LANGUAGE & COMMUNICATION Language & Communication 25 (2005) 39–60 www.elsevier.com/locate/langcom Liminal meanings: sexually charged Giriama funerary ritual and unsettled participant frameworks Janet McIntosh * Department of Anthropology, Brandeis University, P.O. Box 549110, Waltham, MA 02454, USA Received 23 March 2004; received in revised form 23 March 2004; accepted 23 March 2004 Abstract This article analyzes women’s obscenity in Giriama funerary songs, and argues that the unresolved meaning of these songs can be characterized in terms of the contests over their participant frameworks. I begin by suggesting a distinction between metapragmatically explicit rituals with clearly established participant frameworks, and metapragmatically opaque rituals characterized by ambiguous and potentially resistant liminal behavior that tends to lack a clear participant framework. The conflicted social role of the Giriama womens’ songs instructs us about the ways in which the multiple affordances of liminal ritual, in conjunction with destabilizing social change, can give rise to contests over ritual meaning. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Ritual; Liminality; Participant frameworks; Meaning; Resistance; Sexuality 1. Introduction The Giriama villages surrounding the town of Malindi, Kenya see more than their fair share of funerals. Women and men in their thirties lose their lives to tuberculosis, AIDS, or malaria; twenty-year old matatu drivers expire in road accidents; teenaged girls die in childbirth. Despite such a profusion of tragedy, Giriama funerals – like many in sub-Saharan Africa – are structured not so much to dwell on the losses of * Fax: +1-781-736-2232. E-mail address: janetmc@brandeis.edu (J. McIntosh). 0271-5309/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.langcom.2004.03.002 40 J. McIntosh / Language & Communication 25 (2005) 39–60 death as to underscore the richness of life. Hundreds of people from nearby communities gather at the homestead of the deceased, socializing and marking the transition with song and dance. While mourners seclude themselves, tend the body, or prepare the grave, many men engage in prestige competitions and commerce, while women from the community are entrusted with the funerary songs and dances that continue for several days as the corpse lies on an outdoor bed awaiting burial. These performances are particularly striking for the way they invert ordinary sexual decorum and cultural expectations of female reserve. Women shuffle around the body in a great circle, grinding their hips and clapping their hands while giving free reign to song lyrics so sexually explicit and irreverent that many of them refused to recite them to me outside the ritual context. While a few songs at the beginning of their performance remark upon death (‘Fly away, housefly. You can’t land on this body; it’s covered with rashes’), their content becomes increasingly bawdy as the hours pass. The lyrics flaunt female desire, mock the genitals of both sexes, and express frustration with inept male lovers and husbands. Sexual derision is a popular weapon: ‘How will you screw with half a penis?’ One lyric prods jealous husbands while belittling their bodies: ‘A jealous husband will not grow fat – he only has thin legs and long testicles.’ Other songs celebrate the joys of adultery: ‘My husband doesn’t fuck me that way!’ While they flout conventions verbally, the singers enact further obscenities through suggestive dancing and simulated intercourse with whimsically chosen partners, male or female. ‘Traditional’ Giriama culture – by which I mean cultural arrangements considered authentic by Giriama themselves – is characterized by a relatively strict patrilineal structure in which women are subordinated in several ways. Elders and men tend to control marital arrangements, and men are in charge of most political and economic decisions. Women, meanwhile, are essentialized as meek, emotional, and lacking in reason, and receive far fewer educational and economic opportunities than men. These constraints are mirrored by expectations of sexual self-control, particularly on the part of women. Unmarried girls are expected to protect their chastity, and married women who are unfaithful may inflict their children with a supernaturally borne illness termed kirwa. In actual practice, Giriama life in the complex, multicultural settings of contemporary coastal towns and villages is never so neatly arranged, and traditional strictures are constantly violated. Nevertheless, many traditional Giriama values live on in some form, even if their realization is compromised. Among other things, women are expected to speak and act with sexual modesty, to eschew the supernatural dangers and personal shame of promiscuity, and never to initiate sexual liaisons, let alone speak about them publicly. Looked at one way, women’s lascivious funerary songs are an unsurprising accoutrement of the culturally widespread phenomenon of liminality, a time of ritualized ambiguity in which ordinary conventions are overturned in a carnivalesque free-for-all (Turner, 1964; Bakhtin, 1968). Yet the routine permissiveness of liminality should not dissuade us from treating it as significant, nor should the ambiguity of liminal performances provoke us to throw up our hands in despair of locating ‘meaning’ within them. Liminality is not formless; it tends to offer up signs (in whatever medium) that possess multiple ‘affordances’, by which I mean semiotic J. McIntosh / Language & Communication 25 (2005) 39–60 41 offerings that are potentially interpretable in multiple ways. Liminality, in other words, may be ambiguous, but it is not meaning-less. In fact, the Giriama case I describe here highlights the potential significance of liminal performances through the fact that many Giriama themselves are currently invested in asserting and debating the significance of women’s lascivious words. In the Malindi area, two main interpretations of women’s obscenities circulate, in competition with each another. According to one interpretation, women who sing and dance are giving voice to their own desires. According to another, women who sing and dance are simply going through the motions of ‘tradition’ without indexing their private desires or beliefs. In this paper I explore the plural affordances of this ritual text-in-context and the broader social factors that motivate these competing folk interpretations. My approach to this ritual has implications for another potential anthropological reading of the women’s words. It is quite common for anthropologists to interpret oppositional ritual texts as instances of ‘resistance’ against the powers that be – a tempting response to the Giriama case. As texts, Giriama funerary songs are the symbolic antithesis of traditional expectation, for they desacralize sex, flout the order of sexual custom, and defy male elders. But it would be too simple to dub such performances ‘resistance’ on the basis of obscenity in the text; after all, some members of Malindi’s Giriama community pointedly ascribe desire and impudence to the singing women, and others, including the singing women themselves, just as pointedly deny their presence. If we regard meaning as jointly negotiated rather than directly emergent from the singers’ intentions or the text alone, these Giriama disagreements thwart any effort to characterize the meaning of these lascivious texts in any simple way. Indeed, to derive the significance of a ritual purely from its face value would be akin to locating the meaning of a sentence within its semantics and syntax, without considering the contextually grounded indexical relationships that inflect it (cf. Bauman and Briggs, 1990; Gal, 1995; Goffman, 1974, 1981a,b; Silverstein and Urban, 1996). Linguistic anthropologists have long recognized Erving Goffman’s notion of the ‘participant framework’ as a useful instrument for gauging contextual pragmatics. According to Goffman, a participant framework is a jointly constructed interpretive frame that delineates a number of relationships between utterances and people. Among other things, it specifies whether speakers compose their own utterances, whether the sentiments expressed in their utterances reflect their own commitments and/or those of others, which listeners are ratified and which are merely eavesdropping, and so on. The participant framework figures prominently in the meaning of an utterance by situating it in an immediate social context of indexical relationships. Jointly constructed participant frameworks help to establish whether an utterance is taken as earnest, sarcastic, playful, hostile, and so on, while establishing the target(s) of the utterance. In the case of ritual, a participant framework can establish whether community members consider ritual