On the nature of personal narratives of high quality

Abstract

This investigation explores the discourse devices associated with high-quality personal narratives. The study examined normative characteristics of 11 high-quality personal narratives of a frightening experience identified (from a larger set of 72 narratives) for their effectiveness in engaging the audience. Lay ratings and an ethnographic interview with seven of the excellent storytellers further characterized the stories and validated their selection. Narratives of both African Americans and Caucasians were represented, and were similar in nature. The excellent narratives were longer, conveyed more fearful topics, and were more dramatic than average narratives. Drama was achieved through direct speech, prosodic shifts, voice changes, inclusion of multiple characters, repetition, and syntactic and semantic parallelism. Illustrative narrative excerpts are provided. This study illustrates the potential in pairing holistic and analytical approaches to narrative analysis.

Introduction

In search of norms for personal narrative production

A major emphasis in applied linguistic research involves the consideration of norms. The seminal research on orally produced personal narratives established a tradition of isolating patterns of narrative production to relate the formal properties of narratives to their functions (Labov & Waletzky, 1967; Labov, 1972). What has emerged from these on-going efforts (e.g., Bamberg, 1997) is an improved understanding of core characteristics of spoken narratives of personal experience as a genre, such as the differentiation between main event line and evaluation, the importance of features such as reference, emplotment, and coherence, and the dimensions along which narratives can vary, such as chronology and description (Ochs & Capps, 2001).

Establishing norms can be challenging, especially when individual production patterns are diverse. The field of narrative analysis has developed an understanding of what is common to oral narratives and the dimensions along which narratives vary. However, the normal range of oral narrative production, i.e., the differences between good stories, average stories and poor stories as they are perceived by an audience, has not been systematically described. More information is needed about the narrative features that may converge to make oral stories of personal experiences excellent vs. poor. An understanding of what constitutes the average and the above- and below-average are necessary to account for the range of “normal”. This, in turn, is of great relevance and concern to applied studies of discourse.

The current study is an effort toward establishing norms of production for spontaneous narratives that are relatively detached from surrounding talk and activity and are elicited from a single active teller (Ochs & Capps, 2001). Specifically, we focused on narratives at the superior end of a normative continuum. We were especially interested in the discourse devices associated with these high-quality personal narratives. We examined oral stories that “produce in the audience a profound concentration of attention that creates uninterrupted silence and immobility, an effect that continues long after the ending is reached” (Labov, 1997, p. 396) and how this effect on the audience is achieved. The superior end of the normative continuum was chosen as a point of entry because we predicted that excellent stories would display the widest range of narrative discourse devices—both core and optional.

The fundamental nature of personal narratives

Personal narratives have been termed “narratives of personal experience” (Labov, 1972), “personal narrative stories” (Stahl, 1983), and “tales of personal experience” (Goffman, 1974). A range of disciplines recognize the fundamental role of narratives in human experience (Robinson, 1981). From the perspective of psychology, Bruner (1990) notes that the capacity to create and decipher plots is a quintessential feature of the human species. Myerhoff (1978), an anthropologist, characterizes our species as homo narrans. Likewise, Polkinghorne (1996) describes narrative as the primary form by which human experience is made meaningful, because narrative allows us to connect “elements of awareness” into meaningful episodes.

Disciplinary perspectives on narrative vary. Some disciplines, such as linguistics, examine narratives as texts or literary phenomena (e.g., Labov, 1972; Longacre, 1996). Social and cultural psychology emphasize the role of narrative reminiscence for making sense of our lives as we age (e.g., Birren & Birren, 1996; Erikson, 1982; Harris, 1997; Polkinghorne, 1996) for instantiating our social and personal identities (Mumby, 1993), for presenting the self to others (Goffman, 1959, 1974) and for characterizing the human mind within a cultural context (Bruner, 1990). In contrast, sociology and anthropology focus on narrative as performance or on narratives’ socio-linguistic and ethnographic manifestations (e.g., Bauman, 1986; Bauman & Briggs, 1990; Hymes, 1982).

Personal narratives are also a fundamental means of understanding current and historical cultural realities (Finnegan, 1998; Terkel, 1992, 1995). They have been used to characterize the cultures of such diverse groups as African Americans (Champion, 2003; Morgan, 2002; Smitherman, 1977), Americans of Jewish heritage (Myerhoff, 1978), Midwestern Whites (Johnstone, 1990) and Americans in general (Polanyi, 1989). Culture is manifested in conventions of narrative structure and style (Finnegan, 1998), in continuities between personal narrative stories and other genres of folk narratives (Stahl, 1983), and in the choice of which events are remarkable enough to narrate (Bruner, 1990).

The current study draws from linguistics, sociology, and cognitive psychology, to examine the ability of personal narratives to engage the audience by drawing them into the experience being described. Specifically, we explore the linguistic devices used to create such high-quality narratives. The following section addresses the potential characteristics of narratives that may serve to draw the attention and interests of the audience.

