9
Leaning on the Cow's Fat Hump
Medical Choices, Unshared Culture, and General Expectations

Mtegemea nundu hawati kunona: He [who] leans on the cow's [fat] hump does not stop fattening. (The proverb is used when someone is seen to use his connections with others to gain desired ends.)

Shame provides one foundation for the differentiated cultural conformity essential to the effectiveness of statuses. The part played by statuses in evaluation and shame among the Swahili has now received some attention, but their involvement in cultural dynamics has not been specifically addressed. This chapter turns to that by examining one of the ways statuses make it possible for cultural elements to affect group members, including those who do not share them.

This transmission of culture's effects results from a type of relation among cultural elements that has received only limited attention as compared to what can be called the "intrinsic relations" among them. The latter sort of relations include the subset/superset relationship cognitive anthropologists have studied as it occurs in taxonomies (D'Andrade n.d.). Such relations among cultural elements depend on common internal qualities including their being concerned with the same thing. There are, however, many possible sorts of relations that link shared understandings together, and not all of them are directly based in the contents of the understandings themselves.

Cultural Organization

All the relations among cultural elements can be referred to as "cultural organization." There seems to be no limit to the number of kinds of relationships that can exist between understandings, and the same understandings—singly or in complexes—can be present in any number of relationships.

Whatever sorts of relationships there may be among understandings, no

understanding exists in complete isolation from all others in anyone's mind. This is true even as concerns remembering understandings as well as in considering or using them to guide action. For this, they must be related to one another at least with respect to priority, sequence, and whether they always entail one another, can never occur together, or can either occur together or not. In fact, the actual relations among understandings are far more complex than this last suggests, and much of this results, as will appear, from the social mediation of relations among cultural elements.

The shared or cultural organization of understandings, of course, is not necessarily the same as an individual's organization of understandings. Still, for people to behave with the predictability required for social life, they must have some common organization of the understandings that guide their behavior just as they must have some understandings in common. Thus, for example, if senior Swahili men are to greet people in a way acceptable to other senior men using the somewhat elaborate set of greetings they understand as appropriate, they must share not only a fair number of understandings concerning whom to greet and what to say but they must also be guided by them in similar sequences and settings.

The general importance of cultural organizations is not, of course, limited to such simple matters as greetings but extends across all of culture's scope. Below there is an examination of the elaborate understandings about body functioning, illness, and treatment shared by the practitioners of Swahili traditional medicine and the finely dove-tailed relationships among them. This organization unites a complex of intrinsically related cultural elements into an effective guide to behavior for the medical experts who share many or most of the understandings that comprise the complex.^[1]

Limited Sharing of Vital Understandings and Organization

Although a substantial majority of the community share few or none of them, the practitioners of traditional medicine and serious amateurs share a much-elaborated set of understandings concerned with body functioning. The general ignorance of these understandings is perhaps to be expected given the limited nature of cultural sharing in all domains and, especially, laymen's ignorance in technical matters. It is striking, nevertheless, that people submit themselves to treatment, often for conditions they understand to be serious and painful, according to understandings they do not share or, even, for the most part, know about.

In fact, two quite different sources for the influence of these esoteric understandings will be hypothesized. Both are dependent on cultural organizations,

albeit of dissimilar kinds, for their effectiveness. One of these is of the sort students of culture have long pointed out. It stems from the organizational fact that the understandings about the body and illness involve a fundamental cultural element that is also basic to social morality. The two sets of understandings are, or are part of, a "cultural pattern."

A Cultural Pattern: An Intrinsic Organization

For the experts who share both the medical and moral understandings to be discussed, the presence of a common, basic understanding in both gives the medical understandings a force they would not otherwise have. For most members of the community, however, this appeal can hardly be present since, as shown below, more than 85 percent of them do not share the experts' understandings about how the body works. Nevertheless, it will be shown that the "balance pattern" does have a positive, if indirect, effect on the acceptance of medical care from practitioners who follow the balance theory of disease.

Another and more pervasive source of the ability of medical understandings to guide behavior will be shown to involve other cultural elements that often have no necessary reference to medical issues. These other understandings are the general expectations that are characteristic of relationships of the sort referred to in chapter 7 as "multiplex." These provide the basis for an organization of understandings that directly guides many patients' behavior. These general expectations lead the patient to accept advice on medical care, and this advice links the patient's understanding that treatment is needed with understandings about what sorts of treatment are available and desirable. This organization is not based on intrinsic relations among understandings with interlocked contents as some organizations are but is mediated through understandings about actors and the statuses of those actors.

This organization of medical understandings through the agency of cultural elements guiding social relations is broadly important in organizing Swahili culture generally. The organizational contribution of social relations in a different domain, kinship and marriage, will be seen in the next chapter. Organizations based on intrinsic relations among understandings are, of course, also vital. These are the organizations that most readily call themselves to attention and that have traditionally received close anthropological scrutiny. One of them, mentioned above, serves to strengthen the medical understandings held by practitioners and serious amateurs.

To examine the two types of organization closely, it is necessary to consider the understandings about the body and medicine held by experts and, separately, those held by interested laymen.

Expert Understanding of Body Functioning and Illness

The traditional therapists (matabibu , sing. tabibu ) understand the basic cause of most illness to be an imbalance among the four characters or elements (matabia , sing. tabia ) that they take as fundamental to the body's functioning. This disruption is always due to improper diet, although other factors can exacerbate or lessen the disruption. The only effective treatment of the illness that inevitably results from disruption is understood to be dietary change supplemented by medicines, most of them compounded from herbs. The new diet and the medicines are aimed at reinstating the body's essential elementary balance and thereby restoring health.

The fact that the beliefs and values described in this section are shared among experts and not at all among the overwhelming majority of the community is crucial to an understanding of the cultural dynamics in this domain. Most community members have consulted "herbal doctors" (as I will henceforth call matabibu) several times or more, but this cannot be attributed to even sketchy knowledge of the basis for this medical view. Traditional medical care may be slightly less expensive than care from a university-trained physician, but many Swahili can well afford Western-type care. Although the basic understandings in Western medicine are shared no more than are those of traditional medicine, most people utilize Western medicine rather often.^[2] In fact, as an estimate, I would say that a majority of the community uses Western medical services more than traditional ones, even though most people use both.

Fundamentally, the "balance theory," as it can be called, followed by herbal doctors holds that the proper balance of the body's four elements, hot and cold and wet and dry, is indispensable to health. Illness results should any of the elements fail to make its appropriate contribution to the body's operation.

A number of the herbal doctors and expert laymen commented on the fact that the understandings concerning illness they follow come from outside their own society. Many attribute them (correctly) to the early Greeks and, more specifically, to the well-known second-century physician, Galen, who is called "Galeni" in Swahili. The Mombasa Swahili, like their fellow ethnics up and down the East African coast, see themselves as part of the great, worldwide Muslim civilization, and a number of them, including many herbal doctors, know that the cultural heritage of this civilization includes the view of illness held in their own group. The view that balance theory is shared by other Muslim peoples in other areas of the world is, of course, correct (Temkin 1973). There are differences between the Swahili view of balance theory and the same basic theory as held in other Muslim groups (see, e.g.,

Good 1977), but the existence of similarities and a recognized relationship is worth noting.

All the Swahili herbal doctors save one learned their profession in apprenticeships, most commonly with their fathers. The one who has formal schooling went to a Muslim institute in Lamu where he received a certificate for "general studies." He has far more elaborate offices than other herbal doctors, charges higher fees than they do, and told me that a group of visiting American physicians gave him $500 for a one-hour lecture on his approach to medicine.^[3]

Body Functioning and the Bodily Elements

According to the herbal doctors, the body's functioning is to be understood according to what I have called balance theory. Their main attention to discussing body functioning is focused on the four elements (also referred to by English-speaking informants as "complexions" and "characters"): "cold" (baridi ), "hot" (hari ), "dry" (yabisi ), and "wet" (rughtba ). In Swahili belief, the same four elements are present in everyone, but there are important differences among people in their relative amounts and in the balance among them.

