Listening to the City : Difference , Identity , and Storytelling in Online Deliberative Groups

This study examines how members of deliberative groups tell and respond to personal stories during their disagreements. The study provides an in-depth qualitative analysis of some online discussions hosted by AmericaSpeaks in which New Yorkers discussed how to redevelop the site of the former World Trade Center after its destruction on September 11, 2001. Participants’ stories fell into four distinct types: adversarial argument, unitary argument, introduction, and transformation. These stories helped group members build arguments, maintain or challenge their group’s collective identity, display community values, and manage their conflicts. The findings of this study indicate that deliberative theory ought to more explicitly recognize storytelling as a legitimate form of deliberative discourse. Moreover, the study of storytelling during deliberative discussion can hold implications for group facilitators and others who wish to further deliberative ideals.

As a democratic form of discourse, deliberation is premised on the ethical stance that citizens' differing experiences and perspectives should be taken seriously.Indeed, deliberation theorists argue that discussion across difference is crucial to achieve the purpose of deliberation (e.g., Benhabib, 1996;Cohen, 1996;Delli Carpini, Cook, & Jacobs, 2004).Yet, scholars struggle to articulate how deliberative groups should productively manage their differences (Bohman, 1995;Gutmann & Thomspon, 1995;Pearce & Littlejohn, 1997).On a practical level, difference is essential for good problem analysis (Hirokawa & Salazar, 1999), and groups that insulate themselves from disagreement and diversity are likely to make bad decisions (Janis, 1972).
However, research shows that Americans often avoid talking politics with those who disagree with them or come from different backgrounds (Mutz, 2002;Mutz & Martin, 1991;Walsh, 2004).Even in deliberative forums, where group members are put into diverse groups and given interactive norms that could help them to manage their differences, they tend to avoid overt disagreement (Ryfe, 2006).Some recent work posits that one way deliberative group members can manage their differences is to tell stories (Black, 2008a;Polletta & Lee, 2006;Ryfe, 2006).This statement is perhaps counter intuitive because storytelling is not clearly in line with the rational decisionmaking model that pervades much of deliberative theory (c.f.Cohen, 1996;Gastil & Levine, 2005).Yet, the claim makes sense if one understands disagreement as not simply being about argument content but also about group members' identities.That is, when group members disagree about a public issue that is important to them, they not only provide reasons for their positions and engage in argumentation, they also articulate aspects of who they are, what groups they belong to, and what the issue means for their definition of themselves.
Identity is a critical resource that group members draw on and construct through their communication.Johnson and Long (2002) argue that group members experience an ongoing tension between identifying themselves as separate from other members and feeling a sense of attachment to the group.This tension is especially strained when groups are faced with minor disagreement or more deep-seated conflicts based in moral and cultural differences (Black, 2008a).Deliberative scholars have largely overlooked issues of identity, 1 which is unfortunate because the communication processes that group members use to express their identifications, categorize others, and respond to other members' identities are important for shedding light on how diverse groups manage members' differences.
In this essay, I argue that storytelling offers a way for group members to deal with their differences by helping them to manage their identities vis-à-vis one another and their group membership.Storytelling helps people to construct their identity (Bruner, 1990;Gergen & Gergen, 1983;Kerby, 1991;Linde, 1993;Ochs & Capps, 1996;Schifrin, 1990;Tracy, 2002) and manifest the values of a particular group (Meyer, 1995;Philipsen, 1997).In addition, the interactive processes of storytelling can help group members to corroborate on a shared identity (Bormann, 1986;Kellas, 2005), as well as contest each others' identity claims (Hseih, 2004).
For these reasons, storytelling is a good communicative practice to study if we wish to understand how deliberative group members manage their identities.This study examines how group members tell and respond to stories when they are involved in disagreement.Through this examination, the study adds to a growing body of work that aims to build understanding of the role of personal stories during deliberative discussion.

Is Storytelling Deliberative Discourse?
Early deliberative theorists, such as Jurgen Habermas and Joshua Cohen, emphasized that deliberation should be based on rationality, seek consensus, and aim for a decision that serves the 1 for exceptions, see Gastil, Black, Deess, & Leighter, 2008;Hart & Jarvis, 1999;Muhlberger & Weber, 2006 common good (for a review see Mansbridge et al., 2006).More contemporary deliberative theorists have challenged and expanded early theory to acknowledge the importance of plurality and moral conflicts (e.g., Bohman, 1995;Burkhalter et al. 2002;Chambers, 2003;Gutmann & Thomspon, 1995;Gastil & Black, 2008).Yet, the foundation provided by these early theorists remains very influential in deliberative theory.Accordingly, the types of communication that are often considered to be deliberative emphasize problem-solving aspects such as reason giving and analysis of different alternatives.In this view, personal stories would likely be inappropriate because they focus on personal, rather than public, concerns and they may present emotional appeals that are inconsistent with the emphasis on reasoned discussion (see Black, 2008a for a further description of this argument).
I argue that treating stories as somehow belonging outside of deliberation is misguided and relies on a limited, and inaccurate, understanding of storytelling.Several feminist theorists (Fraser, 1992;Sanders, 1997;Young, 1996) have argued that personal storytelling or "testimony" (Sanders, 1997) should be recognized as legitimate in deliberative theory because stories can allow group members to voice their own perspectives and values relevant to the issues being discussed.Stories are highly valued in groups attempting to promote dialogue (e.g., Chasin et al., 1996) because they can help group members articulate their experiences and perspectives and add to the depth of a group's understanding of an issue.Some contemporary deliberative theorists see value in stories as a way to provide information that can be important in deliberative discussion (Gastil & Black, 2008).Moreover, some recent research by deliberative scholars (Black, 2008a;Polletta & Lee, 2006;Ryfe, 2006) emphasizes that storytelling is quite common during deliberative group discussion, and has relevance for understanding how group members manage their disagreements.The following section highlights some key findings from these recent studies of storytelling in deliberative events and provides the research questions for the current project.

