Michael Silverstein has been so influential in anthropology and other semiotically oriented disciplines that almost any topic I might write about must show the effects of his intellectual importance and relevance. It was Silverstein who inaugurated the study of “linguistic ideology,” the politically infused ideas and rationalizations about language that explain—or seem to explain—the forms, uses, and social indexicalities of languages their speakers use, interpret, or hear about (Silverstein Reference Silverstein, Clyne, Hanks and Hofbauer1979). This general theme, ideologies of language, has come to occupy much of my own work for many years. In the present essay I consider an aspect of ideology that I find especially important: its connection with social positioning and point of view.
To illustrate this connection I have chosen two sets of examples from my fieldwork in Senegal. The first set focuses on a narrative in which a social underling displayed his perspective on his social position—a text in which, to draw on Gayatri Spivak’s expression (Reference Spivak, Nelson and Grossberg1988), the subaltern spoke. Before presenting the narrative and to provide some context for it, I reproduce some local commentary in which higher-ranking persons expressed a view of social difference. The second set of examples, starting with a conversation in a different Senegalese setting, comes closer to Silverstein’s own writings because it concerns pronouns of address and, ultimately, Whorfian effects in language.
First, a word about ideology. The term has meant different things to different people (in this way perhaps illustrating the very point I want to make about it), as Kathryn Woolard has shown in her essay on its history and its various deployments in studies of “language ideology” (Reference Woolard, Schieffelin, Woolard and Kroskrity1998; see also Eagleton Reference Eagleton2007). Why retain this term, which carries so much intellectual baggage? One reason is the conspicuous connection it draws between ideas, power, and politics. Related concepts, such as “culture,” are less clearly congenial to those concerns. “Culture,” after all, can potentially encompass everything people engage in or with, from the loftiest ideas down to the proverbial kitchen sink—although, to be sure, culture can be discussed in such a way as to show the role of power and politics in the kitchen or anywhere else. A second reason, more important for the present essay, is that ideology always implies difference. To call some set of ideas an “ideology” is to imply that there is some other set of ideas with which it contrasts. For example, people who wrote of “Soviet ideology” or “communist ideology” during the Cold War purported to describe a set of political ideas with which they explicitly differed, and which they considered false. For those authors who take “ideology” to be something bad, a connotation going all the way back to Napoleon and his disparagement of the French idéologues—the late eighteenth-century savants who proposed idéologie as the study of ideas—“ideology” is something other people have. It is others’ false consciousness, as opposed to one’s own awareness of truth and reality.
I do not equate “ideology” with false consciousness. Doing so would imply that I have a corner on absolute truth, a claim I would not make, much as truth in writing is something to which I aspire. No mere human can have a God’s-eye view. However, what remains important from these various notions of “ideology” is that ideology is always partial, in both senses of that word. An ideologized view of the world is partial in that it is interested (in the political sense of “interest” as having an involvement in something for one’s own purposes, as when one is an “interested party” in a lawsuit or business deal); interest in this sense must include values and the motives that reflect them. That is, one way to understand ideology is as a regime of value, one that supplies subject positions and roadmaps for action.Footnote 1 The other sense of partial is “incomplete,” and an ideology is partial in this sense too. There is always some other point of view, another perspective on the world, taken by someone whose social positioning and experiences differ.
As the Senegalese examples in this essay will show, ideologized views of the world are always differentiated, even when they overlap in major ways. The greater the difference in social and linguistic experience and competences, of course, the greater the effects of ideological differentiation in people’s values, actions, and what they can infer about an interlocutor. Ideological differentiation may consist in divided regimes of value, but it also, to greater or lesser degree, concerns differences in the background assumptions interactants hold about what can or must be said, or not said, in talk.