Each of the elements is centered in a particular part of the body and is associated with a bodily fluid. Hari (hot) is centered in the liver and is associated with blood. People with a predominance of hari tend to be hopeful and courageous in disposition. Hari is more powerful in men than in women and more powerful during youth than in old age. Women rarely get diseases caused by an excess of hot because their menstrual flow protects them from this excess by lessening the blood supply when they are young. When they are old enough for their menstrual flow to stop, they are protected by their advanced age, with its diminution of hot. Men, however, are quite susceptible to hot diseases until aging brings about a lessening in their natural tendency to hotness and makes them more susceptible to diseases of excessive cold as women are all their lives.

Baridi (cold) is in the lungs and is associated with phlegm. People with a predominance of phlegm are inclined to be sluggish, dull, and impassive. Cold is more powerful in women than in men and in old age more than in youth.

Unlike the case for cold and hot, there is no association between gender and the rughtba (wet) and yabisi (dry) elements, but wet is associated with youth and dry with old age. Wet is associated with yellow bile and is located in the bile sac. A person with a preponderance of wet is likely to be proud, quick tempered, and generally given to anger.

Dry is in the spleen and is associated with what the Swahili call maji ,

which is the word also used to refer to water. However, informants specifically deny that this "water" is the lymph that fills blisters, and it is a near certainty that it is the "black bile" Galen associated with the spleen as opposed to the yellow bile he associated with the liver and bile sac (Siegel 1968:258). Those in whom dryness predominates are commonly moody, depressive, and suspicious.

The fact that the Swahili refer to the elements by the same names used for physical qualities of temperature and moisture might lead to the incorrect inference that they are, in fact, directly connected with these qualities. This mistaken idea might be strengthened by the fact that the hot element affects the body, for good or ill, most readily in the summertime, while the cold element is most effective in the winter. In fact, season is understood to affect body functioning indirectly by the influence of ambient temperature on how food is digested. Thus, the cold element tends to dominate the body in the winter and hot in the summer because the foods promoting hot are more effectively assimilated in the summer, while those promoting cold are more effectively assimilated in the winter.

In addition to variations due to sex and age, each individual has his or her own particular balance, which is significantly different from all others. A fanciful balance rendered, strictly for purposes of illustration, in numerical values for the elements might be 3:1/2:2 for one person, while another might be 1:3/3:1. One consequence of these differences is that individuals not only differ in character as a result of the different weightings of the elements in their balances but they are also more or less healthy depending on how stable their balances are. Differences in predisposition to particular diseases and kinds of diseases result from the same individual variations in balance.

When the influence of one or more of the elements becomes excessive, an imbalance occurs and is manifested as an illness. Only reestablishing the individual's customary elemental balance will end the illness. Although the season of the year has an indirect effect on the balance of elements, only aging, sex, food, and drink affect it directly.

The Classification of Food and Drink in the Balance System

All food and drink is classified according to the four elements, or a combination of them, according to the way they affect the body rather than to the nature of the foods themselves. Thus, ice is not cold; it is dry since it contributes to the operation of the element that is given that name. Honey, even if it is taken directly from the refrigerator, is hot because of the nature of its contribution to body functioning.

Edible substances, in the Swahili view, can have another property in addi-

tion to the four components of the body's elemental balance, and this is being either "heavy" or "light." This dimension appears to be similar to another aspect of Galen's system. Galen, like his Swahili successors, believed that in addition to the four primary "qualities," there were "secondary qualities" that modified the primary qualities. As Siegel, an authority on Galen, puts it, "Galen regarded all parts of the body as a combination of primary qualities, but modified by the addition of secondary qualities. Thus, the blood is red; bile is bitter and yellow, and because of some other secondary quality each exhibits a varying viscosity" (ibid., 147).

In Swahili understandings, "heavy" foods facilitate the effect of, say, cold less than "light" foods do. For some foods, being heavy or light is an inherent property. Others, including most meats, are neither heavy nor light in themselves but can become either depending on how they are prepared and on how long they have been stored.

The classification of foods and drinks in this system is exhaustive, with previously unknown foods or drinks being classified by their observed effects on those who ingest them. Most foods are classified according to being either cold or hot and also according to being either wet or dry. Some foods, however, are so strong along one of the dimensions (hot or cold or dry or wet) that their standing on the other is negligible. Cold and hot are more powerful in their effects on the body than dry and wet, with hotness being a definite cause of dryness (i.e., if there is enough hotness, wet foods will be converted to dry), and excess cold can cause otherwise dry foods to produce the reaction of wet ones. Neither dry nor wet, however, can produce either hot or cold. Despite this, wet and dry must be in balance quite as much as hot and cold must be, if illness is to be avoided.

A few common foods and their classifications can serve as illustrations of the system:

Corn: cold and dry, light in the stomach

Wheat: cold and dry, heavy in the stomach

Millet: hot and wet, heavy in the stomach

White beans: dry and cold, heavy in the stomach

Red beans: hot and dry, heavy in the stomach

Beef: hot

Goat meat: hot and dry

Chicken: hot

Fruit: all fruits having juice are hot and wet, heavy in the stomach

Since foods are classified according to how they affect the body, it is not surprising that their classification changes as the food substances do. Thus, many foods are understood to change their effects with time, so that fresh cow's milk is hot and wet, but if it stands for some time, it becomes cold

and wet, and if allowed to sour it is only wet without cold. Rice is hot, but if it is stored for a year or so it becomes dry. In the winter, bananas are hot and wet but in the summer cold and wet.

Similarly, the state, including size, or age of the source of a food is understood as affecting the food's influence on the body and therefore the food's classification. So, for example, the meat from immature chickens, that is, from hens that have not yet laid eggs and roosters that have not yet crowed, is hot, moderately wet, and fairly light in the stomach, but when the birds are older, their meat is hot, dry, and heavy in the stomach. In the same way, the flesh of large fish is understood to contribute to the hot element in the body but that of small fish to the cold. Even within the same food, the constituents can have different elemental standings, if these are taken to have different effects on the body. Thus, the whites of eggs are cold and wet, while the yolks are hot and wet.

Expert Understandings of the Causes of Illness

The herbal doctors and dedicated amateurs say that most community members have a substantial knowledge of "balance" and the effects on it of various foods. As the evidence shows (see below), they are wrong in this, but, for the experts, physical well-being is understood as depending on diet, with all other determinants distinctly secondary.

Chief among these secondary sources are trauma and supernatural agencies, but even these involve the bodily imbalance based on diet. Injury and spirit attack can cause illness, but they can also lead to imbalance, which may bring on further, often more serious, consequences.

For illnesses resulting from either of these nondietary causes, becoming well again requires, as one step, that the effects of the external causes be terminated by overpowering the spiritual agency or overcoming the traumatic injury. This is done, for example, by exorcising the dangerous spirit, pepo , if it is a spirit-caused illness; by manipulating the broken bones if there be such; or by stopping the bleeding if that is present. This termination by itself, however, may not end the patient's distress. What seems to be shock, following broken bones, or various continuing symptoms, following attacks by supernatural beings, are taken as a sure indication that, in addition to the directly observable results, the body's fundamental balance has been upset by the trauma or spirits.

