3
Back to the Elements
Narrative Occasions
(i) Situatedness. Narrative is a mode of representation that is situated in – must be interpreted in light of – a specific discourse context or occasion for telling.
Narratives are both structured by and lend structure to the communicative contexts in which they are told. Ideas from multiple fields of study, including sociolinguistics, social psychology, and narratology, can throw light on narrative occasions taken in this double sense – as communicative environments shaping how acts of narration are to be interpreted, and, reciprocally, as contexts shaped by storytelling practices themselves.
As discussed preliminarily in chapter 1, one of the basic elements of narrative is its situation in a particular discourse context or occasion for telling. In the three-level model of narrative that Bal ([1980] 1997) inherited from Genette ([1972] 1980; cf. Rimmon-Kenan [1983] 2002), the story level (= the narrated content or storyworld) is a mental representation built up on the basis of cues included at the level of the text, and the text level can in turn be viewed as a result (or record) of the communicative processes that occur at the level of narration.1 Acts of narration are grounded in occasions for telling in two senses. On the one hand, interpreters using textual cues to reconstruct a storyworld (see chapter 5) must also draw inferences about the communicative goals that have structured the specific occasion of the telling, motivating the use of certain cues in favor of others and shaping the arrangement of the cues selected. For example, what communicative design informs Hemingway’s use of the (unnamed) male character in “Hills Like White Elephants” as the predominant “reflector” in this story told in the third person – that is, as the center of consciousness through whose perceptions and attitudes the narrator’s account is refracted? How do interpreters of the story factor in this narrative design or strategy when trying to make sense of the text, and what interpretive adjustments would they need to make if Jig were the predominant reflector, or if the story were recast as a first-person account told by Jig, by the male character, or for that matter by the waitress that serves them at the train station? (I return to these questions below.)
On the other hand, processes of narration not only emerge from but also lend structure to particular discourse contexts, molding them into (constituting them as) occasions for telling in a second sense. For instance, Monica begins her story with what Labov (1972) would call an abstract, that is, a pre-announcement of the gist of the narrative, used to clear the floor for the relatively long turn at talk required to tell a story:
(1) So that’s why I say..UFO or the devil got after our black asses,
(2) for showing out.
By announcing that she has something noteworthy to tell, Monica cues her interlocutors (in this case, the two sociolinguistic fieldworkers present when Monica recounts UFO or the Devil) to collaborate in the production of the story by withholding turns at talk that they might otherwise be inclined to take, particularly in the context of an interview where question-and-answer pairs are a typical and therefore expected discourse structure. Note, too, that given the overall structure of their interaction, Monica’s interlocutors would likely feel cheated if she devoted a lengthy turn to a report of unremarkable events (that is, events that are low in tellability in this communicative context). Indeed, if Monica’s interlocutors were her peers rather than interviewers constrained by a somewhat more formal speech situation, they might interrupt a story with no readily apparent point before she ever finished it. Readers of Hemingway’s story, like Monica’s inter-locutors, in effect cede the floor to the narrator, co-creating the narrative by adopting the role of story-recipients. They also use an assumption of tellability to orient themselves to the story: interpreters of literary narratives like Hemingway’s expect the payoff from reading the text to be commensurate with the amount of “floor space” or interpretive engagement the text itself requires – although some (postmodern) works actively manipulate readerly expectations of this sort, frustrating attempts to attribute significance to events sometimes narrated in painstaking detail (cf. the tradition of the nouveau roman, as practiced by writers such as Alain Robbe-Grillet [1957, 1959] 1965).2
In the sections that follow, making frequent reference to “Hills Like White Elephants,” UFO or the Devil, Ghost World (in both its graphic-novel and film versions), and other texts as test cases, I discuss several frameworks for inquiry that can be used to explore narrative occasions defined in both of the two senses just mentioned – as communicative environments shaping how acts of narration are to be interpreted and, reciprocally, as contexts shaped by storytelling practices themselves. My next section reviews sociolinguistic approaches to narrative occasions, drawing on Erving Goffman’s (1981) rethinking of the speaker–hearer pair in terms of what he called production formats and participation frameworks. This section also discusses how research on turn-taking processes, some of which I have already alluded to, is relevant for the study of narrative occasions. Then I move to the theory of positioning (Harré and van Langenhove 1999) developed by social psychologists to suggest how narratives are both a cause and a result of discourse practices, in which participants use utterances to position themselves and others in overarching “storylines” more or less subject to dispute. Finally, I discuss how narratologists have developed yet another approach to narrative occasions; refined by rhetorical theorists of narrative, this approach, sometimes referred to as “the narrative communication model” (Chatman 1978; Herman and Vervaeck 2005a; Martin 1986; Rimmon-Kenan [1983] 2002; Shaw 2005), distinguishes among actual authors, implied authors and narrators on the production side of the storytelling process, and, on the interpretation side, the corresponding roles of actual readers, (types of) implied readers, and narratees (the audience implicitly or explicitly addressed by the narrator in the text).
Goffman (1981) suggested that by decomposing the “global folk categories” of speaker and hearer into more finely grained “participant statuses,” analysts can better understand what talk is and the various ways in which people participate in it. Though focusing on discourse in general, Goffman’s approach also throws light on the situated nature of storytelling in particular, revealing how narratives are embedded in (that is, both structure and are structured by) complex, multidimensional communicative processes. Conversation Analysis, a complementary model for sociolinguistic inquiry with a special focus on the dynamics of turn-taking in communicative interaction, provides further insight into how stories fit within a broader ecology of discourse, with different kinds of stories having different niches within that ecology.
Production formats, participation frameworks, and narrative occasions
Goffman’s rethinking of the speaker–hearer pair in terms of production formats and participation frameworks (Goffman 1981: 124–49) can be traced back to his earlier work (Goffman 1974) on how people make sense of the world by creating frames that channel and delimit the sorts of inferences that need to be made about particular activities and zones of experience. Consider, for example, the different protocols on which people rely to interpret stories told in a classroom, in an experimental literary fiction, or during an argument among family members. We bring different strategies to bear on these storytelling situations because we frame them as different kinds of activities (cf. Levinson [1979] 1992; Wittgenstein [1953] 1958). Likewise, Goffman suggests that participants in talk contextualize what is going on by framing communicative transactions in ways that allow them to assign various kinds of statuses or roles to themselves and others. They then modify both the frames and the role-assignments that those frames entail (or footings, as Goffman calls them) if the nature of a transaction changes in mid-course. Reciprocally, shifting to a different frame or altering one’s footing can change the nature of a given transaction, as when a business person stops making small talk with a potential client and begins presenting the sales pitch he or she was instructed to deliver.3
Thus, in place of older, dyadic models of communication, based on the “global folk categories” of speaker and hearer, Goffman decomposes these supposedly primitive terms “into smaller, analytically coherent elements” (1981: 129). For Goffman, the terms speaker and hearer are insufficiently nuanced to capture the many (and fluctuating) statuses that one can have as a participant in talk. Relevant statuses include, on the one hand, those associated with the various production formats an utterance can have – formats that include speaking as an author, animator, principal, or figure. On the other hand, the folk category of hearer needs to be broken down into a range of other possible participant statuses, including those of addressee, unaddressed but ratified participant (= bystander), or unaddressed and unratified participant (= eavesdropper).
On the production side of things, I may animate utterances authored by someone else, as when a political candidate gives a speech written by a speechwriter; or I may speak for the sake of someone I am defending from an accusation, who then becomes the principal of the utterance. Further, the topic of my talk may be something that I myself said in another place and time, such that the earlier self whose words I animate is a figure, the author of words that my present, speaking self animates (example: someone’s production of an utterance such as Back then I used to tell people that I was from Canada). Meanwhile, on the reception side of things, if I am involved in a conversation with multiple parties, I may have to shift frequently (and rapidly) from having the status of an addressee to having the status of a bystander, and if the interaction splits into two subgroups I may also find myself eavesdropping on a conversation that two or more of the interlocutors intended me not to hear. All these statuses affect the way discourse is designed and interpreted, as participants in talk constantly change their footing, defined as “the alignment [they] take up to [themselves] and the others present as expressed in the way [they] manage the production or reception of an utterance” (Goffman 1981: 128).
But how do Goffman’s ideas bear on narrative communication in particular?4 How can they illuminate the role of context in shaping narrative occasions, as well as the role of narrative in shaping communicative contexts? Both production formats and participation frameworks can be thought of as strategies for formulating responses to questions – questions that we pose as we seek to orient ourselves within the communicative situations we must navigate: am I the intended recipient of this person’s story, or am I an unratified and unaddressed recipient, an eavesdropper catching a gossipy tale to which I was not meant to be privy? Does the narrative text I am reading represent the words of someone speaking in his or her own behalf, as in the case of an autobiography, or the words of someone speaking for the sake of another, as in the case of a Bildungsroman, testimonial, or story told in the context of a eulogy? Further, is the narrative at issue being conveyed by a character within the overarching narrative, a framed tale or story-within-a-story, or is the teller both author and animator of the account? What sort of encounter is going on here anyway? In other words, is my inter-locutor’s story being put to the service of a dispute or a complaint, or are there other communicative purposes at stake, such as memorializing a dead loved one or entertaining an audience of educated readers of fiction? Interpreting stories requires formulating tentative answers to questions like these, that is, attending to how a given story is grounded in a particular kind of narrative occasion.