Narrative features used to engage the listener

Certain aspects of narratives may function to engage the audience. One is the inherent tellability or remarkableness of the event. In addition, certain linguistic devices, such as direct speech, may enhance the vividness and immediacy of the event being described. We consider each of these categories in turn.

Tellability.

Labov and Waletzky (1967) comment that an incident must be unusual, unexpected, or unique to be “tellable”. Van Dijk (1976) refers to this feature as the remarkableness of the event, while Ochs and Capps (2001) refer to this as the story’s tellability. The tellable element or elements are often evaluated as the point of the story. As noted by Polanyi (1989), narratives “are told to make a point, to transmit a message—often some sort of moral evaluation or implied critical judgment—about the world that the teller shares with other people” (p. 12). Such points range from comments on the way the world works, our personal and social identities in the world, appropriate ways to act in the world, and how to evaluate what goes on in the world (Hall, 2002). As Hall (2002) notes, personal stories are often more powerful than direct admonitions and advice as teaching tools. As an illustration of this point, Longacre (1996) comments on how effective narratives can be in gaining the attention and interest of a congregation when the stories are interspersed in otherwise routine sermons.

The tellability or remarkableness of the event is consensually defined, by reference to the norms of the narrator’s speech community (Bruner, 1990; van Dijk, 1976). Thus, narratives function well as instruments of folk psychology (Bruner, 1990). The tellability or remarkableness of the event must also be defined with respect to the context of telling. Complete accounts of storytelling must take into consideration what is being told, to whom it is told, when it is told, and for what purpose a story is being told (Robinson, 1981). In the current study, tellability is operationalized in the form of a salient theme, i.e., the theme of fearfulness and fearful events.

The tellable elements of a narrative are typically verbalized in the form of narrative evaluations (Labov & Waletzky, 1967; Ochs & Capps, 2001). Similarly, direct speech of other characters in the narrative often revolves semantically around the tellable elements of the story; and often expresses alternative perspectives on the unusualness of a narrative event. Direct speech is one device on which the current study will focus. The use of direct speech to highlight the story’s tellability also lends authority to the opinion that the point is tellable, by virtue of the fact that a person other than the narrator is expressing it.

In summary, narratives based on events that are tellable or remarkable with reference to the audience and the culture at large, may hold the attention of the audience. These tellable elements may take the form of narrative evaluation (Labov & Waletzky, 1967), or linguistic devices such as direct speech.

Selected linguistic devices.

Listeners may also be drawn into a narrative through the use of certain linguistic devices that function to engage the audience. Use of these devices may contribute to the overall quality of the narrative as perceived by the audience.

One of the devices that may function to engage the audience is direct speech, or constructed dialogue. As noted by Labov (1972), “the most highly evaluated form of language is that which translates our personal narrative into dramatic form” (p. 396). Direct speech is the reproduction or mimicry of the speech produced by characters in a narrative. This includes speech produced by the speaker-protagonist to others or to himself i.e., self-directed speech or “direct thought” (Chafe, 1994), and the speech of others to others, or to themselves. Researchers in the sociolinguistic and interactionist traditions comment that the details contained in direct speech may be a means of creating images, thereby allowing the listener to imagine a scene (Tannen, 1989). Thus, direct speech may be one means of creating involvement between speaker and audience (Tannen, 1989). This function of direct speech is highlighted among groups with high involvement styles of communication in which the audience maximally participates (Johnson, 2000; Tannen, 1989). For instance, some African American speakers use a narrative tactic called “marking” or mimicry, which is the deliberate imitation of speech and mannerisms of someone else for authenticity, ridicule or rhetorical effect (Mitchell-Kernan, 1972; Morgan 2002; Smitherman, 1977). In addition, the use of African American Vernacular English in direct speech (Ulatowska & Olness, 2001) may also enhance the joint involvement of listener and audience, especially if they are users of the same language variety. Thus, other linguistic and paralinguistic features may combine with direct speech to engage the audience. Direct speech that characterises or individuates the characters in the narrative may increase listener involvement in the experience.

From a cognitive-linguistic perspective, another means of drawing the listener in may be through building a sense of immediacy. For instance, direct speech may contribute to the sense of immediacy in a time-displaced event (Chafe, 1994) by evaluating the information contained in an earlier speech event through use of wording and prosody (Labov, 1972). Another device that may build a sense of immediacy is the use of historical present. Chafe (1994) comments that co-occurrence or symbiosis of direct speech and the historical present may be more effective in creating a quality of immediacy than the use of either device alone. Other devices that may collapse the “now” of the story world with the “right now” of the narration include external commentary (e.g., offering opinions about the events), expressive phonology, and direct forms of address to the audience (Longacre, 1976).

From the standpoint of discourse linguistics, heightened “vividness” in a narrative may be a means of engaging the audience. Again, direct speech may be one device used to create heightened narrative vividness, especially at the peak of action, or climax, of a story (Longacre, 1996). Longacre (1996) speaks of a hierarchy of increasing vividness, from pseudo-dialogue, to dialogue, to drama. This is often associated with a “crowding” of the scene with characters, which involves an increase in the number of characters, often at the climax of a story (Longacre, 1996). Other changes often found in the climax, which may draw the attention of the listener, include accumulation of details, and shifts in person and tense/aspect structure relative to the surrounding text (Labov, 1972; Longacre, 1996).

