of major risk factot~for mortali ty among never-married, married, divorced, and widowed persons in the United States found no differences in risk factor prevalence thac might account for higher mortality rates among widowed than among married persons weiss 1973); thus it seems that differential mortality by marital status is unlikely to be explained by corresponding differences in major risk factors. Some studies show more directly thar che effect of marital status remains even when controlling for risk factors, preventive health care, and socioeconomic status (Hehing, Szklo, and Cornstock 198 Ib). Other research indicates thar the effects o f a person's social environment need not involve direct personal contact. An association has been found between traumatic death or violence in the community environment and subsequent suicide or suicide-like behavior. For example, when newspaper or television stories about a suicide are released, the rate of suicide increases in the following week; the greater the circularion of the newspaper, the greater the increase in suicides (Phillips 1974). When Marilyn Monroe killed herself in 1962, 197 suicides occurred during the following week- 12percent more than the number expected on the basis of "normal"suicide patterns. A recent srudy indicates that teenagers are more susceptible to televised publicit about suicide and thar increases in suicides are greater for girls than bo (Phillips and Carstensen 1986). Although nor commonly thought of as suicidal, motor vehicle crashes fo low a similar pattern. The well-knownJapanesewrirer, Mishima, committe ritual suicide (hara-kiri) on November 24, 1970; Willips (1977) estimate thac in California, 98 motor vehicle fatalities would have normatly be expected i n the following week, whereas 117 occurred. Phillips calculate thar , on average, motor vehicle fatalities increase 9 percent above the pected rate in the week following front-page newspaper stories, and thar, newspapers with greater than average circulation, the increase is 19 perm F inaily, though for ethical reasons, experiments are rarely conducted human beings, the birdday l m e q system depioyed to select men for m tary service in the Vietnam War has been used in a clever study to examine effects of selection for war on selectees (Hearst, Newman, and Hulley 198 As if in a randomized experiment, the long-term health of men select more or less by chance, to serve in Vietnam was compared with the health men not selected. Rates of suicide and motor vehicle fatalities were sign candy greater among those selected than among those exempt. Only

O F SOCIETY A N D CULTURE

89

those selected actually served, however; assuming that all the hs occurred among those who served, their races of suicide would been 86 percent greater and their rates of motor vehicle deaths eater than those nor selected. While selectees and nonselectees r in rates of mortali ty from most other causes combined, mortalcirrhosis of the liver was lower among selectees (for reasons not ). W e thus have evidence that selection for war, and most likely war, increases the susceptibility to subsequent suicide and motor mlity. A plausible explanation is that people internalize their soonment-s, and when these environments are violent, then selfIve violenr behavior may be expressed in response. It remains to cumstances that distinguish persons who are so influm those who resist. e of Biomedicine, "communicable diseases" are sickntracted by the spread of infectious microorganisms, the phrase may apt for conditions that work by means of human symbols and rates of suicide and motor vehicle fatalities increase ents or profound interpersonal experiences, ir seems nicated that trigger a vital (and mortal) response in

Example 4: C u l t ~ r ~ Expectations, E Sicknm, and Healing

Perhaps the most famous o f socioculturai phenomena is the placebo (from the Latin, placwe, "to please"). Placebo medicines and proes, even placebo surgery, are thought to produce desired effects because re believed by patients to have these effects. Placebos (that is, the clnes and/or the procedures) are thought to work independently of fic pharmacological or physiological properties, such as those associared other medical treatments. While, as I describe in rhore detail below, bus play an ambiguous role in Biomedical practice and research, perhaps ique public curiosity because their effects are so widespread yet appatlacking explanation. apiro (1960) has claimed that the placebo phenomenon has been the Inant mode of healing throughout human history. i c might also be ed that in the contemporary nonindustrial world, the placebo phenom.$+~:!i?non .... persists as the major, mosr pervasive force in healing. Summarizing a number of studies, Beecher concIudes that, even in the indusrrial ,"physical"placebo therapies have "an average significant effectiveness 2% +/- 2.2%, " that is, the percentage of patients for whom there is ctory relief (Beecher I95 5 : 1603). e placebo response parallels that af pharmacological1y active drugs. In e course of an enlightening philosophical discussion of the placebo pheomenon, Howard Brody ( 1977 ) summarizes the parallels: placebo effects
: : . ; :
,.:

:..