Consider how Goffman’s ideas about production formats and participation frameworks might be brought to bear on the example narratives included in the Appendix. In “Hills Like White Elephants,” readers have the status of addressees for whom the text authored by Hemingway was intended. But though in one sense the narrator of the story can be viewed as the animator of the account of which Hemingway is author, in another sense Hemingway himself can be described as having the dual status of animator as well as author. His text emerges from prior conventions for storytelling that provide both an orienting context and a basis for narrative innovation, such as the “tip of the iceberg” technique which became one of the hallmarks of Hemingway’s method of composing short stories.5 At another level, Goffman’s ideas can be brought to bear on the conversational interaction, or “scene of talk” (Herman 2006b), represented within Hemingway’s text. Throughout the scene, we readers are unaddressed but ratified participants in the communicative encounter taking place in this fictional world. (If the waitress in the story were to overhear what Jig and the male character say to one another, she could be characterized as an eavesdropper.) Interpreters’ ability to understand Hemingway’s text as a narrative hinges on their capacity to reconstruct the scene portrayed in the story as a coherent whole – a whole whose coherence derives in part from their own specified mode of participation in it.
Within the scene itself, the way the characters frame their own interaction affords them various sorts of footings over the course of their encounter. Hence the text models the process by which people take up and test out the footings that they adopt as participants in talk. For example, the male character appears to be solicitous toward Jig, that is, to treat her as the principal in whose behalf he formulates his remarks about the simplicity of the abortion procedure, etc. However, the male character’s encouraging words and displays of concern for Jig can be read as part of a self-interested attempt to disencumber himself of inconvenient domestic obligations. More generally, insofar as it is a story about the structure and dynamics of participation in discourse, Hemingway’s representation of this scene of talk is metacommunicative, using the resources of written, literary narrative to throw light on processes of face-to-face interaction and the complex, often covert, motives shaping who says what to whom – and how (cf. Herman 2006b).
Further, the production formats and participation frameworks associated with UFO or the Devil and Ghost World can be cross-compared with those structuring interpretation of “Hills”; indeed, differences among the formats and frameworks involved help account for the intuition that these are different kinds of stories. For one thing, as Dorrit Cohn (1999) has argued, one of the key structural contrasts between factual (e.g., historical or autobiographical) and fictional discourse is that in factual accounts the roles of author and narrator (or animator) converge, erasing the distinction between the author and a separate, autonomous narrator that is by contrast one of the signposts of fiction (cf. Lejeune 1989; Ryan 2001b). Definitionally, in fictional accounts, including fictional first-person testimonials such as Ian McEwan’s The Cement Garden (1978) or W. G. Sebald’s Austerlitz (2001), the roles of author and animator/narrator diverge. Whereas readers are warranted in conflating, say, the narrator of Survival in Auschwitz (Levi 1961) with Primo Levi, the biographical individual who authored the text, and can legitimately assume that events recounted by the narrator are ones that Levi himself experienced in the Nazi death camp, adopting the same interpretive strategy for McEwan’s and Sebald’s texts would generate erroneous inferences concerning the life-histories of those two authors.6Although it is conveyed through the medium of spoken discourse rather than written text, the production format of UFO or the Devil is in this respect more similar to that of Levi’s account than McEwan’s or Sebald’s. In other words, we are warranted in assuming that Monica herself underwent the life-changing experience recounted in the story.
Cutting across the different kinds of stories just mentioned, factual as well as fictional, is their use of retrospective first-person (or homo-diegetic) narration, in contrast with Hemingway’s use of third-person or heterodiegetic narration in “Hills.” As a result, in UFO or the Devil, as in the first-person accounts by McEwan, Sebald, and Levi, there are opportunities for the younger experiencing-I to appear as what Goffman would call a figure and what Emmott (1997) would term an enactor, that is, another, earlier version of the speaking agent producing a narrative or of a character included in it. In Monica’s narrative, the figure is a younger version of the self whose words the older narrating-I animates in the here and now of the storytelling situation (cf. lines 18, 25, 31, and 49). Hemingway uses a third-person narrator who is removed from the events of the storyworld – in Genette’s ([1972] 1980) terms, a extradiegetic-heterodiegetic narrator. This narrating agent does not at any point appear in the represented scene and thus cannot take up any sort of footing toward an earlier self constructed as figure or enactor. By contrast, Monica’s self-representation creates possibilities for alignment (and disalignment) between narrator and enactor that do not arise in Hemingway’s text. Indeed, as I discuss in more detail in my account of positioning theory in the next main section of this chapter, Monica uses a variety of expressive resources to signal her alignment with her earlier self and against the traditional or received wisdom animated by the grandmother.
Turning from production formats to participation frameworks, note the commonalities and contrasts between Hemingway’s scene of talk and the analogous though briefer scene represented in lines 48–55 of Monica’s story – that is, the scene in which (as Monica recounts it) Monica and Renee engage in a dispute with Renee’s grandmother concerning the exact nature of their experiences. Though the process of “ratification” is different in each case, both readers of Hemingway’s story and also readers of the transcript (or listeners to the tape-recorded version) of UFO and the Devil are ratified albeit unaddressed participants. Readers of “Hills” are ratified participants because of the conventions surrounding fictional texts, which specify that reading an account of a conversation among fictional characters does not constitute eavesdropping. For her part, Monica consented to have her discourse recorded as part of a larger sociolinguistic research project designed to gather information about the linguistic and cultural traditions of multiple communities across the state of North Carolina.7 Monica’s informed consent is what makes analysts of this recorded interaction ratified participants in the communicative situation. Indeed, academic research institutions now have strict protocols for research on human subjects – protocols designed, in effect, to prevent “eavesdropping” from becoming an institutionalized data-gathering technique.
Finally, in the case of graphic novels such as Ghost World, the multimedia profile of these narratives sometimes entails quite complex production formats, constellations of authorial agents of a sort more characteristic of movies than print texts. Thus, whereas Daniel Clowes created both the text and the drawings in Ghost World, in Watchmen(Moore, Gibbons, and Higgins 1987), a graphic novel which won the 1988 Hugo Award for Achievement in Science Fiction (in the “Other Forms” category), Alan Moore wrote the text, Dave Gibbons served as illustrator and letterer, and John Higgins was the colorist. All three expressive components of the work – textual content, drawings/lettering, color – bear crucially on the process by which interpreters reconstruct the storyworld associated with this graphic novel, just as film narratives result from the combined efforts of cinematographers, screenplay writers, producers of soundtracks, and other agents of cinematic narration. At the same time, insofar as Watchmen draws reflexively on the types of action-sequences, drawing and lettering styles, and bold color choices used in the superhero comics that it recycles and recontextualizes, this authorial collective, as it might be called, also animates prior narrative conventions, graphic styles, and thematic motifs. Given its focus on two teenage girls trying to navigate the transition from high school to post-high-school life, Ghost World stands out contrastively against the backdrop afforded by this same tradition of superhero comics. Far from possessing superhuman powers, Enid Coleslaw8 and Rebecca Doppelmeyer struggle with familial and romantic relationships, resist the stereotypes their peers try to impose on them, and are brought face to face, on more than one occasion, with the fragility and tenuousness of their own friendship.
Both Ghost World and Watchmen were authored in a context defined in part by prior formal and thematic conventions, resulting in a production format in which the authorial agents orient themselves (as animators) to techniques and themes that pre-exist their narrative designs. Yet it should also be pointed out that the convergence of author and animator/narrator roles that functions as a signpost of factual narrative can also occur in comics and graphic novels. In other words, there is nothing intrinsically fictional about the medium of comics, or any other narrative medium for that matter. Rather than being an attribute of narrative media, the status of a given narrative representation as factual or fictional derives from the semantic profile of its storyworld (is the world evoked by a narrative an autonomous or stand-alone domain, or can accounts of situations and events within it be compared with other, competing accounts?) and also from its pragmatic situation (if my interlocutor cues me to interpret his or her propositions about an imaginary domain as claims about how the world really is, then he or she is lying rather than engaging in the production of fictional discourse). Thus Joe Sacco’s graphic journalism, as practiced in works such as Palestine and Gorazde (Sacco 1994, 2000), exemplifies nonfictional graphic narrative, just as print texts can be the bearers of both factual reports and literary fictions.
For another perspective on how practices of storytelling are grounded in (that is, both shape and are shaped by) particular communicative occasions, I turn now from Goffman’s ideas to a different tradition of sociolinguistic research sometimes referred to as Conversation Analysis. A special focus of this tradition are the speech-exchange systems, or processes of turn-taking, that function differently in (and serve to identify) different kinds of sociocommunicative practices – for example, interviews versus ritualized exchanges of insults versus storytelling.
