Relation 1: Performances index existing ideological models

Performances index and depend on the stability of ideological models, while providing opportunities for sudden shifts in ideological position as well as for transformations of those models. These models form part of the ideological context to which speakers orient themselves as they participate in performative as well as nonperformative discourse. Particular utterances index that context, in that they presuppose ideological worlds in which certain propositions are taken – at least by the speaker, but often by the listeners as well – to be true or logical (Silverstein 1979, Ochs 1990). Indexicality is thus an important part of the contextualization process; it allows researchers and participants alike to make sense of the interaction as they interpret their relationships with their interlocutors, as well as the ideological positions of those interlocutors.

F10's “murderer”-on-the-plane story indexes the ideology of nativeness, a linguistic ideology that was firmly established in social discursive space by 1997, when this interview took place, and certainly also by the early to mid-1990s, when the narrated event took place. In answering my question about people who “didn't speak English very well,” F10 quickly indexes the man's foreignness by adding or whatever after a name she apparently knows now (17). Within this ideology of nativeness, mentioning or implying foreignness is sufficient to establish his status as a nonnative English speaker. This association of one nation with a single language is widely held and has been described by Silverstein 1987, Blommaert & Verschueren 1992, and Gal & Irvine 1995. The ideology of nativeness further posits that linguistic nonnativeness makes one incomprehensible. ‘Native’ and ‘nonnative’ are rendered dichotomous, mutually exclusive categories in this discourse. This means that, even if speakers do not use these terms in particular, they exaggerate differences between their own speech – taken to be normal and unremarkable – and the speech of others. In the talk of most international students, even those from countries such as India or Singapore in which English is a primary language, this dichotomous model of the linguistic world appears in, among other discourses, self-denigrating talk of a lack of English proficiency. In the talk of U.S.-born, native English speakers, it appears most clearly in complaints about incomprehensible accents.

Although F10's first telling of the narrative draws on this model of nonnative speakers' linguistic ineptitude and explicitly claims that she was afraid, she has not (at least by line 26) directly argued that it was the man's incomprehensibility that made him frightening. Her emphatic response to my question about why she got scared (28), however, clearly shows that the truth of that proposition should have been obvious. She explains, 'Cause I couldn't understand him! That line reveals her annoyance at having to explain the connection between the man's nonnativeness and her fear. This connection is further highlighted throughout the performance with two additional utterances of scared (40 and 58, the latter uttered by her partner) and two of nervous (51 and 64).

In order for the suggestion of fear to make sense to her listeners in this context, there must exist an ideology that equates foreignness with fear in public discourse. Indeed, F10 seemed to find it easy to draw on xenophobic tendencies in the United States, which were available in public discourse long before September 11, 2001. Inda 2000 analyzes the discourse supporting California's Proposition 187, passed in 1994, and finds a pervasive construction of immigrants as parasites who threaten the social, cultural, linguistic, and economic well-being of the nation. Jaret 1999 similarly notes a heightened fear of foreigners in the last two decades of the 20th century. This xenophobia has long influenced U.S. film and television conventions, which portray nonwhites and/or non-Americans as “bad guys” and link nonstandard or foreign accents with negative – or at least bumbling or humorous – character traits (Lippi-Green 1997). Linking language use with national origin allows speakers like F10 to project a fear of foreigners onto perceived nonstandard language use to produce a fear of linguistic Otherness. The notion of a “standard,” in the “aggressively hegemonic,” emblematic sense which Silverstein describes (1987:286), undergirds F10's naturalized assumption that incomprehensibility and foreignness (and probably also maleness) added together equal threat.⁵