participants to be committed to the sentiments they express during the ritual, or not. It establishes whether the messages encoded in a ritual are directed at particular individuals – such as deities, onlookers, even the participants themselves – or are considered merely hypothetical. It may also establish whether ritual utterances are recognized as 42 J. McIntosh / Language & Communication 25 (2005) 39–60 intertextually invoking other utterances outside the ritual; utterances past, future, or hypothetical (Irvine, 1996). Such jointly constructed interpretations of a ritual event can be crucial in determining whether a ritual text has social bite or merely dissolves because it is construed as irrelevant or incidental to ordinary social life. Instead of inferring absolute meaning from the ritual text on its own, then, one can use the notion of participant frameworks to explore how the ritual in situ is taken to index various relationships, attitudes, and speech acts that extend beyond the bounds of the ritual performance itself. In my analysis of Giriama women’s funerary obscenity, I draw upon participant framework analysis to assess the many indexical complexities of this ritual. At the same time, I will point out some potential complications in the application of participant framework analysis to ritual, complications best illuminated by first addressing the metapragmatic differences between different ritual genres. The participant frameworks of some rituals are highly metapragmatically structured, ‘metapragmatic’ cues being the signals in a speech event that regiment (determine or suggest) interlocutors’ understandings of the significance of that speech event and the tacit rules that govern it. In rituals with highly metapragmatically structured participant frameworks, there is little doubt as to what is being said and what is achieved in the saying of it. Other rituals, however, are metapragmatically underdetermined, and leave room for strikingly divergent interpretations. Ironically, it may be that the obviously liminal and apparently oppositional rituals in which scholars have tended to locate ‘resistance’ are among the most metapragmatically ambiguous of rituals. Rather than presume the presence of resistance, then, I use participant framework analysis to assess the affordances of Giriama funerary songs, but I cannot resolve these into a single interpretation. In fact, I find serious disagreement among interlocutors about how the songs should be framed; more misalignment than alignment, in other words, around any particular participant framework. The multiple affordances of these songs can be traced to several sources, including the indexical ambiguity of the song texts themselves and local shifts in gender roles that change the context of interpretation and inject uncertainty into folk interpretations of women’s liminal license. Here, then, is an apparently oppositional ritual whose unresolved meaning can be characterized in terms of the contests over its participant framework. Its conflicted social role instructs us about the ways in which the multiple affordances of liminal ritual, in conjunction with destabilizing social change, can give rise to contests over ritual meaning. To arrive at a detailed understanding these, I must take a detour through some theoretical material before returning to Giriama funerals themselves. 2. ‘Meaning’ and metapragmatics in ritual ‘Meaning’ in ritual is a vexed category, in part because some scholars have doubted that ritual is best characterized in terms of propositional content. Tambiah (1981), for instance, has argued that the multiple modalities harnessed by ritual – through devices such as music, drumming, and dance – can relegate referential, J. McIntosh / Language & Communication 25 (2005) 39–60 43 denotative meaning to the background. Stoller (1989, 1995, 1996) has also drawn attention to the neglected sensual elements of ritual, suggesting that many aspects of bodily experience that cannot be translated into propositional form are crucial to ritual representations and ritual effects. Bourdieu has argued that the key to ritual lies not in verbal analysis but in the study of practice, including the bodily inculcation of cultural values (Bourdieu, 1977). Others, such as Fernandez and Valeri have emphasized the ambiguity of ritual symbolism (Fernandez, 1974) and the indeterminacy of ritual’s poetic, paradigmatic effects (Valeri, 1985, p. 343). Briggs, furthermore, has convincingly argued that some rituals produce their power in part from ‘witholding denotative meaning from some of the participants’ (Briggs, 1996, p. 220, italics his). There is little doubt that ritual semiotics do not always communicate clear propositional content, or that Western metadiscursive practices have biased many anthropologists to emphasize decodable meaning in rituals (Briggs, 1996; Bell, 1992). Yet it seems just as true that many rituals do convey propositional content, such as articulated prayers, song lyrics, spoken promises, and so on. And when ritual has denotative meaning, then like ordinary talk it cannot be properly interpreted through its semantics alone. We must attend to the pragmatics of particular ritual acts, pragmatics that include the indexical relationships delineated by participant frameworks. The concept of the participant framework emerged from Goffman’s observation that communication is rarely reducible to a simple dyadic exchange between a unitary ‘speaker’ and a solitary ‘hearer’, for interlocutors can be decomposed into more complex participant roles. The speaker, for instance, can be broken down into several roles, including the ‘animator’, ‘author’, ‘principal’, and ‘figure’. The ‘animator’ is the party who produces the message in physical form – through speech, writing, or other means. The ‘author’ is the one who composes the message by selecting and arranging the sentiments expressed. The ‘principal’ is the person ‘whose position is established by the words that are spoken, someone whose beliefs have been told, someone who is committed to what the words say’ (Goffman, 1981a, p. 144). And a ‘figure’ is a persona projected into the audience’s imagination by the utterance. Sometimes, a single individual will fill the first three of these roles – indeed, according to Hanks, in ordinary talk, ‘[u]nless it is indicated or otherwise known, it is assumed that the person who produces an utterance is the source of the proposition stated, the one who chooses the words and the one whose position is expressed. That is, it is assumed that animator, author, and principal are identical’ (Hanks, 1990, p. 153). Very often, though, the functions are distributed across several parties. We can imagine how this happens, for instance, during reported speech, or parody, or when one speaks on behalf of some other person or entity such as an institution. As communicative interactions proceed, furthermore, participant roles can subtly shift on a dime. 1 1 Goffman and others term these minute-to-minute changes ‘footing shifts’. For purposes of simplicity, however, I will restrict myself to the phrase ‘participant framework’, with the understanding that a participant framework can easily shift as interlocutors interact. 44 J. McIntosh / Language & Communication 25 (2005) 39–60 Scholars since Goffman have modified and complicated the model he proposed. Some have refined the possible participant roles, yielding important distinctions between, for instance, ‘addressees’ who are ratified to hear an utterance, and ‘hearers’ who may hear without being so ratified (Clark and Carlson, 1982; Levinson, 1983). Levinson (1988) has expanded the inventory of participant roles by decomposing these into constituent features. Irvine (1996), on the other hand, argues against any predetermined inventory of participant roles while drawing on Bakhtin to direct our attention to the ways in which utterances intertextually presuppose or invoke other utterances past, future, or hypothetical. This observation allows the analyst to derive ‘more subtle types’ of participant roles ‘from a notion of intersecting frames and dialogic relations’ (p. 135). The Insult Poems performed at Wolof weddings in Senegal, for example, get much of their piquancy from the fact that the insults are drawn from gossiping speakers in the past, and will presumably be spread by word of mouth to other interlocutors after the wedding is over. To expand participant frameworks in this way greatly complicates the range of indexical relationships of a speech event. Despite such refinements, a central problem remains for the scholar attempting to apply participant framework analysis to ritual. Some participant roles are moreor-less matters of fact, invariant regardless of the point of view from which they are assessed. For instance, a given person