Conversation Analysis: narrative and the economy of turn-taking
Conversation Analysis is an approach to the study of communicative interaction with roots in the traditions of sociological analysis that emerged from phenomenology (cf. Schutz 1962). Those traditions, pioneered by Harold Garfinkel (1967) under the rubric of ethnomethodology, study how participants display their understandings of an ongoing interaction precisely by making particular kinds of contributions to the course of the interaction itself, and thereby jointly construct it as the kind of interaction that they understand it to be.9 For example, I signal my understanding of an interaction as a conversation versus a formal lecture or a eulogy by performing particular kinds of verbal and nonverbal behaviors in the context of the interchange, which, thanks to my and the other participants’ coordinated performances, becomes (= counts for us as) a conversation.
One of the main ways in which participants display their understanding of a type of interaction, and thereby co-construct that interaction as one that falls within a particular category or kind, is through the methods they use to take turns at talk. In their classic account of “A Simplest Systematics for the Organization of Turn-Taking for Conversation,” Sacks, Schegloff, and Jefferson argued that “[f]or socially organized activities, the presence of ‘turns’ suggests an economy, with turns being for something valued – and with means for allocating them, which affect their relative distribution, as in economies” (1974: 696). As this foundational study revealed, in the case of conversational interaction, as opposed to other types of activity involving speech such as formal debates, professional storytelling performances, or classroom lectures, there is a preference for small turns. Speakers want to produce their utterances in such a way that the risk of being interrupted by others prematurely is minimized, whereas other parties to the conversation are looking for the first available opportunity (what Sacks, Schegloff, and Jefferson [1974] termed “transition relevance places”) to make their contributions to the discourse. Hence, pressure for smallest possible turn size comes from two directions simultaneously: from the direction of the current speaker, who wishes to complete his or her turn, and from the direction of (potential) next speakers, who monitor the ongoing discourse for cues that it is now possible for them to take a turn at talk – in other words, to self-select as next speaker. In consequence, as Schegloff (1981) stresses in a later study that builds on the “Simplest Systematics,” extended discourse in a conversational setting is not tantamount to activity on the part of the speaker and passivity on the part of an interlocutor (or group of interlocutors). Rather, given that participants in conversation must jointly overcome the bias toward smallest possible turn size to produce multi-unit turns at talk, extended discourse productions (including narratives) should be viewed as an interactional achievement.
The process of telling stories thus helps constitute communicative occasions of a particular sort, organizing the turn-taking behavior of the parties engaged in the production and interpretation of narratives – whether in the context of face-to-face interaction or that of viewing a film or reading a literary fiction or a graphic novel. In some storytelling contexts, at least, a hallmark of narrative is an overall preference for the current speaker’s turn at talk to continue, and a dispreference for potential next speakers to truncate that turn by self-selecting. That said, as a number of story analysts have shown, in other storytelling contexts collaborative telling or co-narration is an accepted, even expected, practice (see, e.g., Georgakopoulou 2007: 50–6; Norrick 1992, 2000, 2007; Ochs and Capps 2001; Ochs et al. 1992). Thus, members of a family or other social unit often gain a sense of cohesion or shared group membership through processes of co-telling or co-rehearsing already well-known narratives, which are co-narrated precisely for this purpose of signaling or confirming membership in the group. In any case, both in relatively more monologic and relatively more dialogic (or shared) modes of storytelling, narratives require a dovetailing of sequencing strategies by interpreters as well as producers of the discourse. All parties must actively enable the production of the narrative through a coordinated sequence of behaviors performed and behaviors withheld. Interpreters of spoken narratives make storytelling possible by refraining from taking turns at crucial moments, co-producing the story either by letting a single teller relay the narrative or by intervening at strategic points to collaborate in the process of narration. For their part, readers likewise make narrative possible by an analogous recentering of their attention, at strategic points in time, on a discourse not their own. As Mary Louise Pratt puts it, in literary narrative as well as some forms of “natural” narrative (those in which no co-narration is involved), the role structure of participants in the speech situation remains similarly marked vis-à-vis “the unmarked situation among peers, in which all participants have [in principle] equal access to the floor” (1977: 113).10
Readers of “Hills Like White Elephants” and Ghost World, like Monica’s interlocutor(s), assume the role of an audience ceding its floor rights to discourse producers who must as a result live up to “increased expectations of delight” (Pratt 1977: 116).11 In other words, a common feature cutting across these storytelling situations, and helping constitute them as narrative occasions in the first place, is their reliance on turn-taking procedures distinct from those used in other, non-narrative systems for speech exchange. In stories told in face-to-face interaction, narrators can use abstracts like that contained in the first two lines of Monica’s account to pre-announce their intention to tell a story. Likewise, producers of written narratives can issue requests for “floor space” by using a variety of paratextual as well as textual means. Paratextual cues include the publication of a text in a volume containing other texts labeled as narratives, such as The Complete Short Stories of Ernest Hemingway, or in a graphic novel that includes an array of images among which readers are prompted by the context to assume sequential, narrative-based connections. Textual cues include formulaic openings such as “Once upon a time,” as well as openings that include noun phrases with definite articles (and demonstrative pronouns) right from the start, as in the first two sentences of Hemingway’s story: “ The hills across the valley of the Ebro were long and white. On this side there was no shade and no trees and the station was between two lines of rails in the sun” (Hemingway [1927] 1987: 211, emphases added).12 As I discuss more fully in chapter 5, these sorts of patterns economically evoke the storyworld (or “text world” in Werth’s [1999] terms) to which readers of a fictional text must imaginatively relocate if they are to interpret referring expressions (the hills, the station, etc.) and deictic terms (this side) properly – mapping them onto the world evoked by the text rather than the world(s) that the text producer or the text interpreter occupies when producing or decoding these textual signals.13
Further, some written narratives originating from cultures in transition from oral to literate practices demonstrate particularly clearly how producers and interpreters of literary narrative are caught up in a sociointeractional situation that remains anchored at essential points to the communicative dynamics of face-to-face storytelling (Herman 2001b, 2004). For example, the first word of the Old English epic Beowulfis Hwæt – literally, “Listen” – suggesting a hybridized narrative situation. Here written language is used to issue a request for the floor but in a manner that harks back to traditions of spoken storytelling, in which a narrator might indeed have to ask his or her interlocutors to “listen up.”14 What is more, Beowulf itself foregrounds ways in which storytelling shapes contexts of social interaction. When Unferth and Beowulf publicly recount dueling narratives about Beowulf’s swimming contest with Breca (Beowulf 1993: 33–4), with Unferth casting doubt on Beowulf’s abilities by way of a verbal challenge known as a flyting (Clark 1990: 60; Clover 1980), the poem highlights how telling a story involves not a monologic speech act but a coordination of verbal as well as non-verbal activities on the part of multiple participants.15 Thanks to the surrounding social context in which their exchange unfolds, Unferth’s narrative functions to impugn Beowulf’s valor, whereas Beowulf’s counternarrative reasserts his bravery and obliges him to live up to the heroic self-image presented in his version of the story. In the world portrayed in the poem, then, these stories take on particular forms and functions because of the situation in which they are told and the identities of their tellers. Reciprocally, the alternation between story and counter-story affords protocols used by the participants to make sense of the situation at hand. Stories and storytelling not only emerge from communicative interactions structured as narrative occasions, but can at another level represent the process by which social interchanges assume a narrative profile – and with what consequences and effects.
Likewise, the sample narratives included in the Appendix represent (at a second-order or metacommunicative level) the way storytelling is embedded in communicative contexts. The narratives show how extended turns at talk required to tell a story must be slotted into those contexts in ways jointly negotiated by participants, with the stories themselves shaping such negotiations as they unfold. Hemingway’s story, for instance, suggests that the possibilities for narration are constrained by the exigencies of the characters’ current circumstances, whose profile might very well be altered by acts of narration Jig and the male character could conceivably produce but do not. Thus in the lines marked with arrows in the excerpt below, Jig twice uses questions to invite the male character to produce an account that puts their current situation in a larger temporal context. When he proposes only a very minimal account in response, Jig herself produces an ironically truncated narrative about the happiness of other women who have opted to terminate pregnancies under similar circumstances.
“I’ll go with you and I’ll stay with you all the time. They just let the air in and then it’s all perfectly natural.”
→ “Then what will we do afterwards?”
“We’ll be fine afterwards. Just like we were before.”
→ “What makes you think so?”
“That’s the only thing that bothers us. It’s the only thing that’s made us unhappy.”
The girl looked at the bead curtain, put her hand out and took hold of two of the strings of beads.
“And you think then we’ll be all right and be happy.”
“I know we will. You don’t have to be afraid. I’ve known lots of people that have done it.
→“So have I,” said the girl. “And afterwards they were all so happy.”
“Well,” the man said, “if you don’t want to you don’t have to.
