Telling the Story, Hearing the Story: Narrative Co-Construction and Crisis Research

Karen Taylor, Tulane University; Rita Durant, University of South Florida St. Petersburg; David Boje, New Mexico State University

Abstract

Research on crisis invites "subject" and scholar to engage in joint sensemaking, often through narrative. Invitational rhetoric provides an interview approach to move from inchoate experience to the co-construction of understanding. The movement from indexical narrative fragments to developed coherent narrative requires efforts from both speaker and listener. Invitational rhetoric highlights the salience of the scholars' own experience of crisis.

Keywords

invitational rhetoric, antenarrative, language strategies

Introduction

Modern social science is about finding those conventions, those predictable patterns, with which we can make sense of our lives. Particularly important to the social scientific method is the use of language for identifying, maintaining, and sharing our propositions. "In Western and especially modern consciousness, ... the idea of method draws its power from the fact that certain objects and processes can be experimentally isolated and thereby controlled" (Gadamer, in Grondin, 1994, p. 118).

Our experiences and our data suggest that a valuable skill set for social science research into crises include the following:

1. Invitational rhetoric 2 .Strategies for involvement in discourse
2. Recognition, acceptance, and invitation of antenarratives
3. Ways to work with linguistic resources to develop indexical antenarratives into stable, recognizable narratives.

All of these skills are honed by first-hand experience of crisis and the limitations of language and disruption of expectations imposed by crisis events

Crises are problems that exceed an individual's capacity to respond using already established routines. Crisis in this sense is "a temporary state of upset and disequilibrium, characterized chiefly by an individual's inability to cope with a particular situation using customary methods of problem-solving and by the potential for positive or negative outcomes" (Roberts & Dzieglielewski, 1995, pp. 9-10). Whereas novelty often has the connotation of triviality, we ask the reader's permission to use it to refer to the entirely new set of circumstances that cannot be adequately addressed and overwhelm current systems due to its unanticipated characteristics. When traditional behaviors do not work, the individual or organization needs to go beyond the usual, to reach for more creative solutions, to go from the known to the unknown.

Crises arise when our current tools, largely information and communication language routines, fail to isolate, name, describe, predict, or control a novel situation. Given the close link between language and understanding, accompanying crisis is the failure of language to adequately represent the unforeseen and overwhelming circumstance. The challenge for social science research into crisis is therefore parallel to the basic conflict of language as described by Gadamer in the opening quote. Crisis researchers have the potential not only to gather rich information but also to create an opportunity for agency and authority when they respond to respondents' invitations to talk about the chaotic and anomalous.

Researching Crises

Problems are occurrences that are different from the routine, expected, or "ordinary" (Hopfl, 2000). Within a social science method of objectivity, standardization, prediction, and control, research seeks to avoid problems by controlling or factoring out ambiguity and difference. Because crisis situations are characterized by the failure of the tools of prediction and control, standard methodology may fail to adequately represent in crisis situations. Randomness and errors become the focus of the research, rather than its residue.

Trust is required for someone to divulge sensitive information about one's difficulties in coping (Mayer, Davis, & Schoorman, 1995). Trust is more readily showed to those we consider to be "in it with us" than to outsiders. Insider and outsider status both constitutes and is constituted by discourse practices. The meaning of a class of events is always indexically bound to the specifics of a given situation and is contingent on intersubjectively constructed accounts (Barley, 1991, p. 187). How can we intersubjectively construct accounts of crises, in order both to understand them today and to avoid similar calamities in the future? Our particular interest is in the language tools we might use during research into crisis.

Crisis is a state that many communities will experience, given that, "cultures regularly suffer from contingency; they bump into things they do not expect and cannot control" (Hyde, 1998, p. 105). Therefore the research goals of striving to predict and control crisis may not be a useful goal. Even so, prevention and mitigation are still desirable. And insofar as a goal of modern research is prediction and control, then an understanding of productive processes of crisis research can guide efficient and effective social science research in general. Therefore, it is valuable to identify realistic and productive outcomes of crisis research.

The crisis of language and communication can therefore be imagined to be "betwixt and between," a state which has potential to call forth "new structures, new symbols, and new metaphors" (Hyde, 1998, p. 130, citing Turner, 1977): in other words, new relationships and new understandings. In seeking ways out of the state of crisis, language strategies available lead towards either a reassertion of standardization or an embrace of the new, which in turn requires trust, a sense of humility and/or humor, and faith that an opportunity for creativity lies in the crisis. It also usually requires interaction with others, interaction mediated by inadequate language. As Deutsch put it, "the strategy of mutual problem solving and the tactics of persuasion, openness, and mutual enhancement elicit, and also are elicited by, a cooperative orientation" (1990, p. 190).

