Preferred Citation: Swartz, Marc J. The Way the World Is: Cultural Processes and Social Relations among the Mombasa Swahili. Berkeley:  University of California Press,  c1991 1991.

8 Tongues are Spears Shame and Differentiated Conformity

Tongues are Spears
Shame and Differentiated Conformity

Ulimi arobaini, fimbo arobaini: Forty tongues [are] forty spears.

The term "aibu,"[1] shame, refers to the painful emotional state resulting from the belief that others have or will evaluate one's behavior in an unfavorable way. Aibu rests on those experiencing it having two understandings: that others know, or will learn, of what one has done or may do and that they hold beliefs and values in which the behavior is held to be improper or undesirable.

There is, in fact, no single term heard more than aibu in discussing behavior of questionable acceptability. I asked a room full of prestigious, middle-aged men at a baraza what they thought the main source of concern in the community was. Several suggested that it was the young people, others mentioned the changing world, and still others said poverty, but when one of the men said "aibu," there was a general murmur of agreement.

This chapter will show that aibu is not only a nearly constant concern for virtually every member of the community but that it is also a powerful social and emotional force promoting conformity to the expectations and salience understandings that are at the core of the community's most important statuses. As is true for terms such as kibri (pride, a sin in Islam), aibu refers to understandings that indicate what is valued and appropriate behavior by pointing out the opposite. Unlike other terms, however, aibu refers to a painful emotion everyone is understood to experience.

Shame, Status, and Limited Sharing

The importance of shame in producing conforming behavior has been reported for a number of societies (e.g., Epstein 1984; Obeyesekere 1981; Spiro


1958), and it works in the same general way for the Swahili. An issue that occupies an important place here, however, is how shame based in the sharing of understandings about what is not desirable behavior can function given the limitations in such sharing that have been demonstrated in this community. Further, there is the question of how shame can encourage conformity to the expectations in statuses when those expectations differ, sometimes markedly, from one role to another.

In fact, the hypothesis here is that shame is produced both by the relatively few cultural elements that are shared by virtually everyone and by the values and beliefs shared mainly within limited categories of people (i.e., those resulting from statuses). Regarding the last source, individuals will be shown to experience shame as the result of evaluating their actions according to understandings they attribute to members of a status category. Shame will result even when these understandings are known by the shamed to be antithetical to those shared among members of other status categories. Shame resulting from understandings that are, and are known to be, widely shared does not work in precisely the same ways as does that resulting from understandings that are viewed as being only shared within particular categories and relationships.

One of the differences is in identifying the shaming agents. For the very broadly shared understandings, those whose disapprobation is feared are the same for everyone. These "shamers" are the same group of prestigious men regardless of the status of the individual experiencing shame. With respect to aibu based in understandings taken by those affected as shared by some, but not most, community members, the shamers, obviously, are different according to what behavior is being assessed. The crucial and feared evaluations may be those of anyone occupying a status or statuses qualifying him or her as a likely judge as this is seen by the one concerned about a particular transgression. Those viewed as judges in one sort of transgression may well not be thought of as judges in another.

Double Status Relativity and, Also, Uniform Judgment

The variation in the composition of the set of judges varies depending on the category the actor sees himself in (e.g., soccer player) and also on the categories of those he takes as concerned with what he is doing (e.g., soccer fans and kin but not those without interest in either the sport or the actor's general prestige). The double status relativity with its dependence on the actor's category as well as the categories of those whom he views as possible sources of shame does not apply to all judgments but is limited in its effects by three different considerations. Each of these limits leads to uniform, rather than dually relative, judgments.


First, there are a few broad value violations that are seen as shameful by nearly everyone without regard to whom the violator may be. Second, there is some general agreement on what is shameful resulting from the understood presence in the group of a status whose very high prestige members are taken to make judgments about shamefulness or its absence that are more difficult to ignore than others' are. Finally, an additional status exists whose members are characterized by their willingness publicly to disclose otherwise generally unknown shameful acts, thereby contributing to the operation of the first two considerations.

It will be shown that shame's operation is importantly affected by the fact that much of what anyone does is seen as judged by standards that differ according to the statuses of the shamers, while some of what is done is taken as judged by universally accepted standards. The differential importance of the judgments of occupants of different statuses in producing shame concerning much of what everyone does puts this powerful emotion behind the distribution of culture that occurs through the community's social structure, that is, set of interrelated statuses. At the same time, the uniform importance of the judgments of a single category of high prestige individuals as concerns a relative few of the things people do brings the same emotion to bear as a support for the existence of a single community embracing all its different statuses.

The interest here is mainly cultural and concerned with the functioning of the group rather than primarily psychological and concerned with the minds of individuals. Still, it will be suggested that although the internalization of values may be present, it is neither a necessary nor a sufficient condition for the effectiveness of shame as an influence on behavior.


The locus classicus for the differences between "shame" and "guilt" is the Piers and Singer volume (1971 [1953]). In brief, that source shows that the early idea division was between "shame societies" whose members do not internalize the values of the society and "guilt societies" where the values are internalized. Shame-motivated conformity is based in avoiding disapproval from fellow group members, while guilt motivates conformity through the effectiveness of the internalized values alone. Spiro, together with Piers in the latter's separate but contemporaneous formulation, rejects the notion that shame does not involve internalization of values. He holds that shame is distinguished by the fact that although values are internalized, the socializing agents for those values are not (Spiro 1958:406–422, Piers and Singer 1971 [1953]:25–33).

The position here is closer to Obeyesekere's, which sees "shame" and


"guilt" as both referring to actors' reactions or contemplations of wrongdoing. In his view, the experience of shame is directly and necessarily related to social life, whereas guilt is not. While shame may involve internalization of the values whose violation is the process of interest, it does not depend on it. Obeyesekere (1981:131) puts this as follows:

If . . . guilt were never given language formulation [because it is a primary mental process rooted in infantile experience] this was not true of shame. The [Sinhala] language has a complicated, incredibly large, subtly graded vocabulary of shame and its associated ideas pertaining to honor, status, loss of self-esteem, ridicule, vulnerability to slights, deference behavior, prestige and so forth. . . . Shame is a social emotion, though when it is internalized in a conscience it can act as a powerful mechanism of social control. Fundamentally, shame orients the individual to the reaction of others: he wants their approval and fears disapproval and ridicule.

I will follow Obeyesekere in the use of "shame" as referring to the actor's experienced or anticipated emotional discomfort arising from his understanding of the evaluation of his acts, omissions, or qualities made by others. Definition aside, it is an empirical fact that Swahili informants say that aibu sometimes results from the understanding that others do or will know about an act or quality and will evaluate it negatively. The informants also make it unmistakably clear that aibu can and does result from the actor believing that others view him as acting badly even if he himself considers his acts and qualities entirely proper and acceptable.

Finally, as we will see, informants also say that actors viewing their own behavior as unacceptable experience unpleasant emotions whether or not they understand that others share their views of what they did. These latter emotions are understood within the group as different from aibu, what I am calling "shame."

There is no reason to legislate here about the meaning of "shame" and the similarity between shame in this community and in all others. It is, however, my empirically based opinion that in the segment of American society in which I live, an emotion indistinguishable from Swahili shame exists, and the same is true of the two other societies in which I have done extensive fieldwork, Truk and Bena. Given the universal importance of evaluation in all societies, I find it difficult to believe that there are societies in which people do not experience bad feelings related to actual or anticipated unfavorable evaluations by others, but the concern with and intensity of such feelings probably differs measurably. I am sure that the Swahili experience such feelings, as well as talk about them, and that they influence their behavior. That is what is crucial for this discussion, which, however, may well have some hypotheses that apply generally.


The Power of "Aibu"

In the Swahili language "aibu" most nearly approximates what I have just defined as "shame." "Aibu" refers to dishonor, disgrace, loss of standing, and shame according to the three main Swahili dictionaries (Akida et al. 1981; Johnson 1959 [1939]; Krapf 1882). "Aibu" can be used in the broad and abstract sense in which English speakers use "shame," as in "Everyone tries to avoid shame" as well as in the somewhat narrower and more specific adjectival form, "It is shameful to cheat." The distinction between an act or quality that produces shame and the emotion that is shame is empirically and lexically difficult to make.