As already noted above, more than one device may draw the listener in through enhancing a sense of immediacy or heightening vividness. Co-occurrence of these devices may be especially effective in this respect. For instance, when direct speech is reiterated (repeated) in a given narrative, the repetition may intensify the functions filled by direct speech, because direct speech and repetition are functionally similar in several respects. From a discourse perspective, both direct speech and repetition are forms of evaluation (Labov, 1972). Both are evident in speakers with a high involvement style (Tannen, 1989), such as certain groups of African Americans (Ulatowska, Olness, Hill, Roberts, & Keebler, 2000). Repetition, like direct speech, invites the listener to interpret the meaning of the repeated item more deeply, especially if the item is reformulated or paraphrased (Merritt, 1994). From a cognitive perspective, repetition also focuses the listener’s attention (Johnstone, 1994), as does direct speech. From a sociolinguistic perspective, repetition is similar to direct speech in its potential to engage the listener and create interpersonal involvement (Johnstone, 1994). Similar effects may be achieved through “non-exact” repetitions, such as shifts in wording, intonation, and context of the repetition; syntactic and semantic parallelism; and paraphrases. Like other forms of repetition, these may engage the listener by emphasizing semantic content and delaying the progression of events in the narrative (Gee, 1985).

The goal of this article is to identify whether these devices indeed play a role in the development of narratives of high quality, where high quality is defined by the perceptions of the audience.

Methods

Narrative sampling procedure

The 11 high-quality personal narratives examined for this study were drawn from a sample of 72 personal narratives of a frightening experience. The narratives were produced by 72 adults (half men, half women) who completed a multi-task narrative battery as part of a larger study (Ulatowska, Olness, Wertz, Thompson, Keebler, Hill & Auther, 2001). Half were African Americans, and half were Caucasians. These individuals were primarily middle-class Southerners, who ranged in age from their forties to their late sixties.

The inclusion of the personal narrative task follows in the tradition of studies that elicit personal narratives on emotive topics to sample natural, unmonitored language (Labov, 1972). Such emotive topics have the potential for displaying the optimal productions of good storytellers (Labov, 1997). The participants were asked to tell a story about a frightening event that they had experienced. They were allowed unlimited time to produce their stories. The interviewer acted as an attentive and interested listener.

The personal narrative of each individual was audio recorded and orthographically transcribed. Transcripts alone were used for all ratings. Audiotapes were not used because lay raters were included in the group of raters and because the participants had not provided consent for disclosure of audiotapes to lay raters. Although use of audiotapes for all analyses would have been optimal, the transcripts effectively provided the core information which formed the basis of the analysis.

The transcript of each narrative was rated by the seven individuals on the research team, all of whom were female. The team of raters consisted of three native English-speaking Caucasian Americans, two native English-speaking African Americans, and two non-native but fluent speakers of English (one Pole and one Filipina), with training and expertise in linguistics and communication disorders.

The raters first identified the set of narratives that contained the minimum features that comprise monologue narrative (Longacre, 1996): good coherence (i.e., connectedness of the information in the story), unambiguous reference to the elements talked about in the story (e.g., characters, location, and time), and full emplotment of the event on which the story was based, including the basic narrative structural elements of the story. These elements are orientation, complicating action, and resolution (Labov, 1972). Sixty-five of the individuals met all these criteria. From this sub-set of 65 narratives, the raters then identified those narratives that were the best at drawing the listener into the story. This intuitive judgement of listeners in reaction to a gifted narrator has been described by Labov (1997). He states that “the storyteller is someone who can make something out of nothing, who can engage our attention with a fascinating elaboration of detail that is entertaining, amusing, and emotionally rewarding. From the first lines of such a narrative we know that we are in the presence of a gifted user of language” (p. 396). A total of 11 narratives (or 15% of the stories) met these criteria of a high-quality narrative.

The individuals who produced these quality stories were 11 women (five African American and six Caucasian). Age and education level of the individuals were diverse. Subject ages (across both groups) ranged from 43 to 68 years, with five women in their forties, two in their fifties, and four in their sixties. The individuals’ education ranged from 12 to 16 years: three individuals had 12 years of education, seven had 13 to 15 years of education, and one had 16 years of education. The ages and education levels of the sample of 11 people were representative of those in the larger sample of 72 participants.

Identification of devices contributing to narrative quality

As mentioned earlier, superior stories were of interest for building an understanding of potential devices that contribute to optimal forms of narrative production. After the high-quality narratives were identified, they were analysed for presence or absence of those linguistic devices thought to draw the audience into the story. These linguistic devices were identified using the transcripts. Additionally, paralinguistic devices such as shifts in tempo and loudness were identified by listening to the original audiotape. These paralinguistic features provide additional information that may contribute toward development of suspense and climax, which in turn may engage the audience. The analysis was guided by the literature on potential means of enhancing vividness and drama in narratives. The devices examined included shifts in discourse genre (e.g. shifts to direct speech or dialogue), lexical and semantic devices such as repetition and syntactic and semantic parallelism, the use of specialized morpho-syntactic forms, such as the historical present, and paralinguistic features, such as tempo and loudness.