F10's indexing of the ideology of nativeness is clear in her momentary assertion of the man's inability to communicate at all in English. Such a rejection of any linguistic competence on the part of the Other is widespread and is often encoded in a variant of this claim: “[I/You] couldn't understand a word [he/she/they] said” (Shuck 2001). This categorical rejection is invoked in F10's answer, 'Cause I didn't understand him! Although she reconstructed something he had said (he asked me for a Sprite), she later reiterated the claim of incomprehensibility: I … couldn't understand him. Like period (46–47). Moreover, when one of her interview partners, F9, represented the man's voice as wordless blather – bluh luh luh luh – F10 agreed with her characterization. Such representations are part of a discourse of exclusion, by which only native speakers of English – indeed, only native speakers of certain varieties of English – are admitted as comprehensible, unremarkable members of the community of English speakers. These patterns are relatively stable, accessible, and pervasive, akin to what Bakhtin 1981 might call authoritative discourses that have become assimilated into our ideological consciousness as “internally persuasive discourse” (1981:342–48). They are accepted uncritically as true, at least at the moment of their utterance, and are reanimated in new contexts as though they were originally conceived. (The truth value of these internally persuasive discourses is addressed in greater detail under relation 3.) The creation of this performance, then, depends on the prior existence of an ideological model that provides the speaker with performable themes and linguistic patterns. Moreover, as we shall see in the next section, the relative stability of these patterns allows them to serve as resources for aesthetic transformation in conversational performances.

Relation 2: The set of discourse patterns entailed by a given ideology are readily available, and are drawn upon, as resources for aesthetic displays

Any ideological model produces limited sets of formulaically encoded themes and arguments (Silverstein 1987, Woolard 1989, Hill 1992). This stability makes them readily available for use in verbal interaction. Fixed, or at least moderately formulaic, patterns associated with particular topics are a linguistic resource upon which speakers draw, not only to find the words to make their points but also to make those points in aesthetically interesting ways. The central connection between performance and certain ideological positions has to do precisely with the easy accessibility of discourse patterns associated with those positions. Patterns such as “[I] couldn't understand a word [they] said,” are so readily available that speakers seem to find them hard to avoid, even when the same speakers primarily take ideological positions that contradict the beliefs encoded in those forms (Shuck 2001).

Once we understand the ease with which certain discourse patterns can be accessed by any speaker, we can begin to see how such patterns are used as linguistic resources in performative talk. In most of the interviews in this study, the emergence of certain topics seemed to invite performances, particularly complaint performances: poor instructors, purposeless assignments, unfair grading practices, lazy foreigners, and incomprehensible accents. Within those performances, speakers not only employed fixed phrases; they also poetically transformed them in novel ways. Speakers' reliance on such patterns as resources for aesthetic display is critical to a study of performance because it allows us to examine the human potential for creativity and for “the play of familiar patterns” (Tannen 1989:13). Indeed, the tension between the familiar and the novel provides a foundation not only for the creation of a given performance but also for its acceptance by the audience as a performance.

The incomprehensible accent complaint, one of the most common types of performance in my interviews, illustrates this use of ideological discourse patterns as an aesthetic resource. As I mentioned earlier, this complaint has as a primary component the canonical form, “I/You/We couldn't understand a word he/she/they said.” That sentence, along with variations such as “I couldn't understand a thing,” appeared more than 30 times in nine different interviews. Its prevalence in U.S. discourse practices makes it especially available for aesthetic transformation, as in (2), from another interview:

This image – words coming out of one's mouth – is a clear instance of a poetic transformation of a relatively fixed phrase. A less poetic alternative might simply have been “understand what he's saying.” Instead, F7 chose to provide a much more visceral, visual image. The available pattern, then, became a resource for conversational artistry as well as a moment of participation in the same ideological discourse constructed by the narrator of the “murderer”-on-the-plane story. In that story, the canonical pattern, “I couldn't understand a word he said,” underlies two utterances in particular. First, it is directly indexed by F10's exclamation 'Cause I couldn't understand him! (28). She reiterates the point after my second attempt (41–45) to get her to elaborate on the connection between her fear and his being a nonnative speaker:

Rather than offering further explanation of that connection, she resorts to a variant of the canonical rejection of the foreigner's language proficiency, “I couldn't understand a word he said.”