Current Research in Storytelling, Deliberation, and Disagreement
Narrative theorists define storytelling in a variety of ways (see Langillier, 1989), but in general a personal story can be understood as a series of connected statements in which a speaker recalls some past experience in a roughly sequential order (e.g.Labov, 1972).Stories typically center on some kind of problem, and the events retold in the story describe how this problem was resolved.Stories not only include a description of past events, but they also involve statements that help the storyteller evaluate the events in the story (i.e., the moral).In the context of a group interaction, storytelling should be understood as an interactive process (Black, 2008a;Hseih, 2004;Ryfe, 2006), so studying personal storytelling in deliberative groups includes a recognition that how stories are responded to is important.
Three current studies highlight the importance of stories in helping group members express and manage their disagreements during deliberative discussions.Polletta and Lee's (2006) comparison of reason-giving and storytelling finds that stories can advance deliberative discussion by helping members identify their preferences, put forth unfamiliar or unpopular views, and demonstrate appreciation of other people's perspectives.Moreover, group members seem to find areas of unanticipated agreement by telling and responding to one another's stories.
In this way, group members use storytelling as one way to manage their disagreements.Black (2008a) and Ryfe (2006) also find that stories are important resources that group members draw on as they engage in disagreement.However, unlike Polletta and Lee (2006), these authors draw attention to the relationship between stories and identity.Ryfe (2006) finds that stories help group members to overcome barriers to deliberation, such as a lack of knowledge and the threat that deliberation poses to members' "public face" (p.79).Ryfe also argues that group members use stories to manage conflict by avoiding overt disagreement and simultaneously building a sense of moral community around an issue.By telling and responding to personal stories, individual members can meaningfully participate in, and identify with, a community.Black (2008a), drawing explicit attention to identity issues, argues that storytelling is important because it invites moments of dialogic contact in the midst of disagreement.By telling, hearing, and responding to stories, group members negotiate their identities and come to understand one another's perspective on areas of disagreement.This negotiation is possible because stories allow people to express their worldviews and values and help listeners to take the perspective of others with whom they may disagree.
These three recent works paint a complex and somewhat contradictory picture of the possible functions that stories play in deliberative groups.Taken together, they argue that stories can advance arguments, promote compromise, build community, invite dialogue, manage conflict, and help group members to identify and clarify their own perspectives.But how?
Although none of these authors argue that stories are monolithic, current research has not investigated variations in stories or discerned whether or how stories might fulfill these differing functions.

Research Questions
This study examines how deliberative group members tell and respond to personal stories when they are engaged in disagreement.Because there is little empirical work on this topic, it is appropriate to ask descriptive questions to gain a better understanding of the communication dynamics of storytelling during deliberative disagreement.Thus, this study addresses two research questions.
The first question attempts to understand the characteristics of stories told in such circumstances and examine the extent to which they elicit distinguishable response patterns by asking, "What types of stories do deliberative group members tell when they are engaged in disagreement?"The second research question examines the role that identity plays in disagreement by investigating group members' expressions of identity in stories and responses.
It asks, "How are individual and collective identities displayed in the stories and responses to them?" To address these research questions, I engaged in an in-depth analysis of the stories and responses that group members provided during their disagreements.The data for this project are drawn from two of the discussion groups from an online deliberative forum hosted by the organization AmericaSpeaks.The following section describes the forum and the online discussion format and describes the methods used to analyze the qualitative data.