Social Stereotypes and Divided Values
As some readers of this essay may know, my fieldwork in a rural Wolof community explored the relationships among linguistic repertoires and social hierarchy. Distinctions of rank have been important for centuries in the social traditions of Wolof and other peoples in the Sahel region of West Africa. The distinctions range from the gross to the minute. The gross distinction consists in a division of society into three major social categories, while the minute distinctions involve calculations of seniority and ancestry such that any two individuals can always be ranked (at least, in principle; the criteria sometimes leave room for contestation). There are other differentiations in between: subtypes within categories, clan groupings, and so on. Since at least as far back as the early sixteenth century these Sahelian systems have been described in terms of “caste” because of their emphasis on essential ranked social differences in professional occupation, potential pollution, and endogamy.Footnote 2
The three major categories in these systems are the “nobles” (géer in Wolof: free persons, gradiently high-ranking); artisans (nyenyo: various professions, free but low-ranking); and “slaves” (jaam: low-ranking and, historically, unfree). The géer include clerics and aristocrats, but most are free farmers, the demographic majority of the Wolof population. They are “noble” only in a moral sense relative to the lower ranks, not in the sense of an elite minority. The jaam are menials—or people thought of as menials—whose ancestors were sold or were captured in war. They can “belong” to members of any social category, so the broad category jaam encompasses crown slaves, slaves of free farmers, slaves of artisans, and even slaves of other slaves. The nyenyo artisans include such professions as smiths (several kinds), leatherworkers, woodworkers, and—most relevant for this article—the griots (also several kinds), who are artisans of the word. The griots are a bardic category whose members, male and female, specialize in the arts of communication: praise-oratory, skillful narrative and conversation management, poetry, rhetoric, maintaining (and reporting) genealogies and histories, news gathering, and related activities.
Griots as a category are much disparaged by nobles, who stereotype them as greedy, lazy, untrustworthy, and dangerous people. Griots are greedy, nobles claimed, because they demand money and gifts for their performances; lazy, because they depend on their patrons for their support; untrustworthy, because they are supposed (again this is a stereotype held by nobles) to do or say “anything” for money; and dangerous, because physical contact with them—at least, contact of the “wrong” kind, in the wrong circumstances—is supposed to endanger a high-caste person’s health. In my experience, although nobles maintained friendly relations with individual griots who worked for them as genealogists, entertainers, and messengers, they nevertheless would not sit on the same mat as a griot, or eat food prepared by a griot unless the dish had passed under the eyes of an appropriate high-caste person. Griot families lived in a separate part of town, drew water from a separate well, and could not be buried in the same ground as nobles. In fact, in past centuries griots were not to be buried in the ground at all, in this part of Senegal. Instead, the body would be placed inside the hollow trunk of a baobab tree—because, I was told, the baobab tree sinned in the beginning of time, and God punished it by jamming it into the earth upside down, the roots sticking up in the air. This is why baobab trees look they way they do (not branching out broadly like other trees), or so the story went. Some stories of the origin of griots—stories told by nobles—have a similar theme, in which the first griot did something evil or lazy at the beginning of time, and his descendants were kept segregated forever after.
Caste segregation and its rationales were taken very seriously in the area where I worked. Text 1 comes from a conversation in which two middle-aged high-caste men, ST and MK, were discussing these matters with me. Linking the principle of social differentiation to inequality, caste endogamy (barriers to intercaste marriage), and high-caste rejection of low-caste behavior, they present caste segregation as God-given and based on an ethic of shame:
As another man put it, “Separating people is very important to Wolof.” He was explaining why griots are buried separately from people of other castes. Nobles’ negative valuations could also apply to griots’ linguistic practices. Their full performance style—its loud volume, high pitch, verbosity, and elaboration—are part of what, in the high-caste view, consigns them to low rank (see Irvine Reference Irvine, Lutz and Abu1990). “The Wolof dislike this,” a high-ranking noble told me. “It deforms the voice.” Although nobles want and need the griots’ praises of them to be heard, nobles would be ashamed to speak this way themselves.
