Is Telling Stories Good for Democracy? Rhetoric in Public Deliberation after 9/11

ion of logical arguments come into play on some occasions and not others. The authors appraise this argument by way of a systematic comparison of personal storytelling and reason-giving in public deliberation, the first such empirical study. Drawing upon an analysis of 1,415 claims made by 263 people in 12 discussion groups, the authors show that ordinary conventions of storytelling helped deliberators to identify their own preferences, demonstrate their appreciation of competing preferences, advance unfamiliar views, and reach areas of unanticipated agreement. The ambivalence, however, with which participants generally viewed storytelling as a rhetorical form restricted personal stories to discussions that were seen as without impact on the policymaking process. More broadly, by drawing attention to the evaluative structures through which people’s use of cultural forms is differentially assessed, the authors provide an alternative to both instrumentalist and structuralist approaches to culture. AMERICAN SOCIOLOGICAL REVIEW, 2006, VOL. 71 (October:699–723) Direct all correspondence to Francesca Polletta, Department of Sociology, University of California, Irvine, 3151 Social Science Plaza, Irvine CA 92697 ([email protected]). The authors thank Christopher Anderson, Neal Caren, Caroline Carr, M. Kai Ho, and Joshua Karp for help with coding and analysis, and Edwin Amenta, Nina Bandelj, Martha Feldman, Rodney Lacey, Ann Mische, Calvin Morrill, Marc Weiss, members of the Russell Sage Foundation Colloquium and the Center for Organizational Research seminar at UCI, four ASR reviewers, and the ASR editor Jerry A. Jacobs for valuable comments on earlier versions of the article. The research was supported by a grant from the National Science Foundation (IIS 0306868) and was completed while Polletta was a Visiting Scholar at the Russell Sage Foundation. The authors thank both institutions. Delivered by Ingenta to : Columbia University Fri, 12 Jan 2007 18:21:39 deliberative discourse be oriented toward reasons? Questions like these have spurred a vigorous debate among political theorists, with some arguing for the deliberative value of personal and passionate forms of talk (Young 1996, 2000; Sanders 1997) and others insisting on the necessity of rational argument (Guttman and Thompson 1996; Dryzek 2000; Miller 2002)—but there has been little empirical examination of these questions. Sociologists, for their part, have tended to concentrate on the social contexts and cultural content of a vibrant public sphere rather than on the rhetorical forms that predominate within it. So, with respect to context, they have looked, for example, to the organizational structure of civic life (Skocpol 2003), nationally distinctive routines of news-reporting (Ferree et. al. 2002), and the associational backgrounds of deliberative forum participants (Baoicchi 2003). With respect to the content of a democratic public sphere, sociologists have assessed the prominence of selfor other-regarding orientations in the culture at large (Bellah et al. 1986) and examined the micropolitical cultures that promote arguments based on selfinterest or that privilege one understanding of individualism over another (Perrin 2005; Eliasoph and Lichterman 2003). The existing empirical work on the rhetoric of good deliberation has been either historical (Schudson 1992, 1997; Fraser 1992; Ryan 1992) or based on the ethnographic study of activists or of people in nonpolitical settings (Hart 2001; Lichterman 1996; Eliasoph 1998; Eliasoph and Lichterman 2003). In an exception, Gamson’s (1992) study of talk among groups of ordinary Americans showed that people’s experiential knowledge, typically expressed in personal stories, was a source for a critical understanding of political issues. Gamson, however, did not evaluate storytelling along any other dimensions of political deliberation. The larger gap reflects in part a tendency among cultural sociologists to focus on meaning more than on rhetoric; but it also reflects the methodological difficulty of studying a phenomenon—the public sphere—that has existed mainly in the historical and theoretical imagination. That is, until now. In the past decade, public deliberative forums touted as a realworld incarnation of deliberative democracy have proliferated, some of them convened as part of local or national policy-making processes (Delli Carpini, Cook, and Jacobs 2004; Ryfe 2005; Fung 2003; Baiocchi 2003). These forums provide an opportunity to appraise the conditions for democratic discourse among citizens in a nonexperimental setting. We argue that doing so requires a distinctly sociological perspective. The capacity of rhetorical forms such as arguments, explanations, and stories to foster good deliberation rests not only on features of the form but also on social conventions of its use and evaluation. In this article, we draw on scholarship in sociolinguistics to identify norms of everyday