- 1 In 1972-1973 and again in 1984-1985.
1A. R. Radcliffe-Brown's (1952a) famous paper on "The Mother's Brother in South Africa" sparked one of the many now-forgotten debates in anthropology, this one about the nature and explanation of "joking relationships". No doubt because Radcliffe-Brown was more important as a theoretician than as a field worker, and that he based his argument on ethnographic data collected by other researchers, his account of joking is pervasively abstract and, one might add, humorless. When I conducted field research among the Dyula of northern Côte-d'Ivoire1. I was indeed able, like many of my colleagues, to observe such joking first-hand. Indeed, the obscene joking aimed by worosso, slaves "born in the house", at horon, "free persons", seemed so discordant with Radcliffe-Brown's explanations that I devoted my first publication to the phenomenon (Launay 1977). In retrospect, I do not think that my analysis did justice to the subject I was privy to observe. Radcliffe-Brown and other theoreticians did not ignore the obvious fact that individuals related to one another as joking partners did not, for the most part, actively joke with one another. However, in their analyses, they were all too eager to sweep the significance of this fact conveniently under the rug. This is all the more striking if one bears in mind that joking between certain categories of affines was routinely contrasted to avoidance between other categories. However, affines who were supposed to avoid one another were expected to do so consistently, while those who could joke with one another did so intermittently if at all. The episodic nature of joking in "joking relationships" is, I wish to argue, hardly incidental. Joking has constantly to be instantiated. Someone has to initiate the behavior, and others have to construe it as appropriate. Such instantiation can take quite different forms. Among the Dyula, as I shall argue, certain forms of joking are staged on ritual occasions, while others are enacted on an ad hoc basis. In neither case can joking be reduced to the mechanical application of norms or rules putatively governing social relationships.
2In his remarkable ethnography Naven, Gregory Bateson (1958) analyzed ritual transvestitism among the Iatmul of New Guinea from a series of different perspectives: the cultural premises underlying the relationship between mother's brother and sister's son among the Iatmul; the structural dimensions of the relationship; the contribution of the ritual to group cohesion; the ritual as an expression of Iatmul ethos; and the ritual as an expression of Iatmul modes of thought. Each explanation, taken on its own, seems to elucidate certain facets of the ritual, at the expense of leaving other aspects untouched. But these different explanations do not really articulate with one another, but rather progressively forefront and eclipse different aspects of the ritual. All told, as he admits in his epilogue to the 1958 edition, the ritual remains almost as problematic and opaque as it did at the outset of the book.
3A. R. Radcliffe-Brown's classic discussions of joking relationships (1950; 1952a; 1952b; 1952c) unselfconsciously constitute a similar series of attempts to explain the phenomena in question from different vantage points which do not, in the end, quite fit together. He explains the joking relationship between mother's brother and sister's son among the Thonga of southern Africa in terms of the structure of affective relationships in the patriarchal family: affectionate and easygoing between a man and his mother and by extension (citing the principle of the equivalence of siblings) his mother's brother; authoritarian and formal between a man and his father and, by extension, his father's sister. He also analyzes joking behavior between relatives-in-law, but in a substantially different manner, where joking is an alternative (rather than a polar opposite) of avoidance behavior, where each form of behavior is a means of managing and defusing potential conflict. Joking between grandparents and grandchildren is again contrasted to the conflict characterizing relations between parents and children, an indirect expression of the alliance of both generations against the children of the grandparents and the parents of the grandchildren. In each of these instances, joking behavior is contrasted to formal and constrained modes of interaction, but in essentially dissimilar ways—as an expression of genuine affection between mother (and mother's kin) and son; as an expression of repressed conflict with affines; or as an indirect expression of strains in the relationship between parents and children.
4By something of a sleight of hand, Radcliffe-Brown assimilates joking between pairs of clans or ethnic groups to affinal joking as "modes of organizing a definite and stable system of social behaviour in which conjunctive and disjunctive components... are maintained and combined" (1952b: 95), that is to say, as a means of managing potential conflict between distinct groups or categories. As such, they constitute one of "four modes of alliance or consociation, (1) through intermarriage, (2) by exchange of goods or services, (3) by blood-brotherhood or exchanges of names or sacra, and (4) by the joking relationship [that] may exist separately or combined in several different ways" (1952b: 102). The problem here is, first, that virtually all social relationships combine conjunctive and disjunctive components. However, in all the other instances, joking relationships are essentially constituted in counterpoint to alternatively structured relationships: mother's brother as opposed to father's sister; affines whom one avoids as opposed to affines with whom one jokes; grandparents and grandchildren as opposed to parents and children. Pairs of clans may joke with one another, but there are no corresponding pairs who avoid one another.