...:

:..

90

A N ANTHROPULOGfCAL PERSPECTIVE

peak several hours f0110wing administration; there is a cumulative effecr as the placebo is conrinued over time, a residual effect after the placebo is stopped, and a decreasing efficacy with increasing severity of symptoms. Placebo drugs can also have toxic "side effects," for example, somnolence, insomnia, palpitations, irritabili ry, and rashes. In addition, placebos can be addictive, causing withdrawal symptoms when discontinued. There are reports that the colors of placebos moderate their effects: yellow and white capsules are said to be most effective, red and gray ones to have most side effects (Honzak, Harackova, and Culik 1972). One study suggests, paradoxically, that placebo pills may be effective even when patienn are informed that the pills are thought to be pharmacologically inert (Park and Covi 1965). A painful experiment shows how placebos work by ordinary pharmacological mechanisms. Levine, Gordon, and Fields ( 1978) studied patients whose teeth had been extracted and whose pain had been significantly diminished by placebo medications. The researchers gave these patients naloxone, a compound that blocks the effectso f pain-killing opiates; they found that the patients' pain increased again, almost to the level of patients who had nor initially responded to the placebo. The experiment suggests that the expectation of pain relief causes the release of opiates. Placebo healing is a physiological effect of expectarions-large1 y shaped by the patient's cultural setting (Nahn and KLeinman 1983b). Expectations are not simply logical propositions about future events; they are physically embedded in the brains of those who maintain them and they are thus associated with neurorransrnixters andlor hormones that a&ct physiological functioning. Exfxtcrations are a bridge connecting our cultures and o bodies. Thus, it is reasonable that placebo effects should be found in al cultural settings, and that, with differing expectations in different culture specific placebo effects should vary substantially from one setting ro anothe Evidence for cross-cultural variation in the placebo effect is found Daniel Moerman's insightful reanalysis of thirty-one srudies, carried out sixteen nations, on the efficacy o f cimetidine for relief of gastric and duoden ulcers (Moerrnan 1983). Ail the studies chat Moerrnan reviewed were co ducted in a manner common to drug testing generally in Biomedical se tings: study subjects were inirially evaluated for the presence of ulcers a assigned to treatment or nontreatrnent control groups ; while rhe treatm g o u p was given cimetidine, the control group received a placebo, belie (by the investigators) to have no specific pharmacological effect on ulcers. ensure that results of the study would not be influenced by either the patie or the researchersF knowledge of who was receiving cimetidine and who placebo, the studies were "double-blinded"- neither patients nor resear ers knew which group the patients were i n until drer the patients had be assessed for final results. Four to six weeks after the initiation of the studies, rreaiment and con patients were reexamined for the presence of ulcers. If the 1,692 patie