To help F10 answer my question about why she was afraid, her partner, F9, suggests in lines 54–57 that the man was rude as well as incomprehensible (see relation 5, below, for an analysis of the recontextualization that facilitates such transformations of the antagonist's character). F9 then offers her own dramatic representation of the voice of incomprehensibility: a rapid, machine-gun-like bluh luh luh luh (71). Again, this distorted representation depends on the speaker's easy access to the relatively stable notion that nonnative speakers' uses of English are incomprehensible.

A similarly fixed phrase that appeared several times in my data is “We're in America,” in reference to the use of a language other than English. This trope, along with its variants “This is America” and “I'm/You're in America,” indexes a simplistic ideological equating of English and America, which finds easy entry into a complaint performance about nonnative English speakers in the United States. The following example is taken from an interview with a particularly progressive female student, F6, who expressed interest in other languages and cultures, had had a Palestinian boyfriend, and was among the only students to raise without criticism or lament the possibility that some Americans do not speak English. Her use of we're in America thus came as a surprise, although she and her partner had already been discussing exclusion from a conversation because of the use of a particular language, a discussion about which she had said she had “mixed feelings.” In this excerpt from that interview, she is talking about a Polish friend of hers in the United States. To indicate her sudden shift of position, I have italicized the multilingualist arguments and underlined the monolingualist ones (ignoring for the moment the role of but, just, etc., in constructing a merely concessionary position within a primarily monolingualist argument):

F6 begins with an acknowledgment of the naturalness of the friend's mother's speaking Polish, but at line 16, her speech becomes more involved and dramatic. By line 20, she has fully – albeit momentarily – shifted allegiances, from a more inclusive, multilingualist position to a monolingualist position. Her seemingly sudden use of the trope we're in America can be explained by its accessibility and, by extension, its performability. She seems almost destined to draw on that ideology once she begins to paint an emotional picture of her own linguistic exclusion.⁶

Indeed, the performativity of this segment heightens at the moment she says WEIRD, simultaneously marking a shift in ideological position and an opening of a performative frame that makes the overtly monolingualist ideology more acceptable (see relation 3 regarding the framing of acceptability). That F5 perceives the shift to a more performative frame may be evident in her unfinished line Why are you having… ? Although it is not possible to confirm this, F5 may have been relying on the same perspective-taking strategy she frequently used throughout the interview, constructing the thoughts of the narrative's main character at the time of the narrated event. I interpret her truncated utterance as the beginning of a question like this: “Why are you having this conversation in Polish in front of me?” If this is true, F5 uses the second person pronoun in this question to address the mother and daughter in F6's story, animating F6's voice at the time of the recounted event, rather than using you to refer to the F6 who is currently speaking. If indeed this is what she is doing, she participates in the performance begun by F6, defining conversational rules according to a monoglot ideology. Regardless of F5's intent, F6 makes that ideology explicit as she completes her previously truncated line we're in America. She continues the performance, employing a similar rhythm and intonational contour in each of the three consecutive intonation units beginning with we're in America:

Here, the formulaic phrase itself is not transformed; rather, it provides an opportunity to perform and indeed marks a shift into a performance frame, a shift that F5 seems to have noticed as well. That this commonly heard trope comes so soon after the performance frame is entered is no coincidence. It is, instead, a fairly predictable result of the relationship between performance and ideological elements that are already floating around in social discursive space. The fixity of those elements allows them to be used as raisons d'être for performances because they provide common ground for speakers and listeners, on which novel uses of these patterns can be constructed.

Relation 3: The framing of talk as performance opens up a discursive space for constructing ideological extremes and stereotypes and, more importantly, for momentarily rendering them acceptable

So far, we have examined relations between performance and ideology that assume the prior existence of ideological models. A given performance indexes relatively stable, presupposed components of those models; and certain discourse patterns produced by those models, such as variations of “You can't understand a word he says” or “We're in America,” allow speakers a degree of discursive pre-patterning that provides opportunities for conversational creativity. Supported by that fixity, speakers rely on what they know about performance – their performative competence – to negotiate the tension between the fixed and the novel.

