Speaking Double

Indexicals in Role-Play and Story-World Negotiation

Analytic Framework

This study will employ Barbara Johnstone’s articulations of “structure” (76) and “indexicality” (133) to look at the emergent qualities of fantasy story-worlds co-constructed by children in play. Specifically, this study will look at the way a narrative structure is produced in the “process of interacting” (101) through conscious identity-shifting, indexed by changes in pronoun use, with story-world action performed as “conversational move[s]” in the interactional world of the children’s talk (103). The study will demonstrate the children’s adherence to and manipulation of the “system constraints” (104) of the game (that is, the use of dolls for characters, the excess of said characters in proportion to the number of participants, and the interactional nature of the story being narrated) and the uniquely “generative” quality of the “rules” of this particular stretch of discourse (125).

This analytic framework will allow for an exploration of the children’s talk as a coherent and cohesive sample of discourse, despite the apparent lack of continuity in the narrative that they construct. Treating their conversation in the context of their play will allow this study to explore the structural implications of a story-world that has been mutually created in interaction, rather than experienced, conceived, or fantasized antecedent to the interactional event and concurrently reported in it. In this way, the children’s play can be understood as a “language game” (103) with unique and significant discursive features of its own.

Data and Method

The sample in question is a recording from the CHILDES corpora at TalkBang.org. There are three primary participants, the children Aaron, Katrina, and Jinny, and three secondary adult participants—the supervisors, Bill and Jenny, and Aaron’s mother Sarah. The majority of the recording consists of the children’s interactions with each other as they play.

The interaction centers around a castle in the children’s play area. Though their game begins in an appropriately medieval setting, the context provided by the castle is only a starting point for the story that will emerge. Indeed, the physical setting itself is less important for the progression of the children’s narrative than the mutually imagined world that they produce and inhabit in their talk and continued interaction. The castle functions as an interactional anchoring point for their play, allowing for the uniquely emergent quality of the story-world as a product of their conversation—which is to say, as a product of the “language game” (Johnstone 103) that this study will refer to as role-play.

Through the lens of Labov and Waletsky’s six-part narrative schema, as described in Johnstone (92-93), this study will analyze the “macrostructure” (92) of the children’s story as it emerges, and attempt to identify congruencies and disjunctures with Labov and Waletsky’s model as it is traditionally conceived. Treating the six narrative elements as flexible “constraints” (104), this study will look at the different ways in which each child relates to the constraints and to the other children through the constraints. This study will analyze the various strategies that each child employs to navigate the landscape of their story-world, to develop the narrative in collaboration with the other children, and to arbitrate conflicts and complications. Specifically, this study will concentrate on the shifting use of pronouns throughout the interaction as a vehicle for mediation between the story and interactional worlds.


Recognizing that, structurally, the data is neither a conversation nor a narrative, but both, it follows that this study will necessarily take a hybrid approach. In regard to Labov and Waletsky’s six “elements of narrative” (Johnstone 92)—abstract, orientation, complication, resolution, evaluation, and coda—the study finds that these elements cannot be identified as discrete segments of the interaction, but rather appear as functions or strategies that the children employ in the construction of their story. However, the sample does display the sequential nature of a narrative, as the interaction includes “at least two” of what Labov refers to as “narrative clauses” (Johnstone 92)—events that cannot be reordered without a change in the meaning of the whole. Furthermore, because the narrative is not reported but instead emerges in interaction, the parameters which normally describe a conversation— “spontaneous” and “casual” (101)—and the implications of such, also apply. Significantly, the majority of conversational moves that the children make are either declarative or performative, occurring within the story-world and facilitating the progression of the narrative. The interactional component of their talk is almost entirely mediated through this narrative, and the conversational moves that the children undertake are, in fact, constitutive of the story-world, creating a space for the imaginative embodiment of the children as characters in the story. For the purposes of this study, then, the data can be described as an example of a hybrid or synthetic form of talk that blends the structure and strategies of narrative and conversation into the composite of role-play.