Next, we offer help from ‘invitational rhetoric' as a discursive strategy for crisis inquiry.

Discursive Interactions

We propose that the solution to the uncertainties of language around crisis is the strategic employment of discourse strategies that invite others to contribute to our admittedly partial understanding. Invitational rhetoric, adapted from Foss and Griffin (1995), uses language to express a desire to understand the other, admits the inadequacy of language to express such an understanding, and indicates a willingness to learn from the other in order to build a new relationships, new structures, and new metaphors. As we use it here, invitational rhetoric is a discourse strategy of partiality, ambiguity, and careful attention on the part of the interviewer to what the respondent has to say. Below, we will identify specific rhetorical maneuvers used by some research interviewers to invite input from the respondents' point of view.

The outcome of the use of invitational rhetoric leads to vivid depictions of what the respondents were surprised by and how they felt about what happened. We propose that important observations in crisis situations are indexical narrative fragments, and antenarratives. Narrative fragments lack the beginning, middle, end (BME) coherence of full blown retrospective narratives. Narrative fragments can be so tersely told, that a nod or a word indicate some more complete narrative might be tellable (Boje, 1991). Boje (2001) identified antenarratives as pre-story utterances that "bet" that a more coherent (BME) narrative might emerge. An antenarrative is a ‘prospective' utterance: a kernel, a bet, a seed of a potential story to be co-constructed through interlocution in an act of co-construction. Durant et. al. (2006) built on the concept of antenarrative to suggest that an "indexical narrative fragment" refers to a fragmented utterance that "points" retrospectively to something familiar in an unfamiliar setting or unfamiliar in a familiar setting, in order to invite the interlocutor to join in a sensemaking discourse about the anomaly. In sum, whereas BME and fragmented narrative are typically retrospective (backward-looking), antenarrative is more apt to be prospective (forward-looking).

Given the recursive nature of dialogue, the respondents' indexical narrative fragments are therefore invitations to the interviewer to explore the out of place together in order to co-create or socially construct some sense out of the disordered—the out of place, in the past, present, or (anticipated) future. Taking up the invitation to co-construct sense out of the senseless, the interviewer also accepts an invitation to enter into a discourse community with the respondent—and a fragile, likely temporary, in-group is thereby formed. The interviewer becomes more involved (Tannen, 1987) and the respondent more descriptive.

It is our hypothesis, therefore, that invitational rhetoric leads to indexical antenarratives, particularly in discourse about crises. Partial, non-standard interviewer discourse leads to the partial and nonstandard respondent discourse that better depicts crisis experiences, which by definition are beyond the capability of routines, including language routines, to adequately express the novel circumstances.

Our story, as is often the case with stories, is about a contingency, in which an event from beyond the borders of the focal system intruded, causing a "crisis" of a greater or lesser degree (Hyde, 1998). In the Katrina story, the crisis was of an unimaginable magnitude. Our depiction of it is necessarily partial, so we will emphasize those aspects of the crisis that highlight its novelty. After that, we will describe standardization strategies of interview methodology and contrast these to potential invitational rhetoric strategies identified in the literature. Then we will use our data to illustrate the vividness of indexical antenarratives as depictions of crises and to suggest what happens when a respondent offers an indexical narrative fragment or antenarrative that is declined (or unnoticed) by the interviewer. Next, we give some background on hurricane Katrina, after which we give a brief review of interview rhetoric strategies.

Hurricane Katrina

The most destructive natural disaster to date in U.S. history, Hurricane Katrina killed more than 1300 people, 80% of them in New Orleans. Subsequent stresses of evacuation, homelessness, separation from family, lack of insurance records, difficulty finding jobs, and other conditions of refugee life— Katrina destroyed an estimated 300,000 homes (Federal Response, 2006)—took their toll on the elderly and the poor from the city, causing countless additional fatalities.

Plans had been made and preparations were put into place. Expecting a Florida landfall, officials had sent supplies to staging areas in Georgia, and in anticipation of the Gulf Coast landfall, FEMA pre-staged the largest pre-positioning of federal assets in history. Private sector organizations like railroads and agriculture supply companies readied repair and alternative transports. As the hurricane approached, Alabama, Louisiana, and Mississippi State Emergency Operations Centers (EOCs) expanded their staffing and operations schedules. Even so, Katrina's immensity and New Orleans' flooding overwhelmed response abilities: "Emergency plans at all levels of government, from small town plans to the 600-page National Response Plan...came up short" (Federal Response, p. 1).