either animates an utterance, or they do not, or (in some more complex scenario) they partially animate it. But the determination of other participant roles, such as ‘principal’ and ‘target’, is more complex. Principality, for instance, establishes a speaker as committed to the sentiments expressed in his or her utterance. But the speaker’s beliefs and intentions are private rather than transparent. Who or what, then, determines whether or not this speaker is a principal? How are such potentially ambiguous roles allocated? Goffman’s own work does not readily resolve this matter for us. In his paper ‘Footing’ (1981a) Goffman suggests that participant frameworks are established through the more or less straightforward translation of speakers’ intentions to metapragmatic contextualization cues that are taken up and accepted by other participants. Participant frameworks shift in tandem with the cues contained in the speaker’s tone of voice, pitch, stress, posture, visual cues, word choice, and so on (pp. 128, 145, 147, on). While Goffman does not explicitly say so, it is not a leap to presume from his examples that the speaker has the authority to determine most participant roles in a communicative encounter. At the same time, however, one can locate elsewhere in Goffman’s work the implication that participant frameworks are not wholly determined by speakers; instead, they must be jointly constructed with other interlocutors. In his discussion of ‘The Lecture’ (1981b), Goffman describes how an audience who does not play along with the lecturer’s cues can throw off the lecturer’s attempt to shift from the text-authoritative I to a more informal on-line animator I. Such examples suggest that participant frameworks take shape from the interactive engagement of all interlocutors. The speaker’s intentions and actions alone, in other words, do not determine what game everyone thinks is being played, for participant frameworks are the product of dialogic collusion by multiple actors. Such a dialogic approach to J. McIntosh / Language & Communication 25 (2005) 39–60 45 meaning is increasingly championed by linguistic anthropologists and certain sociolinguists (see, for instance, Clark, 1996; Duranti, 1992; McDermott and Tylbor, 1995; Tedlock and Mannheim, 1995). We see a dialogic approach in Irvine’s analysis of Wolof insult poetry, when she describes the determination of participant frameworks as ‘a participants’ problem, to which there are creative, if often evanescent, solutions’ (1996, p. 136). ‘Participants’, here, include all interlocutors, and the joint construction of a participant framework is ‘part of the pragmatic reasoning by which [interlocutors] interpret an utterance and understand its significance’ (p. 140). If participant frameworks are jointly constructed, there are a number of metapragmatic devices with which they can be established. As Goffman suggests, speakers may signal their stance with the use of contextualization cues – or ‘keys’ (Goffman, 1974; Hanks, 1984) – such as going into a falsetto voice to indicate reported speech or using eye contact to designate targets and/or addressees. Speakers can also use other metapragmatic mechanisms such as an embedded frame that shifts the projected participant roles (‘‘At that point he turns around and yells at me: ‘Never come here again!’ ’’). Interlocutors may engage in metapragmatic repair to rectify potential misunderstandings and establish mutual agreement about the participant framework (‘‘Oh, you mean you felt that way.’’) And interlocutors may use subtle metapragmatic devices to indicate that they accept the participant roles projected in the speech of other interlocutors – as in Clayman’s description of politicians who verbally collude in the construction of journalistic interviewers as ‘neutral’ parties who are not committed to the opinions they have put on the table (Clayman, 1992). All such devices constrain the range of viable interpretations of the participant framework, helping participants to agree, or at least think they agree, about the attitudes at stake. And the general consensus seems to be that in ordinary conversation, when the appearance of agreement over the participant framework falters, there is sufficient improvisational flexibility for repair to take place until mutuality seems to have been achieved. So far, we see that successful conversation is predicated on successful mutual alignment, in which participants are, or think they are, or at the very least act as if they are in agreement about the participant framework(s) that inflect their talk. But how do interlocutors contend with participant frameworks in contexts where alignment mechanisms are harder to come by? In particular, how do participants contend with ritual settings that do not allow for spontaneous repair in the same way that conversation does? As Hanks reminds us, certain genres of talk come with ‘default’ assumptions about the participant roles that attach to them (Hanks, 1990, p. 153, 195 – 6). Weddings, funerals, graduation ceremonies, the Eucharist, and a host of other rituals that performatively effect a change of state while reinforcing the social order typically fall into this category (see Rappaport, 1979 for a discussion of the sincerity conventionally projected onto participants in many such rituals). In contemporary American weddings, for instance, those who take wedding vows are treated as the principals of what they say, while the officiant’s homily on the significance of marriage targets the wedding couple and sometimes, more diffusely, the onlookers. In the Samoan fono meetings described by Duranti (1992), male orators speak on behalf of 46 J. McIntosh / Language & Communication 25 (2005) 39–60 powerful political figures in a conventional context that de-emphasizes of the orator’s mental state of sincerity (and hence, his principalship), but places great weight on another participant role not described by Goffman: responsibility for the perlocutionary effects of what the orator says. Other ritualistic events such as greetings, bank robberies, and theatrical performances also tend to be metapragmatically predetermined according to their genre. In Western theater, for instance, audience members typically assume that actors are merely animating words authored by someone else, and are not the principals of what they are saying. The range of possible participant frameworks in such contexts is regimented by expectations that attach to the ritual genre, and sometimes reinforced by explicit metapragmatic cues along the way (‘This is a stick-up’; ‘Do you, Mary, take John...’). Still other rituals, however, tend toward metapragmatic ambiguity, and oppositional rituals seem especially prone to this. Oppositional rituals are rituals – or periods of time within rituals – in which a liminal space is opened to participants, allowing ordinary social rules and mores to be suspended and violated while providing a space for counter-hegemonic discourse (Turner, 1977). Often these collective violations are enacted by those most subordinated by ordinary social structure. Nevertheless, oppositional expressions in ritual texts do not intrinsically imply a particular socially accepted meaning or functional outcome. On the one hand, aspects of the ritual text suggest subversion; on the other, ritual form has the potential to defuse potentially subversive messages, because ritual is usually sharply bracketed from everyday life by a special space, time, and behavioral protocol, its boundaries serving as quotation marks that set it off from the continuous stream of ordinary happenings. Other confounding factors can include ambiguities in the ritual text itself, or idiosyncratic cues given by ritual performers to suggest – or confuse – particular public interpretations of what is going on. Rituals, then, can be characterized in terms of two commonly (though not inevitably) overlapping continuua: one, of metapragmatic and denotational explicitness; the other, of hegemonic enforcement. On one end of the explicitness continuum, we find rituals with relatively limited and clear affordances, and it is here that we also tend to find those rituals that are maximally geared toward retrenching the status quo. Indeed, Silverstein’s account of ‘the nature of what we recognize to be ritual’ characterizes ritual as both maximally explicit and maximally hegemonic; highly organized and ‘self-grounding’ in the ‘cosmic absolute’, thereby reinforcing folk cosmologies and folk ontologies (2003, p. 203). But toward the other end of the explicitness continuum are metapragmatically and denotationally ambiguous rituals with multiple unresolved affordances. These are rituals of liminality that do not obviously enforce hegemony, rituals in which some may locate the ‘hidden transcripts’ of ‘resistance’ (Scott, 