"Aibu," more than "shame," is used as a trope in which the act that produces the emotion of shame is mentioned but where the reference is actually to the emotion. A person is said to reveal his aibu when he tells of acts that produce shame, but since those who are insensitive to shame can mention acts that for others would be the source of aibu and, their detractors say, do so without aibu, it is clearly the emotion that is the central reference of the word "aibu." I will follow highly literate, English-speaking Swahili and use both "aibu" and "shame," with the latter being mainly reserved for the broad and abstract sense of the term.

So far as I know, there is no word in English that refers solely and specifically to what I have called "the emotional discomfort" that is the core of my definition of "shame," and the same is true of the Swahili language. "Anxiety" in English and wasiwasi or mashaka in Swahili refer to the emotional state, but the words by themselves can refer to bad feelings resulting from many different sources. In the Swahili language, more than in English, the word used for what I have defined as "shame" is used in direct reference to acts. Thus, one can say in Swahili, "His clothes fell open revealing his nakedness. What a shame!" without meaning "Isn't that too bad" or "What a pity." The Swahili usage just alluded to refers to the production of emotional discomfort in the one involved and not sympathy or any other reaction by observers. I will follow Swahili usage here with the understanding that the named acts are a trope for the discomfort they are taken to entail.

Wikans (1984) argues that honor may not be the direct opposite of shame, and this is mainly true for the Mombasa Swahili. As has been seen, fakhri, which can be glossed as "honor," is accorded individuals on the basis of their active, recognized adherence to shared values and cannot be won by the absence of aibu-producing behavior alone. Conversely, being without notable fakhri does not necessarily mean that the individual is filled with aibu. Informants all agree that virtuous but undistinguished people can be low in both shame and honor. People with fakhri are deeply concerned about aibu, and such concern is an essential part of having honor, but they are not free of shame because of their honor.


Those with honor are viewed as only relatively free from aibu. "Relatively" free because, informants all agree, no one is without aibu, and since the people with fakhri are very sensitive to aibu, the fact that they avoid most aibu-producing activities does not guarantee, indeed will not prevent, their having some aibu. The somewhat complex relationship between shame and honor will be seen as a source of power for prestigious men as recognized agents of cultural conformity, but for now it needs only to be clear that "aibu" cannot be taken to refer to the complete absence of fakhri.

Swahili individuals do not readily discuss aibu, their own or anyone else's. Men say that women discuss the aibu of their fellows quite readily, and at least some of my women informants agree. Men and women who consider themselves nobles also say that slaves and their contemporary descendants gossiped about everyone and, specifically, that this gossip was mainly concerned with people's aibu. Regardless of status, actors do not mention their own aibu at all readily, if ever, and all agree it is better not to mention the aibu of others either.

In fact, aibu are treated so circumspectly that their social manifestation is difficult or impossible to see under most circumstances. As among the Omani Arabs described by Wikans (1984:646 et seq.), overt expression of disapproval is not normally manifested. Even adulterers, drunks, and others whose aibu are serious and renowned, are generally treated just as everyone else is.

There are, however, circumstances, especially quarrels, when actors' aibu are openly mentioned. There is also a particular sort of person, a fidhuli , who is believed to have extensive knowledge of the aibu of his associates and who, if even slightly provoked, speaks of them publicly. We will return later to fidhuli and quarrels and their vital role in the operation of shame. For now, it is important to note the general reluctance to discuss aibu since that reluctance plays a part in the way shame operates in this group. The reluctance to mention aibu is part of a general Swahili emphasis on privacy, concealment, and secrecy. This emphasis is itself an important indicator of the importance attached to aibu in this society.

Secrecy and Shame

One of the most cherished ideals for a substantial proportion of the community is privacy or secrecy. "Siri," a noun used where English speakers might use verbs such as "protect," "keep private," "shelter," or "conceal" is a word often used in connection with this ideal. Bushes growing in front of a house's windows are said to provide "siri" for the house and its occupants. When a man dies, his widow sometimes laments that her siri is gone and wonders where it will come from now that the source of her prestige, personal protection, and material support is gone.


The opposite of siri is aziri , which refers to exposure in the widest sense. Aziri is what happens when clothes fall open to reveal the naked body, and it is also the word used to refer to having one's private deeds revealed publicly. Aziri is something to be avoided. Community members speak of it with enough feeling so that to say it is dreaded is not to put the matter too strongly.

Swahili go to considerable lengths to achieve siri and avoid aziri. People whom I know well are reluctant to tell me the names of their children or, even, how many they have. The remarkable reticence to provide information on who lives in which house and how the residents were related is, discussions show, explicitly aimed at maintaining siri and avoiding aziri.

How far this can go is shown by an attempt to get data about the ordinary course of family life and how it was seen to be changing. As part of this, in summer 1984, I paid youthful informants to tell me about their views regarding the differences between Swahili community life as it was in the past and as it is at present. I asked them to focus their attention on family matters and did not press them to talk about quarrels, improprieties, or anything that might cause them discomfort. I had some success with my limited objectives, but experience with two brothers shows something of the extent of secretiveness about family affairs.

The oldest brother came to one session and told me the names of the members of his immediate family but said he thought it would be better to keep the names of his grandparents "private." He did not keep his next appointment with me, and his brother told me that he—the elder—thought he should "protect family honor" by not talking to me. The younger brother (only a year separates this young man from his 18-year-old elder) came to several sessions but eventually failed to keep appointments. We had talked about the various places the family had lived in the course of his father's career as a school headmaster who was transferred around the country and about the ways he and his older brother helped his parents in household duties and in disciplining two younger siblings, ages ten and seven. No other matters were broached, but even these were enough to bring the brothers to discontinue participation.

The fact that one of the most dangerous and undesirable sorts of people in this society is a mdaku , a curious and inquisitive person, indicates the seriousness attached to keeping things private and out of general knowledge. Even the members of the socially disapproved category of boys and young men called "mahuni," whose behavior is generally understood as consistently contrary to accepted standards and whose indifference to the opinions of others is a central defining trait, show the pervasiveness of concern about aibu in the slogans written on the walls of their hideaways (gahden):

Tazama lako : lit. Watch yours, meaning Mind your own business.

Mtana wanatazama, usiku watalala : lit. In the daytime they watch [us], [but] at night they will sleep.

Informants agreed that these slogans show a concern about aibu despite


the fact that the young men in the gahden take pride in being outside their natal community and flouting its standards. Informants say that although people differ in their sensitivity to aibu and in their concern about it, no one is without some concern. The founder of the branch of Sunni Islam followed by the Swahili, Imam Shafi, advises the faithful in a poem given me:

Watch you tongue and
don't mention the aibu of other people.
Everyone of you, from the top of
your head to your toes, is all aibu.
The other people don't have a tongue,
     they have tongues.

If your eye spies out someone else's 
     aibu, say to your eye:
"Oh my eye, look, the other people have eyes."

Whatever else aibu may be, it is ubiquitous and public knowledge of the acts that produce it is what actors see as providing the full emotional force that it has. It may well be, as Spiro says, that undiscovered shameful acts (whether committed or contemplated) continue to torment their perpetrators because the punishment that would end them has not been delivered (1958: 409). Still, the Swahili are uniform in indicating that a central defense against the undesirable consequences of aibu is keeping it secret or, at least, shielding it from public view.

That this may be, as Spiro's argument would suggest, a trading of psychic pain for social gain by concealing the aibu from public view is indicated by several further types of evidence.

One indication of this is the bitterness I have heard in Swahili complaints about false friends who induce you to tell them your aibu but never mention their own. Another and more explicit type of indication of the importance of concealment in preventing aibu is found in proverbs. The proverbs concerning shame that I collected are of two types: one emphasizes the importance of keeping personal matters private, and the other focuses on the readiness of people to gossip and the danger of that gossip. This danger, informants make clear, stems from the aibu the object of the gossip may suffer.

Concerning the importance of concealment, a widely quoted proverb says, Nyumba yasitara mambo : The house conceals [unfavorable] matters.

This proverb is used both to advise people to keep important matters within the family and, more often, to say that because you do not know about a family's aibu it should not be concluded that they have none. The latter point is also made by the proverb, Kila nyumba inaondoa geneza : Every house has a bier come out of it.