Unsurprisingly, griots are aware of this stigma, although they do not share all the values it assumes. They can try to circumvent the stigma by building an individual reputation, cultivating a relationship with particular patrons as a person who can be trusted to keep patrons’ confidences and, in the rhetoric of public performance, skillfully portray the patron in a positive light. In this way a griot can try to distinguish himself (or herself) from other griots, concerning whom the stigma might go unchallenged. However, in private conversations with me, griots challenged the values pertaining to their performances, emphasizing aesthetics, technique, and historical accuracy, and downplaying the importance of whatever remuneration they might receive. The remuneration, some griots told me, is incidental. Offered by patrons who are moved by the performance, gifts are not what motivates the griot to perform in the first place.
Consider now the case of a griot I knew well who illustrated both these approaches—attempting to circumvent the stereotype and challenging parts of it—with varying emphases depending on the context. I’ll call him MM3. In his early thirties in 1975, he was starting to build a reputation as a skilled performer of praise-oratory and as a trustworthy confidant: someone who could be trusted to carry important messages, to accompany a high-caste patron at political meetings, and to make confidential inquiries on a noble’s behalf. In this last capacity, he offered his services to me. In one of our conversations, he produced the narrative presented here as text 2, providing evidence of his good character and professional skill. These qualities are not simply to be taken for granted, especially given the stigmatized stereotype griots must cope with.
An interesting aspect of the narrative, then, is its presentation of the ethics of griot conduct, and a griot perspective on the relations between castes. The story recounts MM3’s experience performing at a noble patron family’s celebration of their sons’ initiation into manhood. While the text reveals some ways in which he shares in nobles’ understandings of the relationship and of the stigma attaching to his caste, it also shows ways in which he rejects these understandings and defies the stigma. Additionally, the text is itself an example of the griot performance style, with its extended and dramatized narrative and its detailed descriptive vocabulary.
In the narrative, MM3 reports that he was called to the household of a noble family in the Sarr patriclan, a family to which his own lineage segment is traditionally attached. That is, the griot lineage segment has a permanent clientage relationship with these nobles and a special responsibility for knowing and maintaining an account of their genealogy and family history. On arrival, he finds that they are celebrating the completion of their sons’ initiation—the rites of circumcision and traditional instruction. MM3 performs praise-oratory for the family, oratory represented in the narrative only in condensed form by recounting its opening line. The patrons, he reports, fall silent and listen attentively. Afterward, they offer him substantial gifts, and they tell him he is an excellent griot whom all noble families would want to patronize. The Sarrs are fortunate, one of the Sarr patrons adds, that MM3 is specially attached to them.
Notice first the ways in which the griot’s good character and skills are stated in negative terms: not being lazy; not coming around perpetually begging; not ignoring your noble patrons; not abandoning your traditional patrons for people who might have more money. Here, the stigmatized image of the griot as greedy, annoying, and undeserving of reward is represented as what this particular griot is not. (And by implication, other griots are—a suggestion MM3 offered in other conversations.) There are positive values and aspirations too: a griot should be serious, polite, and “recognize what a noble is”—that is, acknowledge the patronage relationship and be aware of one’s responsibilities as a griot to praise the noble publicly. Still, toward the end of the narrative MM3 is back to detailing the stigmatized characterization he represents as not applying to him.
Notice too that praise, in this case commendation of the griot himself, is put in someone else’s mouth. Even a griot cannot praise himself. Complimented, in his narrative, by the initiated boys and their mentor, MM3 can boast to me about his good character, but only by quoting other people. However, in showing why these nobles’ opinion is important and why their compliments matter, the narrative has the nobles themselves speak to their own importance and background—“I never run away,” “I’m a prince,” “I’m giving you a thousand,” and so on—utterances they were most unlikely to make in real life. In real life, it would be the griot who would say these things about the nobles, while the nobles’ gifts to the griot, silently yet ostentatiously displayed, would give mute evidence of both the nobles’ munificence and the griot’s merit. Thus the narrative implies that MM3 is worthy of being praised even by nobles. He has recounted a scene in which his narrated self acts within the conventional griot role (as he also does in recounting a dramatic narrative to me), but the end of the narrative reverses these conventional roles. There, the nobles do the praising, while the griot is silent. Only the complex frame of the narrative, and the fact that it is told in a private interaction with me rather than on some more public occasion, permits this reversal.