So, approaching the data from this perspective, coherency and cohesion can be judged not in terms of narrative logic or factual consistency, or even agreement among the participants as to what story they are telling, but rather in terms of persistence. The narrative is coherent and cohesive in that the story world continues to function until a participant leaves the interaction entirely. As role-play, the narrative and conversational qualities of the interaction do not move the interaction toward an expected conclusion, nor does the interaction have clearly marked boundaries or rules. Each conversational move, each narrative clause, feeds back into the interaction, continuing to generate the story-world so long as the children wish to continue playing. The only necessary rule or constraint for the language game of role-play to be successful is that language continue to be produced. In role-play, each narrative element is recursively generative, linked by singularly creative intent. This means, put simply, that the function of role-play is to play. Conversational strategies in role-play are concerned only with play, with the perpetuation of the game, allowing for the enormous expressive freedom that the children exhibit. Like improvisational comedians, the children fabricate the narrative on the fly and in cooperation with each other: the only rule is that they narrate something. Labov and Waletsky’s narrative elements are thus converted from passive descriptors into active functions of the story-world as it emerges in the children’s interaction.

Looking, then, at the beginning of the sample, we can see the child Aaron employ the “abstract” element of Labov and Waletsky’s schema as a function, initiating the language game by “announc[ing] that [he] has a story to tell” (Johnstone 92):

15 *KAT: uh oh
16 *AAR: does anybody want to play at the castle with me?
17 *JIN: why did you say “uh oh”?
18 *KAT: because ...
19 *AAR: anybody want to play at the castle with me?
20 *AAR: I said does anybody want to play at the castle with me?

As a conversational move, this “abstract” takes on a unique form. Aaron’s announcement is actually an invitation. He does not claim “that it will be a good story” nor that it will be “worth the audience’s time and the speaking rights the audience will temporarily relinquish” (92), as Johnstone’s description of the abstract element details, because he is requesting his audience’s participation in the story that he wishes to tell. He goes on to focus his request from “anybody” to particular participants:

21 *AAR: Jinny!
22 *JIN: what [% Looks toward A]?
23 *KAT: what?
24 *AAR: xxx do you [/] &w &o you [/] &wa &o you want to play at the castle with me?
25 *KAT: sure [% leaves easel and runs off-camera]!

Katrina and Jinny are talking (in lines 15, 17, and 18, above) so Aaron directly addresses them, interrupting first with “Jinny!” in line 21, and then with the second person pronoun “you” in line 24. Where in a regular narrative the abstract would summarize in a phrase or two what story the speaker wishes to tell, all that Aaron must summarize in order to begin his story is that he wishes to tell a story—that is, he wishes to “play at the castle” (16, 19, 20, 24). Further, Aaron’s use of the first-person singular objective pronoun “me” (16, 19, 20, 24) indexes the activity he has suggested as a collaborative one. Addressing first “anybody” (16, 19, 20), then “Jinny” (21), and then “you” (24), he situates himself in the object position and the others— “anybody,” “Jinny,” and “you”—as the subjects of his story, the ones doing the playing. Effectively, this shifting of the subject (Aaron’s “I”) into the object position (“me”) opens up the story-world to the input of others as actors. Aaron’s abstract can thus be reduced to simply: play + at castle + with me—these are the parameters of his story. In this way, we can see that the “abstract” of this narrative is functionally invitational and propositional: Aaron indicates a setting and his desired participants, and the stage is set.

In Labov and Walestky’s schema, the narrative element that should follow the abstract is “orientation” (Johnstone 92). This sequence occurs in the data as well, and, as with Aaron’s abstract discussed above, the orientation element of narrative takes on a similarly functional role.

37 *KAT: pretend we’re all Joan of Arc.
38 *AAR: you can be the princess.
39 *AAR: you can be the princess.
40 * KAT: okay and you be [% A interrupts].
41 *AAR: you can be Joan of Arc.
42 *KAT: I wanna be the princess.
43 *AAR: Joan of Arc princess.
44 *JIN: I wanna be Joan of Arc.
45 *KAT: I wanna be a princess.
46 *AAR: Joan of Arc needs to be a fighter long, long ago.
47 *KAT: okay the ballerinas will be the princess [% picks up fuchia doll].
48 *AAR: um hm I’m gonna be all these um xxx turn this over.
49 *AAR: xxx [% manipulating toys].