As a point of contrast, a sampling of 10 stories which were not rated as high-quality stories (but which had good coherence, unambiguous reference, and full emplotment) were also examined for these same devices. Of these 10 stories, five were produced by African Americans, and five were produced by Caucasians. The tellers of these stories were all women, with age and education level comparable to that of the tellers of the high-quality stories.

Holistic evaluations of quality by lay raters

Another approach used to characterize the nature of high-quality narratives was to ask lay raters to comment holistically on the characteristics of high-quality stories. This enhanced the ecological validity in the description of high-quality narratives. Two groups of raters rated the 11 high quality narratives against each other, and commented on the dimensions associated with the relative quality of each story.

Transcripts of the 11 high-quality stories were presented to six middle-aged semi-professional women with an education of 1 to 2 years beyond high school. Three were African American and three were Caucasian. Raters with these qualities were selected to be of the same gender as the research team raters, to avoid a potential confound associated with rater gender. They were selected to be similar in age, education, and professional background to the participants who told the stories, so they would represent audience norms in the narrators’ speech communities. These lay raters were asked to rate each story as “excellent,” “good,” or “just OK” and to explain what made it that way. This provided information about lay perceptions of relative quality within the set of high-quality narratives, as well as the features of each narrative that contributed to its relative quality.

Ethnographic questionnaire for the good storytellers

Story quality is not only reflected in audience perceptions, but also on the narrator’s perceptions of his or her story-telling ability. One female interviewer spoke by phone with seven of the individuals who produced high-quality narratives. This was the total number who could be contacted for interview. The other four individuals had changed residence and could not be located. The seven interviewees were three African Americans and four Caucasians. As part of the conversational interview, the interviewer asked whether the individuals viewed themselves as good storytellers, whether they liked telling stories, whether they tell stories often, to whom they told stories, and how frequently they tell the particular frightening story that they told for the study.

Results

Overview

Findings are presented in four sections. The first presents results pertaining to high-quality narratives. These findings consist of a general description of the stories and detailed illustration of the devices used to generate audience interest. The second section presents the results of lay raters’ perceptions of narrative quality. The third section presents results of interviews with the narrators. Finally, the fourth section presents data on whether linguistic devices found in high-quality narratives also appear in narratives rated as average.

The nature of the high-quality narratives

General nature.

The topics of the 11 high-quality narratives were both emotive and personal, as specified in the theme of a frightening experience. Thus, they are representative of the stories produced in conversations on similar topics. The topics of the stories included potential or actual personal attacks, dangerous driving situations and accidents, and fear of animals. All but one were events that occurred in adulthood, although none were recent events.

The narratives ranged in length from 29 to 108 propositions and minimally contained the basic structural elements of a setting, complicating action with climax, and resolution. The longest structural element of the story was usually the complicating action.

Excerpts from high-quality narratives.

To illustrate the functions of direct speech and other devices used in high-quality narratives, key excerpts are presented. Linguistic devices are often differentially distributed across the narrative structure, with higher occurrence in the complicating action. For this reason, we label each excerpt for the portion of the narrative structure from which it was drawn. A special focus is placed on direct speech and its functions, while other devices are discussed for their co-occurrence and similarity of function with direct speech. Unintelligible sequences are labelled as X’s, and ellipsis of story content is indicated with ellipsis marks (“… . “). Pseudonyms replace actual names. Portions of narratives are referenced through their position in the narrative, indicated at the start of paragraphs in square brackets. Reference to devices such as direct speech and repetition is achieved through bolding in the excerpts proper. Key portions of excerpts are also incorporated in the accompanying discussion. Tempo and loudness are not transcribed in the narrative in order not to detract from the main narrative line, and are referenced in the discussion only where relevant.

(1) Caucasian woman, 68 years old; “Followed by a Stranger”

[excerpt from end of setting] And when I lit this cigarette, there was a light came on across the street, and a man was standing behind a tree and he lit a cigarette. And I thought, “How strange.” And he stepped out into the light and he was dressed in a policeman’s uniform. And my first thought: He was going to arrest me for lighting a cigarette. Then I thought, “No”, this was wartime and there were not enough policemen to go around and they certainly didn’t walk the streets, you know. Cause they just didn’t do that. And I thought, “He’s not really a policeman,” and I thought that “Maybe he’s just off duty and going home or something.”