Case Study Description
This is a case study of "Listening to The City" (LTC), a deliberative forum in which New Yorkers engaged in 2 weeks of online discussions about what should be done with the site of the former World Trade Center after the events of September 11, 2001.LTC was designed to involve "ordinary" New Yorkers in discussion about how the space previously occupied by the World Trade Centers should be developed.These online discussions were preceded by a face-to-face public meeting, hosted by the organization AmericaSpeaks, which was attended by over 4000 people (for a description of this event, see Lukensmeyer, Goldman, & Brigham, 2005).
Eight hundred and eighteen people, divided into 26 discussion groups, participated in the online discussion during the summer of 2002.All groups followed a general agenda that included topics such as "Hopes and Concerns" and "Elements of Rebuilding," such as memorial, design, and economic concerns.Participants could also start discussion threads on topics of their choosing.Topics were followed by opinion polls and the results of those polls were posted on a common web site and included in the final Listening to The City Report that was sent to city officials and the designers planning the physical site (Civic Alliance, 2002).
The LTC online discussions present what Yin (2003) calls a critical case-a situation in which the phenomenon under study is most likely to occur-and, therefore, serves as a site for examining the dynamics of storytelling in deliberation.To qualify as a critical case, the LTC online discussions must be a clear case of small group deliberation and have important features that make it extremely likely that storytelling will occur.
The LTC online discussions share numerous features with ideal models of public deliberation, including small groups of citizens discussing political/public issues that are of great importance in their community.These participants are speaking for a larger group (i.e., "the City"), with the results of their discussions having some influence on decision making about the issues.Moreover, participants were presented with and generally followed discussion guidelines that are similar to models of democratic deliberation (e.g.Gastil & Black, 2008).
The LTC online discussions also have some features that make it extremely likely that storytelling would occur, and a previous study of these data found numerous stories (Polletta, 2006).First, tragic, emotional, momentous events, such as those of September 11, lend themselves to storytelling.People who experience such an event will have a story about it, and those who lost loved ones will have stories about those people's lives.LTC participants were self-selected, and a great number of them were directly affected by the attacks.Several studies of online interactions focused on the topic of the September 11 terrorist attacks (c.f., Cohen & Willis, 2004;Dutta-Bergman, 2004;Dutta-Bergman, 2006;Frank, 2004;Kim, Jung, Cohen, & Ball-Rokeach, 2004;Robinson, 2005) point to the importance of the tragedy in people's lives and show that many people used new media to communicate with others about their experiences, hopes, and fears about what would happen in the aftermath of that event.
In addition, the asynchronous discussion posting format used in the LTC online discussions can facilitate storytelling.One of the preliminary steps in telling conversational stories is gaining the floor (see Sacks, 1995), which can be difficult to do during a face-to-face group discussion.In an asynchronous discussion, in contrast, each person can take the time to write out his or her discussion contributions as fully as desired.This feature makes it likely that people will be able to tell their stories more fully than they would in either an unfacilitated faceto-face conversation or one in which a facilitator may influence speaking turns (see Ryfe, 2006).

Researching an Online Environment
Analyzing the LTC online discussion data presents exciting opportunities and challenges for the discourse analyst.New media forms are growing rapidly, having a large impact on contemporary social life, and being used in increasing numbers in the United States and around the world (Pew Internet & American Life Project, n.d.).Much research on the democratic potential of computer-mediated communication (CMC) asserts that online communication can promote civility and democratic discussion (e.g., Papacharissi, 2004;Weiksner, 2005;Wilkund, 2005), revive the public sphere (Papacharissi, 2002), and encourage civic participation and engagement (Bucy & Gregson, 2001).As CMC becomes an everyday way to engage in public life, deliberation scholars need to investigate online forums.
Another reason for analyzing the discourse of online groups is perhaps less obvious: For an analyst, the online environment offers firsthand access to communication in the same medium that the group members experience it.Because members of online discussion groups view communication from a computer screen, the analyst's access to and experience of that communication is quite similar to that of group members.There is no need to transcribe the conversation (Hine, 2000) and every communication contribution that the group members share in their interactions is readily available to the researcher.However, differences between online and face-to-face communication (Joinson, 2005) present some methodological challenges.For example, emotion is more difficult to assess in a text-based online interaction than in face-to-face communication because of the lack of visible nonverbal cues.Group members may use emoticons (Rivera, Cook, & Bauhs, 1996) or make explicit statements about their intended emotions, but these are not always present.
Finally, structural differences between online and face-to-face interactions pose analytic challenges.Asynchronous discussion is more topically and temporally disjointed than face-toface discussion because group members are not all reading and responding to messages at the same time.It is common for replies to occur much later in the conversation than would be expected in face-to-face settings.Because of these differences, some of the discourse-analytic approaches drawn on here have necessarily been adapted operationally to fit the online environment.

Procedures Study Design
The data included in this paper come from two of the LTC discussion groups.The analytic method used for this study is informed by the tradition of narrative discourse analysis (e.g., Hseih, 2004;Labov, 1972;Maynard, 1988;Polanyi, 1985;Schiffrin, 1990), which provides me with methods for looking closely at the discussions and drawing inferences about how stories function in conversation.The specific measures of discursive aspects are described in the following section.
The analytic method is also influenced by qualitative case study research (e.g., Denzin & Lincoln, 2000;Miles & Huberman, 1984;Lofland & Lofland, 1995;Philipsen, 1982;Strauss & Corbin 1990;Yin, 2003), specifically work done in the tradition of analytic ethnography (Goldthorpe, 2000;Lofland, 1995;Snow, Morrill, & Anderson, 2003).These traditions emphasize the importance of describing the participants' own understandings of their experiences and offering a rich, detailed description of the events studied.Qualitative case study research requires a somewhat open and flexible methodological approach so that the analyst can develop sensitivity to the participants' meanings over the course of the study.Although study design is guided by research questions, analytic rigor in qualitative case study research comes from following procedures that allows the analyst to develop a deep understanding of the participants' meanings and experiences in their own terms.Drawing on these traditions, I entered this analysis with a heuristic framework in mind, based on the research questions of this study, but I expected that the framework would develop and change during the early stages of analysis.
To achieve these ends, I approached the first discussion group in an exploratory manner (Lofland & Lofland, 1995) to discern what aspects of stories and responses might be relevant to the research questions.The second group served as a confirmatory case to test and refine the analytic framework developed from the first group studied, as is encouraged by the methods of analytic case study research.As Philipsen (1982) argues, qualitative case studies do not need to be limited solely to the purpose of describing a unique phenomenon.Qualitative case study research can also be designed to "test the soundness of extant claims" and used to make general "assessments of linkages and for detecting structural relationships" (1982, p. 11).By replicating (Yin, 2003) aspects of the first qualitative study, I tested and refined the conceptual framework developed by the inductive approach. 2s a qualitative, inductive study the goal of this research is to uncover patterns in the storytelling processes of deliberative discussion.A close look at the storytelling interactions in two of the discussion groups can provide a rich description of storytelling that leads to an analytic framework that can inform further research.In a separate study (Black, 2008b), I use the findings of this research to create a content analytic coding scheme to analyze the interactions in the rest of the LTC online discussion groups and test hypotheses about the relationship of storytelling to conflict management outcomes.