Let us now examine the other half of this tension: the novel, the emergent, the discursively constructed elements constituted by the interaction itself. Performance plays a central role in that construction. As I will demonstrate in this and the following two sections, the framing of talk as a performance provides contextualization cues to participants to help them make sense of what is being performed. First, it shapes norms for interpreting the social acceptability or truth value of the performed claims, and it can move ideological boundaries toward the extreme or the remarkable, thereby shaping the ideological model in participants' consciousness to some degree. Second, performance shapes the interactional context such that co-performances or successive individual performances are acceptable and even expected. Third, performances can be recontextualized, increasing the number of people who can reanimate the ideologies embedded in them.

The notion of frame, as developed by Bateson 1972, Goffman 1974, Tannen 1979, and Tannen & Wallat 1993, is fundamental to our understanding of the poetic license that performers take as they create particularly exaggerated representations of others' talk. Framing devices in conversational performance, such as repetition, openings and closings, and rhythm, set a display apart from the rest of the interaction, drawing attention to the form in which the communication occurs. Talk framed as a performance commands special attention and, ultimately, some kind of evaluation from the audience. Because listeners pay heightened attention to performed ideological positions, framing plays a critical role in shaping the norms for interpreting social truths. A shift into a performance frame tells us that it is socially acceptable to sacrifice accuracy, grammaticality, politeness, and so on for the sake of a performance.⁷

An obvious example of the framing of social acceptability can be found in one student's discussion (6) of the benefits of an international first-year composition class, English 107. The student ignores social norms for avoiding racist terms and, without commentary or irony, uses the word Chinaman, which conveniently ends with the same schwa+n ending as German and Mexican:

His saying Chinaman surprised not only me but also a number of students in later interviews who listened to this recording. This student is clearly more interested in the poetry of having three noun phrases all end with -an than in the implications of the word Chinaman.⁸

This reality, at least at the moment of the utterance, is naturalized for the participants, such that the iconic association of an ethnic term with what is thought to be a bounded, identifiable social group is considered natural and unworthy of comment (Gal & Irvine 1995). This naturalization process is particularly evident in claims of the total incomprehensibility of others' accents. Once speakers are performing, Grice's (1975) Maxim of Quality becomes quite malleable, if not altogether irrelevant. Kroskrity's (1992) discussion of non-kiva innovations of Arizona Tewa kiva speech illustrates the possibilities for shifting or suspending norms for interpretation once a performance frame has been invoked. F9's rapid-fire bluh luh luh luh takes advantage of such possibilities. A remarkably similar trivialization of the voice of the incomprehensible Other came from a different interview (7). The student, F1, began by complaining that she did not understand a WORD [her history professor] said. She then added this characterization:

Her attempt to perform overrides any expectation that her imitation be the least bit realistic. Such an absurd, derogatory imitation creates a new truth, a new norm for social acceptability. No one challenges the truth value of either of these students' claims. Indeed, F10, on hearing F9's bluh luh luh luh, goes so far as to claim, That's exactly what he did, explicitly confirming the truth of the representation. Because the notion of the incomprehensible foreigner already exists in the ideological model, such imitations fall within the range of what might count as an accurate representation within that model. Moreover, the performance frames in which these utterances occur virtually eliminate the truth constraints that might otherwise apply to non-performative discourse.