[excerpt from complicating action] But he crossed the street and started walking behind me. And if I walked faster, he walked faster. And if I slowed down, he’d slow down. But he kept getting a little closer and closer. And I got to the corner where to go on I would have walked behind Forest Avenue High School. You have to understand that I was a child (XXX) and I had no training, no teaching, I knew nothing about rape or sex or anything like that. I was very naïve. But apparently I knew danger. And I knew if I walked behind that school, there was nobody would hear me, anything could happen. And when I got almost to the corner I saw a brick lying by the sidewalk. And I picked it up and I turned around and I said, “If you’re following me, I want you to know I’ve got a brick in my hand and I won’t be afraid to knock your head off.” Now this 14-year-old girl talking to a man in a police uniform and the man stood there and never answered me. Then he turned around and walked the other way.

In narrative excerpt (1), the narrator’s use of self-directed speech (“I thought…”) in the setting contrasts with her use of direct speech at the climax of the complicating action (“… and I said, ‘If you’re following me…”). The self-directed speech in the setting expresses the speaker’s realization of impending danger. Its function is to engage the listener. The syntactic and lexical parallelism that immediately follows (e.g., “if I walked faster, he walked faster”) stages and emphasizes the fearfulness. This fearfulness reaches a climax when direct speech is responded to with silence (“…and the man stood there and never answered me.”). The tension produced by this silence makes the quick descent to resolution even more powerful. A flattening of prosodic features throughout the narrative for this older speaker is notably incongruous with the type and number of devices used.

(2) Caucasian woman, 67 years old; “Stranger Breaking into the House”

[excerpt from complicating action] And ah heard a knock on the door so the lights were all on, I walked in, I raised the blind at the front door windows, and a man was standing there. And I looked at him and I said, “What is it? What do you want?” And he didn’t say a word and he just looked at me. And I didn’t like the look. He was either drunk or drugged, um just didn’t look right and it scared me. So um I had I switched the light off so I could see out. So I saw him plainly. I switched the light back on and I lowered the blinds and walked off and wondered what to do about it … so I decided I would just get the rifle out of the closet. I knew where the bullets were, I loaded it…

[excerpt from resolution]: It wasn’t much longer after that that my husband came in and he couldn’t get in the back so he knocked and I answered it with a rifle and he said, “What in the world, what’s the matter?” And I told him I had been terrified.

[coda]: That was frightening. And he said, “Well, you did the right thing except for one thing, the gun was on safety.” But they also say a scared woman never misses.

As previously seen in (1) above, direct speech in narrative excerpt (2) occurs in more than one structural component of the narrative. It occurs first at the peak of the complicating action (“… and I said, ‘What is it? What do you want?”), and again is responded to with silence (“And he didn’t say a word.”). This heightens the fearfulness of the situation. This fearfulness is further supported by all the details the speaker provides to document her realization of the intruder’s threatening nature (“And I didn’t like the look … I saw him plainly.”), as well as the concatenation of sentences with and and so throughout the narrative for building suspense. Direct speech occurs again in the resolution (“… he said, ‘What in the world, what’s the matter?”’), where perspective is shifted by a new character, who evaluates the extremity of the situation in a jocular fashion. This same character offers an alternative perspective again in the coda, when he mentions the irony behind the gun’s function (“And he said, ‘Well, you did the right thing except for one thing, the gun was on safety.’ “)—a perspective countered by the narrator in her closing sentence (“But they also say a scared woman never misses.”). As in (1) above, speaker (2), who is also an older speaker, displays a flattened intonational pattern, despite the relatively extensive use of other devices to produce a quality narrative.

(3) African American woman, 44 years old; “Mice in the Bedroom”

[excerpt from setting] And one night, I kept hearing this noise, schtschtscht, and I didn’t know what it was. I ignored it, schtschtscht, even though a couple of nights I heard it and I ignored it and I, and you know I sleep on through it. And finally, the noise got, started getting louder, schtschtscht.

[excerpt from complicating action] (The noise) wouldn’t really go away but it would quieten [sic] down, and so I like wonder, “What could that be?” I was too scared to get out of bed to go see… . So I opened up the window, and I said, “I don’t know why I’m opening up the window cause I can’t jump” cause I stayed on the second level. I always stay on the second level. And so I like, “What could that be?” … I screamed for (my oldest) to come in there. And she came in and she, and she “What you, what’s the matter?” And I say, “Something’s under this bed.” So she turned the light on and looks up under there—she was always the brave one—looks up there “I don’t see anything.” … It was like 2 a.m. I called this friend. I said, “There’s something under my bed,” cause it had gotten up under my bed, I said, “There’s something under my bed, sorta like a bomb or tick, something tickin, can you come down here?” “This, do you know what time it is?” I said, “Yes, but I can’t go to sleep and I’m scared of it.” And he said, “Sara, do you know what time it is?” “Yeah.”

[excerpt from resolution] (My friend) said “It’s dead, Sara.” I said, “It’s not, you need to get it.” So he took the broom and immediate [sic] got him dead, and swept him out and was going to put him in the trash. I said, “No, that’s not gonna work. This is in the same house. You have to take it out when you go out.”