The idea that one should be careful about telling personal matters to people


is contained in the proverb, Penye kuku wengi, usimwage mtame : Where there are many chickens, don't spill millet.

This last proverb also makes clear that people are understood to relish the aibu of others. A number of proverbs dramatize the danger in other people's talk, and although aibu is not mentioned specifically, informants are unanimous in saying that this danger is in revealing, or inventing, it.

Ulimi arobaini, fimbo arobaini : Forty tongues [are] forty spears.

Ulimi unanuma kuliko meno : The tongue wounds [worse] than the teeth.

Ulimi hauna dawa : [For what the] tongue [does] there is no medicine.

Shimo la ulimi mkono haufutiki : A hole [dug by] the tongue cannot be filled by the hand.

This emphasis on secrecy is evidence of what informants say directly and explicitly: aibu can only do its full harm to a person or family if it is generally known. The fact that even the disaffected young men who are known for their indifference to general standards are concerned about concealing their activities is taken as some indication that concern about aibu is very general in the group, not just to members of a few status categories.

Recognizing Aibu: Different Ideals and Different Agents

In Mombasa Swahili society, what is shameful depends in many instances on both the statuses of the actor whose behavior is being judged and on the statuses of those understood by that actor to be the judges. This implies that there need not be a single list of universally shared values or rules whose violation is aibu to all group members, that culture need not be uniform.

Further, the status-centered view of shame also implies that its cultural foundations include more than the understandings whose violation is shameful. Since the statuses of the judges and the judged are vital, the cultural bases for these status categories must also be included as part of shame's apparatus. Chief among these status constituents are the expectations that are the bases for judgment. Since it is through the expectations in the judging and judged categories and the salience understandings that brings these to bear on a particular judgment, the cultural distribution that constitutes the categories is central to the operation of shame. The nature of social structure, in other words, is a vital part of shame.

The understandings shared in the community provide the broad foundation for shame, but only a small minority of these understandings produce the possibility of shame for everyone, automatically. The rest bring shame only to occupants of some statuses when their actions are judged by occupants of some, but not all, of the statuses in relationships with them. Thus, what is


possibly shameful is in part a result of the community's social structure and the places of the judges and judged within it.

Some Universal Bases for Aibu

I will begin the description of aibu's operation by identifying a few of the ideal understandings that are taken to be sources of aibu for members of this group regardless of who is being judged or who is doing the judging. When I asked informants to tell me specifically some aibu-evoking acts or qualities that they knew about and that were undoubtedly aibu producing for anyone, the following were mentioned: stealing; impregnating an unmarried woman and being impregnated if unmarried; using foul language in the company of "decent people"; being seen naked in public; dressing outrageously; using, selling, or supplying alcohol; making it clear to others that you would like them to give you goods or money; going to people's houses and making a nuisance of yourself by shouting or pounding on the doors, especially at night; and failure of a man to protect a woman of the community from assault by outsiders.

This is almost surely not a complete list of actions or failures constituting obvious and undoubted aibu for everyone, but it is important to note that most of the things mentioned here are of the gross, rarely occurring sort that figure little in everyday life for most people. Even these egregious activities may include some that are undoubted aibu only when taken out of their actual social settings and presented as isolated examples of evil.

So, for example, impregnating someone outside of marriage is aibu, but there are men—probably not a few—who have lovers outside their community-approved marriages and who sometimes impregnate these non-Swahili women. Is this impregnation an aibu for the man? It certainly is so far as the wife of the impregnator is concerned and probably also in the eyes of her parents and siblings. It most often is not for the man's male friends, many of whom would view his liaison as more admirable than shameful.

A Swahili woman being impregnated outside marriage is a great and undoubted aibu for her as judged by everyone including her peers. The impregnator is also shamed, and this event, rare as it is, is admired by no one. However, the same is not true if a Swahili man impregnates his non-Swahili lover. Here the existence of a basis for shame depends on who is doing the evaluation.

Aibu without Personal Belief in Having Erred

Actors surely do not need an abstract understanding of culture's operation for social life to continue any more than they need a knowledge of physiology


to digest food. The fact that they are, willy-nilly, guided by culture makes shame a socially, as opposed to only personally, effective force. The importance of the cultural, as opposed to purely personal, element is emphasized by informants telling me that a person need not believe he has done anything wrong in order for him to have been involved in an aibu. An example will make this clear.

When a marriage takes place, it is usual for a drum to be put in front of the new wife's parent's house. This drum is beaten on the nuptial night to signal that the wife was a virgin and that the husband was successful in deflowering her. Many men view this practice as shameful, an aibu, because, they say, it concerns very intimate affairs about which the community need know nothing. Still, they say, they have no choice other than to allow the drum to be put out because the women—the wives, mothers, and sisters—think it is aibu not to have the drum.

It might be argued that the drum's presence is due to the women's concern about aibu and that the men accept it because of their concern for the sensibilities of their female kin rather than about their own aibu. Taking this view might seem to obviate the need to view aibu as occurring regardless of the actor's own evaluation of his behavior. This would seem to preserve the position that shame is invariably a consequence of the actor's understanding that he or she has violated one or more values, but it does not.

If we accept or reject what Swahili informants say, that aibu result from others' evaluations of your behavior regardless of your own understandings of it, or, as just suggested, we say that aibu influences behavior through intermediates whose aibu is of concern, the result for analysis is the same. This is that we need to know a good deal about understandings concerning who applies what standards to whom, in addition to the content of the values involved, when considering the shamefulness of particular behaviors.

In fact, it is through actors' knowledge of which people apply what standards to whom that a good part of the shame system works. It is not only feeling bad that lets actors know what standards apply but also knowing what is expected of others and what the others expect of the actor. The shame system has the same basis as the social system as a whole: the actors' sharing with one another at least some understandings about what people in different categories should do and actually will do in various circumstances.

These shared understandings enable actors to know not only what consequences their own behavior will have for their prestige and honor but also how they will be affected by the misbehavior of others. A young man, for example, was sitting with a group of friends when his father's brother, a ne'er-do-well, walked by. The uncle was very shabbily dressed, itself a shameful thing in this community, and, worst of all, was barefoot. After the uncle had passed, the young man said to his friends, "My father was the only one of those children who amounted to anything."

The significance of the young man's statement, of course, was that he


knew what standards his friends would apply to his father's brother and what evaluation that would produce for the man. He also knew how that evaluation might affect his own honor and prestige with his peers.

There is nothing surprising about the young man knowing these things. They are the sorts of knowledge that make social life possible. Still, it is worth underlining their existence so that it can be seen clearly that shame involves a more complex cultural foundation than would be the case if it were necessary to consider nothing but what values are shared and what constitutes their violation. The young man's uncle violated a value concerning proper dress, and that is a part of the aibu the young man sought to avoid with his remark. To understand the incident, it is also necessary to examine the cultural elements that establish the nature of the relationship between the boy and his uncle and, given that relationship, the standards for judging the boy because of the uncle's behavior.

To take a more common but still similar problem, informants of both sexes and all social standings agree that the elderly mothers and aunts who often spend their last years living with a child and his or her spouse are a source of concern because of the aibu they might bring. The possibility that these women may go to neighbors' houses and accept tea or food and talk about the poor or scanty food in the houses of their sons or daughters is understood as real and frightening by their children. The feared "begging" does not shame the old women in their children's view; they are seen as being beyond shame. It is the shame that would come to the children from the implication that they do not or cannot care for their aged parent that is frightening.

It might seem that the old women show by their behavior that the children are violating accepted ideals of care and that once that is known, the shame in the situation is accounted for. This, however, misses a vital aspect of how shame works. This is that those who care for an aged parent need to have a full understanding of how others will judge the behavior of the old mother and of how this judgment will reflect on them. The children's concern about their aged kinswoman is based, in other words, in a more complex set of understandings than simply the ideal understanding (value) that may concern neighbors.

Both regarding the shabby father's brother and the resident old mothers, individuals suffer aibu, or fear they might, because of the behavior of other people rather than directly because of their own behavior alone. These instances were adduced to show that simply saying shame comes from the violation of values fails to take account of the fullness of social life and of the complex tasks culture must accomplish if it is to provide a basis for it. Taking a final example from the other side of my contention that feeling you have done something wrong is not a necessary condition for aibu, let us consider a man who felt he did something terribly wrong but who, nevertheless, was not judged to have committed an aibu.