5To Radcliffe-Brown's consternation, Marcel Griaule (1948), whose analysis of "L'Alliance cathartique" between the Dogon and the Bozo stressed its unique derivation from the specifics of Dogon cosmology, expressed considerable skepticism about the possibility of any cross-culturally valid theory which could account for such instances of "joking". Indeed, he suggested that they might constitute so many "cérémonies à cloches", an artificial category based on superficial resemblances, like European church ceremonies—weddings, funerals—in which there is a ringing of bells. ("Cérémonies à cloches" was perhaps not the most felicitous example; one could make a case that this was not as trivial a classification as Griaule implies. The most striking case in anthropological theory is "totemism", a rubric which, as Goldenweiser (1910) demonstrated, encompassed a variety of very disparate phenomena (Lowie 1920: 137-145; Lévi-Strauss 1962). Even Radcliffe-Brown, try as he may, could only analyze different forms of joking in different terms, or alternatively formulate an explanation so broad that it hardly engaged with the specifics of each instance.
6Radcliffe-Brown's various explanations, as well as subsequent analyses of instances of joking (Lévi-Strauss 1945; Harris 1968: 527-30; Launay 1977), all take as their point of departure the specific relationship characterized by joking behavior. In other words, the explanation of joking is presumably to be found in the nature of the relationship of mother's brother to sister'son, husband's brother to brother's wife, grandparent to grandchild, hereditary slave to freeman, etc. But this is not invariably the case. Take, for example, the Mande practice of senankuya, one which has numerous analogies among other West African peoples as well. Senankuya is the joking relationship between different pairs of jamu, patronyms; for example, Keita and Cisse joke with one another. Each jamu has not one, but in fact several, joking partners. While the term jamu has sometimes been translated as "clan" (Launay 1972), it is essential to note that these do not in any meaningful sense of the word constitute a "group". In any specific community, particular descent groups will be identified by their patronym: Bamba, Coulibaly, Fofana, etc. In a large community, it is not uncommon for two entirely unrelated descent groups to share the same patronym. The fact that two groups might share the same name was of minimal social significance, if any at all. However, in principle at least, the relationship of senankuya transcends all locality, linking all people who share one name with all people who share another. Such joking cannot in any real sense establish an alliance. Alliances can only be formed between groups or individuals, not between names. A fortiori, the relationship between names cannot contain "conjunctive and disjunctive components". Senankuya cannot be an epiphenomenon of some other relationship, an amplification of some other content. There is no relationship whatsoever between Keita and Cisse other than the relationship of senankuya itself.
- 2 The "host"/"stranger" relationship is described in detail in LAUNAY (1979).
7Of course, a specific community may include groups both named Cisse and Keita. This was the case in Kadioha, in northern Côte-d'Ivoire. Senankuya was not, in this case, the only symbolic link between the groups; the Cisse were also "strangers" (lunan) of their Keita "hosts" (jatigi)2. In spite of the fact that they acknowledged one another as senanku as well as "hosts" and "strangers", the two groups could not in any meaningful sense said to be allied with one another. The intricacies of local politics were in no way at all constrained by the existence of such relationships. At best, specific actions might receive additional legitimation post hoc by framing them in terms of one or the other relationship, though even here the "host"/"stranger" bond was more likely to be invoked than senankuya.
8There was one and only one category of occasion on which senankuya was regularly invoked. Funeral ceremonies regularly involved the ritual distribution of "alms" (saraka) (Launay 1992: 196-219). The descent group of the deceased gives a small quantity of grain or (increasingly) money to representatives of each of the other groups in the community. Instead of waiting to be granted their share, the joking partners of the bereaved make a show of snatching their portion. Should a particular descent group have more than one senanku in the community, a specific one will be singled out as the principal joking partner for the purposes of funeral ritual. It is perhaps appropriate that a simulacrum of "alms" should be seized in a simulacrum of "theft". Saraka is, as far as Islamic law is concerned, supposed to be a "free" gift, but such ceremonial distributions are, for all intents and purposes, socially obligatory. The mock thievery underscores the fine line between prestation and predation. Be this as it may, the senanku have a scripted part to play, though it is a very minor role and they are little more than extras in the overall production. In their absence, the ceremony would proceed unimpeded, perhaps aesthetically incomplete but hardly shockingly deficient. In any case, structurally, functionally, or symbolically, senankuya might appear to be totally marginal.