Initial hurricane winds and rain cause the lake to overtop the levee system, but the city felt spared. However, the next day, 80% of New Orleans was inundated with 6-20 feet of water from breaches in the levees. Hospitals and nearly all health care facilities were destroyed or rendered inoperable. Firefighters, police officers, and medical personnel across all levels of government, together with citizen volunteers, braved life-threatening conditions to rescue people and animals from flooded buildings (Federal Response, p. 38). Flooded residents flocked to the Superdome, itself now vulnerable to the rising water and wind-damaged roof, lack of resources or utilities, and social tension; to the convention center, never authorized as a shelter; and to raised surfaces, such as the interstate. No post-storm protocols for evacuation were in place. Language routines were inadequate to account for the devastation and responses to it:

Finally, the communications problems had a debilitating effect on response efforts in the region and the overall national effort.  Officials from national leaders to emergency responders on the ground lacked the level of situational awareness necessary for a prompt and effective response to the catastrophe.  This was a recipe for an inefficient and ineffective Federal response. (Federal Response, p. 41)

Interview Rhetoric

Our intent in presenting some detailed rhetorical strategies is not to provide a sharper set of tools with which to pry out useful answers. Instead, we hope to promote an overall approach, a moral orientation, to conducting research interviews about crises. Traditional research interviews fall under the more general class of communicative interaction, which takes place within a more or less shared horizon of pre-understood meanings (Berkowitz & TerKeurst, 1999). Common understandings and identities are often attributed to demographic variables such as race, culture, gender, or nationality, but those are not the only possible bases for shared understanding. Mulligan (2003) identified shared attachments to a place as a source of identification.

In interactive discourse participants employ discourse strategies using available resources to achieve communicative goals (Morales-Lopez et al., 2005). Personalizing interview strategies can be contrasted with a depersonalized, objective research style. Personalizing strategies include using colloquial vocabulary, first-person pronouns, examples or analogies to common circumstances, popular sayings or proverbs, explanations, sympathetic statements, and complimentary attributions. Objective discourse styles, on the other hand, include minimizing the interlocutor's framework; presenting oneself as a medium for the institutional objectives; and employing technical vocabulary, impersonal constructions or clauses in which the subject is an institution, passive voice, "never" or "always," and discussion of institutional rules and procedures without offering explanations.

Willingness to laugh during the interview is another personalization strategy, including in response to a laugh initiated by the respondent (Lavin & Maynard, 2001). For interviewers who are not themselves part of a crisis situation, that laughter can ironically be more difficult to engage with, even in response to following a cue from their interlocutor. Humorous observations, expressions of puzzlement, or repetition, "replay" of respondent phrases (either verbatim or muddled) by an interviewer, invite less structured responses and facilitate understanding, but can also damage rapport, especially with "divergent backgrounds... alternative worldviews" or where "the question may be interpreted as offensive, tactless, or meaningless" (Snow et al., 1982: 292). Other kinds of phrases that substitute for questions and thus serve as invitations to elaborate include descriptive comments, simply noting something the interviewer has observed, and evaluative comments, noting feelings or opinions the respondent might hold (Snow et al., 1982). Tannen (1987), too, noted the ubiquity and usefulness of repetition of phrases in talk.

We suggest that personalization strategies, such as those noted above, are useful because they build and support rapport between respondent and interviewer, rapport that leads to the creation of an "in-group" in whose safety the respondents are willing to divulge their experience of unanticipated and challenging events and feelings. We also suggest that such strategies diverge from epistemological assumptions of modern scientific techniques. Lavin and Maynard (2001: 454), for example, argue that following a highly "standardized" interview "script" is stressed in Western societies where "increments of time, space, volume, weight, distance, and value have all become subject to uniform regulation."

The challenge for the qualitative researcher is how to balance these incommensurate requirements of personalization and standardization in an interview setting. Success in dealing with the paradox is particularly important during interviews about crisis situations because: 1) safety is paramount for those who experienced a crisis and 2) the data sought after is about a novel situation, which will probably best be expressed with non-standard language.

Our main concern is not with better elicitation, it is the story rights of those in the crisis situation that must be first priority. We suggest that Foss and Griffin's (1995) invitational rhetoric provides an appropriate moral, as well as epistemological, and operational guide for identification and selection of appropriate discourse strategies in an interview-elicitation setting. Foss and Griffin (1995) coined the phrase "invitational rhetoric" to suggest an alternative view of rhetoric, one in which the goal is not to bring the other into line with one's own framework but rather to invite communicative interaction of the participants.. Stories "represent an initial, tentative commitment to a perspective that is subject to revision as a result of the interaction" (Foss & Griffin, p. 8).