While the narrative’s representations of the role and image of the griot and of the griot-noble relationship are consistent in many ways with nobles’ representations of them, there are also some differences. Nobles valued plain speaking and taciturnity (“If you really knew what you were talking about you wouldn’t be talking about it,” as one noble put it); griots valued aesthetic elaboration, dramatic performance skills, and verbal fluency. As I mentioned earlier, several griots declared to me that they valued their craft for its own sake, and did not perform just for the money—the opposite of what nobles declared about them. MM3’s text displays these griot values in its dramatic narrative, rhetorical elaboration, and relatively esoteric or unusual vocabulary. But if the griots’ professional prestige and command of material rewards derive from this activity and fluency, those successes do not translate into higher rank in the broader system. Griots are still segregated from contacts, positions, and locales reserved for other ranks.Footnote 7
It is evident that many ideas and assumptions about rank and its behavioral indexicalities are shared by both nobles and griots. The evidence comes not only from statements describing caste stereotypes—statements that confirm the existence of assumptions in the very fact of denying them, or denying that they apply to oneself. It also comes from observable practices that might seem, on the face of things, to undermine those indexicalities altogether, although they actually do not. These are the practices that pertain to within-caste social occasions. So, at griots’ own family events, honorees and senior family members—although griots, of course, by birth but higher ranking than the other griots present—display a relative taciturnity and immobility that indexes a status “higher than thou” and recalls the characterological image of nobles. Similarly, on occasions when only nobles are present, those who rank lower than others take up, in a small way, the roles and practices otherwise associated with griots: telling little narratives, speaking somewhat rapidly and at higher pitch than others, relaying messages, or making requests. (Even in text 1, we can see that MK, who is junior to ST, relays ST’s utterances to me and elaborates them. If a griot were present, he or she would take on those tasks.) These practices show that persons of both high and low rank have some knowledge and competence in how one displays that one ranks higher than, or lower than, some co-present other(s).
This replication of the features of social differentiation, in contexts that involve different scales of social encompassment, is what Susan Gal and I have called “fractal recursivity” (Irvine and Gal Reference Irvine, Gal and Kroskrity2000, and elsewhere). Replication only goes so far, however. The differences that are reproduced, though clearly perceptible, are relatively small compared to either caste’s behavioral extremes, or the special techniques and expertise of the griot’s rhetorical craft. In short, despite such a wide distribution of many ideas, practices, and the characterological images that accompany so many kinds of verbal acts, there remains an important differentiation in knowledge, expertise, and especially values, corresponding to the social differentiation of the caste system. Moreover, while nobles and griots living in the same overall community may know each other very well, their life experiences are certainly not the same. To be a griot, and know that you are stigmatized unless you can prove you are different from others of your family and social category, is unlike being a noble. And to be a noble, and know that you could be accosted at any moment by low-caste people making requests of you, is unlike being low-caste yourself.
The situation for griots, who must always be aware of how they appear in nobles’ eyes (and ears), is reminiscent of what W. E. B. DuBois called “double consciousness”: “this sense of always looking at one’s self through the eyes of others, of measuring one’s soul by the tape of a world that looks on in amused contempt and pity” (Reference DuBois1903). Yet, there is some of the same effect for nobles, too, who depend for their in-caste reputations on the public opinion griots are willing to arouse for them.
Shadow Languages and Whorfian Effects
Now, let me take a step backward and consider my own role in this ethnographic scene. I maintain that after living among rural Wolof townspeople for an extended period, I discerned a patterning among their statements and actions, such that I can describe overlapping but differentiated ideological visions that underlie, evaluate, and effectively regiment their communicative practices and construals. Yet, while sympathetic to the townspeople’s points of view, and learning to behave myself in relatively appropriate ways in Senegal, I do not share either an antiegalitarian ideology or the townspeople’s specific ideas about language, temperament, endogamy, and so on. So there are crucial ideological differences for me too, much as I managed a temporary accommodation. But accommodation did not always work well.