A boy of about eight years very much wanted a bicycle. Many of his friends had them, and he asked his father for one repeatedly, but the father delayed getting it. One day the little boy went swimming in the ocean and drowned. The father was disconsolate. Long after the child's death he mourned the failure to give his son the bicycle he wanted. He obviously understood his failure to provide the bicycle as a grave omission, but it was not, informants agree, an aibu. It was, rather, a clear instance of guilt.

The last case shows that internalized values, in this case manifested in treatment of a son, are not always a sufficient condition for experiencing what is considered shame in this society. With regard to the marriage drum, the converse was seen: that internalized values are not a necessary condition for shame to be felt. This does not imply that internalized values are irrelevant to shame but only that they are but one element sometimes involved in a complex situation. Given the contingent nature of the role of internalized values, it is important to note that they cannot be assumed to be any more crucial or central to understanding shame's operation than such wider considerations as the statuses of those involved as judges and judged.

Aibu and Significant Others: Arbiters and Sanctioners

So far, emphasis has been put on the importance of who people think is judging the behavior at issue. It will be clear that if shame is to have social—as opposed to only individual—influence, establishing who one's judges are in various contexts must have a foundation in shared understandings.

My most general hypothesis about this foundation is that actors view themselves as being judged by two different sorts of fellow group members whom I will refer to as "arbiters" and "sanctioners." The arbiters are quite a small collection of individuals in this society. They are individuals of great prestige and of whom impeccable behavior is expected. Their judgments (rarely, if ever, openly made but often imputed to them by others) are given great weight across a wide scope of behavior, and when their judgments and those of nonarbiters are taken to disagree, the arbiters' evaluations are understood as more significant.

The men, and they are all men, I refer to by the term "arbiters" are not recognized as a class or group by community members, but each of them would be identified as being highly respected and as having a great deal of honor (fakhri). I will return to the arbiters after examining the numerically much larger collection of evaluators, the sanctioners.

Everyone in the society is a potential sanctioner. The term, like "arbiter," is coined here for analytical purposes and is not used by the Swahili. It applies to those who are understood as judging an act by those who perform or con-


template the act. The sanctioners for any particular act may well not include all the members of the society, but everyone is understood as a sanctioner by at least some others for some of the things those others do or think of doing.

Different sets of sanctioners may have, and/or be understood to have by the judged, different views of the same behavior depending, in considerable part, on the statuses of the sanctioners and of the judged. Those who are judged often believe they know more or less clearly what views of their behavior different sanctioners are likely to hold. When these views are believed to differ, the judged are usually well aware of the disparity in consequences for themselves deriving from the different judgments.

To take a clear example, consider the practice of taking a "secret wife" (mke msiri ). Adultery (uzinzi ) is a grave sin as the Swahili understand it, but Islam allows polygyny, and some men have wives who their "main wife" knows nothing about. These auxiliary wives are never Swahili, so far as I could determine, but sometimes are from families that were once Swahili slaves. A husband knows quite clearly both that the judgment of his wife and of her kin will be quite different from the judgments of some, at least, of his cronies and that the consequences of these judgments will be quite different. By controlling information, the judged can attempt to have his activity evaluated only by those whose judgments are likely to be favorable to him, and the favorable sanctioners will often not be swayed by the knowledge that other sanctioners take a different view.[2]

The existence of different sets of sanctioners obviously involves the presence within the community of differences in the values, rules, and beliefs held by different group members. Such differences undeniably exist in the Mombasa Swahili community (as we have seen with respect to sexual adventures and as the study of sharing in chap. 5 indicates more broadly), and these are not only differences between individuals. There are also differences between groups or collections of people such as those between men of certain ages, on the one hand, and wives of men of those ages, on the other. These sorts of differences in shared understandings among various categories of community members are important elements in what I am referring to as "the distribution of culture."

The Distribution of Culture

Despite the fact that sharing within status categories is a great deal less than uniform, group life is best understood as being based on a distribution of the total corpus of the group's shared understandings rather than uniformly shared among them. This distribution includes some understandings that are shared by all group members, others that are shared by many group members, and a substantial number that are shared among a relative few.


The predictability that is essential to social life (see chap. 1) requires that social relationships be based on a certain minimum sharing among participants of understandings concerning behavior in those relationships. This being so, whatever else is characteristic of the distribution of culture among different categories of actors, it always includes some sharing among the people involved of understandings concerning what is acceptable in the interactions involving them. This sharing may be mainly produced situationally by double contingency, which may sometimes be effective because of the exchange of tokens that are not also guides. Alternatively, it may proceed mainly on the basis of understandings commonly acquired through earlier socialization. Whatever its sources, it is essential to social life that there be some kind of sharing to serve as the basis for mutual predictability.

The understandings basic to many relationships, especially if they are multiplex, are rarely situational inventions. More often, they are relatively uniform for occupants of the same statuses through some combination of common socialization and the use of cultural models of the sort examined in the last chapter. To the extent that there is some uniformity, judges who sometimes have the same status as those occupied by the ones they judge will make their judgments on the basis of understandings similar to those guiding the participants. This is what would be expected given a uniform sharing of culture.

But there is no necessity that all others will make their judgments on the basis of the understandings that guide the participants; the relationships could operate quite effectively even if nonparticipants have different understandings about some or, even, all its aspects. In fact, such differences do exist among the Swahili (and, surely, elsewhere), so that some who evaluate relationships, and behavior in general, use different standards from those others do. These differences are a fundamental part of the "distribution of culture."

The basic element in cultural distribution is "status." As noted previously, the term refers to a mental category of individuals grouped together on the basis of understandings that identify what are taken to be significant characteristics that distinguish the category members from nonmembers. Associated with these understandings are additional understandings about what can be expected of category members and others concerned with what sorts of settings and circumstances membership in the status should and will take precedence over other category memberships.

Statuses as the Basis for Judgment

The understandings in statuses of what can be expected of members are the bases for determining aibu both by the members and by those who judge them. That is, each status has as one of its sets of associated understand-


ings a series concerned with how its members should and should not behave and another set concerned with how people in relations with status members should and should not behave. These sets provide the foundation for differential assessments of aibu by different sanctioners, with these assessments equally affected by the statuses of those judged. The acts and their circumstances indisputably are, obviously, what is judged, but the statuses of both the sanctioners and those being judged determine what the judgment will be.

Most commonly, people in interaction with one another agree on the standards that apply to their relationship, and, in that relationship, each is in a position, in the view of the other, to serve as sanctioner for behavior in that relationship. Every now and again, however, relationships come about in which there is little or no agreement about applicable standards. These are instructive in showing how important such agreement, the result of cultural distribution, is. An example of a relationship lacking the results of the usual cultural distribution will make this clear.

Two Swahili brothers married and brought their wives to a house the brothers had inherited. One brother had married the daughter of one of the community's most prestigious men; the other had married a woman from the section of town that is renowned for its toughness, willingness to do manual labor (as fishermen, at least), and contempt for the "refinement" that marks the behavior of Swahili from the other sections of town. The sisters-in-law lived together for a time in an atmosphere that became progressively charged.

The wife from the "tough" section of town responded to any of the disagreements that joint living arrangements produced by saying such things to the more aristocratic wife as "Look at you! What man could love a woman with a behind like yours!" When more seriously provoked by her husband's brother's wife, the lower-class woman would bind up her clothing in a way that is understood as a sign of readiness to fight. The other wife was appalled by this behavior, which was quite foreign to her previous experience and contrary to what she understood to be acceptable. Eventually, the wife with the more delicate sensibilities got her husband to move out of the house and to take her to a place of her own where she was removed from her sister-in-law and the latter's shocking behavior.

As an insightful and perceptive informant told me in discussing this story, what the "tough" wife thought of as prestigious, the high-born wife viewed as aibu. The high-born wife could not bring herself to answer her sister-in-law in kind and was reduced to tears and shame by the latter's behavior. The women of the highest stratum of Swahili society are rarely seen, and for them meekness and gentle manners are a considerable virtue, both in their own eyes and in the eyes of their social peers. The women of the stratum from which the combative and insulting wife came are quite different: they admire bold and aggressive behavior in themselves and in their associates.