I wouldn’t have you do it if you didn’t want to. But I know it’s perfectly simple.” (Hemingway [1927] 1987: 212–13)
In one sense, the brevity of these turns indexes the male character’s and Jig’s understandings of the kind of interaction they are having – understandings which are not necessarily congruent. Faced with a momentous decision that is also likely to be a turning-point in their relationship, the male character’s contributions to the discourse reveal that he orients himself toward their encounter in the way that participants orient themselves to an ongoing dispute, in which turns are moves in a language-game that involves the staking out and defending of positions. By contrast, Jig’s questions might be construed as probes by means of which she tests the male character’s willingness or ability to shift to a different discourse footing – from that of a participant in a dispute or argument to that of a more cooperative interlocutor who engages in the joint construction of a narrative by means of which both parties might jointly think their way through to a different mutual understanding of their overall situation. From this perspective Jig’s own terse contribution, marked with the third arrow, represents her abandonment of any attempt to co-produce with her interlocutor a story about what her and the male character’s future might look like, and a reversion to more dispute-like turn-taking behaviors – that is, to a construal of the interaction as one falling into the category of an argument or dispute. This is not to say that in other contexts, narratives themselves cannot be used as moves in an argument, as when one attempts to undercut an interlocutor’s current position by telling a story about his or her past conduct (Goodwin 1990). But in the scene of talk represented by Hemingway, the characters’ dispute curtails possibilities for narration, discourse possibilities that, if actualized during the encounter, might have enabled them to reconstrue the encounter itself – indeed, to dissolve the basis for the dispute of which the encounter functions as both a symptom and a contributing cause.
Meanwhile, Ghost World remediates the dynamics of turn-taking through the placement and configuration of speech balloons, text-filled spaces within panels that allow utterances (and sequences of utterances) to be attributed to characters whose appearance, demeanor, and non-verbal actions are simultaneously represented through modes of visual narration, in particular, drawing and coloration (Carrier 2000; Eisner 1996; Ewert 2004, 2005; McCloud 1993). For example, in the first two panels of sequence A included in the Appendix, semiotic cues prompt readers to assume that the characters produce a particular sequence of utterances – a sequence that forms part of the narrative being conveyed by Ghost World but that does not itself represent an act of narration, as it would if one of Clowes’s characters were herself telling a story. In the first panel, the musical notes, the proximity of the represented words to stereo equipment, the Ramones album propped up next to the stereo (the words in the first two interconnected speech balloons are in fact lyrics to the Ramones’ song “Carbona Not Glue”), and the positioning of the three “tails” of the balloons (two tails are pointed toward the stereo speakers and the third toward Enid) all indicate that Enid is singing along with the lyrics of the song. By contrast, the separate speech balloon for the woman represented on the TV screen indicates that she is producing (at some indefinite point during the sequence of words contained in the song lyrics) a non-synchronized speech act. Further, the smaller lettering used to convey the woman’s words suggests that the utterance represented as emanating from the TV is not a primary focus of Enid’s attention.16
Then in the second panel of sequence A, the left-to-right reading conventions associated with English-language texts cue the inference that Rebecca produces what Conversation Analysts would characterize as a “first pair-part” of a structure called an adjacency pair, here a question-answer pair, only to have Enid answer with a question of her own. (Note, too, how the use of boldface text allows Clowes to represent emphasized words – that is, prosodic details of the sort that I have tried to capture through other means in my transcription of UFO or the Devil in the Appendix.) In this instance, the unexpected arrival of Rebecca on the scene, and the unexpected appearance of Enid’s hair (revealed to be green in a subsequent panel in this chapter), proves inauspicious for an extended discourse production, whether narrative or otherwise. Instead, the occasion prompts an exchange of questions designed to put the encounter back on a stable footing after the introduction of startling information (Enid’s green hair, Rebecca’s presence) that requires both participants to rebuild a workable framework for interaction.
By contrast, in sequences C and D (panels that are part of the same chapter of the novel but separated by intervening material not reproduced in the Appendix), Clowes does represent communicative situations favorable to (and structured by) storytelling on the part of the characters. Representing events that transpire on different narrative levels, these panels show Rebecca and Enid on the telephone (note the jagged tails of the speech balloons that indicate contributions by the party on the other end of the phone in a given panel); Enid requesting a story from Rebecca; and Enid producing her own story when Rebecca demurs – a story in which Enid appears as a figure (in Emmott’s [1997] terms, an enactor) sitting in a restaurant booth telling a story to Enid’s and Rebecca’s mutual acquaintance, Naomi. Whereas the panels in sequence A represent a type of communicative interaction in which the participants must focus on establishing a basis for further talk, sequences C and D portray scenes in which that basis has already been accomplished. At the first or primary level (termed the “diegetic” level by narratologists), the scene is a drawn-out, late-night phone call between close friends (one of the panels in sequence D shows a blank TV screen and a clock that reads 1:41, the lack of a picture on the TV suggesting that the time is 1:41 in the morning rather than the afternoon). At the second, embedded (= hypodiegetic) level, the scene is a storytelling situation about which Enid in turn tells Rebecca story, a conversation among peers in the context of a shared meal. Both encounters afford opportunities for more extended discourse productions of the sort Enid initiates at both narrative levels, with both Rebecca and Naomi (at their respective levels) enabling these productions by refraining from taking turns when, in principle, it might have been possible for them to do so.
Reciprocally, readers’ assumption that storytelling scenarios are in play at multiple levels shapes how they interpret Clowes’s representation of the characters’ talk in this part of the novel. Thus in sequence C the bounded rectangular text-box containing the words “No, I mean my story...I can’t keep yours straight . . .” can be interpreted as a response to a question from Rebecca (“You told her about me and Martin?”) at the primary or diegetic level, not a turn taken during Enid’s exchange with Naomi in the earlier time-frame. This way of representing the characters’ speech is followed by panels containing unbounded lines of text above depicted scenes; in this way, Clowes suggests that Enid is narrating to Naomi events from a still earlier time-frame preceding their conversational interaction in the restaurant booth. Hence, just as the characters use storytelling practices to frame their encounters with one another as particular kinds of interaction, regulating their communicative conduct accordingly, readers of Clowes’s text draw on their understanding of the protocols for narration to form inferences about the relationship between elements within panels and about the sequential links across different panels.
More generally, in face-to-face interaction as well as other communicative contexts, stories at once emerge from and organize particular modes of practice, enabling participants to collaborate on the accomplishment of extensive, multi-unit discourse productions. Allowing interlocutors to overcome conversation’s interactionally motivated bias toward the smallest possible turn size, stories told face to face promote the creation of carefully structured, pre-planned discourse segments – stretches of talk such as UFO or the Devil, whose production and interpretation require participants to reflect on and evaluate previous, ongoing, or possible experiences. In this way, storytelling contributes primordially to the sense-making activities jointly accomplished by social interactants, affording one of the basic means by which people make sense of themselves, one another, and the world (Herman 2003a). Written and multimodal narratives like Hemingway’s and Clowes’, respectively, build on this same legacy of sense-making, and do so by exploiting medium-specific properties of printed texts (including graphic-novel texts). Because of their deliberate or “worked over” nature in contrast with the relative spontaneity of spoken discourse (Chafe 1994), as well as the longer span of time allowed for interpretation of printed narratives than for stories told face to face, texts like those of Hemingway and Clowes can use scenes of represented storytelling to comment reflexively on the processes of saying and interpreting at work in narrative occasions, among other environments for communication.
My next section draws on ideas from social psychology – specific-ally, the theory of positioning – to outline another approach to the situated nature of narrative representations. This work, too, suggests that narratives emerge from sociointeractional contexts on which their telling in turn has a shaping effect; but it deploys a different descriptive vocabulary than that used by Goffman or the Conversation Analysts.
In the terms afforded by positioning theory (Harré and van Langenhove 1999; cf. Bamberg 1997b, 2004a, 2005), a method of analysis proposed by researchers working in the subfield of social psychology known as discursive psychology (cf. Edwards 1997, 2006; Edwards and Potter 1992; Harré 2001; Harré and Gillett 1994; Harré and Stearns 1995), speech acts are used to assign positions to social actors. Positions, in this model, are places along scales or continua that correspond to polarities of character such as “strong versus weak,” “flashy versus understated,” etc. Over time, self- and other-positioning speech productions help build overarching storylines in terms of which we make sense of our own and others’ doings. Reciprocally, those overarching narratives provide the means for linking particular position-assignments with particular utterances, as when a snide or affirming remark about someone does its work thanks to the way it shores up (or undercuts) a larger story about that person. Positioning theory thus provides another way of characterizing as a basic element of narrative its grounding in contexts for communication. The telling of narratives functions to position both teller and recipient, and in some cases to contest positions associated with competing storylines, while conversely individual speech acts contribute to the formation of more or less convergent or conflicting storylines about self and other. In addition, the process of narration positions characters in storyworlds.17
Positioning in UFO or the Devil
This section draws on research in discursive psychology to characterize narrative occasions in terms of positions – with stories both allowing people to assign positions to themselves and one another and also emerging from that same process of position assignment.18 In Harré and van Langenhove’s account (1999: 1–31), one can position oneself or be positioned in discourse as powerful or powerless, admirable or blameworthy, etc. In turn, a position can be specified by characterizing how a speaker’s contributions are taken as bearing on these and other “polarities of character” in the context of an overarching storyline – a narrative of self and other(s) being jointly elaborated (or disputed) by participants, via self-positioning and other-positioning speech acts. Hence positions are selections made by participants in discourse, who use position-assigning speech acts to build “storylines” in terms of which the assignments make sense. Reciprocally, the storylines provide context in terms of which speech acts can be construed as having a position-assigning force. It should also be pointed out, though, that self- and other-positioning acts are not always intentionally or volitionally performed. An utterance I produce may allow others to position me in ways I neither planned for nor desire – as when Renee’s grandmother uses the girls’ report about their experiences to position them as unreliable narrators (see lines 48–55 and below). Conversely, I may position another person in unintended ways when I produce utterances that connect up with (reinforce, undercut) storylines of which I am unaware, as when I compliment someone on his or her punctuality in the presence of others who have constructed a larger narrative about that person’s obsessive concern with being on time.