conversation that open up possibilities for effective deliberation, possibilities that have been overlooked by democratic theorists. Conversational norms also levy previously unnoted constraints on fair deliberation, however. In particular, we draw attention to popular beliefs about the authority of different rhetorical forms. The “grammars of worth” that structure the evaluation of aesthetic objects and positions in political disputes (Boltanski and Thévenot 2006; Lamont and Thévenot 2000) also structure the evaluation of rhetorical forms. They operate, however, in ways that are variable rather than f ixed. Concerns about the credibility or authority of a particular form are more likely to be triggered by some users and in some contexts rather than others. Teasing out the hierarchies of evaluation through which citizens’ talk is variably assessed can shed light on the conditions for genuinely democratic deliberation. Such an inquiry also suggests a new answer to the age-old sociological question of why it is that even as people use culture practically and creatively, they do so in ways that largely sustain the status quo. Social inequalities may be reproduced not only by people’s uneven capacities to use culture effectively (Bourdieu 1984) but also by the disparate ways in which the culture they use is assessed. In this article, we focus on rhetorics of reason-giving and personal storytelling. Democratic theorists have debated the merits of these two rhetorical forms (Guttman and Thompson 1996; Young 1996, 2000; Sanders 1997; Dryzek 2000; Miller 2002). There also exists a rich literature on storytelling in a variety of institutional settings, for example, legal (Briggs 1997; Conley and O’Barr 1990; Maynard 1988; Manzo 1993; Trinch and BerkSeligson 2002), medical (Clark and Mishler 1992), occupational (Linde 1993), and familial 700—–AMERICAN SOCIOLOGICAL REVIEW Delivered by Ingenta to : Columbia University Fri, 12 Jan 2007 18:21:39 (Ochs and Capps 2002; Johnstone 1990). There has, however, been no empirical comparison of personal storytelling and reason-giving in public deliberation. Here, we study twelve groups of ordinary citizens as they deliberated online about how to rebuild the World Trade Center site in the wake of the 9/11 terrorist attack. In a forum sponsored by rebuilding authorities and civic groups, participants were asked to make recommendations about the design of the site, as well as about housing, transportation, and economic development plans, and a memorial planned for the victims of the disaster. Through a quantitative analysis of the distribution of reasons and personal stories across kinds of claims, claimants, and responses, we identify patterns in who used personal stories in relation to what topics, and with what effect. Through a qualitative analysis of a sample of storytelling exchanges, we probe in more detail the formal features and conventions of use that equipped personal storytelling to gain a hearing for marginalized views, spur deliberative engagement, and advance compromise positions. Finally, through a qualitative analysis of participants’ talk about storytelling in the course of their discussions, we account for a puzzle: participants were enthusiastic about the value of personal storytelling but told personal stories mainly in discussions that were seen as without impact on the policy-making process. In brief, we find that narrative’s conventional openness to interpretation—in essence, its ambiguity— proved a surprising deliberative resource, especially for people with marginalized points of view. Yet the ambivalence with which storytelling was popularly viewed—as simultaneously normatively compelling and politically unserious—ended up disadvantaging just those people. DEMOCRACY AND STORYTELLING Arguments for the virtues of deliberation go back to Aristotle, but they have taken on special force in the context of widespread concern about contemporary democracies’ low levels of citizen engagement and the generally polarized character of political debate. As theorists of deliberative democracy see it, the solution to this state of affairs is to create opportunities for public deliberation (Habermas 1984, 1989, 1996; Barber 1988; Fishkin 1991, 1995; Cohen 1989; Dryzek 2000; Bohman 1996). Public discussion of hot-button political issues can yield areas of unanticipated agreement. Even if participants do not change their minds, they will likely come to recognize a greater range of preferences as legitimate. Once that recognition occurs, people are likely to accept a decision that does not match their preferences exactly (Cohen and Sabel 1997; Shapiro 2002). This means that public deliberation can be integrated with existing electoral, legislative, and administrative processes (Cohen 1989; Fishkin 1995; Guttman and Thompson 1996). The key question, then, is what makes for good deliberation. Most scholars agree that discussion must be open to all, and participants must be unconstrained in the arguments they make, save by the requirement