Both of the brother's wives shared Swahili values about personal dignity, autonomy, and rights of possession (the women quarreled mainly over who cooked what and when and with whose utensils and food). Their understandings about what procedures to follow in defending their rights and dignity and about how to evaluate the behavior shown by the other in that defense were, however, quite different. The differences between them in approach and evaluation were not only personal; each shared her view with many or most of the women of her stratum of the community.

The women of the "refined" status would—and did—judge their peer as having behaved admirably and the behavior of the woman of the "tough" group as having been shameful, while the women of the "tough" group would—and did—make opposite judgments. The question of where shame lies resolves itself into asking who the sanctioners of the behavior are. Their decisions, we see, depend on both their own statuses and the status of the one being judged. In the usual course of events in this society, actors as different in relevant statuses as the sisters-in-law in the above case do not engage in prolonged and close interaction.

Brothers do not often share a house after marriage, nor do they usually marry women of such different backgrounds. Having been brough together, however, the differences between the women's understandings were so great that interaction finally became impossible and they had to separate.[3] Their ability to deal with one another, including their ability to judge one another's behavior along the lines of what was aibu and what was not, was too limited to make continued interaction supportable.

The case of the sisters-in-law, like the example of the judgment of men with secret wives, shows the importance of cultural distribution in the shame process. These two sets of data make clear how there can be a variety of different sanctioners whose views of what is aibu and what is not can be quite different depending on both the statuses of the sanctioners and the statuses of those judged.

Status Differences and Privacy

The fact that sanctioners differ at least sometimes in their judgments helps explain the emphasis on privacy and secrecy found here. If no one knows about an actor's behavior, he has no realistic cause to be concerned about aibu (although he may be concerned anyway), but such secrecy is difficult or impossible in a small, mainly endogamous community of around 2,000 whose members live together in closely packed neighborhoods in a single section, Old Town, of the island of Mombasa whose whole area is only three miles by five miles (DeBlij 1968).

Moreover, some potential aibu, like the shoeless father's brother referred


to earlier, are based in things that occur in public and cannot by their very nature be kept secret. Here actors cannot confidently use secrecy as the sole basis for attempting to avoid aibu once an act known to be taken by some as shameful has been committed or, in some cases, contemplated. What the actor can do is to lessen the likelihood that news of the act will circulate. Frequently, particular attention is given to keeping this news from those likeliest to judge the act unfavorably.

The distribution of culture in this group includes quite sharp differences among statuses in the expectations used in judgments. As we saw in the case of the sisters-in-law and also for those with secret wives, what is aibu for one group of sanctioners can be a source of honor and prestige for another. Complete secrecy, then, would prevent the actor from gaining prestige at the same time it protected him from aibu. This would seem to call for a strategy of careful information control, and that is, in fact, what we see.

It appears that Swahili women commonly are freer and more open in their relations with others, including being willing to discuss more of their own and others' personal affairs. This may be in some part related to the view, held by both men and women, that men are more preoccupied with avoiding aibu than women are. Given men's greater concern with aibu, it could be that men are more willing to forgo honor from a limited group of sanctioners than to risk aibu as seen by a much wider section of the community, while women more often take the opposite strategy.

Whatever differences there may be between men and women in their approaches to aibu and to honor, there can be no doubt that there are differences among sanctioners in their evaluations of others' behavior. These differences are often known to actors and, at least sometimes, used in calculating how to act and what information to allow to pass to whom.

Such calculations must take into account that judgments depend not only on the statuses of the sanctioners but also on the statuses of the judged and the act at issue. Thus, to consider only one of these dimensions, nuclear family members are less likely than outsiders to judge one another harshly as concerns such potentially aibu sorts of activity as eating low-prestige food or employing generally disapproved means to gain advancement in employment or shcool. The same family members, however, can be very severe judges as concerns use of family property, meeting of family obligations, and public displays of unseemly behavior.

Similarly, regardless of the activity being judged, rivals for something are likely to make consistently less favorable judgments of one another than those working together to get it. Generally, it is the specific role, the part of a status's collection of understandings concerned with relations with another in his status, that affects judgment. A son judging a mother proceeds on quite different grounds from another mother judging that mother, so that evaluations of performance in the mother status coming from the mother-son role


View of Old Town, Mombasa.


Two of the Mosques in the Old Town.



Young men praying outside the Basheikh Mosque in Old Town, Mombasa.
Older men are inside where no photographs were allowed.



Girls in religious school (chuo ).


Boys in separate religious school.


Ten important members of the Swahili community, mainly of the contested
ethnicity that forms the basis for chapter 3; these are called "Swahili Arabs."


Men leaving meeting of the Muslim Association; many are not Swahili
since the association includes all Mombasa Muslims. The man in the suit
(foreground ) is the provincial commissioner.


are based in expectations different from judgments in the mother-mother role. Statuses provide the understandings used in judgment. More particularly, evaluations are based in the expectations within the status that governs the relationship with the judged in the status relevant for the purposes of the judgment.

The roles occupied by sanctioners vis-à-vis those they judge affect not only the nature of the judgments they make but also which particular acts they will know about. So, for example, men are far likelier to judge the actions of their male peers in a variety of kinds of activities than they are to judge women of the same age and social standing for comparable activities and vice versa. Members of each of these status groups do judge members of the other as concerns some things such as propriety in sexual seclusion and marital fidelity, but they do not judge members of the other group on some behaviors that are quite important within that group.

Thus, men judge women's clothing and jewelry very little—they have little opportunity to know what unrelated women are wearing since women appear in public only when completely covered by the all-enveloping black buibui—although women use this as a major basis for judging one another. Women judge men very little on how the men follow etiquette, in considerable part because they see little of the relevant behavior since most public expressions of etiquette occur mainly in contexts where the sexes are separate. Men, especially those of the well-to-do sections of the community, judge each other quite closely in this domain. The judgments individuals make vary according to the characters of the individuals and of the relationship with the judged, but this variation is within limits imposed by statuses both as concerns what is judged and the basic criteria for judgment.

Sanctioners, then, differ in a number of ways, including the standards they use for judgment, how they apply these to different sorts of others, and the likelihood that they will make judgments on various kinds of behavior as carried out by people of different types. All of the differences among sanctioners, even the bases for individual variation, stem from the differential distribution of culture, which, as shown earlier, is a feature of the social structure with its differentiated and connected statuses.

In Swahili society, differences between men's and women's understandings about what is aibu and what produces honor, despite some important areas of agreement, are still great enough that men say that women care little about aibu. Wikans (1984:635–652) notes that in many studies of societies where honor is stressed, it appears that women have no honor of their own—or shame either—but only affect the honor of their husbands and male kin.

Such a view would result, of course, from the distribution of culture, leading researchers who work mainly with members of one sex to fail to know about the standards held by members of the other sex if, as is true for Swahili society and for the ones Wikans refers to, gender is used as the basis for


differentiation involving importantly different understandings in the resultant statuses.

Differences between statuses are not limited to the type seen between men and women where members of one category are, or profess to be, ignorant of understandings important to members of the other category. There are also important differences in understandings about to whom and in what circumstances particular standards should be applied. Thus, the members of a particular family were fully aware of the understandings to be used in judging a person who engaged in a business dependent on the sale of alcohol (abhorrent in this pious Muslim group), and they would surely use these standards in judging an outsider, but the family members definitely did not use them in judging their husband-father who had grown prosperous through the sale of fermented palm toddy.

There is still another source of difference in assessing aibu which arises from the distribution of culture. This one results from the fact that everyone occupies a number of different statuses, each of which includes understandings, often different from status to status, about what constitutes proper and acceptable behavior by others depending on the statuses occupied by those others. Because of this, even actors who share many of the same statuses may, on the basis of their relationship to the judged person, use understandings from different ones of those statuses as the foundation of the judgment they make. This choice is not an entirely individual one but depends in part upon culture, specifically, on the type of status-associated understandings that I have called "salience understandings."

For example, some may judge what a fellow group member does according to understandings associated with the status "community member" while others may judge the same person according to the slightly—or grossly—different understandings associated with the statuses "neighbor," "friend," or "enemy." Which will be used depends upon the status the judge assigns himself and this is guided by salience understandings. A number of considerations affect these last including to whom judgment is expressed, but the nature of the relationship between the judge and the judged is often also important. Since the judgments made depend upon the status whose understandings are their base, saliency understandings play quite an important part in the operation of judgment just as they do in the operation of statuses generally.