OLE OF SOCIETY A N D CULTURE

81

11 thirty-one studies are added together, 76 percent of the experimental and 48 percent of the control group were healed, indicating thac dine added 28 percent to the efficacy of the placebo alone, a resulr unlikely to have occurred by chance. Looking ar the thirty-one studies tely, however, only thirteen showed a statistical difference in outcome en experimental and control groups, and eighteen showed no statistical
erman compared the patients in studies showing cimetidine effective the patients in studies showing cimetidine not signi ficantty diEerenr placebo. For parients treated with cimetidine, results were similar in the statistically significant and the nonsignificant.studies; 7 percent in rmer group and 77 percent in the latter group were healed. Bat the nion of patients healed by placebo in the significant studies was far r (37 percent) than the proportion healed in the nonsignificant studies ercent). Thus, what made the significant studies significant was not rhe cy o f cimetidine among those treated, but the relatively lower efficacy of lacebo in the control group. addition, one of the studies reviewed by Moerman (Sonnenberg, Keine, ber 1977) indicated a rernarkabie difference in the results of cimetiand placebo healing. Ulcers recurred among 48 percent of patients ed with cimetidine, but they recurred among only 9 percent of patients ed with placebo, suggesting that "while cimetidine 'heals' ulcers, platreatment can 'cure' ulcer disease" (Moerman 1983: 14). (However, ess pharmacologically active drugs Atrdcf from placebo effects, this result uzzling, since neither patient nor physician knew who was receiving ebo treatment, so thac whatever caused placebo curing should Jso have sed cirnetidine curing. Treatments that are pharmacologically effective uld always have efficacy beyond their own placebo effect .) Among the thirty-one studies he reviewed, Moerman noted chat the efacy of the placebo varied greatly, from 10 percent in one study to 90 pert in another. He also noticed that there was a variation in placebo efficacy country to country; rhe six srudies conducred in Germany had a signifintly higher rate of placebo healing (63 percent) than studies conducted in e other twenty-five countries combined (4 1percent). One might speculate at international di@erences in placebo efficacy correspond to cultural differces in beliefs about the power of medicine and its administration. The influence of cultural expectation is manifest also in the experience of in-of which the pain of childbirth is said to be one of the most excruciatg. Brigitte Jordan has studied be1iefs and practices surrounding childbearng in the Yucatan, Mexico; Holland; Sweden; and the United States (Jordan 993). From one setting to another, beliefs vary greatly abour the nature of birth itself and, more specifically, about the pain associated with labor, rhe need for anaesthesia, and the role of childbearing women in controlling the administration of anaesthesia.

>

'

82

AN ANTHROPOLOGICAL PERSPECTIVE

:.

While the experience of pain may be impossible to compare from one person to another, Jordan observed marked cultural differences in the display and apparent experience of pain. Among women attended (in their own homes) by a local indigenous midwife in the Yucatan, pain is expected, accepted, and responded to with empathy by companions selected by the laboring women. In Holland, where births are thought of as essentially healthy, and where healthy births are also atrended by midwives and a companion of the woman, it is believed that the woman's body "knows best"; anaesthesia is not used. In Sweden, where all births take place in technologically sophisticated hospitais, childbearing women expect anaesthesia, are informed o f available anaesthetic options, and themselves select methods they deem appropriate, as needed. Finally, in the United States, where birth is regarded as a medical event and therefore subject to control by physicians, childbearing women must demonstrate their pain in order to convince amending physicians of rheir need for anaesthesia. Lack of control in the obstetrical setting and anxiety about pain may heighten the experience o f pain itself. Jordan notes that "the experience of pain is observationally more visible i n U.S. obstetrical wards than in Holland, Sweden or Yucatan" (Jordan 199352). (For an analysis of : the development of anaesthesia and its control in U .S . obstetrics, see chap. 8 i in this book.) Cultural expectations have also been shown to affect the timing of major health events; a remarkable example is the postponement of death. David Phillips has recently shown that deaths among the Chinese in the United States from 1960 through 1984 were 35 percent less likely to occur in th week before and 35 percent more likely to ocnu in the week afier the Chines Hawest Moon Festival (Phillips and Smith 1990). Phillips suggests tha determination to see rhe holiday extends the life span. He has also found tha death rates for Jews but not for other groups decline in the week precedin and rise in the week following Passover, again suggesting the postponemen of death for celebration of an important cultural event (Phillips and Kin Z988). At a more personal level, Phillips (1972) has also demonstrated chat th time of year in which a person is born may be connected to the time of year his or her death. A sample of 1,25 1peopk in biographical dictionaries w 28 percent more likely to have died in the same monrh in which they w born or in the two folkwing months than in the two months preceding r month of their birth. Again one may speculate that people postpone ch deaths to witness an event important to them. (It would be interesting compare this phenomenon i n cultural settings where the uniqueness o f i viduals is less valued than the well-being of a social group, and where t events of an individual's life may not be celebrated as they are in the W e [White and Kirkpatrick 19851.)