1990) but others may argue the affordances are too cryptic or too multiple to be identified as ‘resistant’ without exhaustive attention to the local context (Gal, 1995). For any given oppositional ritual, then, how can we approach the locally specified relationship between the ritual and power relations in ordinary social life? One way to break this question down is through the vocabulary of participant frameworks. Are ritual participants taken to be the principals of oppositional ritual content, or J. McIntosh / Language & Communication 25 (2005) 39–60 47 not? Are onlookers who represent hegemony taken to be the targets of this content, or not? These are difficult questions, and scholars have taken varied approaches in addressing them. Few interpreters of oppositional ritual have drawn explicitly on the notion of participant frameworks, but many have debated whether such rituals should be understood as inconsequential play or celebrated as critiques with the potential to alter the status quo. And yet many draw their conclusions about the ‘resistant’ meaning of oppositional rituals without a holistic examination of textin-context. Bakhtin, for instance, has optimistically championed the message of ‘revival and renewal’ (1968, p. 7) in medieval carnivals based largely on the content (text) of such rituals – and not, of course, on discussions with ritual participants and onlookers. Raheja, too, has extrapolated from the oppositional lyrics of women’s songs in northern India, claiming that the songs provide ‘a ‘hidden transcript’ (Scott, 1990) that challenges some fundamental tenets of the dominant discourse’ (Raheja, 1994, p. 122). Babcock, by contrast, seems to privilege the intentions of those who perform oppositional ritual, suggesting that too many Western interpreters have dismissed Native American ritual clowning as mere incomprehensible play because they have focused too narrowly on the ritual texts. Instead, Babcock suggests, we may come to a different interpretation of these rituals, one that sees them as ‘a remark on the indignity of any closed system’ and a ‘significant form of sociocultural commentary’ (Babcock, 1996 [1984], p. 2,12), if only we ‘ask clowns what they’re up to’ (p. 8). Texts and intentions are certainly worth studying, but they do not tell the full story about a ritual’s significance at a broader social level. To ground one’s interpretation of oppositional ritual solely in the text or the intentions of ritual participants is to overlook the role of the performance context and the ways in which society at large frames the ritual event (see Gal, 1995). 2 Fortunately, some scholars have attended to the contributions of these broader forces. Mageo, in her assessment of carnivalesque and possession rituals in Samoa, concludes that cultural hegemony remains ‘tenacious’ despite the expressions of dissatisfaction found in ritual texts (Mageo, 1996, p. 31). The reasons for this failure are several, and include the fact that some of these ritual expressions take place in entertainment contexts, or are considered by their audiences ‘merely funny’ (p. 50). And in his case study of ritual wailing by Warao women, find Briggs that when women criticize social authorities through song, their words are taken seriously by onlookers who believe that women are able to sing only the truth (Briggs, 1992, p. 341). In these songs, he concludes, womens’ voices are treated as ‘potent’; they are ‘accorded the right to reflect critically’ on male acts and utterances (p. 356), and the content of their songs has direct social consequences. The resistant potential of oppositional content, then, can come 2 Indeed, scholars such as Humphrey and Laidlaw have argued that ritual cannot or should not be studied in terms of participants’ intentions, because ritual form is explicitly designed to ‘[sever] the link... between the ‘intentional meaning’ of the agent and the identity of the act which he or she performs’ (Humphrey and Laidlaw, 1994, p. 100). 48 J. McIntosh / Language & Communication 25 (2005) 39–60 to fruition or not, depending upon the socially constructed participant framework that surrounds it. If few scholars have attended to the ways in which social players jointly interpret the meaning of oppositional rituals, fewer still have attended to the potential for ambiguity and indeterminacy in such interpretations. An exception is Laura Ahearn’s analysis of Tij songfests in Nepal (Ahearn, 1998), in which female singers accuse men of misdeeds, complain about in-laws, and otherwise oppose the status quo. Ahearn suggests we not seek ‘definitive interpretations’ of these songs, but instead explore the textual and contextual factors that constrain the indeterminacy of the song’s meanings (p. 61). To this end, she engages in a multi-pronged discussion of contextual factors such as the ‘social organization, personal histories, spatial configurations, and relationships of the performers and audience members’ (p. 61), thus considerably enriching her assessment. She also explores the songs’ participant frameworks, though she seems to locate participant roles alternately in the texts themselves (as in her discussion of the ‘figures’ in the song lyrics) or in the private intentions of women (as in her claim one would need to know a good deal about a particular woman’s ‘background’ to know whether she is a ‘principal’ of a given song [p. 71]). Ahearn’s approach to participant roles, then, does not explore the possibilities for joint construction of – and joint disagreement about – these roles. Hence, while I applaud Ahearn’s thorough attention to context and indeterminacy, my analysis of Giriama women’s songs differs from hers. Whereas Ahearn focuses on constraints upon ritual meaning, I approach the issue of indeterminacy through a focus on the textual and contextual affordances that create metapragmatic ambiguity and confound joint agreement upon a participant framework. In the absence of clear cues within the text or the performance itself, in the absence of the improvisational repair used in conversation to achieve alignment, and in the presence of social upheaval that throws the roles of women into doubt and in so doing confounds genre-based interpretive conventions (if they ever existed for Giriama funerals), there is room for considerable folk disagreement about the significance of Giriama women’s lascivious funerary songs. To understand all of this, let us begin by looking at the song performances themselves. 3. Women’s obscenity in Giriama funerary rituals, Malindi In the town of Malindi and its outskirts, both men and women compose Giriama funerary songs, and some of the songwriters have a degree of notoriety. Yet song authorship retains a kind of flexibility, because songs can be modified during their performance. Typically, the female singers gather around a hired male song leader who selects and initiates the songs. It is not unusual, however, for a particularly bold woman to initiate a song of her choice, and to be eagerly followed by the female crowd. Songs usually follow a call-response pattern in which the leader (the hired man or an adventurous woman) sings solo through a line that is repeated, sometimes with formulaic embellishments, by the crowd of women. This arrangement allows the solo singer the option of altering the lyrics somewhat – changing the names in the J. McIntosh / Language & Communication 25 (2005) 39–60 49 text, for instance, or slightly modifying the narrative – and the group usually adopts these novelties quickly. In some Giriama communities men as well as women dance to funeral songs, but among Giriama close to Malindi it is quite unusual for men to join the women unless pulled by a woman into the dancing circle. Typically, the women who wish to dance form a circle around the corpse while dancing and singing; some also create percussive effects by clapping or shaking winnowing trays filled with glass. 3 The dances involve much hip-shimmying, and at the climax of many songs the singers face the center of the circle, clapping vigorously and stepping in place. On their whim, women may jump into the center to gyrate against and simulate sexual intercourse with one another, the song leader, and anyone else, male or female, they wish to pull into the center of the circle. These improvised enactments of sexual license underscore the lascivious and normally forbidden content of the song lyrics, and vice-versa. I have chosen the following six songs to give a sense of the breadth of their sexual content, their oppositional potential, and their internal pragmatic ambiguity, which I analyze below. (Words and phrases in single brackets are literal translations; those in double brackets are implied but not explicit in the sung lyrics. I describe some of the semantic complexities of these lyrics in Footnote 4.) 