Although everyone is a sanctioner as regards some acts, it is by no means true that everyone is, or ever will be, an arbiter. This is a status assigned to only a few members of the group who meet very demanding standards. As seen above, the judgments of different sanctioners of the same activity and individual often differ. Also different sanctioners, as such, are concerned with different categories of people, so that even when knowledge of some activity is widespread some kinds of sanctioners may not pass judgment concerning it.


None of this is so for arbiters. Within the rather narrow scope of public acts arbiters are understood to notice, judgment is taken as independent of any of the statuses of the person who commits the act. If an arbiter were to judge an act by someone differently from their judgment of another, his standing as an arbiter would thereby be put in question.

The arbiters are accorded great prestige mainly deriving from the understanding that members of this status set very high standards for themselves and live up to those standards. They are all men of middle age or older who can easily be recognized by the spotless traditional clothing they wear whenever their occupations make that costume possible. They are also distinguished by their erect posture, measured gait, modulated voices, and judicious manner. They follow the elaborate etiquette derived mainly from Persian Gulf Arabs, almost every one of them has visited Mecca during the pilgrimage season at least once, and they all have standard places at the front of the mosques where they customarily pray.

These men have a special presence that is obvious even to outsiders. It is, I am told, the men I call arbiters whom people have especially in mind when they say in praise of a young man or boy that he "fears the faces of the 'nobles.'" When young men lower their voices, improve their postures, and look at the ground, the man passing is surely an arbiter.

As the praise for youths suggests and as informants agree, the men I am referring to as "arbiters" are taken by group members as representing what is most honorable and worthy of respect in the community. These arbiters have no special designation. "Arbiter" is a name being used for analysis. Nor do they have a title; they are addressed as "sheikh," as are all mature males in the group.

Despite their importance in evaluation, they do not intercede in the affairs of other people to make explicit judgment. If anything, they are less likely to comment on the behavior of other group members than lesser people are. This is a consequence, informants agree, of their following the ideal that "nobles" do not gossip or discuss the actions of others. What is distinctive of them, and this is something that informants say about them and that they say about themselves, is that they "frighten" (tisha ) their fellow group members.

What "frighten" means in this context is that people are sharply constrained in their behavior when they view that behavior as being considered by the arbiters. Even those given to making cynical and belittling remarks about the arbiters in private, and these are mainly disaffected youths and people whose group membership is marginal, are nevertheless distinctly restrained in their presence—even if they are across the street—and seem concerned about their opinions even as they declare their indifference to them.

There is an illness called mata ya wazima , the eyes of the grown people,[4] that afflicts youths and is believed to be caused by the young people behaving


improperly in their presence. The "wazima" are all the mature men of the community, but informants agree that it is the high-prestige men of the kind I am calling arbiters who are particularly dangerous to youths who misbehave. The danger does not come from the men doing anything active when the youths misbehave. Disease is the direct result, I am told, of the impropriety being committed in front of the mature men, who do not actively participate to bring it about.[5]

The standing of mature men and, especially, the arbiters who embody the essence of their qualities is clearly an important aspect of cultural conformity here. Two anecdotes will illustrate how the members of this status group are viewed in the community and help explain that view.

A Swahili man living in a small city to the north of Mombasa noticed that corn was disappearing from a field he owned, and he resolved to spend the night in his field in the hope of catching the thief. After it became quite dark, he saw a figure enter his field and begin to pick his corn. The owner crept near where the thief was at work and on seeing the man's face recognized him as one of the town's leading citizens. On discovering this, he silently turned and fled. He told no one of his discovery until years after the event.

The second anecdote concerns a highly respected member of the Mombasa Swahili community and me. I had been hurt in an accident while traveling to Mombasa, and I could not leave my room at the Mombasa Club because my injury made it nearly impossible for me to walk. The respected man, a friend and patron of some years standing, lived only about a hundred yards away from the club and had heard, I was told by visitors less prestigious than he, that I could not walk. At first, he sent one of his sons to ask how long it would be before I could go to his house. I sent back the message that it might be several weeks before I could navigate beyond my room. That night he came to the club to visit me and told me that it was the first time he had ever set foot in the place despite its proximity to his house and his friendship with a number of club members.

I thanked him for visiting me and asked him why he had not been in the club before. He said that he knew the club had a bar (it does) and believed that members of the Swahili community sometimes drank beer or spirits there (I have never seen any). He said that if he had seen community members there drinking in contravention of Islam's prohibition, his "respect would be broken." By that, he explained, he meant that the drinkers would never again "fear" him and would "do whatever they wanted to" in front of him. This, he said, would ruin his reputation.

It was clear to me that only his concern for my welfare had brought him to risk this. Both coming to my room and leaving, he used a staircase that did not take him by the club's bar. This route may have been scouted for him by his son when the latter came to visit me the day before.


Self-Reinforcing "Fear"

The heart of both these anecdotes and, I hypothesize, a central element in the effectiveness of the arbiters in the shame process among the Swahili is the "fear" in which these senior members of the community are held. Because of this fear, informants report, people do not do aibu things that the arbiters are likely to find out about, and they certainly do not do aibu things in the presence of these men. I infer from this that anything done in the presence of the arbiters can be taken to be almost certainly free of aibu, and anything that would not be done in their presence has at least some taint of aibu about it. Informants agree with this inference (but seem to find it too obvious to mention).

Since the arbiters' views are taken so seriously in establishing whether behavior is shameful or not, actors have a considerable stake in knowing what these views are. On the one hand, this is made somewhat difficult because instances or examples of their judgments on particular individuals and their behavior are not available, since, as noted, they rarely or never explicitly make such judgments. On the other hand, everyone has a fairly good idea of what the arbiters are likely to think about a wide range of behavior from two related sources.

First, the arbiters are more or less explicitly seen as the representatives of traditional beliefs and values that are, of course, widely understood in the community. Second, the arbiters are all particularly pious Muslims in a community that is generally quite pious, and the views of arbiters are understood to be closely related to the precepts, and there are many, in the Koran. In addition, over the years, various individual arbiters have published their views about desirable and undesirable activities in pamphlets and Mombasa newspapers and presented them on radio and television. But probably more important than the publicly presented general views of the arbiters is the actual behavior of the arbiters themselves.

Arbiters attain their prestige because of the high standards they set for themselves and, importantly, because of their being viewed as living up to these standards. Their standards are understood to be more exacting than those of other group members, and if, given these standards, they can do it, it is not aibu.

The presence of the arbiters when particular behaviors are manifested also serves as an indication of what is acceptable. If they are present and something is done in their presence, the thing is unlikely to be aibu. In part, this is because most group members are strongly constrained by the presence of the arbiters. The fact that young men can actually get a disease from behaving badly in front of the arbiters and their lesser, mature brethren is an indication of how constraining their presence is. This constraint, it should be added,


contributes to the prestige of the arbiters because it shows that people fear them, and that fear is a measure of their prestige. In addition, however, the arbiters try to avoid settings and contexts where aibu are likely to occur.

We saw this in the case of my friend's reluctance to visit me at the club. The high-prestige men in general make a point of avoiding places where men are likely to fight or where men and women are to be seen together as well as places, like the club, where drinking goes on. Aibu rarely occurs in the presence of arbiters, then, both because the arbiters avoid being where it might occur and because the real fear they inspire makes it unlikely actors will be bold enough to misbehave in their presence.

The process is an interactive and self-reinforcing one. The fact that aibu does not occur in their presence, even though partly due to their avoiding settings where it might occur, contributes to the arbiters' prestige, and that prestige is the basis for the fear that makes it unlikely that they will be confronted with others' aibu.

It would seem that in the case of the missing corn, the thief would have lost his prestige and would no longer inspire fear. There is almost surely, however, a generalization of the respect that derives from high standards and their maintenance. This generalization makes it difficult to disassociate from a person all the fear originally vested in him because of his reputation for righteous behavior even when he abandons that behavior. Just this process led the field owner to be silent and creep away when he found out who was stealing from him. The same respect, perhaps mixed with a concern that he would not be believed, led him to remain silent for years after the event. The theft was an aibu without doubt, but the "fear" of the thief made unmasking him impossible.