Infection is understood to operate in a manner quite different from that posited by Western medicine. In balance medicine, although it is diet that is the original cause, the person who has been "infected" is ill not because of his diet but because of that of someone else who has become ill in the ordinary

way. This latter person's diet-based disease is transmitted to the one infected by it. The transmission occurs through small particles that leave the original patient's body and enter the second's through the openings in the skin where hair grows.

Diagnosis and Treatment

Regardless of complaint, the Galenic doctor's first step is to take the patient's pulse. This is done by putting the thumb, normally the right thumb but the left can be used, on the patient's left wrist. It is essential that it be the left wrist because it is understood that the major blood vessels that pass through the entire body are found near the surface only on the left side. By feeling the vessel in the left wrist, three vital diagnostic facts can be established: whether the force of the blood is strong or weak; whether blood flow is normal, fast, or slow; and whether the blood vessel itself is normal, thick, or thin.

A fast pulse and a thick-feeling vein are taken to indicate excess hot, since this condition is understood to involve there being too much blood. A slow pulse and a thin vessel result from insufficient blood and indicate an excess of cold, while too much dry is shown by a weak pulse and a thick vessel. Excessive wet, which is present when there is too much additional fluid (i.e., "maji") in the blood, produces a fast pulse and a thin vessel.

Patients are not asked to remove any clothing, nor is there any physical examination other than the taking of the pulse, but general appearance is noted with attention to paleness, flushes, obvious physical signs such as drooping eye lids, and signs of fatigue. After the patient reports his symptoms and is questioned about the location and nature of any reported pain, he or she is asked about appetite and diet, sleep, bowel and bladder performance, and changes in energy. The data collected in this way are used to establish the type of excess that is the basis of the patient's illness.

Once the type of excess is identified, the next step is to establish where in the body the excess is located so that treatment can be directed to that location. In all cases, therapy consists in overcoming the excess through contributing to the opposite element. Thus, excess hot is treated by increasing cold, excess wet by increasing dry, and so on. The medicines used vary according to the location of the excess, but the basic steps in the treatment of all diseases are similar and follow from understandings concerning the importance of the balance of the four elements.

When excess is located in the head, it is never necessary to cleanse the system since treatment does not involve digestion. The treatment of all other sorts of illnesses, however, often begins with the cleansing of the patient's system in order to ready the body for the changes that are to be made by the medications and dietary regimen that will be prescribed.

One of the herbal doctor informants said that the treatment of all illnesses not based in the head begins with a thorough cleansing in the way to be described in a moment. The other two agreed that cleansing is carried out in the way described, but they understand it to be necessary only sometimes and to be positively harmful when the patient is weak and/or old.

This cleansing is done by a single administration of a laxative in the summertime, when laxatives are understood to affect the body more fully, and of an emetic in the winter, when laxatives are understood to be only partially effective. The two methods of cleansing are seen as alternatives, and patients receive one or the other, rather than both, at the beginning of treatment.

There is only a single emetic in common use. A drink is prepared of one and a half "bottles" (roughly, liters) of water that has been boiled but whose volume has not been reduced and the juice of three limes. The water and lime mixture is allowed to cool and the patient is made to eat something (for obvious reasons it matters little what it is), and the mixture is drunk. The patient then waits for fifteen minutes when a twig 18 or 20 centimeters long and 1 to 2 centimeters in diameter with one end split and frayed into a brush, an mswaki (pl. miswaki ),^[4] more commonly used as a toothbrush, is put down his or her throat. This causes vomiting of a sort that is understood to include the entire content of the system, not just recently eaten food.

When the season calls for a laxative, there is a choice among a variety, with selection depending on the patient's condition. The components of the various laxatives interact and produce qualities in the compounds not present in the components by themselves. Although all laxatives are by their nature hot, there is substantial variation in the particular properties of specific compounds and choice among them is important to therapeutic effect. Like emetics, the basic purpose of the laxatives is to ready the system for medication and a new dietary regimen, but different laxatives have different effects and care must be taken to choose one that will not exacerbate the patient's illness. In some disorders due to excess cold, a properly chosen laxative can have therapeutic as well as preparatory functions.

It might be useful to consider a common illness and its treatment. The illustrative illness is one of a whole class particularly to be found in women and the elderly. It is understood to result from excess cold centered in the lungs. This type of disease results in weakness, loss of appetite, and coughing. Runny nose and difficulty in breathing are associated with some of the specific diseases of this type. Some of these diseases, including the one we will consider, can be fatal, especially in the elderly whose aged systems are understood to respond less well to attempts to reestablish balance.

A useful laxative for beginning treatment for this type of disease is compounded of cinnamon (sanamaki ), tamarind (ukwajuu ), and asafetida (halilaji ). To compound this laxative, the cinnamon, tamarind, and asafetida (using only the small white kernels) are soaked for twelve hours in water and then

boiled for five or ten minutes. The solids, after being taken out of the water, which is thrown away, are squeezed in a cloth producing a liquid that the patient drinks. A few spoonfuls to a quarter of a teacup is sufficient to attain this compound's desired effects.

The cinnamon contributes dryness, the asafetida hotness, and the tamarind wetness and coldness, with the result being a hot and wet laxative. This particular combination is both more effective and safer than others in cleansing the systems of all sufferers from excess cold. At the same time, it has some therapeutic effect in lung disorders, through its influence on balance in the lung area.

After the patient has had his system cleansed with the laxative, the full treatment regimen begins. Medications are given and dietary restrictions (miko ) are imposed. Every disease has its own dietary restrictions, which are viewed as being as important in treatment as the medications are. The restrictions and medications are intended to work together in restoring the body's balance through what might be thought of as "an oppositional strategy." At least part of the treatment of a cold disease, for example, involves giving the patient hot medicines, keeping him or her from eating cold foods, and, perhaps, encouraging the eating of hot foods. Treatment is not as simple as this sketch suggests, since the patient's age and unique personal balance, the season of the year, and the effects of wet and dry also have to be considered, as does whether the foods and medicines are heavy or light in the stomach.^[5]

The common lung disorder to be considered is called balghamu and is understood as resulting from an excess of wet abetted by an excess of cold. A patient suffering from balgamu has a fast and weak pulse in a blood vessel that feels thin. He or she will complain of a headache on the right side of the head only and will have a cough that produces phlegm (makohozi , sing. kohozi ). When the patient awakens in the morning, there will be white matter in the corners of the eyes which is wet rather than hard. If the condition indicated by these symptoms is left untreated, it can result in paralysis.

It is generally characteristic of patients with balgamu to talk about the past and to tell stories of things that happened years ago. This behavioral pattern, like the illness it is associated with, is caused by excess wet. The wetproduced behavior varies across a wide range, from being so mild and similar to the patient's usual behavior that it is not noticed by relatives to being so extreme that it must be treated before or concurrently with the treatment of the physical manifestations of the disease.

The three herbal doctors agree that for these symptoms an emetic is given if the illness occurs in the winter and a laxative is given if it occurs in the summer, but two of them say that the cleansing must be delayed until recovery has begun if the patient has been weakened by the disease's progress before treatment. All agree that following the emetic or laxative, the patient is put on a restricted diet that excludes beef, apples, grapes, and all spices save

vinegar (siki ). In the early stages of the disease, the laxative and the regimen alone sometimes lead to a rapid cure without further treatment, but in more advanced cases, medication is needed.

The medication for balgamu is compounded of three components: three pennies' weight^[6] of fennel seed (shimari ), an equal weight of ginger (sanamaki ), and six pennies' weight of rock sugar all ground to a powder. The three ingredients are mixed together, and the patient is given one soup spoonful three times a day. This treatment is viewed by informants as usually effective save in very advanced or difficult cases that have progressed to paralysis. For these last, different treatments focusing on dry herbs that also have cold properties are called for, with their exact formulations differing from case to case and established on a trial-and-error basis.