As many a fieldworker will find familiar, accommodation to the local scene was easiest when I could stand back as observer more than as participant, and more difficult when I tried to participate actively in some situation I did not create. I was inevitably a foreigner. Consider, then, some situations in which this particular kind of social distance—foreignness—has a bearing on linguistic practices and the assumptions, values, and construals that enter into ideologies of language.
On a warm summer night at the end of my stay in Senegal in 2006, I called a taxi to take me from my hotel in downtown Dakar to the airport. Arriving promptly, the taxi driver greeted me in French, as was usual for Senegalese people addressing white foreigners with whom they had no previous acquaintance. I returned the greeting and continued the conversation in French, initially keeping the vocabulary simple in case his French competence was not strong. I need not have worried much on that score, since French is widespread in Dakar and taxi drivers working for a major company are likely to speak it fairly well. I thought of switching to Wolof, but I was focused on catching my flight and did not feel like embarking on a long explanation of who I was or how I happened to be able to speak Wolof (or why, in speaking it, I sounded like a hayseed; I had spent most of my time in a rural community, not in cosmopolitan Dakar). As we talked, I addressed him as vous, the French plural pronoun—widely considered “polite” or “formal” when used with a singular addressee—expecting him to do the same. But he addressed me with the “informal” tu. What was going on here—why was he using T-forms when I was resolutely using V-forms?Footnote 8 What did our utterances imply?
One possibility is that the taxi driver was simply importing a Wolof pattern into French. There is no convention in Wolof for conveying deference or formality via plural second-person pronouns. Other forms of address offer various ways to display respect, but not the pronouns. Instead, the plural yéen is treated as simply plural, end of story.Footnote 9 I did once hear someone claim that the singular yów was actually a more polite way to address a single person—the opposite of the French T/V pattern—since it would imply that the addressee was the head of a large following, whereas the plural would suggest he or she was merely one of many, such as a member of a family rather than head of household. Although this claim shows that the person making it—a bilingual friend, not the taxi driver—was aware of French patterns and wanted to assert that Wolof-speaking people can express politeness too, I could not confirm the claim in observations of actual usage. Nor did I ever observe this kind of disjuncture between pronoun number and number of addressees.
The only linguistic practice that might be considered to confirm some idea of “deferential T” comes from Wolof praise-oratory, because one of its eulogizing effects (informants pointed out) is to individuate the addressee. Although the performance of praise-oratory links its focal addressee to ancestors and their worthy deeds (in addition to any other laudable aspects of the addressee’s biography),Footnote 10 the fact that the performance details that ancestry on both paternal and maternal sides, and often specifies the source of the addressee’s given name, makes the resulting identification of the addressee unique. This principle, in which singularity distinguishes someone as praiseworthy, could be consistent with a pattern in which T-forms were considered potentially more “polite” than V-forms. On the other hand, like the Wolof plural pronouns, praise-oratory always links its addressee to other people. It does not present the addressee in isolation but instead derives praiseworthy distinction from the addressee’s social relations.
Regardless of what the indigenous Wolof pattern of pronoun usage might signify, however, it is not clear that my taxi driver was following it. Might he, instead, have been adopting a French pattern—and if so, which one? Was it the “formality/intimacy” norm in which the T-form would not ordinarily be addressed to an adult stranger of opposite sex, except to be rude, or patronizing, or flirtatious? Or was it the norm of a certain French egalitarian ideology that advocates widespread use of reciprocal T (somewhat resembling the reciprocal T of the Quakers and “levelers” of seventeenth-century England)? Or was it the adjusted version of this egalitarian pattern, often observable in the actual practice of French university students and intellectuals, who generally use T-forms to address one another—whether previously acquainted or not and regardless of gender—but may use V-forms in addressing elders and working-class people they fear would be offended by T-forms? There is no way to know. Each of these possibilities represents an interactional structure, a possible positioning of the interactants and a set of characterological images; so the effect is indeterminate, because of the ambiguities created by larger patterns and norms of usage. Each possibility, with the social positioning and interactional relation it implies, represents a possible interpretation of the event, an interpretation that is in play whether any participant’s construal actually corresponds to it or not.
