Research on second language acquisition (SLA) examines the nature and development of interlanguage, the linguistic system hypothesized to underlie learner language. But what does it mean to say that interlanguage is systematic? Questions have been raised when variationist SLA researchers have shown that grammatical patterns in second language (L2) learners’ language shift under different speaking conditions. Typically, the speech of L2 learners conforms most closely to a target language standard (i.e., is most “accurate”) when they are consciously paying attention to specific grammatical forms (e.g., adding –s to plural nouns, or correctly using the articles a and the), but may shift dramatically away from that standard minutes later when they are trying to communicate something meaningful. Tarone (1985) documented such shifts in four English morphemes when adult English language learners did three different tasks: a grammaticality judgment task (most target-like), an interview, and an oral narration of a story (least target-like). Subsequent studies showed similar variation to occur systematically in response to changes in other variables, such as interlocutor and topic. The theoretical and practical implications of this kind of variability in learner language in response to external social contextual variables, such as interlocutor, task, and topic, have been extensively documented and explored; however, some (e.g., Long, 1998) have questioned whether such shifts in social context have any impact on the learner’s SLA, or internalized interlanguage.

Ludic language play, as characterized by Bakhtin (1934/1981), has been proposed as one way in which learner language use in social context may influence SLA (Broner & Tarone, 2001; Tarone, 2000, in press). Adult and adolescent L2 learners have been shown to systematically manipulate elements of both their first and second language in the service of imagination, irony, and sarcasm. In doing this, ESL and English as a foreign language learners can exercise and expand their mastery of a range of English registers and dialects (what Bakhtin, 1934/1981, referred to as “voices”). Instances of language play like those documented in Broner and Tarone (2001) suggest that interlanguage variation is not simply a one-time response to a single social context; rather, they show that learners have internalized a range of L2 varieties that they can produce at will in new social contexts as part of ludic language play.

LaScotte (2016) demonstrates that one type of language play involves the speaker’s use of “constructed dialogue” (direct quotations “created” by a speaker) as central parts of unrehearsed oral narratives.Constructed dialogue was a term originally coined by Tannen (1989), who stressed that a speaker does not just act like an audio-recorder, reproducing a protagonist’s exact words; rather, he or she uses imagination to dramatically create and enact that protagonist’s voice to frame information in such a way that it enables the speaker to be directly involved in the dialogue. In other words, the protagonist’s imagined voice comes out of the narrator’s mouth, contrasting sharply with the narrator’s own voice. Bakhtin (1934/1981) would refer to such events as demonstrating heteroglossia, meaning an individual’s ability to speak with a wide range of voices associated with the people they know. These sources suggest that it is important to consider how constructed dialogue may affect the linguistic shape of an L2 speaker’s interlanguage. For instance, what does L2 learners’ reenactment of dialogue between themselves and another speaker reveal about the L2 linguistic varieties they have internalized?

LaScotte (2016) documented dramatic shifts in the complexity, accuracy, and fluency of two bilingual speakers (native French, English L2) in producing the English voices of others in narration. When enacting the voice of a more proficient speaker of English, the accuracy, fluency, and (sometimes) complexity of their own speech improved. One striking finding was that the less proficient (intermediate-low) speaker, Sylvie, never marked present tense verbs with third person singular –s except for the one time she enacted the voice of a native speaker of English. Another finding showed that Sylvie only used the correct word email when reenacting dialogue between herself and a native speaker of English; in all other cases, she used the word mail, understood by her French interlocutor as mél (email in French). Conversely, when her more proficient (intermediate-high) interlocutor, Marine, enacted the voice of a less proficient learner, her speech decreased or downshifted in accuracy and fluency; the only instance in which Marine made an error in question formation appeared when attributing dialogue to her less proficient interlocutor. While shifts in accuracy differed in direction between the two speakers, that shifts occurred at all is key. The internalized voices of these speakers, earlier acquired in social context, were later realized in constructed dialogue as variable patterns of grammar and vocabulary. These variable voices were created and enacted in narratives produced in a different social context in which the physical environment, interlocutor, and task stayed constant. In other words, the voices appeared to have become part of the bilingual’s competence, to be invoked at will by the speaker.