The intrinsic relationships among the understandings concerning balgamu and its treatment are now easily seen. The fundamental understandings involved concern the supreme role of the balance among the four elements in body functioning. These understandings are connected to understandings about what to do when the body is malfunctioning by simple extensions or implications: basically, the nature of imbalance must be discovered and the balance reestablished by adding elements opposite to the excessive one.

Morality, Illness, and Organization

The intrinsic organization of understandings focused around balance is not limited to those concerned with the body and its functioning. It is also found for understandings concerning social morality. A key requirement of proper social relationships is that they be balanced. This balance, like the one in the body, results from different but complementary contributions from the elements in the processes of interest.

The idea that there are important connections between understandings concerning misfortune, including illness, and the values held in the sufferers' group is well established (e.g., Douglas 1975:22–24). In the classical formulation, misfortune is experienced by those who suffer it in terms of the value system of their group (Weber 1922). The classical study of this in anthropology is, of course, Evans-Pritchard's (1937) study of the Azande. His central point is that an explanation of misfortune based on the physical facts alone is considered incomplete by the Azande, despite their being fully alive to natural causation. For them, the moral system embodied in the understandings about witches and their doings is essential to any such explanation.

The type of connection between the Swahili views about illness and their views about an area of considerable moral concern, that is, social relationships, is quite different from the Azande. For the Swahili, the occurrence of a type of misfortune, illness, is not caused by evildoers^[7] as it is for the

Azande. Instead, understandings from one domain of morality are related to understandings about an unrelated type of misfortune, illness, by a shared emphasis on a common desideratum: balance.

My objective here is to show how this cultural organization actually works and to examine the part it plays in the continuing ability of the various understandings involved to guide behavior. Only the herbal doctors and a relatively few serious amateurs share the understandings about body functioning. Nevertheless, it will be shown that, like the unshared medical understandings themselves, the connection between understandings about what is proper in social life and about the body's functioning affect most or all of the group's members.

Balance as a Desideratum in Social Relations

The importance of balance in Swahili social relationships is most evident in the emphasis on reciprocity. In relations between individuals who are not equal, as between a parent and a child or a senior and a junior, the senior must give advice, guidance, and, in some relationships, material assistance. In return, the junior must give respect, deference, and obedience. When each of the participants does the different but complementary things he or she is called on to do, the relationship is understood by Swahili as being a "good" one that is likely to continue and to benefit its participants and those involved with them.

When the junior participant fails to respond to the advice and assistance of the senior or fails to show respect, however, the relationship is seen as "bad," and the junior is thought to harm him or herself as well as the senior and anyone else affected by the relationship. Specifically, the senior is in danger of having the respect in which he is generally held harmed by the junior's behavior, while the youth can lose standing as an upright member of the community and come to be looked on as a sort of hoodlum. No one will respect a senior, informants told me, if he accepts being treated badly by a junior instead of ending the relationship.

The same undesirable situation can arise from the senior failing to behave with proper balance toward the junior. A senior who shows excessive respect to a junior or who treats the junior as though he or she were a senior harms both himself or herself and the junior. The failure to assume the behavior appropriate to a senior puts the senior's own overall status in doubt and endangers the junior by undermining (kumtimba , lit. "to dig [under] him/her") his or her standing and, even, physical welfare. By treating a younger person as though he were older, the senior implies the junior may die first and, through the same sort of supernatural means involved in the evil eye (see

chap. 7), may actually bring about the junior's premature death. To preserve his prestige and physical welfare, a junior in such a relationship must withdraw from it.

Nor are balancing but different contributions expected only between juniors and seniors. Among equals strict reciprocity is called for, with giving by one participant to be followed immediately by grateful taking and, later, by return giving. Whatever one gives the other, whether it is material, social, or emotional, the other should accept it gratefully and return its equivalent in the future. To fail to do so is aibu, shameful, as judged by most of the peers of those involved, just as is failure to conform to prescribed, complementary behavior in relations between juniors and seniors.

In Swahili values, then, the only lasting kind of relationship, the only one beneficial to its participants, is one where each makes distinctive contributions that are complementary to those of the other. This can be seen rather explicitly in several Swahili proverbs that recognize the inevitability of reciprocity, demand it, or bemoan its absence.

Mpaji na mpokezi, mtahamali, nani? A giver and a receiver, who is burdened [more]?

Unajua lete, jifunza twaa : You understand "give," [now] learn "take."

Ulichokula ndicho ununuacho : What you ate is that which you bought, that is, you are bound to return anything given you.

Bure ghali : Free things are costly.

Changumi chakove. Chakove sumu changu mimi : [What is] mine [is] yours. [What is] yours [is] poison for me.

This emphasis on reciprocity, on the importance and (sometimes tiresome) inevitability of participants making equivalent contributions in social relations, is clearly akin to what I have been calling "balance" in body functioning. The understanding that reciprocity is essential to rewarding social relations derives part of its force from understandings learned in early life. This learning concerns such primary issues as controlling the body and winning the approval of the parents and others on whom the infant and child depends.

The child finds that performing his bodily functions as his mother wants him to is more often followed by his mother doing something he wants her to do than is not following her wishes. More generally, the child finds that if he does what he should, those who care for him are likelier to do what they should. This learning has early manifestations, as in quite young children insisting that their playmates "be fair," give them their turn, and return favors. Children are also to be seen pointing out their own good behavior to parents as a reason for their being given what they want.

Swahili have a special term, ngoa , referring to the justified feeling of deprivation a person, particularly a child, has on failing to get what the expec-

tations in his status indicate are rightly his or hers. If parents give one child a gift and do not give a similar one to the others, the others will experience ngoa, which, unlike the hated emotion, husudi, jealousy or envy, is viewed as right and proper. Envy stems from wanting what others have to which you have no right, while ngoa is based in your right to get as much of what is due you as others in your category do. The emphasis on reciprocity and balance has an important part of its foundation in the opposed values that each person should have what is rightly his but should not have what is properly another's.

Balance in Understandings about the Body and about Social Relations

A comparison of the moral values and beliefs that underlie social relationships with those concerning the body and illness shows interesting similarities. The body's elements each make a distinctive contribution to the functioning of the whole, and these different contributions must complement each other, must be in balance, or illness results. When that occurs, the condition is rectified by restoring the proper balance. If that is not possible, the body will eventually cease operating, that is, the patient will die.

The "elements" in social relations are people, not hot, cold, wet, and dry, and their contributions are, of course, different in nature. But the similarities between the understandings concerning the proper course of social relationships and those concerning the body and illness are striking. So similar, in fact, that it is hardly bold to hypothesize that a view of body functioning and illness that stresses the same sorts of relationships and the same kinds of values as those at the heart of social life is likely to be appealing to the holders of the social values. Holding these latter values is neither necessary nor sufficient for accepting the understandings that make up the Galenic theory of body functioning, but sharing the values is conducive to that acceptance.

It is possible that the correspondence of relationship values and body functioning theory is an illusion resulting from imposing on the two cultural complexes a similarity that is alien to the people themselves. That this is not likely can be seen by considering a serious sort of breach of propriety in social relations. This breach involves the same kind of activity harmful to healthy body functioning, and the same term is used to describe it: mizani . This term refers to a scale, either a balance beam or, sometimes, a spring scale, but it is also used to mean "measure," "appropriateness," and "good sense" in social relationships. The social reference of the term is always, so far as I know, to note the absence of the qualities referred to.

A person who says things that are inappropriate to a relationship or social

setting is said to be without mizani. Thus, a man who mentioned sexual activity in the presence of senior men is without mizani, as is a person who discusses family matters outside the family circle. Like the welfare of the body being threatened when an element makes an excessive, that is, inappropriate contribution, the course of social relations are imperiled by a participant who ignores "measure."