Analytic Process
Identifying materials to be analyzed.The first analytic step was to identify periods in the discourse where difference was expressed, and locate stories that occurred during these periods of difference.Rather than examining all stories told by the participants of this group or all instances of the expression of difference, I only analyzed those periods of difference in which at least one story was told.
I define the expression of difference as an interaction that makes salient any dissimilarities between interlocutors.In these groups, difference was most commonly expressed as disagreement; that is, explicit statements that presented argumentative positions that contradicted another speaker.For analytic purposes, periods of difference began with the first discussion post involved in a disagreement-typically, the post immediately prior to the first expression of disagreement.Periods of difference ended when there was a sustained change in topic or the discussion thread ended.
I then scanned the periods of difference to locate stories, following loose guidelines similar to those used in other studies of narrative and deliberation (e.g., Black, 2008a;Polletta & Lee, 2006;Ryfe 2006).Stories are viewed as statements made by group members that refer to a sequence of events or personal experiences outside of the current situation and include standard narrative elements, such as characters, a plot, and at least an implicit moral.Stories must be longer than a single utterance and, for the purposes of this study, need to be introduced into and brought back or related to the current discussion.
Analyzing story structure.The second analytic step was to analyze the structural components of the stories identified.The structural features of narrative are drawn from the work of Labov (1972;Labov & Waletzky, 1967): abstract, orientation, complicating action, resolution, evaluation, and coda.The complicating action and resolutions are what Labov (1972) calls "narrative clauses" because they demonstrate the events of a story.The remaining structural features do not tell about the story events but serve other narrative functions.The abstract and coda mark the beginning and end of the story.The orientation sets the context for the story and evaluations comment on some quality of an event or character in the story; such evaluations can be explicitly stated morals that occur at the end of the story and provide the moral message of the story, but can also be more subtle aspects of the discourse that occur throughout the story (Polanyi, 1985).Evaluative devices are important in stories because they give meaning to the story events and bridge the storyworld with the larger topic of discussion.
Analyzing story content.In each story, I examined the storyteller's expressions of identity.Identity was assessed through participants' use of collective tokens (Hart & Jarvis, 1999) and identity-related self labels (Larkey, Hecht, & Martin, 1993;Martin, Krizek, Nakayma, & Bradford, 1996).Collective tokens are terms such as "we," "us," and "our" that refer to some collective group in which the speaker is claiming membership; identity-related self-labels are terms that identify the speaker with a particular social group, such as "American" or "New Yorker."I paid close attention to how the storyteller referred to him or herself, and the extent to which the collective tokens and self-labels included other group members.
I also examined whether the storyteller was promoting an argumentative position in his or her discussion post containing a story.Because all of the stories included in these analyses occurred during some type of disagreement, I presumed that at least some of the storytellers would articulate their position on the topic under dispute.To count as having an argumentative position, the storyteller's post needed to include a clear statement of his or her position on the issue being disputed.
Examining responses to stories.The next analytic step was to examine other participants' responses to each story.To count as a response, a discussion post needed to explicitly comment on the story or some other aspect of the storyteller's post that contained a story, or mention the storyteller by name.I then noted some content-related features of the response; in particular, whether the respondent expressed a position on the topic of difference and, if so, the extent to which the position agreed with the position expressed by the storyteller.I also looked at the respondent's expressions of identity.
Addressing research questions.To address the first research question, I began by grouping the stories from the first discussion group together for examination, which allowed me to compare their structural and content-related features, and begin to sort the stories into initial thematic categories (e.g., Lofland & Lofland, 1995).This initial categorization of stories was then used to look at the interactive patterns surrounding each story type in their original context.
I examined things such as the number and characteristics of responses elicited by a story to discern interaction patterns that could assess how the different story types functioned discursively.Interpreting these interactive patterns further developed my initial categorization of stories and to finalize the typology.
In studying the second discussion group, I first grouped stories according to the typology developed from the analysis of the data from the first discussion group.Many stories did not clearly fit into the initial framework, and these were grouped together for further analysis.This constant comparison analytic procedure (Strauss & Corbin, 1990) allowed refinement of the initial framework, as explained below.
To address the second research question, I examined the identity expressions evident in stories and responses.This analysis involved looking at each story and its responses to note how many and what type of collective identity expressions were used.In particular, I paid close attention to the extent to which the identification targets in the stories and responses were inclusive of other group members involved in the disagreement.For example, targets such as "our LTC group" or "American" include all the other discussion group members, whereas "my neighborhood" or "victim's family members" exclude other group members.

Results and Discussion
The first discussion group had 6 periods of difference containing a total of 30 stories; the second group had 9 periods of difference containing 33 stories.Analysis uncovered four types of stories told during the periods of difference in these two discussion groups.Table 1 presents an overall description of the number of stories told and how many responses stories of that type received, which may be useful as general background information for the analysis that follows.The following section describes the four story types, organized by the discursive functions they seem to play in deliberative discussion.