of civility. Deliberation should be free of outside control and should aim for agreement that is uncoerced (Bohman, 1996; Habermas 1984; Cohen 1989; Fishkin 1995; Guttman and Thompson 2004). In such a setting, deliberation properly takes the form of reason-giving. Participants justify their preferences by making arguments that others can accept as persuasive (Bohman 1996; Cohen 1989; Habermas 1984; Gutmann and Thompson 2004:3).1 Yet it is precisely this requirement, considered foundational by most deliberative democrats, that has been powerfully challenged by a recent line of criticism. Even if people are granted equal access to deliberative forums, they are not equally able to use the reasoned discourse that is privileged there. Men, white people, native speakers of standard English, and those with cultural capital are both better equipped to formulate abstract reasons and are more likely to be heard as giving good reasons, no matter what they actually say (Young 1996, 2000; Sanders 1997; Bickford 1996; Mansbridge 1999). Further disadvantaging some groups, critics PUBLIC DELIBERATION AFTER 9/11—–701 1 Deliberative democrats differ on whether the values and principles to which participants appeal should be shared by all participants already or whether it is the possibility of unity that motivates them. Most deliberative theorists, however, talk easily about the “universal standards” (Dryzek 2000:69), the “common good” (Cohen 1989), or the “impartial values” (Elster 1998:6) to which deliberators should appeal. Delivered by Ingenta to : Columbia University Fri, 12 Jan 2007 18:21:39 argue, the widely shared values and universal principles to which deliberators are supposed to appeal unfairly universalize the experience of particular, powerful groups. Those with different experiences are easily dismissed as deviant (Minow 1990). In short, privileging reasoned discourse comes at the expense of another deliberative standard, that of equality. The solution to these problems, as critics see it, is to legitimize diverse discourses in deliberation, chief among them, personal storytelling (Young 1996, 2000; Mansbridge 1999). In a deliberative setting, storytelling is equalizing, since everyone has his or her own story (Young 2000; Sanders 1997). By telling their stories, members of disadvantaged groups can gain an empathetic hearing for experiences and values that are unlike those of the majority (Murphy 1993; Young 2000). By showing how their particular experiences elude categories that are supposed to be universal, disadvantaged groups can expose the particularistic character of those principles. That is the first step to crafting more inclusive principles (Delgado 1989; Smith 1998; Sanders 1997; Young 2000). Deliberative democrats in the classical mold remain unconvinced by these arguments. They point out that listeners have no way of knowing how representative a story is, or even what the policy implications of a personal experience are (Miller 2002:217; Dryzek 2000). A story that is told badly may widen perceptions of difference rather than narrowing them (Miller 2002:219), and a story told too well may lead listeners to identify so strongly with the victimization described as to make them prone to vengeance rather than reasoned remedy (Dryzek 2000:69). Absent special effort on the part of storytellers to “appeal to universal standards” (Dryzek 2000:69) or “reasons and principles that are widely shared.” (Miller 2002:221), stories are likely to inhibit deliberation rather than advance it. Disadvantaged groups may have the satisfaction of expressing their needs in their own distinctive voices, but they do nothing to help deliberators move beyond a Babel of competing needs. Theirs is a “politics of futile gesture” that leaves the status quo intact (Gutmann and Thompson 1996:137; Miller 2002; Dryzek 2000). Which side is right? We argue that both sides have tended to treat personal stories as raw accounts of personal experience. This is why even proponents of storytelling have argued that personal stories must be combined with explicit appeals to normative principles to be effective (Young 2000:74). They, like deliberative theorists in the classical mode, have denied personal storytelling the capacity to move beyond registering differences to forging agreement across differences. We argue, to the contrary, that the narrative character of personal stories equips them to do the latter as well. To make this argument, we draw upon scholarship that addresses not only narrative’s literary form but also its use in everyday life; that is, the norms governing how people typically tell, hear, and respond to stories. We define a story, conventionally, as an account of a sequence of events in the order in which events occurred to make a point (Labov and Waletsky 1967; Labov 1972; Linde 1993). In a personal story, the protagonist is the narrator and the events recounted are presented as true. We define a reason (in a deliberative context) as a justification for an opinion that is based on