This story is an extreme one, of course. It is told among the Swahili as an illustration of how powerful the fear of the respected men is and how this fear can inhibit the behavior of the other group members. Even when an arbiter is obviously at a moral, psychological, and social disadvantage, an upstanding member of the community still cannot bring himself to risk his displeasure and disapproval by revealing the respected man's crime.

If this is true, when the ordinary group member has done nothing that could by any stretch of group standards be considered wrong, while the arbiter is personally involved in a serious aibu, how much truer it must be when the ordinary group member sees himself as having aibu and sees the respected man as evaluating the aibu from a position of unmatched moral and social standing.

Group members do not follow the views of the arbiters, as they understand these views, in all instances. The fact that these men receive a good deal of attention in conversation and a good deal of deference in interaction, however, strongly suggests that their understood views are a frequent source of influence on behavior. The arbiters can be looked at from the outside as rep-


resentatives of the group's most respected standards, and the "fear" of the arbiters can be seen as a force encouraging adherence to those standards. This despite the fact that members of different status groupings may have their own standards that differ from one another and, even, from those attributed to the arbiters.

The arbiters' judgments override the differing judgments of different sanctioners in the sense, at least, that they leave in the minds of some of the judged and some of the sanctioners the clear notion—and this can be seen in informants' statements—that the behavior seen as disapproved by arbiters really may be aibu despite others accepting or, even, honoring it. In this respect, and it is a passive one, the arbiters' influence transcends the diversity of standards and judgments based in status differences and acts as a source of moral unity for the community as a whole.

Aibu as a Social and Psychological Process

When the concern individuals feel about the arbiters' judgments actually affects what they do, the influence of the arbiters is manifested socially, although this may or may not be known by the actors involved. Both for sanctioners and arbiters, even if their understood or anticipated judgments in particular instances do not affect overt behavior, they may result in anxiety for the actor and affect other instances or different behaviors.

The difference, however, between the influence of understood or anticipated judgments that results in the modification of behavior and the influence that results only in anxiety is considerable. Actors may be willing to suffer anxiety in return for the benefits they derive from the behavior in question. In fact, I know that members of this group do things—and refuse to do things—even though they understand their behavior to be unambiguously disapproved by arbiters and/or by one or more groups of sanctioners. A few group members, for example, do drink alcohol, engage in generally disapproved sex with other group members, and fail to meet obligations to kin or benefactors.

This happens, I hypothesize, more often when whatever psychological costs in the form of anxiety there may be are not augmented by direct social costs in the form of undeniable disapproval from either sanctioners or arbiters. A main way for actors to avoid these social costs once the disapproved act or failure is begun is to limit the availability of information about the act through taking advantage of the general emphasis on secrecy. However, there are two culturally based processes that militate against and limit this means of avoiding the social costs of aibu.

The first of these is that when people quarrel, they are understood to be


fairly likely to use the proclamation of their adversary's aibu as a weapon. Family members may be reluctant to do this, but even they may become sufficiently enraged to forget their duties to their close kin and their own reputations and proclaim aibu that would otherwise never be known beyond the family home. Sanctioners who do not themselves disapprove of an action or failure know that other sanctioners and/or the arbiters do disapprove, and the former may use their knowledge as a weapon in a quarrel.

A second culturally constituted process that limits the assurance that information can be controlled as a means of avoiding or limiting the social consequences of aibu centers on a status whose members are called "fidhuli." A fidhuli is a person understood to be unusually interested in, and well informed about, the aibu of others and to be quite willing to broadcast his information. I know of no gloss for this status in English, so I will retain the Swahili word "fidhuli" to refer to them.

One such man is reported to have become annoyed at another for a fairly minor slight and to have said to the man publicly that the man should not forget that his grandmother always came to the fidhuli's house when she was sick. This allegation means that the man's grandmother was once a slave in the house of the fidhuli since slaves, and no one else, go to others' (i.e., their forebears' masters) houses for medicine and care when they are ill. For a person who claims to be a "noble" (mwungwana) or full member of the group, having a slave ancestor is a serious aibu that throws his whole social standing into doubt.

The fact that aibu can be revealed by fidhuli whenever they find it convenient to do so and that others may reveal it in quarrels limits the ability of actors to avoid the social consequences of their acts through limiting knowledge of them. Neither quarreling nor being a fidhuli is approved, and, as noted, the same is true even for discussing the behavior of others. However, these disapproved activities, somewhat paradoxically, have as one of their consequences the diminution of the judgmental relativity that results from the existence of different sets of sanctioners whose judgments depend, in part, on their own statuses and those of the judged. The disapproved revelation contributes to moral uniformity by making uncertain the actor's control of who will judge his actions.

Cultural Change, Shame, and Cultural Distribution

An aspect of the cultural dynamics involving the arbiters which has not yet been considered is who judges the members of this status. In fact, the arbiters, like everyone else, have more than one status, and in some of their statuses (father, government official, community member, Muslim, etc.) they


are liable to judgment more or less as anyone else is. Having seen the case of the corn thief, however, it will be clear that separating an arbiter from his status is difficult. Arbiter seems to be one of the statuses, like priest, prostitute, and president in the United States, that "spills over" into other statuses occupied by the same individuals and affects the understandings that apply in those other statuses.

There is little question in the minds of the community members that arbiters have aibu. Community members agree that everyone, even Prophet Mohammed, has aibu, and only God is without it. Informants agree that the men I am calling arbiters are more concerned about aibu—their own as well as others—than anyone else in the society, which, they say, accounts in large part for their scrupulous maintenance of high standards. So long as group members show them deference and inhibit their behavior when they are present, the arbiters retain their prestige. Since no one is free of aibu, it is the arbiters' constant effort to avoid it, as people see it, that is the foundation of their prestige and of their ability to instill "fear" in others (kuwatisha ).

Arbiters and Cultural Change

Earlier, it was suggested that if arbiters countenance behavior, that behavior is almost certainly not aibu. A fortiori, if arbiters actually engage in behavior, it is likely to be viewed as free of aibu. This suggests that the cultural complex centering around the arbiters provides, inter alia, the basis for a culturally constituted means for the acceptance of new understandings about what can and should be done and how to do it; a culturally constituted means, that is, for the acceptance of cultural change.

There is little reason to believe that arbiters can adopt new behaviors entailing new standards without any limits on how radically those standards depart from the status quo, since if an arbiter behaves in ways that make people stop fearing him, he loses respect (i.e., he is no longer "feared"). I would predict with great confidence that in the extraordinarily unlikely event that an arbiter were to convert to Christianity, that would not establish the understanding that such conversion is acceptable. It would only lead to the arbiter's immediate loss of prestige and raise questions about his mental health.

There are actual changes, however, that arbiters are involved in, and a brief look at one of them may be instructive. A senior man of substantial prestige who is undoubtedly an arbiter played an important role in making movie attendance relatively aibu-free some years ago.

As Muslims, the Swahili have, at best, an ambivalent attitude toward pictures of any kind, and activities that are purely for amusement, upuzi , are by no means completely admired. The fact that women in films are sometimes


shown in revealing costumes, that scenes of sexual activities are shown, and that men and women attend the theaters together adds further possibilities for movie attendance being viewed as likely to entail aibu.

The arbiter, however, began to attend selected films in the late 1960s and went several times a year beginning in the 1970s. Sometimes other men of roughly equal prestige accompanied him. Their presence at the films together with the fact that none of them lost respect and all continued to be feared may have contributed to what both informants and I see as an increase in movie attendance during the 1970s by Swahili.

There are still men who report that they do not attend films and never did so. These men are all well into middle age, highly conservative, and generally suspicious of activities associated with the West. Many of them are themselves of the sort I would call arbiters, and it may be that the aura of acceptability of behavior stemming from other high-prestige individuals participating in it is more influential for lower-prestige individuals than for others with prestige comparable to those who participate.

For a considerable proportion of the community, however, film going is now accepted, and even those who do not go themselves seem to indicate no active disapproval of those who do. It is difficult to establish without doubt that attendance by the arbiter and his peers is the only factor involved in this change, but the time of their beginning to go to movies corresponds with the beginning of open and general attendance by increasingly large numbers of male community members.[6] A number of informants have told me that now any group member, even the "strictest," might go to the films without shame, if there is some attention to what films are seen and what theaters are attended.