1. 2. 3. 4. Song One Topola ni topo Atopolwaye ni mche msichana Mkaza mtu ni wari na nyama Ndafa ni ole niriche kutopola 1. 2. 3. 4. [[Let’s]] screw [penetrate]! It is soft. An unmarried non-virgin will be screwed [penetrated] A man’s wife is maize meal and meat I will die and stop screwing [penetrating]. 5. 6. 7. 8. 9. Song Two Charo Baya anamanya kusumba Kodzola haraka namala nikagite Mulumangu khani homba vizho. Charo wa Baya udzanidima wee. Mbolo ni bomu chunu changu Katana sikwadima. 5. 6. 7. 8. Mr. Charo Baya knows how to twist Ejaculate fast; I want to [[get home and]] cook! My husband doesn’t fuck me that way. Charo Baya you’ve won me, you. 3 There are two song varieties, kifudu and kihoma; these have slightly different implications for the dances and the percussion that accompany them. I explain some differences between these forms and their histories in McIntosh (2001). 50 J. McIntosh / Language & Communication 25 (2005) 39–60 9. [[Your]] penis is big; my hips can’t hold you, Katana Song Three 10. Nehomba msichana Goshi 11. Nakuamba hahaa mutumia 12. Yuna tsama mino bichembe 13. Ana mbolo mbidzo muno 10. 11. 12. 13. I fucked a girl at Goshi And she told me – old man, You are sweet, my Bichembe [[Bi ¼ father of]] He has a very nice penis. Song Four 14. Mlume wa wivu khana mwiri wee 15. Magulu mahutsu na machende mare 14. A jealous husband will not grow fat, you – 15. [[He only has]] thin legs and long testicles. Song Five 16. Juzi nakwala chende 17. Yanibwaga were ni bomu sana 18. Yo chende here gogo 16. Yesterday I stamped on a testicle 17. It made me fall down 18. Because it was as big as a log. Song Six 19. Mala ni ronge asena mimi mala ni ronge uganga 20. Funula joho yangire mbolo 21. Thiye yunadza kupiga kambi ana kinyakuzi 22. Uganga ni rero 23. Dzire manya natsekera chende 19. 20. 21. 22. 23. I’d like to train, friends, I’d like to train to be a shaman Lift up your dress to let the penis enter Here she comes to the homestead; she’s turned on [she has parasites]. Sexual intercourse [euphemized by the word for ‘shamanism’] happens today! I have come to know I laugh at the sight of testicles. Clearly these texts invoke bawdy misbehavior, particularly given the backdrop of traditional Giriama expectations for female modesty and fidelity. But what is the broader significance of these songs for those in the community? My first interpretive step is to explore whether the song texts themselves contain particular cues that afford (though they cannot on their own determine) one or more particular partic- J. McIntosh / Language & Communication 25 (2005) 39–60 51 ipant frameworks. This exercise involves taking the ritual text out of context – an extraction to which, as Bauman and Briggs (1990, p. 72) remind us, verbal art lends itself. The point is to assess the text’s potential contributions to a participant framework before expanding the analysis to consider the contributions of other factors. Participant frameworks, it is worth reiterating, cannot be extracted directly from a text, but a text is one of several factors – including embodied cues during performance and genre-based expectations – that can contribute to a jointly constructed participant framework. A close look at the song texts reveals a good deal of deictic ambiguity, a phenomenon noted by Ahearn in womens’ Tij songs as well: ‘[I]nterpretive indeterminacy is built right into the lyrics themselves in the form of multiple voices and subject positionings’ (Ahearn, 1998, p. 71, italics hers). Giriama funerary songs involve multiple deictic shifts, both within and between songs. For example, the first person singular pronoun shifts between male and female voices, the sex of these voices being readily inferred from the fact that Giriama presume a heteronormative social universe in which one has sex with and sexually teases members of the opposite sex. The first person singular in line 4 of Song One, for instance, presumes a male voice by citing the male penetrative role in sexual intercourse, whereas the first person singular in line 16 of Song Five claims to have ‘stamped on a testicle’, and hence presumes a female voice. The first person female voices in some of these songs might suggest a participant framework in which the female singers are taken to be the principals of what they sing. Then again, the next song text is perfectly likely to present a male voice, thus dissolving any straightforward adhesion between song and singer. We see further lyrical indeterminacy in the fact that the deictic relationships within a single song can shift dramatically as a song proceeds. Song Three, for instance, opens with a male voice bragging about a female figure: ‘I fucked a girl at Goshi’ (line 10). In the middle of line 11, the same voice reports the words of this ‘girl’ – thereby switching her into the role of an animator in a frame embedded in the man’s report: ‘old man/You are sweet, my Bichembe’ (lines 11–12). But line 13 engages in an abrupt deictic shift that destabilizes this framing. The third person pronoun in the exclamation ‘He has a very nice penis’ suggests we are no longer hearing the old man report what the girl said to him, but rather hearing the girl’s voice as she addresses someone else altogether, while the old man becomes a mere figure. Similar shifts occur in Song Six. Line 19 opens with a female voice: ‘I’d like to train, friends, I’d like to train to be a shaman’ (among Giriama in Malindi, it is almost always women who engage in such training). But in line 20 there is a shift to what is presumably a male voice that salaciously addresses the female: ‘Lift up your dress to let the penis enter.’ In line 23, the female voice returns to report her delight in or mockery of male genitalia: ‘I have come to know I laugh at the sight of testicles.’ Not only do the voices shift in mid-song, but so too do the addressees implicitly targeted by the voices, and the figures they conjure up. For instance, in Song Six, lines 20 and 21 suggest a male voice, but shifting addressees. ‘Lift up your dress to let the penis enter’, says line 20, suggesting that a male voice speaks to a female 52 J. McIntosh / Language & Communication 25 (2005) 39–60 addressee, perhaps the same female who in line 19 says that she would like to train to be a shaman. Yet in line 21, a male voice announces: ‘Here she comes to the homestead; she’s turned on’ – a line that does not address a specific woman but instead sings about a woman as a figure. In another example, Song Two begins with a female voice complimenting her lover, ‘Charo Baya’ – but by line 9 the narrator seems to be addressing an entirely different man, ‘Katana’, a shift none of my informants could explain. Song One, meanwhile, shifts figures mid-narrative; a male voice in line 2 sings first about sex with an ‘unmarried non-virgin’, then in the next line expresses the satisfaction he takes in ‘a man’s wife’. 4 While we cannot read directly off the voices and deictics of a ritual text to the participant roles as jointly constructed by those in Giriama society, textual deictics can provide keys or cues to participant roles. In the metapragmatically explicit Western wedding ritual, for instance, the vows of the bride and groom are constructed with deictics that suggest a consistent mapping of sentiments described in the ritual text (‘I will love you and cherish you from this day forward...’) onto the wedding couple as they say their vows. But in Giriama funerary song, the vague and fleeting figures evoked by the songs – Charo Baya one moment, Katana the next – weaken the ability of these songs to adhere neatly to any living target, while the many deictic shifts in the voices create an ambiguous relationship between song and singer. The shifts themselves, in fact, might suggest or ‘key’ the notion that these performances are a kind of joke – a series of mock identities fancifully tried on – yet they also return again and again to the voice of the impudent, desirous woman. Overall, the texts neither clearly frame the singers as principals nor clearly provide singers with an alibi to absolve them from the sentiments in the text. 4 The ‘shiftiness’ in these deictics is mirrored by a playfulness in the semantic domain. Funerary song lyrics frequently use words that have one meaning in ordinary conversation and another, more illicit one in the song. The word ‘topola’, in Song One, for instance, ordinarily means to eat a soft food such as a mango; in the funerary context it means to have sex from a male, penetrative point of view. In Song Six, several lyrics imply a homology between the terminology of shamanism and sexual innuendo. When the female voice says in line 19: ‘I’d like to train to be a shaman’, the subtext in the funerary context is that she’d like to have intercourse. (I presume that this double meaning may emerge