Argument Stories
In both discussion groups, the majority of the stories were presented in an argumentative way.My initial categorization of argument stories, based on the first discussion group, was to distinguish stories that were presented as adversarial, persuasive, and emotionally charged from those that were presented in a more respectful, moderate, and collegial way.However, that distinction did not hold true in the second discussion group.On my first pass through the stories in the second group, I tentatively categorized 7 as adversarial, 7 as collegial, and 10 as argument/other, a category created for stories that functioned argumentatively but did not clearly meet the criteria of either argumentative category created through analysis of the data from the first discussion group.
Many of these "other" argument stories were presented adversarially in the sense that they presumed a competition between two or more sides of an issue and were ostensibly being told to persuade others to be more favorable toward the storyteller's position on the topic of difference.However, those stories were framed in ways that seemed friendly, civil, and oriented toward fair and balanced analysis of ideas.Still, they differed from argument stories that emphasized the potential for consensus and the collective nature of the group.
To make sense of this finding and refine the story type framework, I turned to Jane Mansbridge's (1983) influential book Beyond Adversarial Democracy.Mansbridge argues that two different conceptions are at play in "every modern democracy" (p.4).The first ideal is what she calls adversarial democracy, which is characterized by the underlying assumption that individuals have conflicting interests in the issue at hand.Group members operating with an adversarial ideal of democracy will assume that other members' interests conflict with their own, and, therefore, believe that they needs to frame their statements persuasively enough to garner a majority on the topic.
The second ideal is what Mansbridge (1983) calls unitary democracy, which, rather than being based in a sense of conflicting interests, assumes some level of commonality among group members.Unitary democracy emphasizes equal respect, communication among all group members, and the potential for consensus, features that "encourage members to identify with one another" and develops their sense of "common interests" (Mansbridge, 1983, p. 5).

Mansbridge's (1983) concept of unitary democracy captures much of what I noticed
about non-adversarial argument stories.My previous conceptualization of these stories as collegial emphasized that they were told in a friendly emotional tone.However, conceptualizing them as unitary highlights that they are told with an underlying assumption of consensus and common, rather than competing, interests.For these reasons, I re-categorized the argument stories as adversarial and unitary.This refinement in story conceptualization allowed me to parse out the stories categorized as "Argument/Other" and separate them into two distinct groups.

Adversarial Argument Stories
Example and analysis.The following example is a post from a person who chose the screen name "LocalResident."The issue being disputed is whether a major road that separates the former World Trade Center site from one of the neighboring communities (Battery Park City) ought to be put underground to improve traffic flow through the area and make the neighborhood more easily accessible.LocalResident is very active in this conversation, contributing many long posts, and is vehemently opposed to the proposal.His3 post contains two adversarial argument stories.Because some of his argumentative statements that are relevant to the story type that occur in other parts of the post, I have included the whole post here rather than simply displaying the stories out of context.Stories are italicized for easier identification.variety of alternatives available that will achieve the objective that we all have of a rebuilt World Trade Center that will revitalize downtown and be a suitable memorial to those who lost their lives on 9/11.Some of those alternatives will help Battery Park City, some will harm it, and some may do neither.We are simply trying to guide the process away from what seems to be one of the alternatives with the most potential for harm-the burying of West Street (with years of likely disruption due to construction and the other concerns expressed above).
One case in point.Shortly after 9/11, it was proposed that the damaged sculpture from the WTC plaza be placed on public display.So you can understand why, when you say that the underground West Street tunnel will be of benefit for automobile commuters from New Jersey and elsewhere trying to get downtown, I am not very sympathetic when the cost may be the well being of my neighborhood.If you are right, the inevitable result will be to attract more automobile traffic to the area ("If you build it, they will come"), with the attendant noise and pollution.
The content of this post is typical of those containing adversarial argument stories.It begins by responding directly to a particular member of the discussion group, someone with whom the storyteller disagrees.Like other posts containing adversarial argument stories, this one refutes the arguments made by the person with whom the storyteller disagrees and has a very clear statement of the storyteller's position on the topic of difference: that burying West Street will likely "change the character of our neighborhood for the worse" and is one of the "alternatives with the most potential for harm."The content of the stories is also typical of this story type in that LocalResident describes two of his own personal experiences with similar situations to demonstrate his knowledge of and authority on the topic in support of his argument.
The stories themselves were structurally complete.In both stories, LocalResident begins with a brief orientation ("Shortly after 9/11" and "I became sensitized to these issues when I lived in New Orleans") and then provides the complicating action.For both stories, the complications have to do with outsiders proposing some change to the neighborhood about which area residents disapproved.These events are resolved ("following local outcry" and "In both cases neighborhood concerns prevailed") and evaluated positively.The moral that is clear through the evaluation statements present in both of these stories is that the concerns of local residents are not selfish but important and relevant to the public good.Thus, these concerns should be given a great deal of consideration in making decisions to avoid disastrous outcomes.
Use of identity statements.LocalResident's use of collective identities is complex.Before telling his stories, he makes a plea to common goals when he says, "There are a variety of alternatives available that will achieve the objective that we all have."However, his subsequent use of collective tokens, screen name, and statements about his neighborhood clearly demonstrate that the collective group he most strongly identifies with is his neighborhood.
Although he assumes that all members of the discussion group have some shared goals, he is explicit in his argument that there are competing interests and that the interests he supports are those of the residents of Battery Park City.
The majority of his collective identity targets are exclusive because the collective tokens refer to the residents of the neighborhood, which needs to be fought for against "you" (other members of the discussion group who disagree with LocalResident), who advocate putting West Street underground.Although he clearly uses exclusive collective identity references to make an argument and build support for his position, he indicates that he shares some concerns with people who disagree with his position.In his story, he does not argue that the person who suggested putting the statue in his neighborhood was bad.Instead, he says he is "sure" that person "had the best of intentions, but never considered the impact on residents from the probable influx of tourists (not to mention the loss of a lovely little park)."Hence, although the use of collective identity is exclusive, the people with whom LocalResident disagrees are not seen as enemies; they are reasonable people who have good intentions, but need to be persuaded to consider the impact of their decisions on the residents of the local area.
Responses.In the first group studied, adversarial stories were always responded to and they were typically responded to with disagreement, a lack of perspective-taking, similar negative emotions, and other adversarial stories.In the second group, the response pattern was more complex, as 7 of the 16 adversarial argument stories received no direct responses.Six of the remaining stories were responded to with disagreement, did not demonstrate perspectivetaking, and included exclusive collective identity statements.On three occasions, an adversarial story was responded to in a unitary way, with respondents demonstrating perspective taking, using a respectful emotional tone, emphasizing an inclusive sense of collective identity, and including their own stories.