In UFO or the Devil, Monica engages in self-positioning via speech acts concerning racial as well as generational polarities. As suggested above, the abstract of Monica’s story (lines 1–2) functions to situate Monica and Renee within a complex network of ethnic identities. By referring to herself and her friend in terms of “our black asses,” Monica on the one hand can be heard as claiming for herself an identity that is based on skin color and that, in this respect, stands in polar opposition to the identity “white” – even though the composite ethnic heritage of black Appalachians as a group undercuts dichotomous (self-)identifications of this sort (see the background on Texana provided in the Appendix). At the same time, as Mallinson (2006) has shown, Monica is one of the residents of Texana who associates herself with urban black culture and language practices. From this perspective, Monica’s abstract can be interpreted as a means by which she aligns herself with a distinct subgroup of the broader African American population – one that is not immediately present in Texana itself, but that nonetheless constitutes a point of reference for Monica’s strategies for self-presentation. In either interpretation, Monica’s abstract can be construed as a positioning strategy: on the one hand, by positioning Monica and Renee as part of a proximate minority community vis-à-vis the dominant local culture of a county that is more than 98 percent white; on the other hand, by positioning both girls as members of another, larger, and spatially nonproximate minority community vis-à-vis a supraregional culture that is also predominantly white. In either case, the abstract marks the story about to be told as a counternarrative opposed to the narratives circulating within and defining the majority white culture. Yet the minority status that Monica attributes to herself here is, in effect, a double-edged sword: even as it prepares the way for a recounting of experiences to which members of the dominant communities may not have access, because of their association with master narratives or the normative order of discourse, Monica’s counter-narrative also positions her as a kind of self whose experiences may not carry weight or authority when juxtaposed against such master narratives and the assumptions and expectations that they entail.19
This dialectical logic of positioning, whereby authoritative experiential (or “firsthand”) knowledge opposes itself to the discourses that undermine the self’s claims to such authority and such knowledge, also structures the polarity that Monica sets up between herself and Renee’s grandmother toward the end of UFO or the Devil (lines 48–55). The grandmother is represented as dismissing Monica’s and Renee’s experiences by constructing a storyline in which those experiences are in reality the deluded imaginings of overexcited, possibly hysterical, young girls. Specifically, the grandmother attempts to other-position Monica and Renee as unreliable narrators by proposing instead of their supernatural account a naturalistic explanation of the apparition as a formation of “minerals,” distorted somehow by the girls’ own overheated condition (lines 50–4). In turn, however, Monica uses the expressive resources of talk, including prosody,20 to discredit the grandmother’s purported explanation, that is, the storyline according to which Monica’s mind has merely fabricated the big ball. For one thing, Monica uses a slower rate of speech and heightened pitch and volume for purposes of emphasis in lines 46 and 47, where she underscores the effect on her and Renee of the encounter with the big ball. Further, Monica manipulates both pitch and rhythm to construct dismissive, sing-song-like reproductions of the grandmother’s discourse in lines 51 and 53–4; here the downward shifts in pitch are rhythmically timed to co-occur with words that index the grandmother’s purported explanation(s) of events, suggesting the extent to which Monica disfavors and seeks to distance herself from any such account. At the same time, Monica uses nonce words (bah bah ↓ bah ↓ bah) in line 53 to suggest how, in general, the grandmother’s account is discourse devoid of relevant semantic content. She also produces, in line 55, an explicit evaluation (Bullshit) of her interlocutor’s counternarrative about what must have happened. Monica uses a word-internal downward shift in pitch in Bullshit, together with sentence-final intonation at the end of this line of the transcript. Whereas utterances ending with rising pitch (e.g., questions) can be used to implicate various kinds of uncertainty (Ward and Hirschberg 1985), Monica’s utterance in line 53 suggests both prosodically and lexically that she is committed to the truthfulness of her own account in contrast with the grandmother’s.
Thus, exploiting a variety of expressive resources available to participants in face-to-face interaction, Monica other-positions the grand-mother’s discourse as a monolithic voice of authority that in fact has no authority when it comes to this domain of experience. The storytelling process entails a complex embedding or lamination of self- and other-positioning acts, of a kind that narratively structured discourse environments are uniquely able to create and sustain. At a global level Monica uses her narrative as a means for self-positioning even though – or rather, precisely because – at a local level it recounts another person’s attempt to other-position Monica and Renee. Monica embeds a report of Renee’s grandmother’s other-positioning speech act within her own account, critically evaluates it, and thereby puts it in the service of a story about the power of firsthand experience to trump received accounts of the way the world is.
Monica’s mode of narration also positions her interlocutors vis-à-vis the (inter)action unfolding within the storyworld (cf. Bamberg 2004a, 2005). The narrative relies heavily on what narratologists would term internal focalization: once the story is launched, events are refracted through the vantage-point of an individuated participant in the story-world, namely, the younger experiencing-I who undergoes the encounter. That said, Monica uses discourse resources available in the here and now (including tense shifts, deictic references, and prosody) to animate earlier events, whose life-transforming impact thus emerges through the interplay between different time-frames. The same dual positioning is marked lexically in line 55. At this juncture, Monica’s evaluation of the action is ambiguously external and internal, in Labov’s (1972) sense of those terms: Bullshit marks the fusion of the evaluative stance of the younger experiencing-I with that of the older narrating-I. Then, in lines 56 and following, the focalization shifts unambiguously back to the vantage-point of the older narrating-I: Monica speaks summatively about how her behavior changed after the encounter with the big ball.
Overall, though, using the current discourse to stage the main action of the narrative from the perspective of the experiencing-I, Monica aligns her interlocutors (and analysts) with her younger self’s vantage-point. Relevant here is Dorrit Cohn’s (1978) account of the discourse strategy that she characterized as consonant self-narration, where the older narrating-I does not enjoy any cognitive privilege with respect to his or her earlier experiencing self. As the foregoing remarks suggest, however, the term “cognitive privilege” might need to be reformulated as “direction of flow”: is the discourse organized such that the narrating-I animates the experiencing-I’s perspective on the storyworld, or is it organized such that past events become the means for staging current conceptions of self and world? In the case of UFO or the Devil, by positioning her interlocutors with her younger self Monica in effect positions them against other discourses that might claim authoritative status – discourses such as those represented by the grandmother. The repositories of received wisdom, these discourses purport to invalidate the experiences whose formative role Monica’s narrative, by contrast, enacts.
Positioning in “Hills” and Ghost World
Likewise, drawing on the expressive resources of a different medium for storytelling, Hemingway’s mode of narration in “Hills” is communicatively situated in ways that positioning theory can help account for. To reiterate, positioning is a relevant parameter for analysis on several levels: the level of the characters; the level of the reader’s engagement with the text, given the specific narrative techniques deployed; and the level of narrative’s bearing on more or less dominant storylines, or master narratives, about the way the world is. At the first level, “Hills” portrays the unnamed male character and Jig as engaged in both self- and other-positioning acts. At the second level, Hemingway’s mode of narration – in particular, his use of what F. K. Stanzel ([1979] 1984) would characterize as the figural narrative situation (= third-person or heterodiegetic narration in which events are refracted through the vantage-point of a particular consciousness or “reflector”) – positions interpreters of the overarching narrative concerning the characters’ encounter. And at the third level, the story engages with master narratives associated with gender roles in particular.
Although there are shifts of perspective over the course of the story, the male character functions as the main internal focalizer or reflector figure, whose vantage-point provides a window on the action being recounted. Thus, almost all of the nonverbal actions recounted in the story are performed by Jig, confirming that the dominant reflector or perceiver (the one witnessing Jig’s actions) is the male character. The use of the male character as the focalizer tends to align the reader with his vantage-point: we literally see things through his eyes. Yet, for reasons already mentioned, the text ultimately invalidates both the storyline he proposes and his attempt at other-positioning Jig in terms of that storyline. What the male character presents as empathy and concern for Jig can be read otherwise, as a self-interested attempt to sidestep what he views as onerous domestic obligations. The tension between (1) the male character’s status as the main source of perceptual information about the storyworld and (2) his self-centered approach to his and Jig’s situation creates a kind of dissonance in the positioning logic of the narrative. Hence, whatever Hemingway’s conscious stance toward dominant conceptions of gender in the epoch during which he wrote, his text positions readers simultaneously with and against the male character, thereby disrupting sexist master narratives in which men are the repositories of authoritative knowledge and sound judgments while women lack these attributes and are therefore unreliable.21 More generally, to reconstruct the storyworld in “Hills” readers must draw inferences about the larger communicative situation in which Hemich Hemingway’s textual cues are embedded and which the story in turn represents, through its portrayal of the interaction (or “scene of talk”) involving Jig and the male character. The larger communicative situation at issue is thus complex and multi-layered; in it, readers are positioned vis-à-vis the male character’s attempts to position both Jig and himself.