The story excerpts in (3) are similar to excerpts in (1) and (2) for the appearance of direct speech in more than one structural element of the narrative, and in more than one form (self-directed, directed toward another person, directed toward the speaker). However, this speaker also employs “marking” in her direct speech, as her voice, vernacular and intonation patterns change between speakers when direct speech is used. This mimicry enhances the individuation of the speakers, and the involvement of the listener, as does the occasional use of immediate shifts from speaker to speaker, sometimes with no embedding of the direct speech with say (“‘… can you come down here?’ ‘This, do you know what time it is?’ … And he said, ‘Sara, do you know what time it is?’ ‘Yeah.”’) This moves the narrative toward drama, and heightens its vividness. Unlike the stories in (1) and (2), the narrative takes a humourous tone, as characters other than the protagonist comment in their direct speech on the disproportionate level of fearfulness in the narrator, in the face of minimal danger (“Do you know what time it is?”) This alternative perspective is repeated, and the repetition enhances this humour. Repetition of onomatopoetic mouse sound at the beginning (“schtschtscht”) also sets the stage for the speaker’s fearfulness. The effects of direct speech, repetition, marking, and use of vernacular combine in this narrative to provide a good example of heightened vividness.

(4) Caucasian woman, 43 years old; “Suspected Intruder in the Home”

[setting] Pretty close to the end of the movie, I heard him [speaker’s husband] cough, in the area near our laundry room in the house. But I was pretty sure he was in the living room with me. So I looked over, and he was.

[excerpt from complicating action] I said, “That wasn’t you that just coughed.” And he said, “You mean, it wasn’t you?” And we both looked at each other, because we had both heard the same sound at the same time coming from an area nowhere near where we were. Well, I picked up the phone which was right by where I was sitting and called 911. And they searched everywhere. In the kitchen cabinets. Everywhere. Outside. All around our house. In our cars in the garage, up in the attic, down in the basement. Nowhere did we find anybody.

While previous examples show direct speech to occur at the peak of the complicating action, speaker (4) uses the vividness of direct speech at the beginning of the complicating action, to highlight the characters’ realisation of a potential intruder (“I said, ‘That wasn’t you that just coughed.’ And he said, ‘You mean, it wasn’t you?’ “) The simultaneity of the realization is enhanced by the lexical parallelism in these two instances of direct speech. The contrast between the direct speech and the subsequent silence (“And we both looked at each other…”) heightens the tension, as the search for the intruder is initiated. The extensive details of this search (kitchen, everywhere, outside, around house, in cars, in attic, in basement) further enhance this tension. The speaker then makes use of paralinguistic emphasis at the end of the description of the search (“NOwhere did we find ANYbody”) to express the futility of the search, which again emphasizes the fearfulness.

(5) African American woman, 46 years old; “Alone and Followed by a Gang”

[excerpt from complicating action] As I was walking, I heard these footsteps behind me, heavy footsteps, more than one set of footsteps. And ah, I heard this voice inside of me say “Keep walking” so I kept walking. Then I heard this same voice say “Stop and turn around.” So I stopped and I turned around and there was these guys and I don’t know how many of them. It seemed like 50 but I am sure it was but it seemed like it, but anyway there was these guys all around me and I asked a very leading question and I said, “What do you want?” Which that time of night by myself, 14 years old nobody out there but me and these men, and the guy looked at me and said, “Oh we’re sorry, we thought you was somebody else.”

Sample (5) illustrates the use of direct speech with several different devices—lexical and syntactic parallelism (“footsteps … heavy footsteps … more than one set of footsteps;” “‘keep walking’ so I kept walking … ‘stop and turn around.’ So I stopped and turned around”), paralinguistic variation in the form of marking (change in voice and manner of speaking in direct speech), and crowding of the stage at the climax (“I don’t know how many of them. It seemed like 50 …”). However, the direct speech in this sample occurs only in the complicating action. The self-directed speech in the sample is unique in that it is ascribed to a separate “character” (a voice), which speaks at the beginning of the complicating action. The lexical and syntactic parallelism between this self-directed speech and the sentences describing the protagonist’s actions (keep walking, kept walking; stop and turn around, stopped and turned around) delay the action and heighten suspense. The instances of direct speech at the end of the complicating action mark the very peak of the action (“What do you want”) followed by a quick resolution in the benign response of the guy/man (“Oh we’re sorry, we thought you was somebody else.”) The tension is immediately broken as the man’s words offer an alternative perspective on the situation, i.e., an innocent case of mistaken identity. The vividness of this instance of direct speech is enhanced through change in the speaker’s voice and use of vernacular (“… we though you was”). The additional devices that contribute to the quality of this excerpt are the crowding of the stage as one approaches the climax. This crowding and its fearfulness is additionally emphasised through the use of exaggeration (“It seemed like 50 …”). At the peak, the narrator also repeats/summarizes the elements from the setting that contribute to the fearfulness, (time of night, speaker’s age, and isolation), thus drawing the listener into the experience.

(6) African American woman, 47 years old; “Errant Children Chased by a Horse”

[excerpt from setting] We could see this horse down this field, and we were just walking, and we was saying, “Wouldn’t it be funny if that horse decided to just chase us around this field?” So we started walking and he saying, “This horse is getting closer and closer to us.”