9But the very "puzzle" of joking relationships is a feature of repeated attempts to configure them as more or less automatic and epiphenomenal aspects of prior relationships. In fact, such relationships are virtually never automatic. They are either deliberately staged—as in the mock theft of saraka at funerals—or voluntarily instantiated by one or both parties. For there to be a joking relationship, someone actually has to do the joking.
- 3 While I avoid using real personal names in this article, the patronyms are harder to conceal in thi (...)
10Let me provide a concrete example. In Korhogo, I had made the acquaintance of a young schoolteacher with a personal interest in Dyula culture and language, an amateur historian and ethnographer, whose patronym was Konate3. One day, I decided to introduce him to my Cisse "hosts". "Isn't it true", they purposefully inquired, "that Konate is another name for Keita, that they two jamuw are equivalent?". When he answered that this was indeed the case, my hosts then concluded that, as a Keita by another name, he was nevertheless their senanku, their joking partner. I must add that this equivalence between Konate and Keita is not a foregone conclusion, an automatic substitution that all knowledgeable persons would make. The Cisse of Korhogo consider themselves an offshoot of the aforementioned Cisse of Kadioha. In Kadioha, there is a descent group that identifies itself as Keita, another that identifies itself as Konate. The Cisse acknowledge that they are "strangers" of the Keita, but have grown larger than their original "hosts". The Konate, on the other hand, are not only "strangers" of the Cisse, but they remain much smaller and occupy a somewhat subordinate position. The Cisse of Kadioha adamantly denied that their Konate clients were really Keita, much less joking partners.
11There are a number of issues raised by this example. In the first place, the question of which jamu actually jokes with whom is subject to considerable indeterminacy. Lists of the joking partners of any given jamu may vary from community to community, or even from individual speaker to speaker. Moreover, elastic notions of the equivalence of certain patronyms (a notion which can and often does cross ethnic boundaries) allow for the expansion or contraction of the boundaries of joking relationships. In Kadioha, Cisse and Keita, equivalent in size and social importance, were particularly well suited as partners in a reciprocal joking relationship which implied a sort of equality; Cisse and Konate were less appropriate, and structurally redundant, as the Cisse already had "Keita" joking partners close at hand. In the old Dyula quarter of Korhogo, on the other hand, there were neither Keita nor Konate descent groups, so that the equivalency of Konate to Keita carried no wider consequences.
12It is fundamental to note that, in this case, the joking relationship was jointly negotiated by all parties. In a context of interaction where several elders in the community were interacting with a young outsider, it is hardly fortuitous that it was the elders who took the initiative. Once they made the opening move, the young man was free to respond to their invitation to establish their relationship on a joking basis. In this particular case, it was hardly an act of tremendous import. The object was first of all to put the young man at ease, and second to establish a mode of and basis for more sustained interaction.
13This was far from the only instance where individuals in Korhogo mutually framed their interaction on a basis of reciprocal joking. Joking between affines was also a question of deliberate collusion rather than of the automatic enactment of kinship norms. Relatives-in-law are either blan or nimogo. Most affines are addressed and referred to as blan when one is senior to the other. It is a relationship which implies authority on the part of the senior relative, deference on the part of the junior relative. Nimogo, on the other hand, joke with one another because neither is senior or junior to the other; these are defined as one's spouse's younger siblings or, reciprocally, one's older siblings' spouses. However, the case of Aminata and Najara demonstrates the extent to which these affinal categories, too, are subject to negotiation and redefinition. Aminata and Najara were both married to Ahmadu. Co-wives are stereotypically depicted as rivals, though in fact relationships between co-wives in Korhogo was intensively variable and subject to a great deal of negotiation. Some co-wives hated one another passionately and transparently, while others cooperated on an amicable basis. Najara and Aminata were among the latter, consistently friendly to one another. Indeed, to my surprise, I realized that they addressed one another as nimogo, implicitly ruling out the possibility of quarrels and casting their relationship as one of equality and friendship. When I inquired as to how they had "arrived" at such a relationship, I was given the following explanation: Aminata was a distant patrilateral parallel cousin of her husband Ahmadu. Such endogamous marriages were not only common, but normative, among the Dyula (Launay 1982: 61-66, 150-157). As such, Aminata was a (classificatory) "younger sister" of her husband, and indeed the couple addressed each other in the same way as brother and sister. Seen this way, Najara was effectively the spouse of her "older brother", just as she was Najara's husband's "younger sister."