In invitational rhetoric, the goal is to be open to and appreciate and validate the others' perspectives, to learn from the other, and to build an understanding the others' equal value (Ryan & Natalle, 2001). Invitational rhetoric's goal is the articulation of individual perspectives "as carefully, completely, and passionately as possible to give them full expression and to invite their careful consideration by participants in the interaction" (Ryan & Natalle, p. 7). As the ideas and feelings respondents share with the interviewer are received with respect and care, respondents are provided a feeling of safety (p. 10) in the belief that their perspectives are valid and informative, even when those perspectives are fragmented, uncertain, and partial—when they are antenarratives. Before we present our data, therefore, we briefly outline some form and function of antenarratives, with particular attention to indexical antenarratives.

Narrative Fragments & Antenarratives

Retrospective narrative fragments and prospective antenarratives are developed collectively and are subject to revision (Boje, 2001). Antenarrating means you are trying to recontextualize or decontextualized, picking up content or perspective in one context and letting it loose in some other context. Antenarratives are travelers, depending upon an expanding sociometry, a growing network of actors to work in bits of context, to send the antenarrative along its way to becoming a more coherent, less changeable narrative. Antenarratives are by definition "incoherent, collective, unplotted" (p. 1) bets on the future, born from the drive to tell a story before the storyline has been socially agreed upon, yet able to be transformative of social context (Letiche & Boje, 2001). In discourse, antenarratives are characterized by "abbreviated and interrupted story performances that yield plurivocality (Boje, 2001, p. 4) and ambiguity that allows change through social negotiation and conciliation. As admittedly incomplete, antenarratives, as well as retrospective narrative fragments, can identify boundary conditions of one's own perspective and so reveal partiality, uncertainty, and admitted lack of control—all characteristics of crises. In doing crisis inquiry, narrative fragments may not mean there is a tersely told more coherent narrative a skilled interview can strategically uncover, such as in a detective novel. Rather, the interviewee may be in trauma, and the crisis has not passed from reenactment (reliving events with raw emotion) into storyable experience. To story is, by definition, to make events storyable, more willful control of the memory of events, shaped into the wisdom of experience. Interviewees, in order to make some crisis discussible, need language and images that can be shaped into meaningful stories. The interviewer can facilitate or impede this process.

As they signal the existence of a break or rupture in the expected, narrative fragments and antenarratives can indicate the triggering of initial sensemaking activity. Fragments may be sortable into a coherent narrative. Or, antenarratives can locate the points at which the expected processes and outcomes not yet present may yet emerge. Specifically, we suggest that "indexical narratives" provide a useful tool for inviting others to help us make sense out of events, of particular concern in crises, that have yet to become coherent experiences. If the interviewer is merely prompting an interviewee to recite some often retold tale, then fragments have already been sorted, and the antenarrative map to the future is not very ephemeral. Sometimes our world is turned so upside down that all we can do is point out the strange. There are times and places in which we cannot give reasons for events, nor weave them into a coherent narrative with a beginning, middle, and end (BME). In such cases, we may only be able to utter "indexical narrative fragments" or in special cases, "indexical antenarratives," pointing out something still out of place, or likely to remain so. Indexical narrative fragments, either point to a more fully formed narrative coherence or to the possibility one can be con-constructed. "Index" denotes the index finger gestured toward an object, saying "Look at that!" Indexical antenarratives emerge from the phenomenological, wordless experience of a rupture in the expected. As respondents use indexical antenarratives to describe their observations of "the strange, the out of place, the unexpected" to the interviewer, they are simultaneously inviting the interviewer to join with them in co-constructing a shared appreciation for the importance of what they saw.

Next, we summarize our research approaches and operational procedures.

Research Methodology

Our example indexical narrative fragments and antenarratives were gathered during the course of research which sought to assess the effect of public and private information sources on evacuation and relocation decisions in the context of Hurricane Katrina. Most interviews were conducted within two months of the disaster (the time it took to acquire IRB approval) in four states by both communication faculty members and graduate students taking a research methods course. The earliest interviews (10) were conducted by faculty in South Carolina, through a shelter established for evacuees. Parallel to evacuation patterns for those fleeing Katrina, the bulk of the interviews (84) were conducted in the Baton Rouge, Louisiana, area. Louisiana interviewers were given cards for distribution to the interviewees with toll-free numbers for local counseling services that had been set up specifically for Katrina evacuees. Consent forms, protocol forms, demographic forms, cassette recorders, microcassette tapes and incentive discount store gift cards were also distributed at the same time. Some interviewers who seemed to have less confidence were initially paired with more experienced interviewers for support and further mentoring. Sensitivity to the interviewees was stressed.