The general hypothesis being advanced here is that the Swahili distribution of culture with its particular set of understandings included in the status "arbiter" results in the members of that status being in a special position compared to other group members in different statuses. They can engage in behaviors that might lead others to be judged as committing aibu, without being so judged, at least initially. If following their engaging in these behaviors, they retain their ability to "frighten" others and to be generally respected, the behaviors appear free from aibu and thereby become more acceptable to other group members than they might otherwise be.

Shame, Behavior, and the Distribution of Culture

Earlier, it was shown that individuals are exposed to a variety of judgments of similar acts with those of some sanctioners approving and those of others


disapproving what was done or contemplated. Clearly, the different assessments cannot affect behavior equally in a given situation, so guidance is needed in choosing which to accord most weight. One of the factors in this is what has been internalized and another is who the holders of the different views are.

The weightiest assessment and the actor's own views may not coincide, but the variety of standards available through the multiplicity of sanctioners with different views offers the possibility of flexibility as well as uncertainty. With the guidance of both internalized beliefs and values and with knowledge, not necessarily explicit and formulated, of the status system, a person can satisfy his or her personal needs without either refraining from desired behavior or being negatively evaluated by any of a diversity of sanctioners.

The ability of the individual to respond to the diversity of standards associated with different types of sanctioners by controlling information about his activity and by inhibiting his behavior in some, but not necessarily all, social settings suggests strongly that he or she has not internalized all of the different—and perhaps conflicting—values concerning that activity with equal strength. This would be so, at least, if internalization usually results either in behavior in accord with what is internalized or in discernible anxiety when engaging in behavior contrary to the standards.

My data concerning how different individuals experience aibu and how it is involved in motivation are regrettably sparse as, so far as I can establish, are the comparable data from studies of shame in other societies reported in the literature. Swahili informants tell me that aibu is an unpleasant feeling that no one wishes to experience. People suffering from serious aibu are said to be unable to look others in the face, and anyone who customarily looks at the ground instead of at those around him is generally understood to be experiencing chronic shame. I am unable to determine, however, how effective aibu is as a force in motivation.

Informants agree that people, probably everyone including the worst, try to avoid aibu and that "good" people, like those I call arbiters, try harder than others. I do not have information that allows me to make statements about how far people are willing to go in avoiding aibu, and I cannot even approximate how experiencing it is weighed against the benefits and social costs of doing desired, but aibu, things. Nor can I comment on the emotional cost of not doing desired things because they are aibu.

Balancing Shame and Contrary Forces: A Little Case

A brief case suggests that at least sometimes group members will endure such psychic pain as they may experience from behavior that they agree is


somewhat shameful, if the behavior has social benefits and acceptable social costs.

A young Swahili man told me that although he enjoyed the company of his male companions, they often caused him concern because of their drinking, smoking marijuana, and talk of sexual activities. He said that a "true Muslim" must lead others away from sin and he, although he saw himself as a true Muslim, did nothing to influence his companions to behave in more acceptable ways.

I take this to suggest that he felt some shame at his failure to do what he thought he ought to do. His views about what he should have done to influence his friends are in accord with what the Swahili say is true Muslim doctrine, and, although he never said so to me, I am sure he knew that. Still, he did nothing to correct his friends and continued to see them despite his misgivings. He never mentioned to them his disapproval of their actions but only, he told me, remained silent when they spoke of their activities and when they drank and smoked in his presence. However, he steadfastly refused to drink or do other things contrary to Islam despite the urging of his friends.

Insofar as my informant's account can be taken at face value (and the possibility that he sees me as a sanctioner vaguely aligned with men who are arbiters cannot be dismissed out of hand), his behavior appears to be more directed to gaining the benefits of associating with people who amuse him than with avoiding the feelings of aibu deriving from that association. It is to be noted that his associations had little or no social cost. The arbiters neither knew nor were likely to find out that he associated with the particular youths who were his friends, since their gatherings are in "cold houses" (cafés where cold soda and snacks are sold) where arbiters rarely or never went, and anyway, most or all of these youths were from good families and had done nothing sufficiently public and notorious to gain bad reputations. My informants knew of their aibu, but arbiters and others from sanctioner groupings likely to disapprove did not. From a personal perspective, he could—and did—console himself with his refusal to participate actively in his friends' sinful ways.

The youths themselves are, of course, sanctioners, and there can be no doubt they would have disapproved of any action by my informant aimed at "correcting" their behavior. Whether or not my informant and other group members would consider a negative judgment of this sort as a source of aibu (and I suspect they would not call it that), it would seem likely to be the functional equivalent in being a stimulant of feelings of being disapproved and losing prestige among the judges. At the same time, my informants did gain the social benefit of having usually amusing companions. The informant's aibu was, perhaps, not very great by his own standards, although he did feel he was behaving wrongly by not attempting to stop his friends from sinning or, at least, by continuing to associate with them when they did not stop.


Such emotional pain as this may have caused him, however, seems to have been outweighed by the social and personal gains of continued association and, perhaps, his personally virtuous behavior.

I do not mean to suggest that all cases result in feelings of aibu being overpowered by other considerations. My informant believed it was aibuproducing to drink alcohol, and he never did it despite teasing from his friends. I only mean to suggest that the existence of the feelings associated with aibu do not necessarily prevent the behavior that is identified as causing that aibu. Social considerations seem very weighty in determining aibu's effect on behavior, and many of these considerations involve an understanding by the actor that different "significant others" have different views of what is right and proper. Epstein (1984:40) says of the views of both experimental psychologists and psychoanalysts,

common to these diverse approaches is the way shame is held to be intimately linked to threat to the image or negative evaluation of the self. [On the negative side are] . . . feelings of inadequacy, worthlessness, and the like . . . elicited by the exposure of some act or quality . . . one perceives as reflecting discredit on the self. . . . More positively, shame may also be seen as providing, at the level of the individual, a major thrust towards the development of a sense of identity, serving at the same time, at the cultural level, to protect and maintain basic social values.

I agree with Epstein. I would add, however, that shame is not a monolithic process in Swahili society but that it "protects and maintains" not only "basic social values" (those represented by the arbiters for the most part) but also the cultural distribution that is as basic to the existence of the society as are the fundamental values. The individual is motivated to behave in accord with widely held values because of the shame he would experience and the social costs he would incur if he ignored the judgments of the arbiters and such sanctioners as might support these values as they applied to him. The individual, however, is also motivated to act in accord with values different from, even opposed to, the fundamental ones because of the shame he would experience and the social costs he would incur if he did not.

Shame as a Support for Cultural Diversity

It is shame's operation as a support for diversity based in differences among statuses that is of particular interest in dealing with the questions raised by the fact that not all of culture's elements are shared by all of a community's members. The broadly shared values applying more or less equally to everyone, such as not going naked in public, serve more importantly as tokens than as guides. By following them, and it is effortless for almost everyone,


one affirms group unity, and members show each other that they behave as decent people do. However, the varying values and beliefs of the sanctioners, applying selectively but predictably, support a conformity to a culture that is differentially distributed among the statuses that make up the social structure and is not even fully shared within the status categories.

Like the models based in terms considered in the last chapter, shame not only promotes conformity to understandings shared by group members, it does so in a differentiated way. This differentiation is based on status differences with both the status of the judged and that of the judge greatly affecting the kinds of conformity expected of the judged. The sort of universally applicable pressure for conformity to universally shared understandings is limited to understandings whose importance for most sorts of behavior is rather slight. It is the status-dependent, differentiated pressures that promote much of the behavior that is vital for individual adaptation and community life. This sort of pressure for conformity allows substantial flexibility in the sorts of behavior found within the group and does not depend on broad sharing of large numbers of specific understandings.

The interest in statuses and their role in cultural processes in this chapter and those preceding it has focused more on how statuses function and their general properties as seen in Swahili culture and social life than on their operation as systems of interconnected elements. In the next two chapters, some of the effects of interconnected sets of statuses, that is, of social structure, will be examined.


8 Tongues are Spears Shame and Differentiated Conformity

Preferred Citation: Swartz, Marc J. The Way the World Is: Cultural Processes and Social Relations among the Mombasa Swahili. Berkeley:  University of California Press,  c1991 1991.