This interpretation is strengthened by the fact that when herbal doctors refer to the body's balance, they use either the Arabic word muutadil or, more commonly, that same term, "mizani." When I mentioned the common role for balance in the two domains, experts in Galenic medicine found the idea slightly surprising but quite plausible.

Patterns, Nonsharing, and Cultural Organization

The common presence of the understanding that balance is vital for proper functioning unites the social and bodily domains in what can be thought of as a single cultural organization. Benedict's "patterns" (1934) as well as Opler's "themes" (1945) are based mainly or entirely on just this source of relationships between different cultural domains. A well-known example of this is Benedict's finding of common elements in the understandings that guide ritual and in those concerned with personal gratification. The same sort of organization of cultural elements has been identified by more recent researchers, as, for example, Geertz's tracing of the understanding tjotjog (to fit) in quite different complexes concerned with medicine, arithmetic, and a number of other domains of Javanese culture (Geertz 1973a :129–130).

This "common element" organization seems likely to increase the probability that the understandings in the different sets will actually guide behavior. The fact that an understanding is taken as important in one area of life does not necessarily affect its standing in another area, but it probably gives it an inevitability it would not otherwise have. When the understanding has sacred or value-laden connections in one of the areas in which it occurs, as with the Swahili understanding that balance is essential to proper social relationships, its obviousness and importance in another area is probably increased. This seems to be what Benedict, Opler, and others imply when they identify patterns and suggest that cultural conformity—whose presence is implied by a ubiquitous "style" of behavior in different domains—is enhanced by their presence. It is surely true that for those who share the Galenic views about the body as well as the commonly held values concerning social relations, the importance of the balance understanding in both gives the Galenic complex an appeal it might otherwise not have.^[8]

Laymen's Understandings about Illness and Body Functioning

As table 13B shows, more than 85 percent of those interviewed had not heard of any substantial part of the Galenic scheme. Eighty percent were without most of the understandings that are basic to the Western scheme. This being so, an organization of the understandings connecting those concerned with the body to those concerned with social morality through the common presence of a belief in the positive effects of balance is not possible for most of those interviewed.

The forty people in the survey provided information through conversational interviews. That is, they were not asked questions with multiple choice responses but were encouraged to discuss freely the issues brought to their attention. After saying that what was wanted was their views on how the body worked and what was mainly responsible for illness, they were encouraged to say whatever they wished.

If they had not mentioned the four bodily characters or elements fundamental to the Galenic view, they were told the names of the elements and asked to comment on them and what, if anything, they had to do with body functioning and illness. Those who did not mention the elements and how they worked in their "free" responses and did not identify two or more of the elements when their names were mentioned were taken not to have "heard of the elements in the Galenic scheme." If, however, their discussion of illness showed that they had a clear idea that particular sorts of foods had particular kinds of effects on the body, in the general way Galenic understandings indicate, they were taken as having heard of the elements of the Galenic scheme even if they did not mention or recognize the names of the elements.

In discussing the Western view, a basically similar approach was taken following from the same question about body functioning and illness. Informants who did not mention them in their spontaneous remarks about the sources of illness (even if they did mention the Galenic elements) were asked to identify presha (high blood pressure), tell what organs were involved in it; identify ambuzika or wito (infection) and comment on its working or origin; discuss afia nzuri (good health) with attention to what promotes it and what interferes with it; and discuss food with attention to its role in health and body weight. Those who spontaneously mentioned any two of the following six terms or otherwise indicated some knowledge of them were taken to have some understandings from the Western scheme. These six were "vitamins," "calories," "cholesterol," "virus," "bacteria," and "germs."

Whether they mentioned two or more of the six terms or not, understandings from the Western scheme were taken as present in those who were able to discuss blood vessels (mishipa ya damu ) in a way that suggested a dis-

tinction between arteries and veins or, in discussing the heart, indicated they were aware of the heart's pumping action or muscular composition. All those who participated in the study were given as much time as they wished to take to discuss illness, health, and the body's mechanics, and all were encouraged to comment on both Western and Galenic views.

Despite this encouragement, less than 15 percent of the sample indicated familiarity with the understandings basic to Galenic medicine and only 20 percent with the Western scheme. Given this lack of familiarity, a reasonable inference is that they could not choose among types of medical care on the

basis of selecting therapists who shared their views of the body and its workings.^[9] Not only were most informants immune to influence by the common importance of balance in social morality and the Galenic scheme of body functioning but they also could not be affected by whatever understandings in the Western approach might make it attractive since they did not share the elements of that scheme either.

In fact, patients usually consulted therapists without regard to whether or not the patients' own understandings were taken to have any correspondence to those they might attribute to those therapists. Table 13C shows that of the thirty-two patients and former patients who have consulted an herbal doctor at least once, twenty-seven (84%) share few or no understandings about the Galenic scheme followed by those therapists. Of the thirty-seven interviewed who have consulted a hospital doctor, thirty (81%) shared few or none of the understandings involved in Western medicine.

The situation regarding knowledge of medical understandings is similar in an important way to what Willis (1972) found among the Fipa in southwestern Tanzania. There the medical practitioners have an elaborate set of understandings about illness resulting from injury to social relationships as a result of the activity of spirits or sorcerers. Laymen understand illness differently as being the result of sorcerers poisoning food and drink because they are jealous of their victims. As with the Swahili, Fipa laymen and experts do not share understandings about illness. There is, however, an important difference between the Fipa and the Swahili laymen: Willis reports the former as having a fairly elaborate set of understandings about the causes of illness,^[10] while the Swahili laymen have no such set.

In some ways, the situation is closer to that Keesing (1987b ) found among the Kwaio, where experts have a rich set of understandings about dying and going to the land of the dead, but those not especially concerned with such matters do not share these. The differences he found between experts and laymen led him to say, "Such diversity of knowledge in religious matters is perhaps not surprising, but it seems to me to render deeply problematic premises about culture as system [sic ] of shared understandings" (ibid., 163).

This is not to suggest that the Swahili patients interviewed had no understandings at all about illness and the working of the body. They did have what Roy D'Andrade suggested was "default knowledge" in that they rejected some suggestions as to how disease occurred and accepted others as possible (pers. commun.). For example, I mentioned to several young and early middle-aged informants that during my first year of visiting Old Town, twelve years prior to talking to them, I had been told by elderly Swahili that many people believed that paving Mombasa's streets with asphalt caused illness by reflecting the sun's heat up and into people's bodies. None of those whom I told about this thought it was likely, and several of them thought the idea was laughable. At the same time, my account of having been told by the same elderly

informants that oven-baked white bread (bofulo ) and other Western foods caused illness was heard with interest. All agreed that such food might be harmful, although none save the few who shared Galenic understandings offered explanations of how it might produce disease.

Cultural Organization and Guidance: Choice without Understandings

For more than three-quarters of those interviewed, the decision about what help to get must be made on a basis other than a personal set of understandings about what sort of treatment would be most likely to produce desired results since they share few understandings about treatments and how they work. There is a good deal of sharing, however, of understandings about the signs of being ill. These include pain, sleeplessness, weakness, fevers, loss of appetite, changes in eliminatory activities, and, occasionally, aberrant behavior. There is, people say, some judgment involved in deciding one is ill because sometimes these signals can occur and persist without anything consequential following. Most people say that they come to the understanding that they are ill from the severity of the initial symptoms, from their persisting for an extended period, from the similarity of the symptoms to those leading to familiar illnesses, or because someone—usually a member of their household—convinces them that they are ill.