This is precisely how performance is simultaneously presupposed by and constitutive of ideology. Such an extreme representation of another's accent in English as equivalent to “a foreign language” is possible only if the ideological discourse already supports the premise that foreign accents are incomprehensible. The extreme then becomes part of the ideology at the moment of its performance. Its vividness – its particularity (Becker 1984) – reifies the incomprehensible foreigner for the audience. The performance further establishes a context in which nonnative English speakers are classed as others and subject to derision. This further creates contextual conditions that allow for the derision of other Others; in the interview with F1 (7) and her partner, those also included American Indians, authors who write about the beauty of the desert, and people who keep quiet during class discussion even if they disagree, namely the one Hispanic student in an English class full of Anglo students. Talk that is framed as performance places all these Others on the same contrasting level in a taxonomy of the human world: bad teachers, people with accents, myopic writers, taciturn students, and American Indians are thus rendered equivalent and therefore subject to equal, and socially acceptable, degrees of ridicule. Although non-performative talk may similarly render disparate groups or phenomena equivalent, performance frames in particular provide a kind of safety that makes such comparisons less likely to be challenged.

The contextualization cues in the “murderer”-on-the-plane story enable a particularly offensive transformation of truth, beyond the simple claims of the man's incomprehensibility. Together, the two performers take the canonical image of the incomprehensible foreigner and embellish it until it is acceptable to think that this nonnative English speaker, merely by virtue of his nonnativeness, is dangerous.

By the middle of the first telling, F10 has claimed that the main character was scary. This can be seen as the first of three transformations of his character. With F9's performative contribution, he becomes rude as well (8). F9 attempts her own answer to my question about the connection between fear and incomprehensibility by suggesting, without apparent irony, that the man must not have been polite, although she knew little more about his behavior than that he tried to talk to F10 and asked her for (or possibly asked if she wanted) a Sprite:

F9 contributes to the collective performance with some intonational and rhythmical parallelism in 54–56, in which the intonational contours end on approximately the same pitch. This rhythm sets up the next phase of her performance, in which she contrasts a hypothetical polite, comprehensible stranger with the rude, incomprehensible stranger she imagines the man to be:

The second major transformation of the man's character – from nonnative speaker to frightening nonnative speaker to rude, frightening nonnative speaker – is now complete.

Finally, encouraged by this positive response from F9, F10 underscores the message that foreigners with incomprehensible accents are scary by transforming the man from ‘scary, rude foreigner’ to ‘murderer’:

Although the responses that follow her performance (86–90) might be interpreted as signs of embarrassment or disbelief, it is not likely that F10 has fallen over the edge of an ideological, performative precipice. Within an ideology that easily supports fear of nonnative English speakers, and within a performance frame that encourages the construction of extreme or absurd representations, the notion that this man might be a murderer is not an enormous leap.

It is this last transformation that directly connects this ideology to a more widespread anti-Arab racism, heightened in the period following the hijackings on September 11, 2001. In this narrative, the “foreigner” is on a plane to the United States from Saudi Arabia and has a name that F10 regards as foreign. At the end of the narrative, F10's quiet utterance represents what she thought when she woke up and found herself sitting next to him: This foreigner, whom she could not understand, took [her] mom [and] killed her. We must not dismiss this simply as an absurd exaggeration that is meant to enhance her performance. The ideological construction of the racialized Other underlies the same fears that allowed Arab, Pakistani, and other nonwhite passengers to be ordered off of U.S. flights following the attacks and that have resulted in the detaining without charge of hundreds of nonwhite immigrants (American Civil Liberties Union 2002, CBS News 2002). It is particularly significant that the fear expressed in this performance – and indeed, F10's suggestion that the man could have been a murderer – arises in direct relation to a question about language. Linguistic otherness is associated with ethnic otherness, and both are associated with fear. In recent years, the widespread association of Arabs in particular with terrorism, made more intense since September 11, 2001 but existing long before that (Shaheen 2001), increases the possibility that an Arab man in particular would be constructed as dangerous.⁹

Relation 4: The collaboration that performances invite create opportunities for the proliferation of ideological extremes

The capacity to create ideological extremes lies not just in the attempts at drama in a given performance. It also lies within the fundamentally dialogic nature of conversational performance. F10 is unquestionably the primary performer of this narrative, and yet F9's role in the whole performance is critical, particularly in transforming the nonnative English speaker into a rude stranger. Because performance increases involvement in the interaction, it heightens the possibility of shared participation in the making of ideological meaning. Collaboration in performances enables the repetition of ideologically salient discourses and encourages participation in transformations of truth such as the one discussed in the previous section. Performances, then, are jointly accomplished and culturally valued means of rationalizing ideological premises in the collective imagination.