from the fact that both shamanism and sex have potentially creative and destructive potencies.) In line 22 of the same song, the term ‘uganga’, which normally refers to ‘shamanism,’ is taken instead to mean ‘sexual intercourse’. Another double meaning is invoked in line 21, when a male voice says of the woman ‘in training’ that she has ‘kinyakuzi’, a type of parasite that can make women scratch their vulva – the implication being that this scratching reflects and/or enhances female sexual arousal. Some of these double meanings are so obscure that not every singer is aware of them, although they do realize that much of what they produce is as semantically peculiar as (in the words of one informant) a ‘word salad’. The double meanings of song lyrics certainly provide a contextualization cue that what is being said is different from ordinary conversation. But do they afford any particular participant framework? On the one hand, the apparent playfulness and salaciousness of some of the code might suggest that funerary songs are mere fun and games, and so should not be seriously yoked to the sentiments of the singers. On the other hand, the use of polysemous code invokes another special context: that of shamanistic language, which often engages in word substitutions during divination. Shamanism is regarded by most with high seriousness. Hence, the semantic domain contributes to the affordance of multiple, potentially contradictory participant frameworks. J. McIntosh / Language & Communication 25 (2005) 39–60 53 What about the potential keys offered by the singers’ behavior as they perform these songs? Initiating a song, singing with a particular demeanor, and moving one’s body through space in particular ways could contribute to the emergence of a participant framework. In Giriama funerals, though, the singers’ behaviors create further ambiguities. Songs are often initiated by the male song leader, which could suggest that men instigate these violations of female modesty, and provide an alibi for singers. At the same time, however, it is not unusual for women to override the male song leader and to launch songs themselves. As for the singers’ demeanors, some women look down while singing, a sign of shyness that might conceivably be used to dissociate themselves from the songs’ content. Still others look up and laugh, pull men into the dancing circle, and flirt with men when they are not singing and dancing, suggesting enthusiastic involvement. Even the same woman may not behave consistently from one song or funeral to the next. It is quite possible that individual women stand in individualized and shifting relationships to the content of these funerary songs. Obeyesekere (1981) has pointed out that different individuals may project different meaning onto the same ritual symbols, depending on their particular social and psychological histories. Analogously, the content of funerary songs might give rise to a range of stances on the part of individual singers. One ritual participant may feel committed to the sentiments expressed in an oppositional ritual text, and might imagine directing the most cutting critiques at her husband, while another individual may sing the same song with uncertainty or indifference about its application to herself and the men in her life. It is impossible to know how many gradations there are in between these stances. Still, I did talk to quite a few singers about their attitudes towards funerary song lyrics. Most of the twenty or so singers I spoke to about this matter saw me as a familiar and friendly face, but they may have wanted to protect their privileged moment of expression by denying that any real expression goes on at all. What I found was that all the singers I spoke to defended their performances with alacrity by strictly denying that they endorsed the sentiments in the dances or song lyrics, and denying that the words were intended to apply to any men in their lives. I can’t know whether these women felt they were telling me the truth – but what is important for present purposes is that they apparently hoped they would not be considered principals of sexual songs. The most common locution I heard from them was that funerary songs are ‘just our tradition’, a phrase that largely absolves the woman of desire or agency in the act of singing. Many women explained the ‘dirty words’ (matusi) of funerary song as an innocuous, playful means of distracting the bereaved from their grief. They also freely acknowledged that what they sing is ‘bad,’ but argued that the corpse provides license for such words by transforming the funerary space into a ‘bad place’; obscenity, in this formulation, emerges as part of the liminal and unsettling context created by death, rather than from internal desires of the singers themselves. According to the women, then, the men they know are not the intended targets of the song lyrics, and they themselves are not the principals of the song content. The voices in the songs that rupture the sexual order are merely those of imaginary theatrical figures. 54 J. McIntosh / Language & Communication 25 (2005) 39–60 In ordinary conversation, a speaker could engage in repair to make these intentions (or ostensible intentions) clearer should they be misunderstood. Ritual, however, does not offer the same improvisational opportunities for repair of alignment. Should an onlooker misconstrue the women’s ostensible intentions as they sing, the structure of ritual singing and dancing does not offer this onlooker a clear opportunity to express their view, nor the singers a clear chance to rebut, so the misalignment between performers’ intentions and onlookers’ interpretation goes unaddressed on the spot. The women I spoke to apparently hoped that an interpretive convention – the weight of ‘tradition’ – might secure alignment around a common, genre-based participant framework, one that exonerates them. And yet, there is no agreement within their community about the participant framework at play. This became increasingly evident to me as I overheard cynical remarks about the singers and listened to men debate whether they would allow their wives to sing and dance at funerals. Realizing that this was a contested matter among community members, I spoke to dozens of individuals who do not sing – including men of all ages and women who sit the songs out – asking them about the significance of funerary song lyrics, with a focus on whether the songs are taken to reflect anything about the singers themselves. I received two types of responses, each in contradiction with the other. On the one hand, many were quick to echo the singers with a stock response that these songs are ‘just tradition’ or ‘just an old custom’ and hence inconsequential. As one man put it: ‘There’s no harm in singing funeral songs,’ while another said, ‘A husband cannot get mad for what happens at a funeral because women are just expected to sing those songs. It [the singing] doesn’t matter.’ Like the singers themselves, many explained that the songs are a diversion, an antidote to tragic emotions. Such onlookers construe the singers as animators of a benign, traditional form of entertainment designed to startle or entertain the mourners out of their sorrow. Others linked the sexual imagery to the possibility of fertility, or explained that the ‘bad words’ in funerary songs symbolize the dissolution and chaos brought by death – ‘Things have already turned to shit, so why not let it all hang out’, as one informant put it – but still did not believe that the singers themselves endorse the sentiments within the songs. Despite these dismissals of the possibility of real subversion in the songs, I also found currents of profound disapproval and alarm running through the community, particularly among male and female elders and younger men, both married and unmarried. Many of the older individuals I spoke to suggested that the young girls who participate in funerary songs and dances ‘don’t respect their elders anymore’ and are prone to promiscuity. Some of the women who don’t participate told me their husbands try to protect their reputations by forbidding them from doing so. A number of men confided to me that the women who perform at funerals are licentious; as one put it, ‘These rituals are danced by whores.’ Indeed, a group of unmarried men jokingly referred to the singers as ‘crown land’; like the terrain formerly owned by the colonial government, they are host to multiple ‘squatters’. Other disapproving spectators tell me the singers ‘don’t care’ about their reputations and fail to give men the respect they should. A number of young men say that while they might be tempted to sleep with a woman who enjoys performing the songs and J. McIntosh / Language & Communication 25 (2005) 39–60 55 dances, they would not marry her – suggesting that the free expression in the performances is sexually attractive, but forebodes inferior virtue outside the ritual setting. The tales of young men pulling singers away during funerals to have sex with them also suggests that such spectators read the desire articulated in the songs as expressions of loose mores. Strikingly, one young man explained to me that female singers are the principals of even those song lyrics that are enunciated through a male voice in the text. ‘When a woman sings ‘nakuzika kitu’ [‘I’ll bury something in you’ – a lyric that plays on the funerary and sexual meanings of ‘bury’]’, he insisted, ‘she’s pretending to be a man, but in her heart she wants to have something ‘buried’ in her’. These onlookers, then, presume that the female singers are the principals of the lascivious texts, and real men in their lives are their targets. They draw an intertextual connection between the sexual lyrics in the ritual context and sexual events (and, perhaps, utterances) in ordinary life before and after the ritual. Whereas the singers themselves hope the relevance of the ritual texts will be taken as restricted to the ritual itself, some of their interpreters regard the songs as all of a piece with, even a training ground for, enduring promiscuity. 4. Social change and unsettled participant frameworks Where might these antithetical pragmatic interpretations – one benign, the other projecting danger – emerge from? Is this duality a mirror of the ambiguous affordances in the song lyrics and the ambiguous behavioral cues offered by the singers? Probably in part, for the semiotic ambiguity of the textual voices and the singers’ behaviors could support either projected participant framework – and, as I have said, a ritual like this one offers little chance for on-line ‘repair’ that might assure alignment. But there is a broader motivation for the divergent folk interpretations of this ritual, having to do with overarching shifts in Giriama patriarchy and Giriama sexual behaviors. This portion of my analysis, then, is an effort to address what Bauman and Briggs call the ‘perennial micro-macro problem of how to relate the situated use of language to larger social structures’ (1990, p. 79). My argument here is that the disagreements over the pragmatic significance of these ritual texts emerge in part from the disruption of conventional patriarchal structures and the emergence of a fraught sexual ethos in society-at-large. As I have mentioned, traditional Giriama culture was (and, in some places, is) structured in a fairly strict patriarchal order. Homesteads are made up of several polygynous households headed by a senior man who makes decisions about property allocation, conflicts, and patrilineal rituals. Parents and other elders oversee marital arrangements, including bride-price transactions, and women’s rights to resources are severely constrained by a patrilineal inheritance system. Sexual courtship and behavior are meant to be highly discreet, while sex itself is believed linked to supernatural powers, including the kirwa illness caused by adultery and mathumia: controlled ritual sexual intercourse between spouses (often, senior males and their first wives) that brings prosperity and purification when performed at moments of significant 56 J. McIntosh / Language & Communication 25 (2005) 39–60 social transition (Udvardy, 1990a,b). Udvardy argues that these rites often depend on the leadership of male homestead heads, and therefore re-inscribe their seniority (Udvardy, 1990a, p. 87). Today, members of the younger generation believe in kirwa and follow mathumia restrictions only erratically, but women are still expected to resist advances, defining their decency through modesty. I was repeatedly told, for instance, that unlike men, women should not initiate sexual liaisons verbally. ‘A woman can say it [express desire] with her body’, my informants tell me, ‘but not with her mouth’. The female body’s language, furthermore, should be subtle and indirect: a glance, a fleeting halfsmile, or a slight, sinuous motion. If sexual expectations continue their policing function, sexual behavior in Malindi town constantly threatens to violate these proscriptions. While Giriama communities everywhere are faced with changes, Malindi’s economic and cultural situation has created exceptional tensions. The cost of living has risen steeply, and many young families are driven from rural areas toward Malindi so that men can seek wage labor and women can sell agricultural surplus in the markets. Having left traditional socioeconomic arrangements, and being subject to new pressures and new opportunities, some families rapidly dissolve. In the small villages flanking Malindi town, household arrangements have an improvisational feel, and many divorced, abandoned, or single young women live on their own, supporting themselves and their children rather than returning to their natal home. As traditional arrangements break apart, male homestead heads no longer orchestrate sexual conduct through arranged marriage and mathumia. As women and girls improvise new sexual identities, fears of adultery mount. Women’s potential promiscuity spells a drop in male control over female sexuality (and, by implication, paternity), giving men reason for discomfort. Meanwhile, in recent decades, numerous upcountry Kenyans have moved to the coast for economic opportunities, bringing versions of modernity in tow, including nuclear family structure, consumerist aspirations, and ‘Western’ sexual behavior with demonstrative pre-marital passion. European and American tourists flock to the coast in search of a ‘sex safari’, and prostitution has filtered into the villages that surround Malindi, providing a financial safety net for women with limited economic opportunities. Giriama women who don’t sell sex are also changing their sexual habits. One Giriama man in his late thirties told me that he had recently propositioned a woman for sexual relations, but was so shocked and revolted when she accepted on the spot (rather than putting him off time and again as he expected) that he withdrew the suggestion. How to relate these gendered changes to the folk interpretations of women’s salacious funerary songs? I don’t want to idealize the past with the suggestion that Giriama societies were ever free of internal conflict about gender roles, or ever entirely secure in thinking that women’s sexuality was well under control. However, contemporary social forces in and near towns like Malindi have placed new strains on patriarchal structure, making it very difficult indeed to enforce. And surely the anxieties particular to this historical moment are relevant to contemporary interpretations of women’s acts and utterances in funerary songs and dances. The shifts in J. McIntosh / Language & Communication 25 (2005) 39–60 57 homestead organization, gender roles, and sexual activity seem to have fomented currents of suspicion that motivate some ritual onlookers to make intertextual links between ritual expressions of desire and desire outside the ritual context. And if there was ever an historical moment in which women were more firmly under patriarchal control, it may have been easier then for onlookers to frame the songs as ‘mere custom’; a means of celebrating life while distracting the bereaved. While some women may have taken private delight in these songs, the notion that they might be iconic of actual female desire or sexual behavior may have been more readily dismissed by society at large. Such a participant framework would construe women as mere animators of fanciful narratives that have no concrete principals or targets and no important relationship to acts or utterances outside the ritual context. Today, however, the new possibilities for female sexual behavior seem to inform and motivate the conflicts over which projected participant framework should hold sway. 5. Conclusion Apparently, the same oppositional text can potentially have multiple significances in situ, meanings that cannot be read directly off the text or off performers’ intentions. Understanding the broader social context is vital to understanding the way in which a ritual is framed, and as the Giriama case suggests, shifts in social structure that give rise to widespread anxieties may have a crucial impact on the participant frameworks projected onto a ritual. In some respects, my conclusions resonate with those of Gluckman, who suggested that where there is little threat of misconduct outside of the ritual context, ritual rebellion can occur unobstructed. But, he predicts, ‘where the [social] relationships involved are weak, there cannot be license in ritual’ (1965: 132). Contemporary Giriama funerals provide an example where the social relationships in question are not secure, yet