The Louisiana interviewees were recruited from a Salvation Army shelter, a church clothing distribution center next to a Red Cross shelter, and a church-affiliated food distribution facility in the City of Baton Rouge. Each of these facilities had been set up specifically for Hurricane Katrina evacuees. Additionally, snowball samples were conducted in the community; that is, students were asked to recruit additional interviewees based on word-of-mouth referrals

An additional nine interviews were conducted in Tuscaloosa, AL, by two graduate students of the University of Alabama under the general supervision of a faculty member from Tulane who was herself an evacuee with visiting professor status at UA. Approximately nine additional interviews were conducted in Texas by the same Tulane faculty member, primarily through the West Houston Association of Ministries, a non-denominational relief organization that assisted evacuees with housing, food, clothing, and other basic necessities. The remainder of the Texas interviews were conducted at the temporary offices set up in Houston by Tulane University with the purpose of broadening the range of interviewee demographics.

The interviews ranged in length from ten or fifteen minutes (generally characterizing inexperienced interviewers' initial experiences) to over two hours. A half hour to 45 minutes was typical. A few of the interviews were lost due to failures of the recording equipment or human error, but most were successfully captured on tape and transcribed for analysis. The results reported here depend primarily on the transcripts, but also on comments and observations supplied by the faculty supervisors.

We turn now to the kinds of interview strategies that either elicited, circumvented, or ignored indexical narratives and antenarratives.

Findings

In our procedures, we defined the presence of invitational rhetoric as marked by open-ended questions, participatory responses such as backchanneling, and explicit acknowledgement of listening and hearing the other's perspective. We expected that invitational rhetoric would result in more instances of antenarratives being developed into narratives characterized by the normally expected BME markers of narrative coherence. We also expected that more conversational turn-taking behaviors, more signs of involvement, and greater willingness to admit to partiality and uncertainty, would yield better evidence of the co-constructedness of indexicals, and thus of a shared reality.

Below, we introduce a variety of examples of formal, "depersonalized" interviewing techniques, in which roles are clearly demarcated and the goal is standardization of accounts. Six examples of the way in which such interactions resulted in undeveloped antenarratives and reduced co-construction are provided. Subsequently, we provide eight examples that illustrate invitational rhetoric and co-construction, showing how these interactions effectively work to stabilize both listener and speaker in a more complex way. In these cases, we show how different rhetorical strategies supporting invitational rhetoric lead to both antenarratives and the shared development of antenarratives into narratives.

Standardized Rhetoric Strategies

Pre-set codes. One way to standardize interviews is to concentrate on the facts and to record those facts into pre-set codes. Example One illustrates the impact of such standardization. Asking for facts like "where and when" resulted in no narratives, and the use of questions with outlines resulted in a short response "another person." This answer, even though it seemed to beg for a follow-up of "what person?" was merely translated into the standard code "word of mouth." A similar tactic for objectivity is to double check on the accuracy of the responses. Example Two illustrates how testing response accuracy shuts down co-construction. In this example, the open-ended question resulted in a relatively long narrative, which included an internal rhyme scheme "boarding... loading" that gave the story a sense of increasing momentum. The opening question, phrased "I would like," used a more invitational structure that construed answering as a "favor." The follow-up probe questions that challenged the point and the accuracy of the responses ("So your mom really watched") resulted in almost monosyllabic responses. The more ambiguous "situations like that" was likely an attempt by the interviewer to re-invite greater respondent participation.

Double-checking accuracy. Also important to standardized research is sticking to the script. The interchange in Example Three complicates our hypothesis in shared laughter would suggest good rapport. However, the interviewer used relatively closed questions in order to keep the interview on-track, and did not invite more participation around the ante-narrative "after the water level started rising."

Laughter unreciprocated. Because laughter often points to the ironic, it is often seen as inappropriate when clarity and pointedness are important. In Example Four, the respondent is the one who invited the interviewer to get more involved in the dialogue, but neither instance of laughter was reciprocated. Similarly, the respondent's indexical antenarrative "I watched the water come up fast, because I didn't believe it ‘til I seen that" was completely ignored by the interviewer, who stuck with the script and continued down the list of questions on the interview protocol.

Quantification. An essential tool in standardization is quantification, as seen in Example Five. Our ability to assign numbers allows for specific valuations and broader comparisons. The respondent resisted assigning a number to the experience; she wasn't clear on what the number would mean and didn't trust that a number would be able to do justice to the "living experience."

Playing the bystander role. Standardized interview methodology is based on the power dynamics of structuralism, whereby the "knower" stands outside of and therefore impervious to "the known." In Example Six, the respondent's discourse of being knowledgeable and witty with ironic constructions of academic terminology is ignored by the interviewer. Our respondent's narrative fragment of "semi-normal" begged explanation, as did his passing reference to "contra-flow." The respondent's bid for authority with the use of technical terms was not acknowledged. There are two possible explanations for how this co-construction worked; perhaps both interlocutors fully understood the meanings of both terms (though less likely, since there was no affirmation or acknowledgement) or perhaps the interviewer was focused on constructing the research project while the interviewee was working to construct the stability of post-crisis "new normal."