There is also fairly uniform sharing of the understandings that indicate a variety of types of medical help is available. Even though all believe that expert medical care for the sick is often needed and that many kinds are available, most lack a schema that directs them in what choices to make once they decide they are ill. Many have no understandings of their own which make it obvious that one sort of care is to be preferred above another.

In situations where there are no perceived alternatives, where all medical help is viewed as "the same" or different but in unclear ways, the individual need only understand that competent help is needed and where it can be got. When, however, there are sources of help that are taken to differ in important ways, the individual must be able to choose among them to avoid suffering the fate of the fabled donkey between equidistant and equally attractive bales of hay. As table 14A shows, there is a strong basis for believing this problem is present since the Swahili are aware of the availability of quite different types of medical care but understand little of their nature or specific differences.

Not much help in choosing among available therapists is provided by language. Many of the most commonly used names for illnesses are quite inclusive. Homa can be glossed as "fever" and applies to all conditions in which the patient has an elevated body temperature as evidenced by sweating,

being warm to the touch, and, sometimes, having chills. Kohoa refers to a cough and is used for all illnesses involving that symptom. Kitwa , a word meaning "head," is used for any disorder involving head pain; pua , nose, is used for the symptoms often described as a cold in the United States; and all chest pain or difficulties in breathing can be called kifua , chest, or pumzi , lungs.

It is notable that none of these disease terms carries with it any intrinsic suggestions for action other than to seek medicine identified by the name of the illness. Thus, if one has a headache, one may buy dawa ya kitwa from a Western drugstore or one of the two herbal medicine shops. Even this, however, depends on deciding between the two kinds of medicine, and some basis for this decision is necessary. The decision could be—and sometimes is—made on the basis of the distance to the shops selling the different types or on their relative price. This is likely to be effective, however, only if the medicine chosen performs as the patient hopes it will. If it does not, the patient must decide what to do, and this may lead him or her to seek guidance. Lacking understandings about how the illness comes about, this guidance must be derived from sets of understandings of other kinds including, especially, the advice of those whom the patient trusts.^[11]

There are names for specific conditions, such as baridi kuu (lit. great "cold" with the last word referring to the Galenic element) for stroke, wito for infections, presha for high blood pressure, kipindupindu for seizures and cholera, and moyo (lit. heart) for all heart diseases. These last, however, are experts' names and become known to many only after they or those close to them have undertaken treatment for them with herbal or hospital doctors.

Illness and Treatment

Once an individual has decided he or she is ill and that the illness is not going away by itself, the obvious issue is what to do about it. This second decision is almost always a choice among the following: do nothing and continue to wait, buy medicine from a shop, see a medical practitioner of some kind, or seek advice from someone. If the decision calls for anything other than waiting, further decisions are called for: what type of medicine, advice, or consultation to seek; which particular dispensers of these to use; when to make these visits; how much to expect from them; how much money will be called for; and, probably, others.

The decision about seeking help is partly based in the fact that a wide variety of kinds of medical care are available in Mombasa, and almost everyone knows about most or all of them. As table 14A shows, every one of those interviewed knew that there were at least two sorts of medical care available: that from herbal doctors and that from hospital doctors (daktari ya hospatali,

as they are called). Most also knew about "spirit doctors" (usually referred to as waganga, sing. mganga ) who, most informants agree, are also part of the Swahili "tradition." These practitioners control their own majini (sing. jini ), that is, spirits mentioned in the Koran, who help them stop the bad spirits that are causing the illnesses. As of 1988, there were at least three such spirit doctors practicing in Old Town.

Another generally recognized type of medical practitioner is from other African ethnic groups and is also called mganga, as spirit doctors are, as well as mchawi (pl. wachawi ). Such practitioners seem generally, but not exclusively, to be consulted by those who suspect sorcery and, thus, seem to be chosen according to intrinsically organized schemata.

A final category of curers can be formed of the more specialized practitioners, including midwives (mku'nga, pl. waku'nga ), bone setters (mkandaji, pl. wakandaji ), and teachers who treat illnesses with prayer (Mwalimu, pl. Walimu ).^[12] I did not systematically ask the forty informants about these specialists, but more than a quarter of them mentioned them and eight reported consulting one or more of them.

Medical Care and Advisers

Choosing among this variety of kinds of care is, unavoidably, on the basis of the organized understandings the individual has. Despite limited sharing of cultural elements, neither the understandings nor their organization is completely idiosyncratic. Table 14B shows that the original decision to visit either an herbal doctor or a hospital doctor was usually the result of advice rather than direct guidance by personally held understandings. Of the thirty-two who have visited herbal doctors, only two report that they did so on the basis of their own knowledge of what was wrong with them and what sort of practitioner was likeliest to be able to deal with the problem.

A substantially larger number of patients decided to visit a hospital doctor on the basis of their own understandings of what to do for their illness, but it is still only six out of the forty in the sample. Again, the largest part, around eighty percent, made the decision to seek this type of care on the basis of advice received.

Advice in Multiplex Relations with General Expectations

As shown in table 14C, the adviser was a parent for a substantial majority of those who lived with or close to one or more parents. Next most common as a source of advice was a spouse with neighbors, kinsmen, and persons with

similar illnesses (three of these were co-workers) taken together as the final, and smallest, category of advisers. The important common feature these advisers all have is that their relations with the patients were of the sort I have been calling multiplex. The responses to my questions about why the advisers were chosen and/or listened to all indicated that the patients accepted the advice because they viewed those who gave it as being truly interested in their welfare.

In almost all cases, the advisers had relationships with the patient that went far beyond medical matters and involved mutual expectations of a broad sort in many domains. This is unquestionably true of the parents and spouses, of course, and is hardly less true of the assorted kin and neighbors informants mentioned as advisers. Even the co-workers, in the few cases where they were the advisers, seemed to be understood by the patients as having a general concern about the patients' welfare, and it is probably more accurate to refer to them as friends. In the terms used in chapter 4, the expectations in all these relationships were general rather than specific and the relationships were multiplex rather than simplex.

In addition to being subjects of general expectations, the advisers were often characterized as "knowing about illness." This special knowledge was often seen as deriving from the adviser having had a similar illness or being familiar with it through a third person who did. In some cases, the adviser was characterized as having a good deal of knowledge about medical matters generally, and a few of the advisers actually gave medicine or other treatment to the patient as an early step in treating the illness.

The belief in special knowledge or experience does not obviate the importance of the generalized expectations in the relationship between the patient and her adviser; it supplements or focuses them. That is, patients usually have generalized expectations of help and concern of a number of people and they choose among these according to which of them presents herself (the majority of advisers are women, as were the majority of my informants about this) as willing to help and as particularly able to do so.^[13] As would be anticipated given the importance of generalized expectations, in none of the cases was specialized knowledge or experience the sole basis for seeking and/or accepting medical advice. Thus, several of the patients reported that in addition to their advisers, there were people (almost always women) in their neighborhoods with whom they had no close ties but who had the reputation of knowing about medical matters. None of the informants reported going initially to these knowledgeable, relative strangers for advice, although several seem to have indirectly been influenced by them through their direct advisers talking with them.

It might seem likely that advisers had at least some medical understandings that many patients did not since the patients viewed them as qualified to give advice in this area. As seen in table 15B, however, only about half of those

who had advised someone to consult an herbalist had substantial understandings about the Galenic view of the body's elements and in what ways illness was helped by the sort of treatment the herbalists gave.^[14] Only a third of those who recommended Western medical care had substantial understandings about that view of the body and treating illness. The special knowledge the

advisers had, in many instances, was the result of their experience in consulting practitioners rather than their having understandings of their own about how the body works and how illness comes about.

The interviews with advisers included discussion of all the advice they recalled giving rather than only the advice they had given the patient I interviewed. Thus, the twelve advisers told me about twenty instances of recommending therapists rather than only twelve.