license in ritual remains; instead, it is the interpretation (or interpretations) of the ritual that may shift. Not only can oppositional ritual endure in the face of tenuous social relationships; it can also recur again and again in the presence of serious social disagreement about its ‘meaning’ – misalignment of a sort that does not tend to occur in ordinary conversation without being addressed. This study of the situatedness of a ritual text, then, is hardly intended to hammer down ‘the’ significance of oppositional rituals, for such a thing is not possible given the many different ways that context can inflect them. The potential for ‘resistance’ may be present in – afforded by – a ritual text, but its ability to unsettle hinges on its perceived significance, and this, in turn, may hinge on broad aspects of social context. In the Giriama case, if each new performance of funerary ritual creates unease for some in the community, the result seems to be the policing of female sexuality, as seen in the cases of men who forbid their wives to participate. We also see incipient signs of a model of womanhood in which those who dance (and are suspected, correspondingly, of being sexually loose) might be considered desirable but are too sexually sullied to be suitable candidates for marriage. If funerary song lyrics contain ‘resistant’ potential, then, this does not have a clear liberating effect on the ground. 58 J. McIntosh / Language & Communication 25 (2005) 39–60 The distance between ritual text and social outcome, it seems, is a long one, mediated in large part by participant frameworks that may answer to and enhance social anxieties, thereby perpetuating the very constraints the ritual text militates against. Acknowledgements Thanks to Gaye Thompson, Kainga Kalume Tinga, and especially Maxwell Phares Kombe for their insight into Giriama culture and language, to Jim Herron and Bruce Mannheim for invaluable conversations about participant frameworks, and to Paul Manning for his helpful reading. I am deeply grateful as well to the Giriama friends and acquaintances who so generously aided me in this and other projects. This research was carried out with a grant from the Fulbright-Hays Foundation. References Ahearn, L., 1998. ‘A twisted rope binds my waist’: locating constraints on meaning in a Tij Songfest. Journal of Linguistic Anthropology 8 (1), 60–86. Babcock, B.A., 1996 [1984]. Arrange me into disorder: fragments and reflections on ritual clowning. In: Grimes, R. (Ed.), Readings in Ritual Studies. Prentice-Hall, Upper Saddle River, NJ, pp. 1– 21. Bakhtin, M.M., 1968. Rabelais and his World (Helene Iswolsky, Trans.). MIT Press, Cambridge. Bauman, R., Briggs, C.L., 1990. Poetics and performance as critical perspectives on language and social life. Annual Review of Anthropology 19, 59–88. Bell, C., 1992. Ritual Theory, Ritual Practice. Oxford University Press, New York. Bourdieu, P., 1977. Outline of a Theory of Practice (Richard Nice, Trans.). Cambridge University Press, Cambridge. Briggs, C., 1992. ‘Since I am a Woman, I Will Chastise my Relatives’: gender, reported speech, and the (re)production of social relations in Warao Ritual Wailing. American Ethnologist 19 (2), 337–361. Briggs, C., 1996. The meaning of nonsense, the poetics of embodiment, and the production of power in Warao healing. In: Laderman, C., Roseman, M. (Eds.), The Performance of Healing. Routledge, New York, pp. 185–232. Clark, H., Carlson, T., 1982. Hearers and speech acts. Language 58, 332–373. Clark, H.H., 1996. Using Language. Cambridge University Press, Cambridge. Clayman, S.E., 1992. Footing in the achievement of neutrality: the case of news-interview discourse. In: Drew, P., Heritage, J. (Eds.), Talk at Work: Interaction in Institutional Settings. Cambridge University Press, Cambridge, pp. 163–198. Duranti, A., 1992. Intentions, self, and responsibility: an essay in Samoan ethnopragmatics. In: Hill, J.H., Irvine, J.T. (Eds.), Responsibility and Evidence in Oral Discourse. Cambridge University Press, Cambridge, pp. 24–47. Fernandez, J., 1974. The mission of metaphor in expressive culture. Current Anthropology 15 (2), 119– 145. Gal, S., 1995. Language and the ‘Arts of Resistance’. Cultural Anthropology 10 (3), 407–424. Gluckman, M., 1965. Custom and conflict in Africa. Basil Blackwell, Oxford. Goffman, E., 1974. Frame Analysis: An Essay on the Organization of Experience. Harper & Row, New York. Goffman, E., 1981a. Footing. In his Forms of Talk. University of Pennsylvania Press, Philadelphia. pp. 124–159. J. McIntosh / Language & Communication 25 (2005) 39–60 59 Goffman, E., 1981b. The Lecture. In his Forms of Talk. University of Pennsylvania Press, Philadelphia. pp. 162–196. Hanks, W.F., 1984. Sanctification, structure, and experience in a Yucatec Maya Ritual Event. Journal of American Folklore 94 (384), 131–166. Hanks, W.F., 1990. Referential Practice: Language and Lived Space among the Maya. University of Chicago Press, Chicago. Humphrey, C., Laidlaw, J., 1994. The Archetypal Actions of Ritual: A Theory of Ritual Illustrated by the Jain Rite of Worship. Clarendon Press, Oxford. Irvine, J., 1996. Shadow conversations: the indeterminacy of participant roles. In: Silverstein, M., Urban, G. (Eds.), Natural Histories of Discourse. University of Chicago Press, pp. 131–159. Levinson, S., 1983. Pragmatics. Cambridge University Press, Cambridge. Levinson, S., 1988. Putting linguistics on a proper footing: explorations in Goffman’s concepts of participation. In: Drew, P., Wootton, A. (Eds.), Erving Goffman: An Interdisciplinary Appreciation. Polity Press, Oxford, pp. 161–227. Mageo, J.M., 1996. Continuity and shape shifting: Samoan spirits in culture history. In: Mageo, J.M., Howard, A. (Eds.), Spirits in Culture, History, and Mind. Routledge, New York, pp. 29–54. McDermott, R.P., Tylbor, H., 1995. On the necessity of collusion in conversation. In: Tedlock, D., Mannheim, B. (Eds.), The Dialogic Emergence of Culture. University of Illinois Press, Urbana and Chicago, pp. 218–236. McIntosh, J., 2001. ‘Tradition’ and threat: women’s obscenity in Giriama funerary rituals. In: Brettell, C.B., Sargent, C.F. (Eds.), Gender in Cross-Cultural Perspective, third ed. Prentice-Hall, Upper Saddle River, NJ, pp. 409–422. Obeyesekere, G., 1981. Medusa’s Hair: An Essay on Personal Symbols and Religious Experience. University of Chicago Press, Chicago. Raheja, G.G., 1994. On the uses of subversion: redefining conjugality. In: Raheja, G.G., Gold, A.G. (Eds.), Listen to the Heron’s Words: Reimagining Gender and Kinship in North India. University of California Press, Berkeley, pp. 121–148. Rappaport, R., 1979. The obvious aspects of ritual. In his Ecology, Meaning and Religion. North Atlantic Books, Berkeley. Scott, J., 1990. Domination and the Arts of Resistance: Hidden Transcripts. Yale University Press, New Haven. Silverstein, M., 2003. Indexical order and the dialectics of sociolinguistic life. Language and Communication 23 (3–4), 191–229. Silverstein, M., Urban, G., 1996. The natural history of discourse. In: Silverstein, M., Urban, G. (Eds.), Natural Histories of Discourse. University of Chicago Press, Chicago, pp. 1–17. Stoller, P., 1989. The Taste of Ethnographic Things: The Senses in Anthropology. University of Pennsylvania Press, Philadelphia. Stoller, P., 1995. Embodying colonial memories: spirit possession, power, and the Hauka of West Africa. American Anthropologist, New Series 96 (3), 634–648. Stoller, P., 1996. Sounds and things: pulsations of power in Songhay. In: Laderman, C., Roseman, M. (Eds.), The Performance of Healing. Routledge, New York, pp. 165–184. Tambiah, S., 1981. A performative theory of ritual. Proceedings of the British Academy 65, 113–169. Tedlock, D., Mannheim, B. (Eds.), 1995. The Dialogic Emergence of Culture. University of Illinois Press, Urbana. Turner, V., 1964. Betwixt and between: the liminal period in Rites de Passage. In: The Proceedings of the American Ethnological Society: Symposium on New Approaches to the Study of Religion. University of Washington Press, pp. 4–20. Turner, V., 1977 [1969]. The Ritual Process: Structure and Anti-Structure. Adline, Chicago. Udvardy, M., 1990a. Gender and the Culture of Fertility among the Giriama of Kenya. Unpublished Doctoral Dissertation in Anthropology, presented to the University of Uppsala, Sweden. Udvardy, M., 1990b. Kifudu: a female fertility cult among the Giriama. In: Anita Jacobson-Widding, A., van Beek, W. (Eds.), The Creative Communion: African Folk Models of Fertility and the 60 J. McIntosh / Language & Communication 25 (2005) 39–60 Regeneration of Life. Uppsala: Academiae Ubsaliensis. Distributed by Almquist and Wiksell International, Stockholm, Sweden, pp. 137–152. Valeri, V., 1985. Kingship and Sacrifice: Ritual and Society in Ancient Hawaii. University of Chicago Press, Chicago.