In the case of Ghost World, the graphic-novel medium affords still other expressive resources by means of which interpreters of the text can be positioned – and through which, in the storyworld evoked by the narrative, characters’ own attempts at self- and other-positioning can be represented. Likewise, aspects of the text serve to position Clowes’s account vis-à-vis dominant storylines or master narratives circulating in the culture at large. Consider the positioning logic at work in sequence B, for example. The verbal-visual organization of the sequence aligns readers with Rebecca and Enid, while distancing them from the backgrounded male characters about whom the two friends converse (or argue). Clowes deploys here the multimodal equivalent of a print text’s use of third-person or heterodiegetic narration that moves along a spectrum from relatively more external to relatively more internal views – that is, from external focalization, where the vantage-point on events is not associated with a character in the storyworld, to internal focalization, where the vantage-point is in fact a character’s.22 For instance, in the third panel of sequence B, readers can use the communicative context established by the visual design of first two panels as a basis for drawing an inference concerning the status of the image represented in this panel. Specifically, they can infer that this image of the former bass player is mediated through the perceptions of one of the two main characters – most probably Rebecca, given her physical location and the orientation of her torso and gaze in the preceding panel. That inference is reinforced by the absence of a speech balloon in the third panel, even though the bass player is shown talking on the phone. Readers can assume that, because of the male character’s location at the far side of the restaurant, Rebecca cannot hear what he is saying on the phone. By contrast, in the case of the (self-incriminating) utterance that is represented by means of a speech balloon in the second panel, readers can assume that this remark (“You guys up for some reggae tonight”) was made within Rebecca’s and Enid’s perceptual range and is therefore included in the report of their perceptions at this point in the unfolding action.23 Both the design of individual panels and sequential links among panels thus align readers with particular vantage-points on the storyworld, and prevent or at least inhibit other identifications and alignments.
The net result of this logic of identification, enacted through the progression of the text’s word–image combinations, is that readers are prompted to construct a storyline concerning Rebecca’s and Enid’s difficulties with identifying potential romantic and life partners – and, by extension, a larger storyline about the similar difficulties that any intelligent, independent-minded young woman is likely to experience in this connection. The verbal and visual details found in individual panels and panel sequences cue readers to attach local textual details to this (emergent) storyline, while that storyline in turn provides context for interpreting the actions, postures, and speech productions of characters represented within a given panel or across panels. Further, as the screenshots also included in the Appendix suggest, Terry Zwigoff’s film version of Ghost World picks up with and indeed amplifies this same global storyline. Screenshot 1 reproduces Rebecca’s and Enid’s perspective on the reggae-lover’s self-incriminating remark. Meanwhile, portraying scenes not included in the graphic-novel version of Ghost World, screenshots 4, 5, and 6 bolster the storyline concerning the dearth of potential romantic or life partners – given the unpromising conduct and demeanor of the patrons registered by the camera (which corresponds here to Enid’s gaze).
At the same time, sequence B exploits the modulation between relatively more internal and relatively more external perspectives to present alternating views of Enid’s and Rebecca’s table as the primary vantage-point on the situations, events, and characters inhabiting the storyworld. Panels 4 and following prompt readers to pull back from the internalized view of the ex-bass player in panel 3 and adopt shifting perspectives during Rebecca and Enid’s dispute concerning what Rebecca characterizes as Enid’s impossibly high standards for men. (Apparently, only a “famous cartoonist” named “David” Clowes would be up to snuff!) In a manner reminiscent of the shot/reverse-shot technique in cinematic narratives, the text first provides, in panel 4, an over-the-shoulder view of Rebecca from Enid’s perspective, followed in panel 5 by an over-the-shoulder view of Enid from Rebecca’s perspective. Then in panel 6 the perspective shifts again, to a more externalized view that captures Enid’s angry expression as she defends her preference for the cartoonist over the “guitar plunkin’ moron” (= ex-bass player) whom Rebecca had alluded to favorably. By showing both Enid’s angry reaction and the now discredited male characters in the restaurant, and by attributing to Rebecca the utterance “Still, I just hate anybody who likes cartoons,” panel 6 aligns readers with Enid’s position, fracturing the global storyline concerning the lack of viable male partners into competing storylines about life choices for young women in Enid and Rebecca’s position. Making sense of both individual panels and panel sequences thus requires situating them in a broader logic of positioning and counter-positioning, thanks to which Clowes’s panels serve particular communicative functions and which in turn takes shape because of how the panels themselves are sequenced.
In short, the idea of positioning, although originally developed by discursive psychologists for the purposes of analyzing everyday, face-to-face communicative interactions, can also throw light on the communicative strategies by which readers are positioned vis-à-vis literary narratives like Hemingway’s and multimodal texts such as Clowes’s. This work suggests that framing inferences about the stance one is being prompted to adopt toward particular positions represented in a narrative is a fundamental part of the process of reconstructing that narrative’s storyworld, no matter what the medium in which it is presented. Reciprocally, concepts and methods originating in narratology, such as internal focalization, consonant self-narration, and the contrast between authorial and figural narrative situations, can lead to finer-grained analyses of positioning logic, as can research on modes of narration made possible by the medium-specific properties of comics and graphic novels, for example (cf. Baetens 2002; Carrier 2000; Eisner 1996; Ewert 2004, 2005; Groensteen 2007; McCloud 1993).
I turn now to a third broad approach to the study of narrative occasions – one “homegrown” within the field of narrative theory itself. In parallel with Goffman’s work on production formats and participation frameworks but using a different analytic scheme, this third, indigenous, approach aims to identify the parties to narrative transactions and how different relations among them affect the process of telling and interpreting stories, which can in turn affect how these parties orient to one another.
The Narrative Communication Model
In this final section of the chapter, I discuss how narratologists have created their own indigenous vocabulary for describing and analyzing narrative occasions. This approach, sometimes referred to as “the narrative communication model” (Booth [1961] 1983; Chatman 1978; Genette [1972] 1980, [1983] 1988, [1991] 1993; Herman and Vervaeck 2005a; Leech and Short [1981] 2007: 206–30); Iser 1974; Phelan 2005a; Prince 1982; Rimmon-Kenan [1983] 2002; Shaw 2005), focuses on the constitutive factors of narrative communication and explores how those factors come into play differently in different kinds of storytelling situations. In the version developed by rhetorical theorists of narrative, the approach distinguishes among actual authors, implied authors, and narrators on the production side of the storytelling process, and, on the interpretation side, the corresponding roles of actual readers, (types of) implied readers, and narratees (the audience implicitly or explicitly addressed by the narrator in the text). According to this model, telling and interpreting narratives must be situated within a multi-layered process of narrative communication, in which an implied author, for example, might communicate something to an implied reader by having a narrator tell a particular kind of story in a particular way to a specific narratee. Reconstructing a storyworld thus requires framing inferences about the dynamically unfolding relations among the parties to a given narrative transaction – relations that are both an emergent result of textual designs and a basis for understanding how those designs form part of larger narrative occasions.24
The narrative communication model has roots in both structuralist narratology (Chatman 1978; Genette [1972] 1980; Prince 1973) and the rhetorical approach to narrative pioneered around the same time by Wayne C. Booth ([1961] 1983). As discussed in chapter 2, the narratologists, influenced by Saussure’s emphasis on the linguistic system versus particular speech acts made possible and intelligible by that system (Saussure [1916] 1959), focused not on particular authors and readers but on the system of structural options available to producers as well as interpreters of narrative texts. This general approach led narratologists to create a taxonomy of narrators and narratees, agents of narrative production and reception about which inferences can be formed based on features immanent to the text. For his part, Booth took issue with then prevalent Formalist (New Critical) approaches to literary analysis and conceived of fictional works not as autonomous artifacts that had to be bracketed from their contexts of production and interpretation to be understood but rather as elements of a purposeful communicative process in which authors and readers participate, but in ways that are mediated by the rhetorical designs that structure the text.