What is evidenced, here, is the importance of the orienting function to role-play. Because role-play collapses the interactional world of conversation into the story-world of narrative, identities are performed simultaneously in each, and must be indexed accordingly. Before the action can begin in earnest, the children must decide who is “playing” which characters, and whether one or multiple. As in the first lines of this excerpt, this is mediated through the use of pronouns. First, Katrina uses “we’re” (37), the first person plural, and names a historical figure, “Joan of Arc” (37) that the children should “all” be. But the children cannot all be Joan of Arc, because Joan of Arc is a definite and singular character. Aaron notices this immediately, moving to correct Katrina (whether or not Katrina used the “we’re” in error or because she thought that they could actually all play Joan of Arc), using the “you can be” second person construction twice in close succession (38, 39). This separates out Aaron’s “I” from Katrina’s “we,” marking her as “you” and so, distinct. In lines 41 through 46 Aaron, Katrina, and Jinny negotiate the identity and role of Joan of Arc, culminating in Katrina’s nomination of the “ballerinas” as the “princess” (47). Though there appears to be a pluralization issue here, Katrina is, in fact, identifying the indistinct quality of the unnamed dolls, giving the ballerina dolls the story-world type “princess.” As Katrina simultaneously picks up the fuchia doll, her use of the singular “princess” indicates that of all these ballerinas who are now princesses, she will be this one ballerina-princess in particular. Katrina’s language, scanned for “proper” grammar may be wrong, but as a function of the collapse that occurs in role-play, her congruent syntactic collapse is, effectively, data compression. There is no evidence that she, or either of the two other children, are confused by what she says. Aaron has already demonstrated a willingness to correct Katrina, after her use of “we’re” in line 37, but he does not correct her after line 47. Instead, he continues to orient himself to the story, satisfied that Katrina has identified both the ballerinas as princesses and herself as a particular princess in the world. Through the orientation function, the story-world, situated in and around the castle by Aaron’s abstract, has characters introduced to it and its narrative expanded.

Because the story-world of role-play is mutually constructed and generated in process, the orientation that occurs in the sample is not reported in the past tense but negotiated in the present and projected through the future tense verb constructions “can be,” “will be,” “wanna be,” and “gonna be.” In lines 38, 39, 40, and 41 the construction “you can be” and “you be” is used, in lines 42, 44, and 45 “I wanna be” is used, and in 48 the similar “I’m gonna be” is used. Additionally, “needs to be” and “will be” are used in 46 and 47 respectively. These orientational phrases function as statements of being, performatively indexing each child’s identity within the story-world. By saying who they will be, in the future of the story-world, they identify who they are now, in the interactional present. Through the verb “to be”—in each construction listed—the children orient themselves to and inhabit a story-world character who they will be playing as the narrative progresses. The shift between the interactional world and the story world, and the appropriate shift of identity from self to character, occurs in the linking of a subject—we, I, you, the ballerinas—to an object—Joan of Arc, princess, ballerina, these—through this “to be” construction, conflating the two worlds and their respective identities in the act of play.

The shift from the interactional proposition of the abstract to the mediatory orientation of the interactional world to the story-world is marked by Katrina’s use of the word “pretend” (37). After each child has identified with a doll or dolls, Katrina asks, “now what do we have to do?” (50). Aaron replies, “only pretend it’s raining” (51). His repetition of the word “pretend” marks the end of the orientation phase, and, as an answer to Katrina’s query, begins the narrative proper with a “complication”—the rain—the third of Labov and Waletsky’s narrative elements (93). The division between orientation and complication, as that between orientation and the abstract, is not clear cut. In the lines that follow Aaron’s response, he and Katrina continue to orient to and negotiate the parameters of the story world:

53 *AAR: and these three soldiers are marching.
54 *AAR: and they hafta keep their heads dry but they have_to spoil in the rain so they’re putting a xxx board over their heads &=laughs [% looks at K as he talks].
55 *KAT: okay but what can the ballerinas do so they don’t get wet?
56 *AAR: they always run and um: they hafta go to the castle.
57 *KAT: okay.
58 *KAT: I think I’ll go in the jail so I can be more comfortable in there [% gets up, walk around J and relocates doll in castle].