The LaScotte (2016) study raises a number of questions: Does this phenomenon occur in other speakers’ narratives? Is the ability to produce a variety of voices restricted to proficient bilingual speakers? At what stage of development can English learners also produce voices in constructed dialogue characterized by clear shifts in complexity, accuracy, and fluency? What might such results suggest about the nature of the interlanguage competence that such speakers have acquired? The present study builds upon LaScotte (2016), exploring the role of constructed dialogue in language play in the oral narratives of 10 adult learners of ESL, studying in an intensive English program at a large Midwestern university in the United States. The participants are demonstrably at several stages of development, ranging from high-beginner to advanced proficiency, and come from a variety of language backgrounds (Arabic, Chinese, Japanese, Korean, Russian, and Spanish). This study thus partially replicates and expands upon the smaller case study (LaScotte, 2016) to build the generalizability of those findings.

During the summer and fall of 2017, we audio recorded and transcribed speech samples from these 10 adult L2 learners of English. In response to a video prompt and interview questions, the English learners narrated stories (depicting past experiences in their home countries and the United States). At the time of this writing, data analysis is still in progress. However, preliminary results show that all participants produced in their narratives episodes of constructed dialogue that imaginatively recreated the English voices of characters in their stories. Our presentation at the TESOL convention in Chicago describes those narratives, exploring the degree to which the episodes of constructed dialogue, when compared with the rest of the narratives, demonstrated shifts in complexity, accuracy, and fluency of the learners’ speech. Such learners’ ability to produce a range of voices in constructed dialogue supports a variationist and heteroglossic theory of SLA and a view of ludic language play as facilitative of the acquisition of an interlanguage containing a range of internalized English language varieties that can be invoked at will by L2 learners, no matter what the social context. Clearly, SLA theories with more unitary and simplistic views of learners’ competence cannot account for such findings.

References

Bakhtin, M. (1981). The dialogic imagination: Four essays by M.M. Bakhtin (M. Holquist, Ed.). (C. Emerson & M. Holquist, Trans.). Austin, TX: University of Texas Press. (Original work published in 1934)

Broner, M., & Tarone, E. (2001). Is it fun? Language play in a fifth grade Spanish immersion classroom. Modern Language Journal, 85, 363–379.

LaScotte, D. (2016). ‘So please be nice in class!’: An analysis of the complexity, accuracy and fluency of two English learners’ language through a heteroglossic lens (Unpublished master’s qualifying paper). University of Minnesota, Minneapolis, MN.

Long, M. (1998). SLA: Breaking the siege. University of Hawai’i Working Papers in ESL, 17, 79–129.

Tannen, D. (1989). Talking voices: Repetition, dialogue, and imagery in conversational discourse. Cambridge, England: Cambridge University Press.

Tarone, E. (1985). Variability in interlanguage use: A study of style-shifting in morphology and syntax, Language Learning 35, 373–403.

Tarone, E. (2000). Getting serious about language play: Language play, interlanguage variation and second language acquisition. In B. Swierzbin, F. Morris, M. Anderson, C. Klee, & E. Tarone (Eds.), Social and cognitive factors in SLA: Proceedings of the 1999 Second Language Research Forum (pp. 31–54). Somerville, MA: Cascadilla Press.

Tarone, E. (in press). Voices in learner language: Language play and double voicing in second language acquisition and use. In M. Haneda & H. Nassaji (Eds.), Perspectives on language as action: Essays in honour of Merrill Swain. Clevedon, England: Multilingual Matters.

Darren LaScotte is adjunct faculty at the University of Minnesota. He has also taught courses at Hamline University and at Minneapolis Community and Technical College.

Elaine Tarone is professor emerita and retired director of the Center for Advanced Research on Language Acquisition (CARLA) at the University of Minnesota, and she has published research on interlanguage variation.