the opinion’s consistency with a more general principle (Baumeister and Newman 1994; Bruner 1991; Elster 1998). Storytelling differs from reason-giving in ordinary conversation in at least four ways. Stories integrate description, explanation, and evaluation; they are detached from the surrounding discourse; they are allusive in meaning; and they are iterative in the sense that they elicit more stories in response. We elaborate on each of these features as we describe how they work to meet three critical challenges of deliberation. One challenge is to get deliberators to listen as well as speak (Barber 1988; Bickford 1996). Research shows that the quality of people’s opinions is improved when they take into consideration other people’s opinions (Price, Cappella, and Nir 2002; Wyatt, Katz, and Kim 2000; Stromer-Galley 2005)—but it is a distinctive kind of consideration that is required. If one danger is that people hear other perspectives as incomprehensible or threatening, another is that they so assimilate other perspectives to their own as to miss what is different about them. In addition, deliberators must not only hear other perspectives but also communicate that they have done so. This is important both because understanding is improved by checking one’s perceptions against those of the person whom one is trying to understand 702—–AMERICAN SOCIOLOGICAL REVIEW Delivered by Ingenta to : Columbia University Fri, 12 Jan 2007 18:21:39 (Goodin 2000) and because deliberative legitimacy depends on participants’ belief that their preferences have been recognized as legitimate even if they are not ultimately acted upon (Cohen and Sabel 1997). Proponents have argued that personal storytelling conveys the particularities of people’s experiences in a way that reasons do not (Henderson 1987; Murphy 1993; Young 2000). Why, though, would anyone want to submerge themselves in the details of another person’s experience? Because when that experience is structured as a story rather than as an account or description, listeners anticipate that it will make a larger point, one that is relevant to their own lives. To understand the story is to grasp its moral implications (Bruner 1991; Labov and Waletsky 1967; Linde 1993; Ochs and Capps 2002; Polanyi 1979; White 1987). This is different from saying that people can combine their stories with an appeal to shared values. Rather, the values are built into the story itself. It is also different from saying that what is important about personal stories is their personal, particular quality. Rather, stories integrate the particular and the general. They do this by way of plot. Plot is the structure of the story; it orders events such that those peripheral to the central causal thread are excluded. Plots are conventional in the sense that they are drawn from a common, cultural stock. An account of a personal experience makes sense against the backdrop of similar stories (Bruner 1991; Polletta 1998; Polkinghorne 1988; Johnstone 1990). A story is recognizable as a David and Goliath story about the little guy triumphing over the big one or a Pride Before a Fall story about the little guy biting off more than he can chew. The plot lines available at any one time, however, are numerous, and stories never hew to familiar plotlines exactly (Bruner 1991; Ewick and Silbey 2003). This means that people can convey genuinely new perspectives in the reassuring form of generic plot lines. In a deliberative context, personal stories can help listeners comprehend experiences both as unlike and analogous to more familiar experiences. Turning unfamiliar perspectives into compelling personal stories obviously takes skill, but narrative’s very form prepares audiences to listen empathetically. People understand an argument by evaluating the consistency between claim, justificatory principle, and evidence. By contrast, they understand a story by tacking back and forth between the events described, the internal emotional states of the protagonists experiencing those events, and the larger whole to which the story adds up (Baumeister and Newman 1994; Bruner 1991). They reject a story not because it lacks evidence but because the protagonist’s experiences or insights are not “believable” given what they know about the protagonist and her world (Johnstone 1990; Linde 1993). Audiences are prepared from the very beginning of a story to suspend disbelief. Even in ordinary conversation, speakers rely on a variety of linguistic devices to effect a transition to the separate time and place of the story (the equivalent of “Once upon a time .|.