As evidenced by the use of “they” and “these three soldiers” and “the ballerinas” the collapse of interactional world and story-world has not yet completely occurred. There are still details to be worked out, details necessary for narrative progression. Aaron’s use of “pretend” in lines 51 and 52, and then his explicit explanation of what the soldiers are doing (“marching”) and what they “hafta” do (“keep their heads dry”), further orient the children to the narrative as agents in the narrative. By stating what he is doing in the interactional world, Aaron simultaneously does what he says he is doing in the story-world. When Katrina replies “okay” in line 57 and then proceeds, in line 58, to speak in the first person, the collapse is complete. Her actions and interactions are entirely mediated by her story-world identity. The interactional world remains embedded in the story-world, with only minor gaps as the children shift between characters and confront complications in the narrative, for the next 274 lines. Future-performative verb constructions “I wanna,” “I’m gonna,” and “I’ll” are used consistently throughout this duration, along with present-descriptive constructions like “I use,” “I say,” and “I am.” Looking at every instance of such is beyond the scope of this study, but it is clear from a survey of the data that, aside from the actual voicing of story-world characters by the children, “I am” and “I do” statements, in present and future constructions, are regularly used to perform story-world “complicating action” (93) in the interactional world of role-play.

It is not until line 333 that a participant moves fully back into the interactional world. In line 270 Aaron takes Katrina’s doll’s “sister” and makes her “fly[] way up in the air” and, in the process, damages the doll. At first, the children attempt to negotiate the complication in the story-world:

328 *KAT: hey are you all right [% kneels down by A]?
329 *AAR: yes smash your belly.
330 *KAT: soldiers soldiers.
331 *AAR: your dress is smashed.
332 *AAR: your dress is smashed, because one of the xxx knights broke your belly off [% looks up at K].
333 *KAT: it’s one of the leaves [% gets up and walks away from A].

Aaron provides an explanation for why the “sister” doll is broken: the “knights broke [her] belly off” (332). But, for some reason, Katrina is not satisfied: “it’s one of the leaves” (333). Whether this is intended as a contradiction, correction, or simply a statement of the facts is unclear, but what can be seen is a movement on Katrina’s part out of the story-world and into the interactional world to deal with a “complicating event” that cannot be remedied in narrative. Katrina goes to Bill, one of the adult supervisors, and tells him that “Aaron broke off one of the leaves” (338). Bill determines that he cannot fix the doll and so goes to get another doll for the children. Meanwhile, Katrina shifts back into the story-world:

354: *KAT: King!
355: *KAT: I’m sorry the Queen got messed up but we’re gonna have another one.
356: *KAT: the other one’s gonna be really nice to you.

After briefly breaking character to solve the problem of the broken doll, Katrina returns to the narrative and integrates her solution as a further complicating action. This return is marked by her direct address to Aaron as “King” (354), his story-world identity, followed by her use of first-person voicing in lines 355 and 356. What such an event makes clear is an incredibly sophisticated awareness of the grammatical use of pronouns to shift between identities and voices, and by extension, to shift between worlds. Through continuous pronominal orientation, the children are able to actively construct the sequence of their story-world, in conversation.


As a unique genre of discourse, role-play blends the features of conversation and narrative into a complex hybrid form, implementing narrative structures in the interactional world as tools for the generation and perpetuation of a mutually constructed story-world. As evidenced in the data, children engaging in role-play with one another demonstrate a high level of competency in navigating this linguistic landscape, implementing specific pronouns and verb constructions in regular, predictable ways to shift between worlds and identities. The discourse exhibits a high level of compression, with sophisticated conversational moves taking place in direct, simple syntactic formulations. Despite their young age, the participants demonstrate a great degree of awareness in regard to the constraints and possibilities of the genre. This study is far from comprehensive, and further analysis of the data in question would certainly yield more compelling features of role-play as a discourse genre. But, for now, this study fills the role of a preliminary inquiry into this adaptive and intriguing mode of discursive interaction.

Works Cited

Johnstone, Barbara. Discourse Analysis. Malden: Blackwell Publishing, 2008. Print.

MacWhinney, Brian. “Eng-NA/ErvinTripp/linked/bowyer02b.cha.” The CHILDES Project: Tools for Analyzing Talk. Third Edition. Mahwah: Lawrence Erlbaum Associates, 2000. Web.

Previous Post Next Post

« Trauma and Remembrance Absolute Particularity »