[complicating action] And all of a sudden, this horse is starting to trot. And he trots a little faster and a little faster and he’s coming towards us. And it’s like ah I told my brother, I said, “That horse is coming toward us,” and he said, “No girl, no it’s not.” Mmm and then all of a sudden it’s like this horse is coming toward us, running towards us. So we started running and we ran, CaSaundra [name of interviewer], I never ran so hard in my life. We ran and we were screaming and hollering for our parents. And this horse mmm, this horse was running after us. This horse was like gonna get us.

[excerpt from resolution] … And we was just screaming and hollering and went up there and told our parents about it. And she said, “See, that’s what you get. Cause that one, that’s one of our bad horses and he does that to people. And he will run after you.”

The instances of direct speech in this narrative occur across more than one structural element of the story. Each instance refers to the main source of fear, i.e., the horse, and this thematic emphasis is also realized in use of determiner this with horse in its highlighting function, at the beginning of the setting. In the setting, the direct speech presages the actions of the horse. In the complicating action, direct speech is used to provide two different perspectives on the potential threat of the horse (“That horse is coming toward us,” and “No girl, no it’s not.”) The speaker uses voice change to indicate change in speaker. Extensive repetition (of words such as faster, run, and screaming and hollering) produced with emphasis extends the action associated with the chase, which increases suspense, and the use of the vocative (CaSaundra, the name of the interviewer) further serves to draw the listener into the action, as does a shift to historical present in this sequence (“… is starting to trot.”) Accelerations in rate of speech and increase in volume are also employed as iconic means of displaying the pace of the chase, at the climax of the complicating action. The direct speech in the resolution (“And she said, ‘See, that’s what you get …”’) offers the parental perspective on the incident as the consequence of straying away from home.

Summary.

There seems to be little difference between the narratives produced by African Americans and Caucasians in the linguistic devices investigated in this study, even though participants in the study could not be assumed to be part of the same cultural group. Ethnic group differences were limited to vernacular forms of pronunciation and morphosyntax. The eleven stories were constructed in the genre of drama. Many involved multiple actors, and direct speech occurred in 10 of the 11 narratives. This involved both self-directed speech of the narrator and direct speech of other characters in the narrative, highlighted through the use of voice and vernacular changes. Instances of direct speech often offered alternative perspectives of other characters on the overall fearfulness, or sometimes humour, behind the event, which either confirmed or contrasted with the perspective of the narrator. In two of the narratives, direct speech of the narrator was followed by silence from other characters, at the climax of narratives in which the narrator/protagonist was being threatened. This silence contributed to the overall suspense of the story.

Use of linguistic devices was often guided by plot development. The location of occurrence of direct speech varied from narrative to narrative, although it was usually found in more than one structural component of the narrative, and it was consistently found in the complicating action of the narrative. This was often accompanied in the complicating action by crowding of the stage with characters, and with shifts in tense and syntactic complexity, such as a shift to the historical present, as previously noted by Longacre (1996).

Of the 11 good storytellers, three of the Caucasians and one of the African Americans used minimal tonal semantics, even though the topic of the story was suspenseful. These four individuals were the oldest individuals in the sample. The striking feature of the other seven speakers was the use of tonal semantics, manifested as changes in tempo and loudness, and strong stresses for emphasis. Also, in instances of direct speech, these speakers imitated the vocal characteristics of particular characters in the story. This is similar to the phenomenon of mimicry of characters in stories, cited by Smitherman (1977), Kochman (1972), and Morgan (2002). There was also acceleration of speaking rate that corresponds to the acceleration of action in the story, especially in the complicating action and the climax.

Repetition and syntactic/semantic parallelism were also pervasive, and served to highlight information for the listener. Some speakers also addressed the listener/interviewer by name during the interview (i.e., used the vocative), which further served to draw the listener into the story.

Holistic evaluations of quality by lay raters

Ratings by lay raters revealed that even among the stories rated as excellent, there was a hierarchy of narrative quality. Those that were ranked the “highest of the high” in quality were typically those that involved a direct physical threat to the narrator/protagonist, such as a house break-in attempt, or being followed by gangsters or a rapist-in-disguise. Those that were ranked lowest by the lay raters were stories with less physical threat to the narrator, such as fear of mice and driving in the snow. For the stories rated by lay raters as excellent, the justification for such ratings often focused on how well depicted the story was, and how well the rater was drawn into the story. Such comments included, “Very detailed,” “You can feel her panic,” “This story makes you think,” “Very frightening. This happens daily,” and “The fright concept is excellent.” For those stories with moderate to low ratings, comments often focused on the listener’s lack of empathy for the fearfulness in the event. Such comments included, “You don’t really feel any of the author’s sense of fright” or “The author tells the story and you get all the facts—just nothing to make you feel their fright.”