Narratological foundations
In the narratological framework, narration can be conceived as a communicative process in which information about the story level is conveyed by a particular kind of narrator to a particular kind of narratee. Gerald Prince defines the narratee as “[t]he one who is narrated to, as inscribed in the text” ([1987] 2003: 57), and contrasts this participant role with both the real reader and the implied reader (see below). Hence, in contrast to the biographical critics who pre-dated them or the reader-response theorists who came later, the narratologists focused not on authors and readers and their role in narrative transactions but on narrators and narratees viewed as communicative positions correlated systematically with identifiable textual markers. In other words, whereas authors and readers are dimensions of “narrative parole,” particular instances of narrative discourse, the narratologists’ main concern was narrative langue, or the underlying semiotic system that makes narrative production and understanding possible in a given case. Hence, adapting folk models of the communicative process, based on the three components or participants of sender, message, and receiver –
sender → message → receiver25
– the narratologists developed an analogous model of the participants involved in the process of narrative communication:
narrator → narrative message → narratee
As noted by Rimmon-Kenan ([1983] 2002: 95–106) in her exposition of Genette’s ([1972] 1980) foundational work, narratologists have classified narrators as well as narratees according to narrative level, extent of participation in the storyworld, and degree of perceptibility; further, under the influence of rhetorical approaches to narrative that can be traced back to Booth’s work and that I discuss more fully in my next subsection, degree of (un)reliability has come to constitute an additional parameter for comparing and contrasting narrators in particular.26
With respect to narrative level, narrators are extradiegetic if they do not inhabit the storyworld evoked by their discourse (as in Hemingway’s text); intradiegetic if they are characters within the storyworld and tell a story within the story (as in lines 48–55 of UFO or the Devil, when Monica narrates the narrative that she and Renee told to the grandmother); or hypodiegetic if, within an embedded narrative, a character narrator tells yet another story (as in sequences C and D from Ghost World, where within Clowes’s narrative Enid tells Rebecca a story about how she told her loss-of-virginity story to Naomi on a previous occasion). Whereas Hemingway’s story does not feature an explicitly characterized narratee, that is, a textually evoked recipient of the narrator’s discourse, successful interpretation of Monica’s narrative and of Ghost Worldrequires sorting out who is the narratee at a given point. Thus, I will misconstrue UFO or the Devil if I fail to realize that two sets of narratees are in play in lines 48–55: on the one hand, the two fieldworkers to whom Monica is conveying her account (and to whom she addresses the three instances of the locution you know27 found in these lines); on the other hand, the grandmother, to whom in the earlier time-frame being recounted here she and Renee tell their story about the encounter with the big ball. Not only do interpreters need to keep distinct these different narratees; what is more, they can infer that Monica is communicating something to the fieldworkers (the extradiegetic narratees) by constructing the grandmother as an unsympathetic audience (intradiegetic narratee). Likewise, in Ghost World, readers must sort out who is being addressed at what moments in Enid’s account, and frame inferences about (1) what Enid seeks to communicate to Rebecca by telling her the story of her narration of her loss-of-virginity story to Naomi, and (2) what Clowes seeks to communicate to the reader by having Enid attempt to communicate this to Rebecca.
Shifting from the question of narrative levels to that of extent of participation in the story, analysts have drawn on the narrative communication model to distinguish among autodiegetic, homodiegetic, and heterodiegetic narrators. Autodiegetic narration constitutes a special case of first-person or homodiegetic narration in which the narrator does not only participate in the action being recounted but is also the main character in the storyworld evoked by the text. To put the same point another way, a homodiegetic narrator is one who has participated (more or less centrally) in the circumstances and events about which he or she tells a story, with completely central participation yielding the autodiegetic mode.28 In heterodiegetic narration, by contrast, the narrator has not participated in the circumstances and events about which he or she tells a story. Furthermore, narrators can be extradiegetic-homodiegetic, like the older Pip who narrates his earlier life experiences in Charles Dickens’s Great Expectations; extradiegeticheterodiegetic, like Henry Fielding’s narrator in Tom Jones, who comments evaluatively on but does not participate in events in the storyworld; intradiegetic-homodiegetic, as when Marlow, in his role as a character narrator in Conrad’s Heart of Darkness, tells about his experiences in the Congo; or intradiegetic-heterodiegetic, as when in Chaucer’s Canterbury Tales the Miller tells a bawdy tale (specifically, a fabliau) centering on events in which he himself did not take part.29
In these terms, Hemingway’s narrator is extradiegetic-heterodiegetic, whereas UFO or the Devil involves both extradiegetic-autodiegetic narration (when Monica tells the interviewers the story of her earlier encounter with the big ball) and intradiegetic-autodiegetic narration (when she tells the story of how she told a previous version of that same story to Renee’s grandmother, just after the events in question transpired). Ghost World, meanwhile, features extradiegetic-heterodiegetic narration (because the narrating agent who produces the primary diegetic level does not inhabit the storyworld evoked by the graphic novel, and does not figure as a participant in the narrated action30) along with instances of intradiegetic-autodiegetic and hypodiegeticautodiegetic narration (as in sequences C and D). Also, in the film version of Ghost World, Enid is herself a fledgling comic artist, and her sketchbook images of Seymour (a character not included in the original graphic-novel version) constitute further instances of intradiegetic narration, further stories-within-the-story, some of them heterodiegetic in focus.
The broader point to emphasize here is that, as was the case with textual cues prompting interpreters to draw inferences about who is communicating with whom at what point and on what narrative level, recipients of narrative discourses and texts make sense of storyworlds in part by figuring out a narrator’s degree of participation in or involvement with narrated events – how a narrator fits into the overall participant structure of the communicative transaction in which they engage while interpreting a story. Indeed, as Lubomír Doležel (1998) has argued, modes of narratorial participation bear crucially on the process by which circumstances and events in storyworlds are “authenticated” (or not) as fictional facts – as things that must be assumed to be true about the storyworld if one is to build up a coherent interpretation of it. As Doležel puts it, “entities [and events] introduced in the discourse of the anonymous third-person narrator are eo ipso authenticated as fictional facts, while those introduced in the discourses of the fictional persons are not” (1998: 149, quoted in Margolin 2005b: 33; cf. Doležel 1980). In other words, fictional facts reported by third-person narrators have an authority, or mark a degree of certainty, lacking in first-person reports given by characters or character narrators occupying specific positions in a storyworld.31
In sequences C and D from Ghost World, for example, the content of the phone conversation between Enid and Rebecca is authenticated in a way that Enid’s own report of the story she tells to Naomi is not. The phone conversation framing this story-within-the-story can be taken as fictional fact, whereas Enid’s assertion that she did not tell Naomi about Rebecca’s own loss-of-virginity story remains a (disputable) assertion made by Enid in her role as an intradiegetic (or character) narrator. Indeed, this assertion seems to be undercut by the panel that precedes the one in which Enid makes this claim in sequence D; here Naomi is shown reacting (“Oh God . . .”) to Enid’s remark that Rebecca’s “first time was with this completely fruity guy she met on a computer bulletin board!” – as if, in a jump back up to the primary diegetic level on which Rebecca and Enid are having their phone conversation, Enid is recalling what Naomi looked like at precisely the moment when she (Enid) did tell her the story at issue. In the dialectical interplay at work in narrative occasions, then, the truth status of a given narrative act (or narrator’s report) can be determined only in light of the larger context of telling of which it forms a part, and, vice versa, localized acts of storytelling shape interpreters’ understanding of the overall narrative sequence to which they contribute.
Rhetorically oriented theorists have also explored the structure and dynamics of what Doležel called narrative authentication, or the process by which narrative texts prompt interpreters to sort propositions about the storyworld into factual and nonfactual assertions – and conversely to build up a global understanding of the storyworld precisely through that sorting process. As I discuss in my next subsection, though, rhetorical narrative theorists have approached this issue from a different direction, attempting to link the semantics of narrators’ statements about fictional worlds with the pragmatics of audience participation. To construct this link, they have proposed new inflections of the narrative communication model, and thus new ways of grounding narrative itself in communicative processes.
Rhetorical inflections
When it was integrated with the ideas of Wayne Booth and Wolfgang Iser that came into prominence around the same time that structuralist theories of narrative were being developed and disseminated, the basic model of narrative communication acquired additional nuances, as suggested by the form it took in Seymour Chatman’s study, Story and Discourse (1978: 151, reproduced in Rimmon-Kenan [1983] 2002: 86; a slightly simplified version of the diagram is presented here):
Real author ···→ Implied author → Narrator → Narratee → Implied reader ···→ Real reader
In this version of the model, the dotted arrows indicate how individual authors and readers are not in themselves targets of this approach to narrative analysis; rather, they are relevant only insofar as they adopt communicative roles (compare Goffman’s “participant statuses”) in order to take part in narrative transactions. On the production side of things, authors of fictional narratives create a persona that Booth called the implied author and that can be defined as “the governing consciousness of the work as a whole, the source of the norms embodied in the work” (Rimmon-Kenan [1983] 2002: 87– 8). In other words, the implied author can be defined as a “streamlined version of the real author, an actual or purported subset of the real author’s capacities, traits, attitudes, beliefs, values, and other properties that play an active role in the construction of the particular text” (Phelan 2005a: 45); this streamlined version of the implied author can be assumed to be “responsible for the choices that create the narrative text as ‘these words in this order’ and that imbue the text with his or her values” (Phelan 2005a: 216).32
Shaped in part by the New Critics’ strictures against what they called “the intentional fallacy” (Wimsatt and Beardsley [1947] 2001), according to which the attempt to use an author’s assumed intentions as a yardstick for literary interpretation results in a version of the genetic fallacy, or arguments based on origins, Booth’s ([1961] 1983) approach avoids committing the analyst to claims about the intentions of the actual author. Instead, Booth suggested, authors adopt a “second self” (if an author produces multiple works, this persona becomes a “career author”) for the purpose of engaging with readers in communicative acts mediated by texts featuring particular kinds of rhetorical designs. In turn, readers scan fictional texts for these designs – that is, for cues that enable them to pick up on the attitudes, norms, and values associated with the implied author, the persona adopted by the actual author for the purposes of engaging in such rhetorically situated acts of narrative communication. Interpreting a narrative entails searching the text for signals that convey information about these underlying norms and values, which in turn enable recipients to detect favored versus dis-favored character traits, modes and degrees of unreliable narration, and other dimensions of the broader communicative occasion in light of which particular textual details acquire specific narrative functions.