|.”): for example, indications that a story is about to be told; an orientation to the time and place of the story; or a shift in verb tense (Jefferson 1978; Labov 1972; Linde 1993; Polanyi 1985). These devices, which detach the story from the ongoing conversation, encourage listeners to suspend their skepticism about the credibility and relevance of the story and strive to grasp the motivations of the characters and the unfolding logic of events. In other words, when audiences enter the story-world created by the narrator, they know from the beginning that they are making a projective leap. In a deliberative context, this should prime them to empathize with the narrator without misrecognizing his experience as their own. Finally, by retelling other people’s stories or by explicitly imagining the experiences associated with a particular position, deliberators can demonstrate their understanding of an opinion with which they do not agree, or at least show their appreciation of the experiences that gave rise to it. In sum, narrative’s conventionally normative and discursively detached character equip it well to communicate, rather than merely express, people’s distinctive preferences. A second challenge in deliberation is that the connections among preferences, principles, and practical solutions are rarely obvious, even to those who profess them. The assumption that preferences are alterable stands behind all arguments for deliberation: if preferences were fixed, then bargaining or voting would be better mechanisms for sorting among them (Cohen 1989; Elster 1998). Preferences, however, are often inchoate as well as open to change (Button and Mattson 1999; Conover, Searing, and Crewe PUBLIC DELIBERATION AFTER 9/11—–703 Delivered by Ingenta to : Columbia University Fri, 12 Jan 2007 18:21:39 2002:53). People’s experiences may have given them a feeling about an issue rather than a clear opinion; they may have an opinion but see no relevant principle with which to justify it; they may have a principled opinion but not see any policy option that matches it. All of these are especially likely to be the case for deliberators whose experiences or opinions put them in the minority. Telling personal stories in this situation is a way to get other people’s help in formulating one’s opinions, crafting justifications for those opinions, and defining the options that are available. Again, people listen to a personal story because they trust that it will make a normative point; it will have relevance to their own lives. But if all stories have a moral, the moral is rarely announced explicitly. Rather, audiences accept that they will have to interpret the story to extract its meaning (Bruner 1991; Iser 1972), and, indeed, that the story’s meaning may not even be obvious to the person telling the story. Here, we draw attention to narrative’s allusive character. Of course, all discursive forms require interpretation; but audiences expect good stories to be interpretable more than they do good reasons or good reports. Conversational analysts have found that when people tell their stories, their listeners often participate in interpreting and even telling the story (Goodwin 1986; Ochs and Capps 2002). The point of the story may be offered by the narrator, then modified or amplified by her interlocutors. Or the narrator’s interlocutors may supply the point of an account that the storyteller presented as ambiguous (Robinson 1981:69; Maynard 1988:451; Manzo 1993; Ochs and Capps 2002). People may tell stories in deliberation with just this possibility in mind. By interpreting the narrator’s story, or by telling one of their own, deliberators can help illuminate a larger point. If solving problems requires identifying what the problem is and what solutions are available as well as matching one to the other, personal storytelling may be a valuable resource. Finally, and this is the third feature of deliberation that we want to underscore, deliberators, at least some of the time, must be able to persuade each other of the merits of their views. This is notoriously difficult to do. Research suggests that while people like being exposed to other points of view, they are keenly sensitive to efforts at persuasion. They worry that they will be cowed into changing their opinion or that the discussion will become awkwardly argumentative (Button and Mattson 1999; Conover, Searing, and Crewe 2002; Eliasoph 1998; Mansbridge 1983). They may be especially worried when participating in an online forum, given such forums’ reputation for hostile discussions (Sunstein 2001; Wilhelm 2000). The question, then, is whether people can reason together without antagonizing each other. Telling personal stories may be helpful, not because they offer a clear moral but because they do not do so. We draw attention once again to the allusive character of storytelling but also to its iterative character. Reasoned argument invites assent or dispute. Telling a story implies an invitation to tell a story in return (Ochs and Capps 2002; Sacks 1992). “That reminds me of something that happened to me .|.