Invitational Rhetoric Strategies

Generally speaking, invitational rhetoric is a discursive strategy that values the interlocutor's point of view, particularly because it is invaluable to a co-constructed interpretation of events.

Repeating phases. Repeating phrases used by the respondent gives evidence of careful listening and acceptance of the respondent's point of view. Example A illustrates invitational rhetoric resulting in greater conversational involvement and movement past the initial sense of crisis. The first and second narratives here each ended with invitations towards co-construction. The shift from first to second person indicated turn-taking and was a move towards a present-oriented shift instead of past-oriented. The response moved the conversation back into past-tense, and used two separate question forms, one open-ended, and the second suggesting alternatives. The second retrospective narrative fragment opened with specific alternatives raised in the question format, but then returned to a pragmatic (action-oriented if not present-oriented), and prospective antenarrative focus. The second narrative fragment ended with a humor bid, which was not followed up on.

Repair work. After these two fragmented narratives, the next question was answered monosyllabically, so the interviewer did repair work by repeating the response and then clearly indicating that it was still the other speaker's turn with "okay." The next question used phrases taken directly from the second narrative such as "first off" (compare to "started off") and "go back." The questioner also mirrored the shift to present and future action-orientation in order to elicit further narrative development.

Empathy. Empathetic and encouraging comments support disclosure and participation. Comments such as "amazing" and "wow" illustrate a willingness to appreciate and try to share the perspective of the other. Colloquial phrases and slang signal informality and equality, as we see in Example B. In this example, there was a mirroring of linguistic elements: for example, "and" in the first narrative got repeated in the next question. Characteristic southern "y'all" was echoed by "gone into the neighborhood somehow and knocked, known he was there and knocked." The narratives were more fully developed and included markers of sequentiality ("well in the meantime" "about a week later") and rising action ("that was the interesting part" "everything came together").

Backchanneling. Backchannelling as an indicator of acknowledgement is another way of co-constructing, and the listener's role should not be discounted as less active in the narrative work here just because they say less. The listener in Example C here shaped the response. The initially short, factual (and somewhat negative) indexical narrative fragments got developed into longer retrospective narratives. By the time the next full question was asked (following up on the question about how the respondent got to Baton Rouge), the respondent included dialogue, the word "like" (indicating an informal and analogic schema), and expressions of uncertainty. Repeating "the kindness" and "I never knew" lent a moral and mythic quality to the story which transitioned into a story about the affirmation of life and community rather than chronicling isolation, with a forceful conclusion about the reality of the newly constructed understanding by ending with "actually."

Seeking clarification. Asking for clarification with comments and repetition (with and without rising intonation) signals interest in being clear on the details of the respondent's experience. The narrative here in Example D was co-constructed through roughly equal relatively short turns. Both participants used affirming and backchannelling responses such as "right" and "mmm hmm." Questions and answers repeated segments from previous talk turns for cohesion and clarity. The brevity of responses indicates a fast rate of turn-taking, indicating high involvement of both parties.

Interpersonal support. Invitational rhetoric is inclusive and recognizes that individuals co-create experiences and stories about them, as seen in Example E. Here there were two interviewees, supporting by repeating one another's indexical antenarratives: "No people," "Nothing man"; "All that's down," "All sorts of cars, cars flipped upside down." The interviewer's repetition and interpersonal support created a safe place to let humor represent tragedy through irony and expression of emotions: "gonna make a joke" "you gotta make a joke," You know, to keep from crying." The repetition carrying over from one turn to the next broadened to include a depiction of the community: "Like a ghost town" "It's a mess. It's a ghost town." Because the repetition of indexical narrative fragments indicated a potentially shared perspective, there was little need to develop them into more elaborate accounts. The longest narrative, about the bus, was the least shared and thus required the most description to convey.

Exploring boundary conditions. Invitational rhetoric takes for granted that one's own perspective is not shared. Therefore, explanations and elaboration of the interviewer's own observations establish the boundary conditions of similarity and difference with the respondent, as we see in Example F. This narrative included relatively more background information on the interviewer, and invited in turn detailed narratives from the respondent. Repeated segments tied the perspectives together into a coherent interlocution.

Superlatives. Superlative semantics, including adjectives and adverbs such as "fantastic" and "wonderful," reinforce the uniqueness, rather than the standardization, of the respondent's experience. Unique perspectives develop into vivid depictions as both participants use emotional and evaluative language in Example G. Here, the interviewer offered a situational definition that confirmed the construction offered by the respondent: "That's okay [to talk a lot]. Especially in situations like this." Vivid verbs and adverbs included "single-handedly," "rescued," "tracked down," and "literally." The prefacing of "pointed questions" with the open-ended "anything else you'd like to add?" signaled that the perspective of the respondent was of equal or greater value than the interview protocol.