Unitary Argument Stories
Like adversarial argument stories, unitary argument stories were also used to present support for an argument, but they emphasized the collective nature of the group and focused on building consensus rather than gathering support for one side of a debate.These stories are characterized by a prevalence of inclusive collective identity statements and a tentativeness around their position statements.
Example and analysis.The following example occurred near the end of a discussion thread about transportation issues relevant to the proposed renovations.The posts immediately preceding this story included negative statements about "tourists" who were behaving "disrespectfully" when visiting the site.In her post, Bunnymusic comments on both the issue of transportation and her experience with tourists.
Example 2 To wrap up, West Street seems to be a bone of contention.Since I'm completely unfamiliar with its geography, I can't offer any personal insight.I do see that we all agree public transportation should be intelligently improved so that businesses and workers would be encouraged to travel down there.I don't think any of us want to encourage more vehicular traffic daily in this over-congested area.As for tourists, I have not been to the WTC area recently.When I did visit I was impressed by the quiet, dignified behavior displayed by New Yorkers and out-of-towners; all seemed to be silently honoring those lost.If it has changed so dramatically I am truly saddened and dismayed.I believe we want to make it relatively easy for people to use public transportation to pay their respects at the Memorial, not to come there as though it's a side-show.
Bunnymusic begins by indicating that although West Street is a "bone of contention," she sees agreement within the group about the larger issues related to transportation.Her position, indicated in the final sentence of her post, is in line with those points of agreement.These statements about agreement within the group and her particular position display her orientation toward group consensus.Despite differences of opinion about the particular transportation proposals, Bunnymusic sees consensus in the group on the key underlying issues of the controversy.
She also indicates that "personal experience" is important in making a decision.She opts not to comment on the proposal for West Street since she is "completely unfamiliar with its geography" and "can't offer any personal insight."However, she does have personal experience with tourists visiting the site of the former World Trade Center and offers a comment on that topic.The topic of her story is her experience, and the emotions that she describes are fairly positive.She describes the visitors to the site as "quiet" and "dignified" as they were "honoring" those lost in the attacks.Her statement, "If it has changed so dramatically I am truly saddened and dismayed," demonstrates a connection with the emotions stated in previous posts (saddened and dismayed), but her use of "if" notes that the situation described in previous posts is different from her experience.Her final statement expresses her position on the topic, which is one that demonstrates agreement with both sides of the disagreement about transportation.
Like many of the unitary argument stories, this one is fairly brief.It contains the structural elements of orientation ("when I did visit") and an implied complicating action ("if it has changed so dramatically"), which is evaluated negatively ("I am truly saddened and dismayed").Her position on the topic serves as a resolution for the story: "I believe we want to make it relatively easy for people to use public transportation to pay their respects at the Memorial, not to come there as though it's a side-show."This structure is typical of the unitary argument stories found in these groups.
Use of identity statements.Throughout her post, Bunnymusic emphasizes an inclusive sense of collective identity.She uses the collective tokens "we all" and "any of us" to refer to discussion group members.In her story, she describes two groups of people who were visiting the site: "New Yorkers and out-of-towners."However, she goes on to say that "all" of the people (including both groups) were demonstrating respect.Hence, rather than using the label "tourist," which is given a negative meaning by this group, to describe people who do not reside in New York, she calls them "out-of-towners" and includes them in the collective group "all." Responses.There seemed to be two distinct patterns of unitary argument story interactions.The first consisted of unitary stories that occurred in relative isolation from all other stories.These stories occurred during periods of difference, but they were not made in response to, nor were they responded to with, other stories.Instead, these isolated unitary argument stories, like the one provided above, received no responses at all.
The second group of unitary argument stories occurred in chains.They tended to be told in response to other argument stories (both adversarial and unitary), and they were almost always responded to in unitary ways.The responses to them exhibited perspective-taking, had a respectful emotional tone, and emphasized connections within the group or at least between the respondent and the storyteller.Responses also often included stories.

Non-Argument Stories
The two remaining types of stories both occurred during disagreements, but did not contain explicit argumentative positions on the topics being discussed.For this reason, I classify them as "non-argument stories."However, despite their lack of explicit argument, the stories seemed to serve important discursive functions in the disagreements.These two distinct story types are introductions and transformation stories.