Thus, in a text like Robert Browning’s My Last Duchess, a dramatic monologue told by a Renaissance-era Duke who has had his wife murdered (one assumes) because of his own pathological jealousy and possessiveness, reconstructing the storyworld requires recognizing the gap between the norms and values of Browning’s implied author and those displayed by the Duke in his role as narrator. Or, in the case of Hemingway’s short story, which involves not unreliable narration but rather a clash of attitudinal and practical stances vis-à-vis a conflict-causing turn of events within the storyworld, invoking the concept of the implied author would allow the rhetorical theorist to reformulate in yet other terms some of the interpretations that I have used Goffman’s ideas, accounts of turn-taking strategies in discourse, positioning theory, and narratological work to advance in previous sections in this chapter. For example, I have suggested that Hemingway’s male character frames his utterances such that Jig appears to be the principal for whose sake the utterances are spoken, when in reality they are spoken for his own sake; that he treats his and Jig’s interaction like a dispute in which each turn at talk constitutes a move within a competitive game, whereas Jig makes tentative efforts to move beyond this dispute-like discourse environment to one in which she and the male character can work together to envision a future for themselves; and that there is a tension between the male character’s status as the source of most of the perceptual information in the story and the self-interested storyline that he attempts to foist upon Jig. Recasting these claims in the language used by rhetorical theorists, one could argue that Hemingway’s textual designs indicate a disparity between the norms and values informing the male character’s words and actions and those that contribute to the profile that interpreters assign to the implied author, viewed as the source of the work’s underlying norms and values. Interpreters work to reconstruct this profile in order to make sense of the communicative strategies motivating the use of particular textual choices to represent the characters’ relations to one another in the storyworld, even as the choices themselves provide grounds for this process of reconstruction.33
Reciprocally, the implied reader is Iser’s (1974) term for the communicative role that must be adopted by actual readers if they are to discern the gap between an implied author and an unreliable narrator like Browning’s Duke or a less than ideal partner like Jig’s interlocutor in “Hills” – or, for that matter, between the governing norms and values of Zwigoff’s film version of Ghost World and the male characters who come across as grossly inappropriate candidates for romantic relationships in screenshots 1, 4, 5, and 6 included in the Appendix.34 In the terms afforded by the (rhetorically inflected) narrative communication model, only to the extent that readers occupy the role of implied reader will they recognize the irony in Browning’s treatment of his unreliable narrator, whose words provide at best oblique cues for making sense of what has transpired in the world of the narrative, or the irony associated with Hemingway’s and Zwigoff’s treatment of their respective male characters. In other words, the implied reader is the intended addressee or target audience of the implied author. This reader knows that the Duke’s words are not to be taken at face value, that Jig’s inter-locutor does not necessarily have her best interests at heart, and that Enid would be ill advised to attempt a relationship with any of the male characters glimpsed in those screenshots.
But what is more, rhetorical theorists of narrative seeking to improve upon earlier versions of the narrative communication model have retained the distinction between actual readers and narratees but divided implied readers into two kinds, the authorial audience and the narrative audience, to use terms proposed by Peter J. Rabinowitz ([1977] 1996, 1998). The authorial audience can be described as the implied author’s target audience, the hypothetical reader who is able to pick up on all the norms, attitudes, and values that are inferable (in principle) from every textual design included in a narrative text. By contrast, to the extent that they take up a position within the narrative audience, readers construe as truthful a narrator’s reports concerning what is going on in that world. On this model, further, to engage fully with fictional texts, actual readers have to enter both audiences simultaneously, maintaining an awareness of the characters as serving the larger design of the work even as they get caught up in the characters’ situation, as if they were real-world individuals. The model is meant to explain why actual readers can be “taken in” enough to empathize with the characters and experience curiosity, suspense, and surprise (Sternberg 1990, 1992) on the characters’ behalf, but not so taken in that they jump onto the stage during the performance of a play to “rescue” a character being threatened by a villain, say. As a member of the narrative audience of Clowes’s Ghost World, I experience suspense about how Enid’s and Rebecca’s attempt to make the difficult transition to adult life will turn out, as well as curiosity about the situations and events that have led them to where they are for the time-span covered by the narration. But as a member of the authorial audience, I construe these characters’ utterances and actions as signals designed to cue inferences about the constellation of norms and values informing the storyworld that has been constructed by Clowes. Thus, in sequence B, I recognize that Clowes has used Rebecca and Enid to stake out different positions vis-à-vis the possibility of forming romantic attachments with their male peers, and in this light I read their conversation as a dialectical interplay between the conflicting stances represented by the characters interpreted as theme-bearing textual designs.35
After these refinements are factored in, along with others described by Martin (1986: 154) and Herman and Vervaeck (2005a: 22), something like the following picture emerges:
Sender: author → implied author → dramatized author → (un)dramatized narrator
⇓
Narrative message
⇓
Receiver: (un)dramatized narratee → narrative audience → authorial audience → real reader
A dramatized author uses “I,” unlike the implied author, which is the persona or belief-set adopted by an actual author for the purpose of creating a particular narrative text; a dramatized narrator is an intradiegetic or character narrator, like the younger Monica who, in the account given by her older, narrating self, tells Renee’s grandmother the story about the big ball in lines 48–55 of UFO or the Devil; a dramatized narratee is a characterized recipient of a story, like Rebecca and Naomi (at different narrative levels) in sequences C and D in Ghost World, or like Renee’s grandmother in Monica’s story-within-the-story
Not all of these participants roles will come into play in every narrative transaction; indeed, as already discussed in the first section of this chapter, one of the signposts of fictional narrative is a disparity between author and narrator that does not obtain in the case of nonfictional narratives (biographies, autobiographies, histories, witness testimony in court) (Cohn 1999). But the rhetorical theorist would argue that the communicative roles that are pertinent in a given storytelling context will have the structure indicated here. Admittedly, some ludic, experimental narrative texts toy with this linear arrangement of communicative roles, as when Denis Diderot in Jacques the Fatalist uses an elaborate structure of narrative frames, coupled with techniques of frame-breaking, to create uncertainty or at least hesitation about what sort of communicative agent (dramatized author? dramatized narrator who is himself a character within a higher-level story?) is narrating the tale and in what context. But such disruptive effects become palpable against the backdrop afforded by the expectation that narrative transactions will generally have the kind of structure suggested by the diagram. For example, in a fictional narrative featuring a dramatized narrator (like Enid telling Rebecca the story of how she told her own loss-of-virginity story to Naomi), the default assumption is that this narrator is a textual design created by the actual author rather than the other way around. (The biographical individual named Daniel Clowes conjured Enid Coleslaw and Rebecca Doppelmeyer, but not vice versa.) By the same token, the way a fictional text deploys a dramatized narrator is likely to generate inferences about that narrator’s relation to the implied author. For example, in the case of sequences C and D in Ghost World, Enid comes across as a partly unreliable narrator, given that she appears to tell (that is, to experience a memory of telling) to Naomi Rebecca’s loss-of-virginity story, despite reassuring Rebecca that she did not. In Rabinowitz’s terms, this discrepancy translates into a disharmony between the roles of the narrative and the authorial audiences – between the belief-set readers adopt in order to engage fully with Enid’s story-within-the-story, and the belief-set they adopt in order to make sense of the functions of her embedded narrative within Ghost World as a whole. The rhetorical approach thus suggests that “authentication” (or not) of particular situations and events as fictional facts depends not only on whether they are presented through first-person, third-person, or figural modes of narration, but also on the audience positions those narrative modes invite readers to occupy.
More generally, the approach complements those used by socio-linguists and positioning theorists to study the dialectical interplay between specific textual features and storytelling occasions, narrative texts and the larger communicative contexts in which they are told and interpreted.
The sociolinguistic, discursive-psychological, and narratological approaches reviewed in this chapter stem from quite different traditions of inquiry, and practitioners have up to now worked largely independently – some concerned mainly with natural-language narratives told in contexts of face-to-face interaction, others focusing chiefly on stories conveyed through literary or cinematic art. (A fully developed narratology of graphic narratives, despite the promising beginnings made by scholars such as Baetens 2002, Ewert 2004, 2005, and Groensteen 2007, remains a goal for the future.) Taken together, however, these investigative frameworks provide complementary tools for studying the structure and dynamics of narrative transactions – while also revealing the extent to which common storytelling processes unite narratives presented in different semiotic media. More than this, despite their different disciplinary origins and distinctions among the kinds of texts on which their practitioners characteristically focus, these approaches afford convergent insights into what I have described as the first basic element of narrative, namely, its status as
(i) a mode of representation that is situated in – must be interpreted in light of – a specific discourse context or occasion for telling.
As all of the approaches suggest, although it may be possible to identify the bare propositional content of a storyteller’s utterances without factoring in the context in which his or her narrative is told, detaching the utterances from that surrounding context is like focusing on the semantic content of a compliment without stopping to consider whether it is being said earnestly or ironically (compare What a great guy! said of a humble philanthropist versus someone convicted of stealing medicine from the elderly). I may issue a compliment, or tell a story, either to praise or to shame; and just as storytelling shapes the discourse contexts in which it unfolds, those contexts are what give any story its point or reason for telling.36