Poetic repetitions. Invitational rhetoric values poetic forms as ways to depict an experience of crisis intensity. Given the limitations of language, repetition is therefore used not only as a rhetorical device for clarity and cohesion, but also for aesthetics and emphasis. In Example H the repetition of the number three (three weeks, three days, three times) lent the narrative a mesmerizing mythic quality. Each additional narrative fragment, punctuated by backchannelling in a nearly call-and-response form, linked both forward (antenarrating) and backward (retrospecting). Devices such anaphora ("we had no phone, we had no electric" "...left there... got here" "contacted here... stay here... been here") had variation within the repetition, as did the lexical shifts from automobile to car to vehicle.

In sum, we have offered examples of interview techniques that either facilitate or shut down inquiry into the nature of the story. In the case of full-blown BME narratives, a fragment only indicates that a full narrative has been already told, but is being kept out of the interview. In the case of fragmented narratives that do not have a previously performed narrative map, acts of co-construction on the part of interviewer and interviewee can allow events to become shaped into more coherent memories of experience. Finally, while most research has been about retrospective sensemaking the past, there are antenarrative moves between interlocutors that are prospective sensemaking, dancing in the present to make the future seem less ephemeral.

Conclusion

Crisis by its very nature asks us to find new ways to imagine meaning in our worlds, to engage in the intersubjective process that draws upon the work of many. An awareness of invitational rhetoric techniques can help in sorting fragments of previously told full blown narratives from those that have yet to be shaped. We suggest that this is quite important in crisis research. Further, invitational rhetoric can set the stage for antenarrative inquiry. Expressions of openness, and invitation, and support are our offerings for those confronting crisis, an offering identified in both scholarly literature and community norms. We honor the narrative fragments, and antenarrative indexicals, as fragmented pieces of stories about shattered lives and expectations, as offerings that help us recognize and respond to the early stages of sensemaking.

At the same time, we are reminded that the construction of meaning is a joint effort, and therefore the question of who is doing the listening is as important as who is doing the narrating. Those who have lived through crisis are able to participate in the co-construction of meaning in a way that is often richer, and more sensitive to the uncertainties and limitations of language when routines are disrupted. Co-construction plays a vital role both in telling the stories about what has happened in this specific crisis, and also a vital role in research about crisis. It matters not simply whether one is insider or outsider in relation to a specific community struck by crisis; what matters is the skill set brought to the role of researching, interviewing, and listening.

The sharing of meaning is dependent on and supportive of the relationship among the parties, individuals, or systems involved. Because in research interviews, the interviewer initiates the moves in this particular language game, it is up to the researcher to model the appropriateness of non-standard and novel language forms and expressions and content. Accepting the respondents' invitational narrative fragments requires the researcher to both recognize and acknowledge the fragments that refer to anomalies. Researchers' shared frameworks, a pre-understanding of the story's context, are therefore valuable to the research process. During times of crises, when pre-givens are rare, a simple "I know what you mean" can provide a sense of stability and community to the respondent. While research on narrativity has tended to focus on the role of the speaker, the role of the listener has been acknowledged as critical (Labov etc.). Interview research depends on the active listening skills of the researcher. Due to the recursive and turn-taking conventions of conversation, the antenarratives of the respondent are openings for the researcher to invite the respondent to "tell me more." Researchers who themselves have experienced the crisis are thus uniquely situated to recognize and respond to such invitations.

Of course, others can also develop the skill sets needed to conduct crisis research, and those who have experienced crisis should not feel pressured to have to do crisis research simply because they have a unique and useful perspective. But one way to bring meaning out of the uncertainty of crisis, is for the community of "outsiders" to a specific crisis to recognize and help co-construct narratives that value what comes out of the experience of crisis. It is our hope that seeing the difference created by that difference in terms of research interactions as illustrated here, can help in that regard.

References

Barley, S. R. (1991). Contextualizing conflict: Notes on the anthropology of disputes and negotiations. In M. H. Bazerman, R. J. Lewicki, & B. H. Sheppard (Eds.), Research on negotiation in organizations: Handbook of negotiation research (pp. 165-199). Greenwich, CT: JAI Press.

Berkowitz, Dan, & TerKeurst, James (1999). Community as interpretive community: Rethinking the journalist-source relationship. Journal of Communication Vol. 49, Iss. 3; p. 125-137.

Boje, D. M. 1991. “The storytelling organization: A study of storytelling performance in an office supply firm.” Administrative Science Quarterly, Vol. 36: pp.106-126.