Introduction Stories
Many of the stories fell into what can be called "introduction stories," and most of these occurred during the earliest discussion thread.These stories typically reported the teller's experiences of September 11, emphasized the teller's individual identity, and often demonstrated the teller's connection to a specific place.Although it may seem odd that stories of introduction occurred during a disagreement, they were frequent in these discussions.Primarily the introduction stories occurred during a participant's first post to the discussion group.These stories did not directly engage the topic of disagreement, but perhaps were seen as necessary to provide a participant with the right to participate in the discussion.The fact that these stories occurred during a disagreement could just be an artifact of the asynchronous design of the online discussion format.In any case, they were common in the periods of difference that occurred in these two groups.
Example and analysis.The following example is one of the longer introduction stories, but it clearly demonstrates the content-related and structural features that are typical of this story type.This story occurred in the introductory thread, during a disagreement over whether the towers should be rebuilt to their original height.
Example 3 I attended Listening to the City at Javits and am here to observe the decision making process.And to see if there's some way I can help.
I try to follow international affairs a bit more these days and spend time contemplating a bumper sticker I picked up recently on a trip to Charlottesville, Virginia that states, "Terrorism is the symptom, not the disease." I'm Todd.
Todd's story chronicles his experience of the events of September 11 and subsequent days, and includes the structural features indicated by Labov (1972).For instance, his first paragraph serves as an orientation by describing where he was and what he was doing on the day of the event.The bulk of the post contains a series of complicating actions ( such as "the first tower collapsed," "the school slowly moved to a state of near pandemonium," his "son [was] holding his head and bemoaning 'those are my buildings, this is my city, how can this happen?'" and "I tried volunteering my service at the site").Although many of these complicating actions have resolutions ("all the children were picked up" and "I knew he [the son] was scarred," and "but [my help] was not needed"), the story, as a whole, lacks a clear resolution.
This structure is typical of introduction stories, which tended to be complete narratives, with the occasional exception that shorter stories sometimes did not include an explicitly stated complicating action.Because these stories were typically about a participant's experiences on September 11, the complicating actions alluded to were the attack on the Twin Towers and the subsequent mayhem in the city.The omission of complicating actions is not surprising given the gravity of the situation and the teller's understanding that everyone participating in the discussion had shared knowledge of the events.
The content of Todd's story is also typical of an introduction.He describes his experiences on the day of the event, demonstrates a connection to place by noting where he lives, and indicates that he watched the event directly from his rooftop.Despite his evaluation statements that indicate a highly emotional negative response to the actual event, Todd's post contains no explicit position on the topic of the disagreement-whether to rebuild the towers to their original height.
Use of identity statements.The introduction stories tended to contain expressions of both individual and collective identity.In Todd's story, he distinguishes himself by noting aspects of his individual identity, such as his role as a "minor public official."He uses collective tokens to refer to a range of groups: some small (such as his family or the employees of the school where his wife works) and some larger (such as the United States or New York).Although Todd does not use collective identity labels, some introduction stories did.The most commonly used labels were "New Yorker" and, to a lesser extent, labels that demonstrated their connection to a particular part of the area.This type of connection with place demonstrates some sense of social  1, all were given in response to the three stories that were told by people who had lost family members in the attack.These responses included statements such as, "I'd like to express my sorrow to those of us who have lost loved ones," which drew attention to an aspect of the teller's individual identity and also used collective tokens, such as "us," to refer to the discussion group.Thus, responses to family members highlighted both the individual and collective identity aspects of the teller.Introduction stories tended to happen in chains, but those stories typically did not refer to stories other people told.Respondents tended not to take issue with the statements made in introduction stories (as they did with argument stories); instead, the common response was to give one's introduction.

Transformation Stories
The final story type can be called "transformation stories," which included vivid emotional language and centered on some personal or social change.Transformation stories expressed conflicting emotions and described situations that were unsettled and unfinished.
Transformation stories often lacked the structural element of resolution or the resolutions stated were negative, which implied that some further resolution is necessary.These stories typically did not include an explicitly stated position on the topic of difference; the focus of the story was on how some kind of process would bring resolution or redemption to the situation.
story concluded, "To watch the city continue after the 11th is a testament to the power of lifewe just worked out new ways to do things which is why this discussion is most valuable."For these group members, participating in the deliberative process was an important way to heal from the pain caused by the events of September 11 and was crucial for their personal transformation.
Use of identity statements.The transformation stories included frequent use of collective tokens and identity labels.Although the content of the stories was about a personal experience, transformation stories typically contained strong statements praising a collective identity group, as evident in this storyteller's statements "our spirits will not be broken" and "we will move on." It was common for transformation stories to end with patriotic statements about American values or what it meant to be a New Yorker.
Responses.In the first group, people always responded to transformation stories, but the second discussion group only had four transformation stories, with two of them receiving responses.Responses were generally positive, celebrated the collective identity meanings provided by the storyteller, and used those meanings to acknowledge healing processes.For example, one response included: Someday we can all say we took part in the rebuilding-in a small but very crucial and "American" way.Whatever is built downtown, everyone of us who has participated can turn to our grandchildren and say, "I had a voice in this."Is this worth protecting with every fiber in our being?You bet it is.Are we in service to our country and to the ideals we believe in?You bet we are.
This response reinforces the meaning of "American" that had been provided in a transformation story and uses it to celebrate the deliberative process as something definitive about being American.In response to a transformation story, this group member argues that deliberation is what can enable the social transformation and personal healing.