This collectivity manifests itself in a number of ways, each of which differs essentially from the others in participant structure (see Philips 1983). (i) Listeners can take minimal turns during and after another participant's performance, usually indicating positive evaluation of both the performance and the performer's ideological position (Yankah 1985). (ii) Speakers can collectively create a performance, whether there are primary and secondary performers or there are equally participatory co-performers. (iii) A successful performance by one speaker effectively invites a subsequent performance by an audience member. It is in the second and third participant structures that we see the greatest potential for joint construction of ideology; these, therefore, will be the focus of analysis in this section.

We have already seen how F9's contribution to F10's man-on-the-plane story encourages F10 to take her performed fear of the presumably nonnative English speaker to a new level. As F9 supplements F10's performance with imagined details about how the stranger spoke, she firmly aligns herself with the ideology of nativeness, particularly with its basic premise that nonnative speakers are incomprehensible. My challenges to F10's claims, however dialogic they may be, elicit only some slightly more specific retellings of the story but do not elicit the transformations that F9 did. In contrast, F9 is able to transform the stranger and get F10 to confirm that new representation of him: That's exactly what he did! F10 explicitly agrees that his speech sounded like gibberish, despite her earlier claim that she understood that he was asking about having a Sprite. The two speakers draw upon each other's claims in order to create a representation of the stranger that was quite different at the end of the performance than at the beginning.

Such co-performances allow speakers and listeners to switch roles in meaningful, well-coordinated, and poetic ways, contributing directly to the interactional organization of the experience (Müller 1992, Erickson 1992, Haviland 1996). They serve two primary purposes. First, they constitute a form of positive audience response. Because my questioning F10's fear clearly did not meet her expectations as a response, F10 did not see it as an opportunity to embellish her story as much as she did after F9's performative contribution. This response, in turn, encouraged further performances with the same theme. Second, the opportunity to jointly create an aesthetic display facilitates the expression of ideological solidarity. In the case of the “murderer”-on-the-plane story, although the two speakers had already expressed ideological solidarity by the time this performance emerged, the collaboration allowed them to confirm the “truth” of this developing representation of the Other.

In some cases, however, speakers who did not consistently express similar positions moved in the direction of ideological unity once they began performing. Such is the case with two other interviewees, one of whom (F6) oscillates between ideological positions frequently during the interview but who largely expresses positions of linguistic inclusion rather than exclusion. This is the same student who began a complaint about her own exclusion from conversations between a Polish friend and her mother by claiming we're in America (4). Once again, at the moment her performance begins (see ex. 11), she plants herself in a monolingualist position. Her partner's contributions to her performances, rather than being an opportunity for F6 to argue for her often-expressed, multilingual-friendly views more firmly, only ground F6's utterances more deeply in the dominant monolingualist model.

The two interviewees had been talking both about nonnative English speakers in the United States and about a student exchange trip F6 had taken to Slovakia. In this segment, they jointly decide that obligations to learn a particular language depend on whether one is living in or visiting a country in which that language is dominant. F6 has already suggested that she had felt some degree of obligation to learn Slovak while in Slovakia, saying, If I had the time, I would. However, after she tells a story of being told she should speak Slovak while in Slovakia, and once she and her partner have established an ideological affinity with each other regarding visiting vs. living in foreign countries, her utterances express increasingly monolingualist perspectives. The less performative first half of this segment reveals a considerable degree of ideological solidarity, evident in the degree of lexical echoing (underlined), overlapping speech, and repetition of yeah. In the course of their continued agreement, F6's earlier, albeit limited, sense of obligation to learn Slovak diminishes rapidly until it disappears:

As F6 begins performing, with the evaluative opening I was insul[ted], this lack of obligation becomes outright disdain (22–23). F5 contributes to this performance by constructing dialogue as though she were F6, using the first person pronoun and imagining how much Slovak F6 might have known (lines 25 and 27). The two students are now almost bragging about F6's – and, with the deictic shift (25), F5's – not knowing Slovak.¹⁰

In particularly collaborative co-performances such as this one, speakers move closer to ideological solidarity, in part because the “sanctuary” of the performance frame invites simplified premises and exaggerated representations. Furthermore, performances demand responses, which themselves frequently appear in the form of subsequent performances, such that speakers take turns creating related performances that rely on similar performative strategies and ideological themes. The implied invitation to perform can even result in dramatic contradictions between the ideological stances taken during performative and non-performative talk (Shuck 2001). Buttny & Williams 2000 maintain that one performative strategy in particular, reported speech, enables speakers to repeat particularly memorable, quotable words of others. They also suggest that speakers rely on and repeat structured themes, such as discourses of “respect” (see also Hill 1992), as means of establishing their authority in the interaction. My study further suggests that performance frames are especially effective in providing speakers with opportunities to comment on previous performances by drawing on each other's words, thus strengthening ideological cohesiveness.

Relation 5: Performances' capacity for recontextualization allows ideologies to be shaped by greater numbers of speakers and therefore to be subjected to multiple transformations

Because performance frames are actively constructed and because they provide linguistic resources by which listeners interpret talk as performative, they are intricately linked to their immediate situational contexts. However, performances are paradoxically “susceptible to treatment as self-contained, bounded objects separable from their social and cultural contexts of production and reception” (Bauman & Briggs 1990:72). They can then be reperformed or reported upon or discussed, or otherwise reframed in other contexts. This potential for decontextualization and recontextualization is due in part to the aesthetic effects of certain linguistic elements in a text that create rich textual coherence as well as particular connections to audiences and to the rest of the world outside the text (Becker 1984, Tannen 1989, Buttny & Williams 2000). Because some texts are framed as performances, their aesthetic effects contribute to their memorability, and this helps them to become part of the expressive repertoires of particular listeners/readers/viewers. Performances, then, forge a link between a discursive past and a discursive present (Becker 1979), and also between that same discursive present and a discursive future, as ideological premises become encoded in a recontextualizable form.

Within the “murderer”-on-the-plane story, the performer herself recontextualizes her own narrative twice, each time heightening the drama of the first telling. In doing so, she moves further and further from an ideological core that characterizes the nonnative speaker as an incomprehensible Other, toward an extreme representation of the nonnative speaker as dangerous enough to be capable of murder.¹¹

Her next version of the story (13) was apparently sparked by my question in the last line above, Why'd you get scared? In it, she repeats the same narrative event structure as the first version:

1. a man that she didn't know sat next to her (line 30, below);
2. he tried to talk to her (lines 33–34);
3. and she got scared (40).

The new opening (line 28) serves both to answer my question and to offer an external evaluation of the narrative, which attempts to clarify the ideological relationships between fear and a foreigner's accent. Her tone, however, is more dramatic in the new version, starting with the first line (28) and continuing through the rhythm of lines 30–32 to the new, more rhythmical version of her reported thought, with two syllables per intonation unit (36–38):

Immediately following this reperformance is my own reframing of my question about what accounted for the fear (ex. 14). She seems to think it is not worth another performance or even an explanation for an interviewer who is apparently missing the point anyway, and so she ends her next turn before letting it develop fully into a performance. She repeats her overall evaluative comment (lines 46–47) and begins another description of the scene (49) but truncates her own utterance (50) before offering her closing statement, a reiterated claim that she was frightened (51):