Q F SOCIETY AND CULTURE

93

e suggests that cultural expeccarions can lead not only to the ofmajor health events but, under different circumstances, to elerntion. To distinguish salubrious placebo effects from noxious cultural expectations, Kissel and Basrucand ( 1964)coined the term m the Latin nucwe, "to hurt"). Nocebos are expectations of harmful 1 events that lead to the fulfillment of those expectations. 0 s infamous ~ of nocebo effects is voodoo death. In 1942, the physlter Cannon, better known for his discovery of the "fight or flight , surveyed the variery of voodoo death phenomena reported in somdicional societies (Cannon 1942): from Australia, where, among original peoples, pointing a bone at someone was said to induce & o f that person, ro Latin America and Africa, where belief that ewitched was also said to lead to rapid death. Cannon explained rrality as a result of prolonged and heightened emotional stress ivation of the sympathetic nervous system. He believed that the menon was far more likely to be found among "primitive people, of their profound ignorance and insecurity in a haunted world, than ducated people living in civilized and well-protected communities" n 1942:174). In contemporary Western societies, however, Engel 197 1) has described and analyzed the similar phenomena of the up-given up complex" and "suddenand rapid death during psychostress." A review by Gomez (1982) indicates the breadth of such ena in industrialized society, from faith healing to medical rounds in
"

unlikely that belief in the healing power of large doses of arsenic transform this chemical into a healing agent: yet I submit, though I not attempt to prove it, that such a belief would retard its lethal Likewise, lack of faith in antibiotics may diminish the potency of rugs, and faith or skepticism about materials or practices thought of lomedicine) as pharmacologically inert may shift the results in expected
acebos are the self-fulfilling prophecies of healing, nocebos the selfprophecies of sickness. . . The hypothetical 'effect is iliustrated in 1 . Expectations are represented on the vercical &s, outcomes on the tal axis. Expectations chat healing will occur are represented above horizofical axis, the stronger the higher; expectations that sickness will r are represented below the horizontal axis, the stronger the lower. ilarly, healing outcomes are represented ro the right of the vertical axis, more powerful the further to the righr; and pathological outcomes are resented to the left of the vertical axis, the more p o w e h l rhe further co left. The placebo/nocebo phenomenon predicts that, because outcome nts are dected by expectations in the expected direction, events of sicks and healing in human life will concentrate in the upper right and lower
"

Ethnolnedicd Belief and Expectation

(+I
Faith Healing

Outcome

(-1
[ Paradoxical

Voodoo Death

Fjga* 4.1 T h e placebo thesis: Rdations between expecrarion and outcome (From Hahn and Kleinman 1983b. Reprinted by permission.)

left quadrants of this graph. Events in the other two quadrmrs might be ./.:. .. regarded as placebo/ nocebo side effects.
.... ... ... .. . . ... ..... ... .. ..: . .... . .. ... .. ..... . ... .... .... ... ..... ...:.. .. .. .. ..... .... ... ..... . . . . ...... ..:: .... ..... . .. . . ... ....... ..... .. : , ...... ........ ....... .....;. ...... . .1:. ........ .. ..:
i

.....

BIOMEDICINE A N D THE CONSTRUCTION, MEDIATION AND PRODUCTION OF SICKNESS AND HEALING A central tenet of Biomedicine i s that human sickness is, in the~v,:$!i ...... if not in practice, essentially reducible to disturbances i n human biology a that healing i s reducible to the correction of biological disturbance. dominant model . . assumes disease to be fully accounted for by devi tions from the norm of measurable biological (somatic) variables" (En 1977:130). SociocuJtural effects on sickness and healing, if considered, regarded as peripheral and secondary---perhaps caused by but wit bout i portant causal role in human biology. Of the three sociocultural effects on sickness and healing i have scribed-construction, mediation, and production-only mediation is cepted in Biomedicine. Patrerns of human interaction and social organizat