Conclusion
When engaged in disagreement, the group members involved in this study told four distinct types of stories.The distinguishing characteristics were the story's argumentative position, the extent to which the storyteller's collective identifications include other members of the discussion group, and the stories' discursive functions.
Argument stories were, by far, the most commonly told story type in this research, and they elicited a high number of responses.Both adversarial and unitary argument stories furthered the group members' deliberation by presenting evidence to support argumentative claims.The responses to adversarial argument stories show that group members engaged with the topic of the dispute, which furthered the analysis and evaluation of options.The argumentative function of stories clearly aligns with the emphasis in deliberative theory on reasoned argument (Burkhalter et al., 2002).
Unitary argument stories presented a group member's personal experience and perspective on the issue but also emphasized the values and connections that were shared within the group.This type of story can help participants to move beyond the limitations of seeing their differences as simply a two-sided debate.By demonstrating both some level of individuality and a common shared identity, unitary stories can potentially help deliberative groups to find opportunities for consensus.
The non-argumentative stories are intriguing because they occur in the midst of a disagreement but do not directly deal with the issues under dispute.Introduction stories helped establish the legitimacy of individuals' experiences, which then gave them authority to participate in the discussion.Because these stories emphasized individual identity, they set the stage for further deliberative interaction simply by bringing people's experiences and perspectives to the table.They may also have helped group members to establish rapport early in their time together.This relational function is perhaps the most expected result of hearing stories in groups, and is in line with the concerns of scholars who emphasize the importance of personal experience in deliberation (Fraser, 1992;Sanders, 1997;Young, 1996).
Finally, storytelling can potentially be transformational, describing some personal change that mirrors a social transformation and having a strong emphasis on group members sharing a collective identity.Transformation stories can help group members to get through times of crisis and look forward to new directions together.In terms of deliberation, these transformation stories could help group members to increase their identification as civic participants.By telling and hearing stories that connect diverse experiences to common goals and visions, group members may come to see themselves in a new way.This transformation from isolated individual to active civic participant is likely to be gradual, if it occurs at all.However, transformation stories could aid in this process.
deliberative practitioners.At a minimum, facilitators should develop an ear for stories, and think about the variety of purposes stories can serve in deliberative groups.Further, it might be important for some facilitators to resist the urge to treat personal stories as evidence that the group is getting "off track" and instead view stories as an opportunity for furthering discussion in a different way.
For example, if facilitators recognize the difference between adversarial and unitary argument stories, they could help groups understand that the way participants frame their personal experiences holds implications for the way the group manages conflict.Facilitators can also help highlight the morals of stories told by group members and ask follow-up questions that help group members investigate how well those morals are linked to community values implicated in the decisions.
In sum, this study shows that stories can provide information, clarify values, encourage a sense of collective identity, and even provide group members with a vision of how the future could be.All of these functions align well with aspects of deliberative theory and can further productive conflict management in groups.However, there could easily be dangers inherent in storytelling in deliberative groups.Adversarial stories could be used to divide group members into factions that hinder the group's ability to work together.Unitary argument stories or transformational stories could potentially create a sense of false consensus by overemphasizing inclusive collective identities and covering over real differences among group members.
Deliberative scholars and practitioners alike should be cautious of either vilifying or valorizing storytelling in deliberative groups.The story types discerned in this study provide a foundation for future study that can help us better understand the conditions that enhance the positive aspects of storytelling for conflict management and help us recognize specific dangers and limitations of stories told in deliberative forums.

Limitations and Future Directions
In addition to substantive implications, this study makes a methodological contribution to the literature on group deliberation through emphasizing the discursive practices that members engage in when faced with difference.The focus on how groups of people actually communicate when they deliberate is different from the ideal stance taken by most deliberative scholars (c.f., Buttom & Mattson, 1999;Ryfe, 2006).The two-stage study design is useful for this project because it provides a more refined framework than does the analysis of a single group.
The results of this study, however, have to be understood within the context of its limitations.As a qualitative case study, the ability to make generalizable claims is limited.The story-type framework provided here needs to be refined and tested through further research.By further operationalizing the framework, scholars could investigate the extent to which these story types occur in other groups.The response patterns found here could lead to hypotheses about interactional issues, such as conflict management and decision making.
For example, it seems reasonable to expect that unitary stories would be more likely to lead groups toward consensus than other story types, and that adversarial stories would promote more issue analysis than other types.This study shows that unitary argument stories elicited more responses when they were told in response to other stories than when they occurred as the first story in a period of difference, which is worthy of further investigation.Future research could improve our understanding of how storytelling functions in group contexts and also provide more nuanced recommendations for facilitators and forum organizers who hope to improve the ability of groups to manage conflict, weigh options, and engage in the analytic and social aspects of deliberative discussion.
to respond briefly to a few of your comments from last night and then I think I will have beaten this Transportation horse pretty much to death.First of all, I do not think those of us in Battery Park City are being either parochial or hypersensitive, nor are we ignoring the greater public good, by expressing deep concerns about plans with the potential to change the character of our neighborhood for the worse.There are a

Table 1
Number of Stories and Responses in two Listening to the City Discussion Groups identity, but it does not necessarily indicate a shared identity that includes other group members.Responses.Introduction stories elicited almost no direct responses from other group members in either of the discussion groups studied.Although there are eight responses noted in Table