My own use of V-forms with the taxi driver represents some sort of compromise between the French “formality/intimacy” norm and the adjusted egalitarian one. I was studiously avoiding T because of the possibility that it would sound patronizing; historically it was the form used by French colonizers to their African servants. I also wanted to avoid T because of its possible suggestion of intimacy. As it was, I felt mildly uneasy that evening because we were alone in the taxi and the route we took to the airport was roundabout, through poorly lit industrial areas and construction sites without many people around. True, even though it was late at night, the likely purpose of the unusual route was to avoid traffic and roadwork; indeed, we arrived at the airport without incident. Still, even if uneasy, I wanted to be polite. And in the back of my mind was the excessively “formal” T/V pattern I was taught—previous to any exposure to Parisian student egalitarianism, much less to any Senegalese usage—by my eighth-grade French teacher. She, a French aristocrat who was slumming, in effect, by living for a while in America, refused even to teach us the T-forms because, she claimed, the only persons with whom they were appropriately used were one’s French lovers and one’s own French children under the age of ten. Since we Americans were not going to have either of these kinds of persons in our lives (she declared), we did not need to learn how to address them.
Yet, might I have sounded more patronizing to the taxi driver by using vous than I might have with tu? Did the vous form imply that there was no possibility he could be my equal? Or did it resemble the excessive, and hypocritical, politeness noticed by a Senegalese business executive soon after Senegal’s independence, who reported that his European colleagues treated him with “added politeness” yet, all the while, ignored his opinions, as if he were only a messenger with no authority to discuss the message’s contents (R. Cruise O’Brien Reference Cruise O’Brien1972, 242)? Again, there is no way to know.
The fact remains, then, that the pattern of asymmetrical T/V address that emerged between me and the Senegalese taxi driver—a pattern that was not actually limited to this particular occasion, although this was one of the most extended interactions in which it was manifest—is not clearly interpretable according to anyone’s conventional practice. What indexicalities we each knew about, and what we each intended, must have been opaque.
The indexicalities of T/V pronouns have been explored by Silverstein in papers that reanalyze Brown and Gilman’s (Reference Brown, Gilman and Sebeok1960) classic study and place the English-language developments in historical perspective (Silverstein Reference Silverstein and Mertz1985, Reference Silverstein2003)—that is, the developments that resulted in the virtual disappearance of T (thou) and of “deference” values for V (you) from most English-language utterances. As Silverstein noted, a very important move Brown and Gilman made was to consider reciprocal usage of the person forms in interaction, not just the pronouns in isolation. Ironically, however, although Brown and Gilman focused their analysis on interaction, they nevertheless interpreted the “meaning” of the pronoun forms cognitively, in terms of an individual’s understanding of their semantics. This interpretation, as Silverstein pointed out, falls in line with a prevailing European and American ideology of language as a structure inhabiting the individual speaker’s brain, prior to and independent of social action. In contrast, his own analysis points to tropes and their historical evolution. The plurality of V, for a single addressee, constructs a trope from the contrast between the scene at hand and an imagined scene in which I (speaker) and you (addressee) are not alone together, a scene lacking the intimacy of sole face-to-face interactants. This imagined scene—a shadow conversation, as I have called such things (Irvine Reference Irvine and Silverstein1996)—becomes available as the background of a trope of social distance; hence, in some circumstances, of deference or formality. In many European traditions this background understanding of the tropes has become so thoroughly established that they have become cultural convention, taken for granted as “meanings” inherent in the forms.
Once established as conventions indexing deference and formality, Silverstein argued, these “meanings” made the pronouns available for next-order indexicality when speakers began to diverge in their usage. As we know, in seventeenth-century England, Quakers and other “levelers” refused to use “deferential” V-forms for a single addressee, resisting what they saw as the superficiality of social convention, and preferring an egalitarian literalness in address: one addressee, therefore singular pronoun forms. The next-order indexicality arises from the stereotypes of social identity, once the pattern of generalized T use became associated with its typical users (and avoided by other people, who wanted to emphasize their politeness and avoid sounding like Quakers). In short, T and V became indices of speaker identity and no longer indexed deference.