14Aminata and Najara were, to my knowledge, the only co-wives who construed their relationship in this way, as affinal joking. In fact, in a survey of Dyula marriages in Korhogo in 1973, I found that over 40% were cases of descent group endogamy. In other words, most co-wives could, if they so chose, have construed their relationship as nimogoya. Clearly, such a strategy was anything but automatic. Nevertheless, everyone in Korhogo concurred that Aminata and Najara were perfectly entitled to construe their relationship in joking terms. They perfectly respected the logic of the system, even though they applied it in a voluntarily idiosyncratic way.
15In fact, relationships between individuals who were more unambiguously classified as nimogo were by no means always free and easy. Lansina was the youngest of nine brothers; when I first met him, he was twelve, and his eldest brother was fifty. His father and the mother of his five half brothers was dead; his mother and three full brothers were still alive. Several of his half brothers were quite prosperous, but the only one of his full brothers who was successful was living in another town. His half brothers were quite convinced that he was spoiled rotten by his mother, but they could not decently flaunt her authority. One day, Lansina protested to his half brothers that Fatoumata, the wife of one of them, had quarreled with him and struck him in the face. Although Lansina and Fatoumata were nimogo, she was at least ten years older than he, and they were neither social equals nor in any way on joking terms. The accusation, however, was a serious one. Joking partners should not quarrel and, above all, should never use physical violence against one another. However, there were apparently no witnesses to the incident. Lansina's brothers consequently decided that Lansina was lying to get Fatoumata in trouble, on the grounds that she knew better than to hit her nimogo, and he rather than she was punished.
16The incident demonstrates, unsurprisingly, that joking relationships do not in and of themselves alter the dispositions of individuals towards one another. On the contrary, Lansina attempted to use the fact of the relationship as a weapon to punish his elder brother's wife for her demonstration of hostility (no one seemed to doubt that they had quarreled, though the blame was not made to rest with Fatoumta). The outcome ultimately depended on the way in which the senior kin of both parties chose to construe the incident.
17Indeed, joking relationships could be construed by the parties involved in ways that were only tangentially related to "joking" itself, for example in the case of "marriages" between grandmothers and their grandsons. I only learned about such marriages by accident, when I fortuitously ate a papaya which belonged to an adolescent boy in the household in which I was living. When he informed me his wife had given him the fruit, I replied that I knew very well that he was not married. He proceeded to remind me of a trip we had recently made to a nearby village, where he had stopped to talk to an old woman. "That was my wife", he asserted, "and also my (classificatory) grandmother". His father, who was sitting nearby, not only confirmed the story but pointed to an eight-year old boy passing by who, he added, was married to his real grandmother. In fact, I had been ardently collecting data on marriages between kin, and, needless to say perhaps, such marriages were never once mentioned.
- 4 By the 1980s, this was no longer unusual, though many still thought it reprehensible (LAUNAY 1995).
18These marriages are, in one sense, an extreme manifestation of the mild sexual joking that takes place between grandparents and grandchildren of the opposite sex (again, it is important to bear in mind that many grandparents were not in the least inclined to indulge in such behavior). However, in other respects, such marriages have a perfectly serious side to them. In the 1970s when the incident occurred, it was considered unseemly for women of marriageable age to remain unmarried4. Widows and divorcees were expected to remarry after a decent interval, though sometimes marriages of convenience were arranged for old women. Elderly townswomen, for example, might be married to elderly villagers who would thus have a place where they would be fed and lodged during their infrequent visits to town. Even so, many elderly women were not averse to putting a definitive end to their marital careers and not to have to be bothered, even occasionally, by the demands of cranky old men. By marrying their grandsons, such women signaled the end of their married life. Even though elderly women had very considerable leeway in deciding their own marriages, grandmothers could not arrange such marriages entirely on their own initiatives. They required the approval, indeed the active collusion, of their male relatives, especially their adult sons, who invariably took an active interest in the remarriages of their mothers.