Table 15C shows that the advisers recommended hospital doctors slightly more often than they did herbal doctors. Those who reported more than one recommendation—four told of two each and two of three—divided them almost equally between herbal and hospital doctors, with only two of the six multiple recommenders referring people to only one kind of therapist. A major appeal of hospital doctors, the interviews make clear, is that such care is free and provides treatment quickly. Herbal doctors have to prepare the medicines they give, which sometimes takes hours or days while the herbs are found (generally in the same shops where laymen buy them) and compounded. In contrast, as informants say, the hospital doctor simply fills his syringe and uses it.

When the advisers recommend herbal doctors, they report, it is either because hospital doctors have already been tried and have not produced the desired results or, for some of the advisers, because they view the Galenic approach as superior in some or all kinds of illnesses. One adviser told me that hospital doctors get faster results but that their cures do not last. "Real" cures, in her view, came from herbal doctors, and she recommended hospital doctors only for quick relief.

Advisers' Understandings and Experience

What is particularly notable is that more than half of the advisers (see table 15B) shared few or none of the Galenic understandings and three-fourths shared few or none of the Western ones. Despite this, most of the advisers had recommended both sorts of treatment at one time or another, with hospital doctors being recommended slightly more than herbal doctors despite the slightly greater sharing of understandings about herbal doctors.

In fact, as table 15D shows, about half of the practitioners recommended were originally discovered through someone telling the adviser about the type of care in question. Advisers report that they discovered the usefulness of hospital doctors on their own as often as they discovered herbal doctors in that way. This finding, seemingly at odds with the fact that more advisers shared some understandings about the Galenic scheme than had comparable understandings about the Western scheme, suggests that advisers credit their own

experience even if they have little basis for understanding why that experience was positive.

There can be little question that getting desired results is a central element in recommending that therapist to someone else. The question remains, however, how patients choose therapists in the first place, before they have experience with them. The existence of hospitals and doctors' offices with signs plays a role quite apart from elaborate understandings about the body and illness. People can decide to visit a hospital doctor because they have seen the hospitals they work in and, perhaps, because they are impressed by the number of people going in and out, the seriousness of the enterprise as indicated by erecting a building, or simply because they know that something concerning the treatment of illness is available. Similarly, they can visit an herbal doctor because they have heard it said that the care given is traditional, because the visit is less intimidating than going to someone in a white coat speaking English or standard Swahili,^[15] or some other such consideration not necessarily based on the understandings involved in the Galenic scheme followed by the herbal doctors.

Social Relationships and Plans of Action: Cultural But Extrinsically Organized

For there to be a plan of action, or schema, enabling sick people to get medical care, they must be able to connect the understanding that they are ill with their understandings about the existence of a variety of types of medical care. We have seen that people share understandings about being ill and also about the existence of a variety of kinds of medical care. For the plan of action in dealing with illness to be cultural rather than idiosyncratic, however, the connections between the understandings about being ill and those concerning the existence of medical care must also be based on shared understandings.

Were the connection based on intrinsic relationships among the shared understandings, as are those, for example, that govern the playing of chess, the schemata would be culturally constituted in all its respects. In chess, the movement of the rook and the protection of the king are connected through understandings of how pieces move and how the game proceeds. For most people, such an intrinsic relationship between understanding one is ill and seeking help for the illness is not possible, since, as seen, they lack the medical and physical understandings such connections require. But this does not prevent them from having culturally constituted schemata, nevertheless.

These schemata are produced by connections between the understandings of the presence of illness and the availability of care which are shared but

not mainly concerned with medical care. For most community members, the advice of others is the basis for forming a schema that takes them from understanding they are ill to choosing among therapies and therapists. It is also true that roughly half of the advisers interviewed were themselves advised about the kind of care they recommended. The advisers got their advice, as the patients did, from people with whom they had multiplex relations. Thus, for many who give advice, as for those who receive it, the decision about what kind of medical care to choose involves a social relationship as a crucial link in the organization of understandings that guide their behavior in giving advice.

Some of the patients and advisers formed intrinsically based action plans after they had had experience with various medical experts. Their understanding about what was currently wrong with them contained within it an indication of what sort of therapist to consult on the basis of similar symptoms having led in the past to successful therapy from particular practitioners or kinds of practitioners. This, however, does not change the fact that in many cases, the original plan of action was founded on advice sought and/or accepted on the basis of general expectations in the relationship with the adviser, with these expectations not being limited to medical matters.

The use of advice to form plans of action led to what might be thought of as social relationship-based schemata. Nor were these social relationships always limited to those between the patient and her adviser. I did not investigate where the advisers' advisers got their introduction to the type of care they recommended, but it is very likely many of them got it in the same way the advisers and the patients did: from consulting people with whom they had multiplex relations.

General Expectations and Cultural Organizations: It isn't What You Know

The cultural elements concerned with medicine have just been seen to affect people who did not share them. A substantial proportion of the patients sought treatment from therapists whose understandings about the body and illness were utterly unknown to them. Even those who advised patients as to what care to seek did not, in a majority of cases, share the medical understandings of those whom they recommended. The direct effect of the medical understandings, of course, comes through the behavior of the therapist who is guided by them in giving treatment, so the key issue concerning how culture provides means for dealing with the effects of illness is how these therapists are chosen.

The choice of therapists was on the basis of the patients' expectations of others (mainly parents, spouses, and neighbors) rather than according to un-

derstandings of how the therapists understood illness and treated it. These others, in turn, at least sometimes gave their advice on the basis of their expectations of yet others with whom, almost always, they participated in multiplex relationships.

The transmission of culture's effects in the medical domain, then, is a social structural phenomenon. Patients submit to treatment as a result of the understandings that connect them in their statuses to others in theirs. Getting medical care is dependent most directly on the effectiveness of the understandings that connect patients and advisers, and also advisers and their advisers, rather than directly on the understandings that direct treatment itself. What gives these connecting understandings the ability to operate as they do is an obvious and important question.

Why Advice is Followed

The advisers are older than the patients in almost every case. On this basis alone, the formers' advice is to be taken more seriously than that of younger people because of the understandings concerning the wisdom of those who are older. Two proverbs state this quite clearly:

Aliyeona jua kabla yako hata kwa siku moja ana akili kuliko wewe : He who sees [the] sun before your [seeing it] even for one day has sense more than you.

Asesikiza la mkuu kuvundika 'guu : He [who] does not listen [to the words] of [an] elder [is] breaking [his] foot.

There is evidence from other societies that older people may generally share more medical knowledge with one another and with medical specialists than younger ones do (Garro 1986), but it is not just knowledge that makes them important in the Swahili community. The influence of the understanding that those who are older are wiser is enhanced by the fact that the advisers are not just older than the patients, in many cases they are also their parents.

The parent status includes the widely shared general expectation that its occupants are genuinely concerned about their children, and this is surely closely related to the fact that patients say they "trust," "believe in," or (in a few cases) "obey" their adviser. In the numerous cases where the patient follows advice in getting medical care, the connection between his or her understandings about being ill with understandings about what sort of help to get and from whom derives from the general expectations just mentioned and derives its strength from the strength of the relationship.

A few patients and a slightly larger proportion of advisers do share either Western or Galenic medical understandings with each other and with the doctors. For these people, the organization of understandings that leads to getting medical care, and to advising about it, is intrinsically organized and not dependent on general expectations in relations with others for its effectiveness

in guiding behavior. These fairly unusual individuals have multiplex relations with others who do not share the medical understandings, so they are in a position to meet the expectation that they show concern and provide help by providing a link between their partners' view of themselves as ill and the understandings concerning what to do about it.