Silverstein’s analysis of the indexical order emphasizes shared tropes, deployed against a common cultural background—although, as in the case of the seventeenth-century Quakers, there could be differences in the values attaching to the tropes. That is, the Quakers abhorred (and avoided) tropes of deference in human social hierarchy, while other people appreciated them and applied them more and more widely.Footnote 11 In my interaction with the Senegalese taxi driver, however, very little common background can be assumed. This was an interaction between strangers, each speaking a language (French) that was not their native tongue. Our first languages, Wolof and English, respectively, lack a T/V deference pattern, at least today. One could certainly say that each speaker in this taxi-riding pair was informed by some cultural/ideological background, and some sense of the possible tropes—but not the same background, and not necessarily the same tropes.
A second-language speaker may, as I did while speaking with the taxi driver, work from late-acquired notions about the tropes that are supposed to apply to interactions in that second language. These acquired notions must operate as a sort of ideological overlay. Just as I did with rural Wolof antiegalitarianism, you can know about, and work with, a linguistic ideology to which, or to portions of which, you do not subscribe. Often, however, the assumptions and tropes embedded in your first language prevail. When they do, as could happen sometimes when understandings from American English infiltrated my spoken Wolof, the results illustrate some of the Whorfian effects to which Silverstein, in a number of essays, has called renewed attention (Silverstein Reference Silverstein, Clyne, Hanks and Hofbauer1979, Reference Silverstein and Duranti[1981] 2001, Reference Silverstein and Kroskrity2000, among other writings).
One example from my Senegalese experience concerns kinship terminology. Among the ways Wolof kinship terms differ from English ones is in regard to the gender of the referent. Basic Wolof terms do not distinguish the gender of children (providing only doom ‘child’ rather than ‘daughter’ and ‘son’). Siblings are distinguished as same-sex and opposite-sex of speaker (or propositus), rather than by maleness or femaleness as such; and if same-sex, the terminology focuses on relative age. Thus the term ràkk could refer equally to a woman’s younger sister or a man’s younger brother. Only in an opposite-sex relation is the sibling’s gender effectively specified: a woman’s brother is her cammëñ, and a man’s sister is his jigéen (regardless of age). Wolof pronouns and person forms do not specify gender either.Footnote 12 Yet, because the relevant English terms do specify the gender of the referent, I felt a strong urge to specify it when speaking Wolof. This is the kind of effect that Boas described in relation to obligatory grammatical categories (Reference Boas1911) and that Sapir called “language habits” (Reference Sapir and Mandelbaum[1929] 1949); Whorf, writing about this very example—gender in English—called such things “covert categories” (Reference Whorf[1937] 1956, 90) and linked them to “habitual thought and behavior” (Reference Whorf[1939] 1956). When speaking Wolof and talking about my own family, I often used expressions like sama mag bu jigéen ‘my female older same-sex sibling’, to the confusion (initially)—and snickering (later)—of my interlocutors, to whom this construction was redundant and ridiculous.
These “Whorfian” effects are more noticeable in languages other than one’s own, in which obligatory distinctions and sedimented tropes are so easily taken for granted as features of the world rather than of language. Speaking another language, however, can bring out such distinctions and tropes when they are imported into the other language. Complications and misunderstandings in a cultural contact situation can result. In her work on linguistic evidentials—the very grammatical feature Boas focused upon in his argument about obligatory grammatical categories—Alexandra Aikhenvald cites examples from interactions between Americans and Jaqi Aymara, when both are speaking Spanish (Reference Aikhenvald2004, drawing on Silver and Miller Reference Silver and Miller1997). Since Aymara has obligatory evidentials specifying the source of a speaker’s information for a proposition, and English has not, the interactants interpret each other’s Spanish utterances in unfortunate ways. The Aymara speaker sees people who do not precisely specify the sources of their knowledge as either untrustworthy or stupid (or both); the American sees people who insist on expressions like “I am told …” or “You say that …” as obnoxious, unimaginative, and suspicious. When the American says (in Andean Spanish) “I am from California,” and the Aymara responds, “You say you are from California,” Aikhenvald comments (Reference Aikhenvald2004, 336), “The outsider [the American] is then likely to feel that they are being accused of telling a lie. And yet for a Jaqi speaker the issue is simply one of ‘accuracy’ and not of ‘morality.’”