|.,” speakers often say. The longer conversational turn that stories require comes with an obligation to cede the floor once the story is done. The second story is expected to relate to the first one but may do so in any number of ways (Arminem 2004; Ochs and Capps 2002; Sacks 1992; Polanyi 1979). For example, the second story may offer an alternative perspective on the events related in the first story or may recast the first story’s point. In a deliberative context, a speaker may offer his personal experience in the vein of an earlier speaker’s story but use his story to advance quite a different opinion. Again, that the point of the story is usually implied rather than stated explicitly means that it is more difficult to reject as being wrong or irrelevant. Storytelling’s allusive and iterative qualities allow deliberators to advance even competing opinions in a way that is not perceived as combative. If the norms of storytelling in everyday conversation point to neglected benefits of telling personal stories in deliberation, they also point to at least one important risk. Some people’s stories may be more credible than others, not only because of the skill of the storyteller but also because concerns about the credibility of the genre itself may attach more to some people than others and may arise on some occasions and not others. Storytelling’s value in deliberation is shaped by popular assumptions about how stories work: assumptions about how audiences respond emotionally to stories, who is equipped to tell credible stories, how stories convey or cir704—–AMERICAN SOCIOLOGICAL REVIEW Delivered by Ingenta to : Columbia University Fri, 12 Jan 2007 18:21:39 cumvent the truth, and what stories are good for. We call them “popular assumptions” to distinguish them from the institution-specific storytelling norms that scholars have investigated, for example, in courtrooms (Conley and O’Barr 1990), plea-bargaining sessions (Maynard 1988), and hospital intake interviews (Clark and Mishler 1992). The assumptions that we are referring to cut across institutions. For example, the popular expectation that true stories remain the same in their retelling has hurt women who have made rape charges in court. Juries have fixed on the fact that elements of the victim’s account changed from the time of her first agonized call to police and, as a result, have questioned her credibility, despite the fact that this is not an atypical occurrence for victims of sexual violence (Scheppele 1992). Assumptions about how stories work and how credible they are relative to other rhetorical forms vary historically and culturally. They are also contingent. Different and, indeed, contrary assumptions may be triggered depending on the speaker and the setting. Sociolinguists have drawn attention to the first kind of contingency, noting that although personal narrative is generally denigrated relative to forms of discourse that are considered more scholarly or scientific, higher-status storytellers still have authority. As Hymes puts it, “[O]nly the anecdotes of some would count” (1996:113–14; see also Blommaert 2001; Briggs 1997; Ewick and Silbey 1995). We emphasize the second kind of contingency. The anecdotes of all may count, but only in relation to some topics or on some occasions. This reflects the fact that popular views of narrative are ambivalent rather than dismissive. People tend to see narrative as authentic if also deceptive, as normatively potent if also politically unserious. We suspect that these views reflect the symbolic codes of analogy and difference that produce cultural meaning more generally (Lévi-Strauss 1963; Alexander and Smith 1993). Just as people know what reason is through its relation to other binary oppositions—reason is to unreason what man is to woman, cognition is to emotion, and culture is to nature—storytelling makes sense when ranged along culturally familiar oppositions. Insofar as storytelling is understood in terms of oppositions of concrete/abstract, emotional/ rational, female/male, personal/public, informal/formal, and folkloric/scientific, it is denigrated by its association with the negative pole of each opposition. “Scientific,” however, is also commonly contrasted with “commonsensical,” and narrative is favored by its association with the latter. Similarly, narrative’s association with morality as against strategy and emotion as against rationality may make the form seem more trustworthy than, say, logical arguments, at least in settings that are perceived as private or informal. What this means for deliberation is that when disadvantaged groups use narrative to challenge the status quo, they may be especially vulnerable to skepticism about the veracity, authority, or generalizability of the form. When advantaged groups use narrative, they may be less likely even to be heard as telling stories. Alternatively, the use of narrative on some occasions may raise doubts about the form’s value. If personal stories are commonly seen as appropriate during discussions that are personal, casual, and social, they may raise such doubts during discussions that are public, policy-oriented, or technical. This perception may disadvantage already disadvantaged groups since it is in such discussions that supposedly neutral standards, universal principles, and technical criteria tend to operate with special force to marginalize