19So far, the instances of joking I have cited have all been amicable (though not necessarily, as we have seen in the case of Lansina and Fatoumata, the relationship itself between the individuals involved). In these cases, joking amounts for all intents and purposes to a form of teasing, a kind of behavior which, however inappropriate it may be in certain situations or between certain categories of persons, is hardly limited to joking relationships. Indeed, joking partners were hardly more likely than anyone else to tease one another; only when, for whatever reason, they did choose to tease one another was such behavior ascribed to the prior existence of the joking relationship. But the literature is full of examples of joking which goes well beyond the confines of mild teasing, notably involving insults and obscenities which constitute blatant instances of extreme verbal aggression, behavior which, on the part of anyone but certain joking relations, would constitute a serious violation of the rules of ordinary sociability, to put it mildly. Such words, uttered by anyone else, would conceivably provoke a fight, whereas inappropriate teasing would at worst provoke a sharp reprimand. Functionalist and structuralist explanations of joking relationships tend systematically to anesthetize the verbal violence of some of these exchanges. Arguably, Freudian explanations would be better suited to address this aspect of joking; "ambivalence" has more teeth than Radcliffe-Brown's antiseptic combination of "conjunction and disjunction". But in many cases the act of joking would seem to create, rather than reflect, the state of ambivalence.
20Among the Dyula, the most overtly aggressive form of joking was the obscenity which slaves born into the family (worosso) could direct at free-born (horon) members of the community. Obscenity was simultaneously the privileged license of slaves and a mark of their shamelessness. Free men would listen to the obscenities of worosso with apparent amusement, but would demonstrate real embarrassment if they had to explain them to me. But obscene joking could be used by slaves to put free men in their place. For instance, I once observed an elderly woman at a wedding singing and dancing what I suspected were obscene lyrics. I asked a friend who was present whether she was a worosso; I knew her as a hajja, a woman who has performed the pilgrimage to Mecca, and consequently (on most other occasions) appropriately dignified. "Worosso lo (She's a worosso)", my friend replied quite loudly, clearly within earshot of the woman in question. "Forosso lo (He's a penis-so)", she retorted immediately, at the same time demonstrating her status and indicating that my friend had been at fault for pointing it out so conspicuously.
21I was more than once the target of obscene joking, and though it is impossible to gauge the extent to which I was singled out, I made a conveniently conspicuous target. For example, when a worosso (a Qur'anic scholar in ordinary life) was clearing a path with a machete as part of the preparations for a wedding, he rushed over to where I was sitting, making obviously suggestive gestures with his machete menacing to cut off my private parts if I did not compensate him properly. The free men who were with me on the occasion immediately instructed me to give him a coin to buy him off, the appropriate response on such occasions. On another occasion, right before a funeral sermon, when I was seated among a group of venerable and very proper-looking marabouts, a worosso came up to me and began to ask me obscene questions. The scholars were rather amused at my bewilderment and, despite the solemnity of the occasion, hardly scandalized. "He's a worosso. Give him a coin and he'll leave you alone", they counseled. In both cases, the slave jokers were testing me to see how I would react, literally cutting me down to size in the first instance, and deflating my pretensions as an anthropologist to be in certain measure an "insider". It is not coincidental at all that these incidents took place during weddings and funerals, ritual occasions where worosso are often expected to act the part. Elderly men and women who, in other circumstances, are staid and respectable members of the community are, precisely on such occasions, expected to behave shamelessly as a mark of their social inferiority. However, they can use this very shamelessness as a weapon to deflate the pretensions of putative superiors.