In Aikhenvald’s example, the Jaqi Aymara and the Californian are speaking what is a second language for both of them. The Whorfian effects work in both directions, across the language difference. And in these cases, just as in my overspecified gender usage in Wolof kinship expressions, the Whorfian effect comes from a language not actually spoken in the situation at hand. It is a language lurking in the background, in the interactional shadows—and from there, affecting a speaker’s usages and interpretations of other people’s utterances. Gendered lexemes and grammatical evidentials may be features of linguistic structure; recall that the V-form’s conventionalized trope of “deference” seems, to its speakers, to be a feature of linguistic structure too. A shadow language brings with it not only a grammatical structure but also the tropes, the practices, and the ideologized projects and construals to which the speaker has become habituated.
Conclusion
My two sets of examples have afforded different analytical emphases, but the examples are less different than they may seem. To begin with, second languages and the indexicalities they might imply are found in the rural Wolof system too, not only in relation to me as a foreigner temporarily living in a Wolof community. Notice that there are a few French words in both texts: in text 1, baraas (probably from French barrage ‘blocking, barring’)Footnote 13 and alors; in text 2, guerre, balle, and kano (from French canon). French and other “foreign” languages fit into the local ideologized visions of linguistic practices and values, being considered linked to the fluency and rhetorical elaboration associated with griots and other low castes. Text 2’s mention of Lawbé wailing and Sarakholé noise are part of this picture, since Lawbé (itinerant woodworkers, an artisan caste) are locally known to speak Pulaar as well as Wolof, and Sarakholé (an ethnic group from southern regions, locally considered low-ranking and insufficiently Muslim) are speakers of Soninke, a west Mande language. High-ranking nobles in this rural town tended to deny knowing any foreign language other than, perhaps, a written form of Arabic, while griots might proudly add foreign languages to their repertoire.
Meanwhile, in the second set of examples, the taxi driver and I both understood the relevance of competence in French for urban Senegal, especially in such “modern” activities as taxi trips to the airport. Most of our interaction went smoothly. I was familiar with many things about Dakar and its inhabitants, while he was evidently familiar with the needs of international travelers. We were strangers, but not in everything.
So although the first set of examples focuses mainly on values and experiences, and on people who are residents of the same town and know one another well, and the second set focuses mainly on linguistic elements and second-language speakers who are strangers to each other, these are differences of degree, not of kind. Both sets concern social differentiation, the indexicalities that accompany linguistic forms, the interpretations people draw about their interlocutors, and the projects and values that accompany linguistic practices.
Decades ago, Voloshinov wrote, “wherever a sign is present, ideology is present” (Reference Vološinov, Matejka and Titunik1973, 10). One might add, where there is ideology, there is difference. Some portions of an ideologized system might overlap, perhaps even many portions; but differences that relate to social differentiation and point of view must remain.
A Note on Transcription
The Wolof words and transcripts in this article are rendered in the standardized Wolof orthography established in the 1970s by the Republic of Senegal and further developed by Senegalese linguists. See Fal (Reference Fal1999, Reference Fal2000) for description; Faal is how this linguist spells her name when she writes in Wolof. Fal, Santos, and Doneux (Reference Fal, Santos and Doneux1990) and Diouf (Reference Diouf2003), the most authoritative dictionaries of Wolof, use this orthography. The present article is not concerned with any phonetic issues that might require major departures from standardized spellings. In the two texts, words of French origin are in italics.