The interview data suggest that sometimes there are chains of individuals linked together by the general expectations in particular roles and that one or a few participants in the chain have elaborate sets of understandings about medicine and use them to provide advice to a less-versed partner in a multiplex relationship. The latter, then, in time, passes on the same advice to others who may, in their turn, pass it on to yet others. In most cases, the advice will be effective even though none of those currently associated with it has a developed set of medical understandings. This is because effectiveness derives directly from general expectations of concern and help rather than from specific understandings about particular things.

Another group of advisers do not share the experts' medical understandings in any substantial number, but in the past, they have had personal experience with the therapist or kind of therapy (usually both) they recommend. These advisers form chains as well as the more knowledgeable ones do, except that these advisers often originally went to the therapist they recommend on someone else's advice. There seems always to be some kind of link between the therapists and the patients which does not depend solely on general expectations of help from people whose cultural competence in the area is not very different from the one needing help. This link, however, can be mediated through a substantial number of relationships, none of whose participants need have either much experience or much knowledge of the medical understandings at issue. A little knowledge may or may not be a dangerous thing, but through social transmission, it can go a long way.

Looking at this from the perspective of the social structure of the Swahili community, it is that structure, rather than the cultural elements concerning illness and treatment acting directly as a guide to behavior, that accounts for the choice of therapy and therapists in many cases. This social structure, the connections among statuses based in the mutual references in the understandings that constitute the expectations and salience understandings in those statuses, is a product of culture. Through statuses having different expectations, culture is distributed among actors so that some have understandings about Galenic or Western views of medicine, while others do not. Some have generalized expectations in their relations with particular others, and others do not have them in those relations.

But the social structure does more than distribute culture. It provides, as we have just seen, a basis for organizing elements of culture and providing schemata for individuals who have no means of forming them on intrinsic grounds. Through the mediation of social relationships—most specifically,

the general expectations that are part of the statuses of those involved in the relationships—understandings are linked together so that people have a basis for action even though they do not share the understandings essential to that action.

Nor is this social structurally based cultural organization limited to medical understandings or to the Swahili community. It seems likely that the same sort of organization operates for Fernandez's West African participants in ritual who share few understandings about that ritual with ritual specialists in their group and Keesing's Solomon Islanders who know little of what their community's specialists know about what lies behind the ways they act toward the dead.

Illness, Nonsharing and "Patterns"

Earlier, it was shown that there is a "pattern" in Swahili culture of the sort Benedict made widely known. This pattern is the result of a common element, the understanding that balance is essential to proper functioning, that is present both in the complex of understandings concerned with proper social relationships and in the Galenic view of body functioning. The relations between the two are similar to those Evans-Pritchard demonstrated as present regarding the Azande's views of morality and their views of affliction. As for the Azande, the understandings concerning affliction are more acceptable and forceful because they are linked to broader moral understandings commonly shared throughout the community.

The effectiveness of this link as an influence on the behavior of community members, however, is put in serious doubt by the fact that many, even most, of them do not share the Galenic understandings. The understandings about social relationships are widely shared, in part because of the constant exposure group members have to the cultural models contained in relationship terms (this is a central topic in chap. 7), but the desired state of these relationships, balance, can hardly be connected to the Galenic understandings about health for those who do not share those understandings.

This does not mean that the pattern is without influence even on the behavior of those who do not share Galenic understandings. As seen earlier, for many people, advice from someone of whom they have generalized expectations is the link between the understandings indicating that they are ill and the understandings about what action to take to alleviate the effects of the illness. Many of the advisers, however, were themselves originally led to the kind of therapist they ultimately recommended by advice they had themselves received earlier.

Still, some of the advisers did share at least some understandings about Galenic medicine. For these individuals, the fact that both the body's healthy

functioning and proper social relationships depend on balance would, it is worth hypothesizing, tend to make balance seem an obvious and attractive force in keeping things "right." If this is so, and I have only the limited evidence concerning the acceptance of explanations to be examined in a moment, the "common element pattern" involving balance would encourage using doctors whose approach to illness was based on reestablishing the body's balance.

More than this, those who share the Galenic understandings may well be readier to recommend Galenic doctors than those who do not, because, in part at least, they take the approach to illness these doctors use as self-evidently correct. In fact, all those who shared Galenic understandings recommended herbal doctors, while all those who shared Western understandings were almost as likely to recommend herbal doctors as they were hospital doctors.

Granting that sharing Galenic understandings increases the likelihood that herbal doctors will be recommended, the pattern would affect even those who do not themselves share the Galenic understandings through making it likelier that advisers would recommend herbal doctors. That is, even if those who give advice in particular instances do not themselves share the Galenic understandings, their advisers—those at a second remove from the patient whose illness is being considered—might. This would lead the pattern to affect those who do not share it in just the same way that understandings about the body affect those who do not share them: by being transmitted through social relationships in which the patients (or the advisers of the patients) have generalized expectations of benefit from those they consult.

And that is not quite all. The existence of the balance pattern may well give balance explanations of illness a certain validity even when those hearing the explanation had not previously known about the four elements in the body and the significance of their being in balance. Only a few patients told me that the herbal doctor they consulted explained to them why they were ill and why he was giving them the medication he did. But those few uniformly reported that they found these explanations satisfying and easy to accept. One of the reasons I was given for the popularity of the leading herbal doctor in Old Town was that he explained things so clearly and well. The similarity between the widely shared belief in the importance of balance in social relations may well be involved in making such explanations of illness easy to accept.

This "pattern-based" process was not seen for those who consulted hospital doctors, although they shared few or no understandings of the Western view of the body and illness. These few reported that the accounts they got were difficult to follow and not couched in terms they found familiar. This is surprising since the Galenic scheme also involves terms and understandings that are not encountered in ordinary affairs. There is no reason to believe hot, cold, wet, and dry as body processes are more obvious or initially convinc-

ing than infection, immunity, and vascular blockage. But it may be that the former scheme receives a friendlier reception than does talk of microbes, immunities, and cholesterol because health, like proper social relationships, results from the balanced participation of the different involved "elements."

Conclusion and Summary

General expectations in social relationships of the multiplex sort provide a connection between the elaborate sets of understandings shared among experts and the health needs of the majority of the community among whom the experts' understandings are not shared. There are a variety of types of medical care available to the Swahili. Galenic medicine, viewed by some as "traditional" Swahili medicine, and Western medicine are the most common and popular.

The experts' understandings in the Galenic approach hold that a balance among the body's four elements is essential to health, and this is strikingly similar to the balance in social relationships called for by Swahili ideals. This common element organization, or pattern, involving two different domains may serve to promote the acceptability of the balance understanding as a guide to behavior for those who share the balance understandings in both domains. Most community members do share the understandings about social behavior, but they do not share those concerning the body and illness. Nevertheless, the pattern may affect them through making it more likely that Galenic medicine will be recommended to them and through its having an intrinsically appealing structure.

When community members understand themselves to be ill, and these understandings are made on very similar grounds by almost everyone, they consult—or are given advice by—people they trust and of whom they generally expect help. These advisers as often as not direct the patients to practitioners on the basis of earlier advice they have themselves received. The schemata that lead patients to take advantage of the understandings shared by the medical practitioners are organized around the advice that has been given rather than around intrinsic relationships among the understandings concerning illness and medical care. Since the advice is accepted on the basis of the general expectations of help and concern that are part of the statuses of the advisers (most of whom are parents, spouses, kin, and neighbors), statuses are seen as playing a key part in the relations among understandings as well as among people.

In chapter 10, the part of social structure in culture's operation will be examined further. It will be shown that statuses affect the relations among understandings in ways quite different from constructing schemata through being vehicles for "importing" them.