01; SOCIETY AND CULTURE

95

ledged to play a pan in outbreaks o f infectious disease, in diseases "life-styk,"in clusters of environmental disease, in inreninjuries, and in differential access to and utilization are within segments of the population. Marcia Angell, a physician itor of the prestigious N e w Efighz~d]oumal o f Medicine, writes, s overwhelming evidence that certain personal habits, such as srnokettes, drinking alcohol, and eating a diet rich in cholesterol and fats, can have great impact on health, and changing our thinking se habits" (1985:1572). recognized risk facton, such as smoking and cholesterol, howlational eEects of human behavior and social organization may be to mockery and to unusual scrutiny in Biomedicine. Physician and wis Thomas comments in the New Engk~ndJournal o f Medicine n aa early analysis of the Alameda study described above. The n (Belloc and Breslow 1972; Belloc 1973) found that seven faceating breakfast; not snacking, smoking, or drinking excessively; ing regularly; maintaining normal weight; and regularly sleeping outs a night-each predicted enhanced longevi ry. Of these factors, accepts only not smoking or drinking excessively, because they fit e knows contribute ro several diseases and to automobile crashes. He s the other risk factors indicated in the study (which he refen to as g tennis" and "goingoEqn family picnicsp') as well as "thescience that ced this illumination.'' omas asserts thac the Alameda study fails to control for the participants' healrh status, which might account not only for the differential morbut also for the distribution of risk factors in the study population; for ple, people sick when initially interviewed might not ear well, sleep or exercise b ~ d % k they ~ e are sick. But Thomas does nor consider whether nitial health of the study population might affect the habits he believes mental as well as those he questions. Moreover, the researchers do in fact ol for the initial healrh status of study subjects, at least insofar as red by the subjects themselves. The seven behaviors are still associated lower mortality rates for all age groups; and the greater the number of e behaviors a person has, the longer he or she i s likely to live. he mockery of non-Biomedical beliefs is common. An editorial i n the England Jounral o f Medicine asserts that "holistic"approaches to healing be divided into those that are adaptations of traditional medical praces in other societies-Chinese, Navaho, and so forth-and those that re invented, so to speak, the week befote last by some relatively successful nk. In contrast, "medicine in industrialized nations i s scientific me&e" (Glymour and Stalker 1983:960). It is what I have referred to as the ~ociocultural production of sickness and aling , a more direcr causal effect of social relationships and beliefs, that urs the greatest skepticism in Biomedicine. Angell writes, "it is rime to
"

e :
:.

/'

AN ANTHROPOLOGICAL PERSPECTIVE

acknowledge that our belief in disease as a direct reflection of mentd state is largely folklore ( 1985 : 15 7 2). Attitudes toward and uses of placebos in Biomedicine are indicative of Biomedical disdain. Moerman points out (1983) that, if, as he found fur placebo efficacy studies, a drug were discovered to have an eflicacy ranging from 10 percent to 30 percent, Biomedical researchers would rush to discover what accounted for this range and, in particular, for rhe 90 percent success. However, Moerman points out, placebo responses are examined i n order t o control or minimize their effect in research settings; medical research focuses ( on efficacy beyond the placebo effect. Placebos are a nuisance. The Biomedical attitude toward placebos is also notable in clinical practice. A study of placebo use by physicians and nurses in a teaching hospital shows chat the placebo phenomenon is commonly misunderstood, its powers underestimated (Goodwin, Goodwin, and Vogel 1979). Placebos are often ,administered to show that a patient's pain is "all in his (or her) head"and that ithere is nothing "really" wrong. Sixty percent of physicians report giving placebos to ascertain whether the patient's pain is "real. Placebos may also be given to patients for whom standard treatments are not working or to patients thought undeserving of pharmaculogicaHy active drugs. It may be the combined power of the placebo phenomenon and its anomalous position in Biomedical theory that make it a highly charged matter. Physician and philosopher Howard Brody gues so far as to define the placebo phenomenon as an anomaly for the Biomedical perspective. In format fashion, he proposes, for example, that "the change in C {patient's condition] i s attributable to I [active intervention), but not to any specific therapeutic effect of I or to any known pharmacologic or physiologic properry o f I" (1977:d 1). The paradoxical consequence o f this definition is to make the placebo phenomenon forever incompatible with known pharmacotogical or physiological properties. As the phenomenon is recognized within pharmacological or physiological science, so it disappears; as it is known, it ceases to exist. It seems strange to define a phenomenon in terms of current ignorance, and more reasonable to regard placebo (and nocebo) phenomena as inexplicable by contemporary physiological and pharmacological knowledge, yet necessarily compatible and requiring revision of current physiological and pharmacological thoughr.
"