Although she ends this performance abruptly, her interest in performing is renewed when F9 chimes in with her own performance. Indeed, that performance might be seen as a further reframing of F10's narrative (ex. 15). F9 takes the basic narrative propositions (i) that the man talked to F10 (lines 54–56, 61–63, 69–71) and (ii) that F10 got scared (58), and the ideological propositions (i) that he was incomprehensible and (ii) that his incomprehensibility made him scary, and she reframes them in contrast to a hypothetical scenario:

Once F9 has affirmed the truth of F10's position on incomprehensible foreigners with this scenario (69–71), F10 begins to participate once again in the now collaborative performance. Her renewed performance (ex. 16) is clearly more an expression of ideological solidarity with F9 than an explanatory answer to an interview question. F10 repeats F9's you're like as she resets the scene (73), but then immediately responds to F9's performed imitation by laughing and agreeing with its accuracy (75). Finally, she uses the opportunity gained by her having taken the floor loudly and retells the story once more, this time with highly exaggerated constructed thought, during which she offers her most extreme transformation of the nonnative English speaker (lines 83–84):

Interestingly, in this final version she does not mention that he talks to her. Rather, she replaces his speech with her thought (81–85, 89) and then ends the performance with a focus on her own fear, however ironically expressed.

Each successive rendition of this story is a repetition of the essential narrative events and ideological propositions of the first version with a performed transformation of one or more of its elements. Because performances can be recontextualized, the ideological premises inscribed in them can easily and quickly become increasingly outlandish. However, at the same time, some of those premises become increasingly fixed: Foreigners are scary. People with accents are incomprehensible. Stories about scary, incomprehensible foreigners are performable.

Herein lies a paradox: Because performance carries the potential for dramatic transformations, the very fixity that indexes an ideological core, providing speakers with the regularized linguistic resources they need to create unique performances, can simultaneously serve as a springboard for constructing extreme, peripheral positions – positions that were not so clearly shared in this case by F10's interlocutors. Performance's capacity for recontextualization makes performed ideological discourse all the more susceptible to multiple interpretations and multiple voicings.

The notion of recontextualization (relation 5) is thus clearly related to the collective nature of performances (relation 4). To the extent that performances provide opportunities for collaboration in the making of ideological meaning, they can also create new contexts in which to reframe ideological premises, which are often exaggerated as involvement in the performances builds (relation 3). Dialogic relations multiply as speakers move from interaction to interaction. This capacity for recontextualization plays a critical role, then, in the social construction of ideological models. It is central in mediating between discourse at the level of the single interactional encounter and at that of the entire discourse community. Such reframings, and thus extension of performances' lifespans, can occur with very long intervals between performances (Bauman 1986), or they can occur immediately, as in the multi-performance of the “murderer”-on-the-plane story.

A number of factors influence the potential for recontextualizing that a given performance may have, including the aesthetic skill of a particular performer and the “sound bite” effect of an especially concise performance. Equally important, or perhaps more important, is the ideological salience of the performance. Buttny & Williams 2000 refer to this salience as a particular resonance of others' words. In their analysis of students' use of repeated themes and structures, particularly surrounding the notion of “respect,” they suggest not only that certain words had a performative power that led students to repeat them frequently, but also that the argument the students heard on a video gave them powerful and memorable ways of articulating their ideas about race. Such resonance is particularly available with dominant ideologies because they are so closely associated with performable linguistic forms. The incomprehensible accent, for example, is so readily available as a trope that it is highly likely to be performed and reperformed when speakers provide even the smallest space for such performances. Indeed, the dominant ideology contains a number of narrative prototypes, such as the angry immigrant or the first encounter with an international instructor, that are particularly susceptible to recontextualizing. We thus see a unifying force in language, one that encourages the reiteration of certain ideological positions, partly because of their inseparable relation to pre-patterned, performable linguistic structures and partly because of an ideological context that accepts the truth of those positions.