Boje, D. M. 2001. Narrative Methods for Organizational and Communication Research. London: Sage

Boje, D. M. 2007. Storytelling Organization (forthcoming). London: Sage.

Deutsch, M. (1990). A framework for teaching conflict resolution in the schools. In B. H. Sheppard, M. H. Bazerman & R. J. Lewicki (Eds.), Research on negotiation in organizations (pp. 189-203). Greenwich, CT: JAI Press.

Durant, R. (2002). You mean the world to me: Story telling and leader listening in organizational learning. Unpublished dissertation, The University of Alabama.

Durant, R., Gardner, C., & Taylor, K(2007). Indexical antenarratives as invitational rhetoric. Tamara Vol. 5 No. 4: 174-182,

The Federal Response to Hurricane Katrina: Lessons Learned (February 2006) http://www.whitehouse.gov/reports/katrina-lessons-learned/ Accessed 7/29/2007

Foss, Sonja K. & Griffin, Cindy L. (1995). Beyond persuasion: A proposal for an invitational rhetoric. Communication Monographs, Vol. 62, March (2-18).

Gadamer, Hans-Georg (1976). Philosophical Hermeneutics. Translated and edited by David E. Linge. University of California Berkeley

Grondin, Jean (1994) Introduction to philosophical hermeneutics. Yale University Press New Haven, quoting Gadamer "Language as Dialogue." 117-120

Hopfl, H. (2000). On being moved. Studies in Cultures, Organizations, and Society, 6, 15-34.

Hyde, L. (1998). Trickster makes this world: Mischief, myth, and art. New York, NY: Farrar, Straus, and Girous.

Lavin, Danielle & Maynard, Douglas W. (Jun., 2001). Standardization vs. Rapport: Respondent Laughter and Interviewer Reaction during Telephone Surveys. American Sociological Review, Vol. 66, No. 3. pp. 453-479.

Letiche, Hugo & D. M Boje (2001). Phenomenal Complexity Theory and the politics of organization  Emergence: A Journal of Complexity Issues in Organizations and Management. Vol. 3 (4): 5-25.

Lindsley, D. H., Brass, D. J., & Thomas, J. B. (1995). Efficacy-performance spirals: A multilevel perspective. Academy of Management Review, 20(3), 645-678.

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709-734.

Morales-Lopez, Esperanza, Prego-Vazques, Gabriela, and Domingues-Seco, Luzia (2005). Interviews between employees and customers during a company restructuring process. Discourse & Society, Vol 16, No. 2, pp. 225-268.

Mulligan, Melinda (2003). Displacement and identity discontinuity: The role of nostalgia in establishing new identity categories. Symbolic Interaction Vol 26(3), 381-403.

Roberts, A., & Dzieglielewski, S. (1995). Foundation skills and applications of crisis intervention and cognitive therapy. In A. Roberts (Ed.), Crisis intervention and time-limited cognitive treatment. Thousand Oaks, CA: Sage Publications.

Ryan, Kathleen J. & Natalle, Elizabeth J. (2001). Fusing horizons: Standpoint Hermeneutics and Invitational Rhetoric. RSQ: Rhetoric Society Quarterly, Vol. 31, Issue 2. (69-90).

Snow, David A., Zurcher, Louis A., Sjoberg, Gideon (1982) Interviewing by Comment: An Adjunct to the Direct Question. Qualitative Sociology. Winter82, Vol. 5 Issue 4, 285-311.

Tannen, Deborah (Sep., 1987), Repetition in conversation: Toward a poetics of talk. Language, Vol. 63, No. 3, 574-605.

About the Authors

Karen Taylor teaches organizational communication and does research primarily in rhetoric of science. Her focus is on the ways in which institutional structures shape how knowledge is produced, used, and shared. She finished her Ph.D. at University of Pittsburgh in 2003, and her MA from Texas A&M.

Rita Durant learned about the importance of high quality relationships from Jim Cashman, of rapport from Kim Sydown Campbell, and of taking the high ground in discourse from Ron Dulek. She learned about stories from Steven Hobbs and about interpersonal systems of meaning from Joanne Terrell. She has published in the Journal of Organizational Change Management, Emergence, Tamara, Journal of Business Communication, and Organization Studies.

David M. Boje holds the Bank of America Endowed Professorship of Management, and is past Arthur Owens Professorship in Business Administration in the Management Department at New Mexico State University. Professor Boje is described by his peers as an international scholar in the areas of narrative, storytelling, postmodern theory and critical ethics of answerability. He has published nearly 100 articles in journals, including the top-tier journals such as Management Science, Administrative Science Quarterly, Academy of Management Journal, Academy of Management Review and the international Journal of Organization Studies.