A SOCIOCULTURAL PERSPECTIVE ON SICKNESS AND HEALING The sociocultural perspecrive on sickness and healing suggests an understanding very different from that of Biomedicine, different also from Western thought in general. Nevertheless, wen within Biomedicine, there is a range of views, some of which are much closer than others to the socio-

F SOCIETY AND CULTURE

97

specrive. The specialty of family medicine, for example, recogportance of interpersonal relationships in pathology and healing; of psychiatry similarly recognize and treat patients for interpere fields of epidemiology and public health also t s in the distribution of sickness and health in populations. es similar to the one I propose have also been formulated by physicians in response to rheir own tradition. Brody and Sobel applied systems theory to the understanding o f sickness and ierarchical levels of interacting phenomena, incells, organs, and human bodies, but persons, biosphere. The elements within each of rhese ract; different levels also interacr with one anorher. Brody and e "disease"as "a failure to respond adaptiveiy t o environmental s resulting in a disruption of the overall equilibrium of the system" ). They define "health"as "the ability of a system (for example, cell, family, society) to respond adaptively to a wide variety of environllenges (for example, physical, chemical, infectious, psychologi' (1979:92-93). Brody and Sobel do not explain what they mean o what end, defined by whose goals? But, they al values and personal judgment in determining

systems approach has also been adopted by physician George Engel. in ark article, "The Need for a New Medical Model: A Challenge for icine" ( 1977). Engel outlined the fundamental assumptions of Bione and noced that, though the Biomedical model has yielded enorenefits, it is inadequate and has become a dogma in Western society. e now faced, " Engel wrote, "with the necessity and the challenge to to disease to include the psychosocial without sacrifice enormous advantages of the biomedical approach" ( 1977:13 1). In nse to this challenge, he proposed an alternative , "biopsychosocial" 1: "Toprovide a basis for understanding the determinants of disease and ing at rational treatments and patterns of health care, a medical model t also take into account the patient, the social context in which he lives, the complementary system devised by society to deal with the disruptive is, the physician role and the health care system" cluded psychosocial effects on susceptibility to verity, and course, as well as the parienr's experience toms. Medicine, Engel recommended, had to take
e biopsychosocial model with the case of a patient, r. Glover, " who suffers a myocardial infarct ion (Engel 1980). Engel examned the multiple events occurring inside and around Mr. Glover on chemial, physiological, psychological, social, and symbolic levels. He demonaonal quali ties--he was a hardworking, driven

man-affected his own response to early symptoms as well as the responses of chose around him, including his supportive employer. Mr. Glover's own response may have worsened his condirion, while his employer's response led Mr. Glover to seek medical help. Engel also described how the behavior ofrhe physicians in training who treated Mr. Glover exacerbated Mr. Glover's condition by nor listening to him or responding to his anxiety. Engel then showed how the biopsychosocid approach, in addition to responding to Mr. Glover's physical condition, would make use of information about his personal characteristics and his social circumstances to minimize his symptoms, respond to his personal concerns, and enhance his recovery. In this chapter, I have displayed a range of sociocultural effects on sickness and healing. I have not presented a comprehensive review (which would require several volumes) but rather have focused on evidence from diverse studies. The evidence indicates the power of cultural constructions in mediating the social distribution o f both exposures to sickness and resources for i t s relief. Evidence also indicates the more direct power of social relationships and cultural expectations in the production of events of sickness and healing. Societies and their relationships and beliefs sicken, kiH, and heal as well. The sociocultural perspective suggests that human societies, interpersonal relationships, and cultural beliefs are nor simply outside of and in the space surrounding us, but rather are embodied, literally, as a part of our anatomy and physiology (Hahn and Kleinman 1983b). There is a physiology of belief and of interpersonal relationships. Similar1y, sociocuttural phenomena suggest that the body is literally mindful, moving and functioning not only from biological but from cultural and social imperatives as well.