Narrators’ perceptions of their storytelling ability

Of the seven narrators who could be reached for a phone interview, all stated that they view themselves as good storytellers based on their audience’s reactions while they are telling stories. When asked if they tell stories often, all responded affirmatively. Two of the seven stated that they did not start telling stories until they were adults, and the other five stated that they have enjoyed telling stories from childhood. All stated that they told stories for purposes of entertainment, five stated that they sometimes told stories to preserve family history, and three stated that they sometimes told stories to teach or help other people. Six of the seven individuals stated that they do not frequently tell the particular frightening experience story they told for the study. Many stated that this was because the event was a defining or important event in their lives, but the appropriate time to tell about the event does not occur often. One person said that she did not tell it often because everyone close to her has already heard it.

Occurrence of linguistic devices in average stories

Although the focus of this article is on high-quality narratives, a brief comparison with a sample of average stories highlights relevant features of narrative production. This comparison also contributes toward the overall motivation for this research, i.e., to develop norms of storytelling. Recall that 11 people out of the 72 individuals were rated as excellent storytellers, and 54 were rated as average, but still having good coherence, reference, and emplotment. The analysis of a representative sample of 10 of these stories rated as “average” revealed several differences compared to stories rated as “excellent.” Most of the stories rated as average were considerably shorter than those rated as excellent. Length of the excellent stories ranged from 29 to 108 propositions, with a mean of 57 propositions. Seven of the 10 average stories were shorter than 29 propositions, i.e., the shortest of excellent stories. Two more were about the same length as the shortest excellent story (30 and 31 propositions respectively). Only one story rated as average, with a length of 58 propositions, was similar in length to the larger set of excellent stories. Associated with this shorter narrative length in the average stories was considerably less use of direct speech. In instances where it was used, it did not serve the purpose of development of suspense, or elaboration of the theme of fear. As a result, the direct speech used in the average narratives also made little use of prosodic shifts or voice changes. Although there was no basic difference in the choice of topic between average and excellent stories, none of the stories rated as average described extremely threatening events such as near-death experiences or attacks.

Discussion

This study was designed to gain a better understanding of the nature of high-quality narratives, i.e., those at the high end of the normative continuum. These high-quality stories, all personal stories of a frightening experience, were initially selected by the raters on the research team for their ability to draw the audience into the story. This criterion for a high-quality story was confirmed by a group of lay raters, who evaluated the “best of the best” stories to be those that allowed the listeners to feel the fear associated with the event. A potentially informative extension to this research in future studies would be direct comparison of rating of clinical and lay raters on the same set of stories, and the dimensions that contribute to high-quality stories.

Both the content of the stories and the devices used to express them appear to contribute to the stories’ ability to engage the listeners. The stories rated as excellent were narratives of more threatening events, such as attacks and stalking. Among the 11 stories rated as excellent, those rated as the very best by lay raters were those that involved more personal physical threat to the narrator/protagonist. The devices used to express the events in the excellent stories were those that are often associated with high-involvement styles (Tannen, 1989), such as direct speech, repetition, and varied intonational contour. These devices, from several linguistic levels, enhance the vividness of the narrative for the listener. Another device used to heighten vividness was the inclusion of details (Labov, 1972, 1997). However, it is not simply use of these devices, but rather their use in combination at the optimal locations within the narrative, such as the climax, that allow speakers to achieve high-quality narratives by producing a dramatic form, as opposed to a simple narrative form. Similar devices were employed, but were less pervasive, in narratives rated as average. Overall length of the narrative also seems to be associated with narrative quality—those higher in quality with relatively more length, and those of less quality with less length, although this was not an absolute. Presumably, a narrative of a certain length is required to effectively use quality-building devices, and vice versa.

Thus, topic, linguistic devices, and length all seem to contribute in some combination toward the production of high-quality stories. Whether these findings can be generalized to other means of narrative elicitation, i.e., beyond theme-elicited personal narratives, remains to be seen. Different elicitation methods may yield more elaborated story-telling, and different use of devices. However, the informal style and vernacular forms used in these stories, and the fact that they are on emotive topics told to an interested listener, seem to indicate that they are probably representative of natural story-telling abilities.

Even though the devices used by these narrators are associated with high-involvement style, the members of the excellent storyteller group were both African Americans and Caucasians and included individuals from groups not necessarily identified as high-involvement cultures. Thus, the group of good storytellers may not be culture-specific. High involvement style may be associated with specific cultures, or with specific individuals who consistently demonstrate the ability to engage their audience. It would also appear that excellent narrators are a relatively small group. However, these characterisations of good storytellers cannot be generalized beyond the group sampled here, namely, women who are telling personal stories of a frightening event and who are characterized as good storytellers by other women.

One contribution of this study may rest in its pairing of two complementary approaches to discourse analysis—one holistic and incorporating audience perceptions, and the second linguistic and analytic. Such an approach may be particularly useful for those aspects of discourse production that cannot be easily quantified a priori, such as narrative quality or the ability of certain personal narratives to engage the audience. As suggested by Patry and Nespoulous (1990), key discourse properties such as unity and informativeness are not yet amenable to empirical investigation. Combination of global discourse quality ratings with inductive exploration of the linguistic devices associated with certain ratings may provide insights into the complexity of narrative production and its potential influence on an audience.