22Nowhere is the aggressive potential of joking more obvious than at the burial of elderly free persons. Once prayers have been said over the corpse and it is being transported to the cemetery, "grandchildren" of the deceased (a broad category including everyone in the appropriate generation) and worosso accost mourners with requests for money, ostensibly to help console their grief. Such requests may also be accompanied by clowning; on one occasion, a young girl mimicked an elderly woman hobbling about with a cane—presumably the deceased—to the general amusement of the crowd. Indeed, the standard rationale for such joking is that it is normal that the children of the deceased are sad, and that in this way the grandchildren can cheer them up by joking. The explanation is not very convincing. As in the instances of worosso joking discussed above, the requests for money are a virtual form of blackmail, more characterized by aggression than jocularity. Indeed, while this behavior is cast in the idiom of the joking relationship between grandparents and grandchildren, demands for money are made by children to members of their parents' generation, with whom they do not normally "joke". Nonetheless, in this one context, their behavior is virtually indistinguishable from the "joking" of worosso with horon. Each request need only be met by a small coin, but at an important funeral mourners are literally mobbed by children and slaves asking for handouts. Once, after I had been divested of all my loose change in a matter of seconds, I was rescued and whisked away in a car which some of my friends were using to drive the short distance to the cemetery, not in order to avoid the walk but in order to dodge the crowd. As the fact that they possessed a car suggests, they were relatively prosperous and consequently, like me, particularly apt targets as sources of handouts. While the cost of distributing such largesse is not huge—each payment is no more than a token amount—large funerals are already a heavy financial burden, and any extra expense is usually unwelcome. In effect, such joking temporarily turns large funerals into carnivals, ritualized events where relatively powerless inferiors—children and slaves—are licensed to harass those who normally exercise authority over them. However, the occasion can be co-opted by the rich and powerful. Prominent and wealthy mourners who wish to accompany the corpse on foot will go to the bank beforehand to purchase rolls of coins which they can ostentatiously distribute to the grandchildren and worosso. The overall outcome of the occasion ultimately depends on the strategies of participants.
23I have constructed this discussion around a series of anecdotes of specific instances of joking for theoretical reasons. Joking relationships, I wish to suggest, exist largely in their instantiation. In important respects, the exchange of jokes is not entirely unlike the exchange of gifts, as analyzed by Bourdieu (1977: 4-9):
"[. . .] cycles of reciprocity are not the irresistible gearing of obligatory practices found only in ancient tragedy: a gift may remain unrequited, if it meets with ingratitude; it may be spurned as an insult. Once the possibility is admitted that the ‘mechanical law' of the ‘cycle of reciprocity' may not apply, the whole logic of practice is transformed."
24The practice of joking is, of course, different in important respects from the practice of giving. The reciprocity inherent in gift giving is, of necessity, deferred. The significance of the gift only fully emerges in retrospect, in light of the nature of the response (or the failure to respond) of the recipient. In joking, the reciprocity (if indeed the relationship is reciprocal) is generally instantaneous. The meaning of the act emerges directly in the context of the event, not, as with gift giving, in its sequel. Still, in both instances, the "rules" provide at best an incomplete, at worst an entirely misleading explanation of the actions involved specific relationships do not in any mechanical way generate joking behavior; more frequently, it is the joking itself that creates, or at least reframes, a particular relationship.
25As we have seen, even within the confines of a single society, instances of joking seem so dissimilar in nature and tone that one is tempted to concur with Griaule that these are indeed "cérémonies à cloches". However, these dissimilarities shrink, if they do not entirely disappear, once one considers "joking" as a modality of social interaction rather than as a feature of specific relationships. From this point of view, I would still distinguish between ceremonial joking and ad hoc joking. Among the Dyula, weddings and especially funerals of elders are ritual occasions during which joking is a predictable feature. Senanku, worosso, grandchildren all have scripted parts to play, though they always have a certain amount of latitude in which to improvise. But the outcome of the drama is always uncertain. It may serve to deflate the pretensions of the relatively rich and powerful, or alternatively an occasion on which to demonstrate their largesse and their savoir-faire.
26Outside such ritual contexts, joking is even less subject to rules or norms. Rather, it represents one of several possible ways in which two persons can interact in any particular situation. Joking is in such cases the outcome of informal negotiation, much as two individuals may come to address one another as "tu" rather than as "vous". Such joking requires the complicity, not only of the individuals involved, but of those around them. There are instances where joking is an acceptable basis for interaction, others where it is quite unacceptable. In this respect, Radcliffe-Brown was, after all, right to contrast affinal joking with in-law avoidance as complementary paradigms of interaction. Reciprocal joking, in particular, sets a tone for a relationship—equality, amity, informality.
27All in all, the theoretical focus on joking relationships instead of joking behavior tended to put the cart before the horse. In any case, as we have seen, relationships themselves are not simply the product of the mechanical application of logical rules, but rather the product of strategies about whether, when, and how those supposed rules can be invoked. Indeed, as we saw in the case of Lansina and Fatoumata, these very rules can be invoked to pursue quarrels instead of precluding them, as the "theory" of joking would imply. One way or the other, neither the practice nor the meanings of joking are ever predetermined. They retain their potential to create and to subvert relations of hierarchy or of equality, to stoke or to subdue aggression. As such, it is hardly surprising that they remain a powerful idiom in African social interaction.
28Northwestern University, Evanston (Illinois).