ANALYSING THE CHARACTERISTICS OF A PAIR TALK ON AN EDITING TASK ON TWO INTERMEDIATE INDONESIAN LEARNERS OF ENGLISH

Recent studies have paid attention to peer interaction as a means of second language (L2) learning. Interaction among learners in carrying out task is believed to mediate the process of L2 learning. The study therefore observed the nature of a pair talk in an interaction while completing a grammar task. The interaction was recorded and analysed by using language related episodes (LRE) in terms of form, lexicality, and mechanism. LRE was used in order to understand the characteristics of a pair talk at homogenous (intermediate-intermediate) level and to examine the participants’ accuracy in completing the task. The results suggest that even though the pair solved the questions in the task using form-focused LRE (F-LRE) and resolved the task interactively, they were not able to make correct decisions over grammar errors.


INTRODUCTION
Several studies suggest that peer interaction is very beneficial in promoting second language (L2) learning (Storch, 2007;Watanabe & Swain, 2007).During peer interaction, learners use and explore the L2 as well as work collaboratively to solve the linguistic impasses.This study therefore attempted to analyse the characteristics of a pair talk at intermediate level when completing an editing task.
Further, Storch (2007) found that learners working in pair could reach more Risa Fitria Englisia Vol. 4, No. 1, NOVEMBER 2016 | 43 accurate grammar over the learners working individually.Thus, it also examined whether working in pair can help intermediate learners make correctly grammatical decisions when solving the given task.Given the lack of study focusing on intermediate proficiency dyad, the study sought to understand the characteristics of learners' talk that was produced by the same degree of homogenous level, which is intermediate-intermediate level.

Literature Review
This article examines the characteristics of a pair work between two intermediate English learners during completion of a grammar task.In L2 classroom, group work and/or peer interaction are commonly carried since it can facilitate the process of L2 learning (Gass & Mackey, 2007).The interaction process that occurs during completing task is believed to be the stage where learners use and exchange their knowledge and information of the L2.This view is supported by the psycholinguistic and sociocultural theory.Despite their ontological differences, both acknowledge the importance of interaction for L2 learners to assist each other in developing their language proficiency.In sociocultural theory, cognitive development can be established via interaction between people.This development occurs during process of imitation and zones of proximal development (ZPD) where the more advanced facilitators such as teachers, tutors, or peers assist the lower learners (Vygotsky, 1987).In the imitation process, learners exchange languages and might fix their utterances that have been corrected by more proficient peers.Meanwhile, Psycholinguistics posits the theory of interaction as a means of language transfer.Long (1996) in his revised theory argues that language learning is transferred from feedback, input, and output.The process of giving feedback, receiving input, and producing output might occur when there is a communication failure.
In analysing what occurs in the interaction and how language is learned, many studies have focused more on the interaction between peers with heterogeneous level.
A study conducted by Leeser (2004) investigated how the different proficiency had an impact on learner proficiency towards the production of LREs (Storch, 2007).
This study revealed that higher proficient learners were able to produce more LREs compared to their lower counterparts.Interestingly, the higher pairs produced more on grammatical form of LREs whereas the lower ones gave their attentions more to the lexical items given the difficulty in understanding the task.Likewise, Watanabe and Swain (2007) investigated the effect of proficiency differences on the production of LREs and the participants' post test results among the Japanese participants who were divided into four groups where four intermediate learners worked with their lower and higher English proficiency partner.The study revealed that although the intermediate participants produced more LREs when working with their more capable peers but they were able to achieve higher post-test score when paired with their less capable counterparts.This then suggests that lower competent learners can contribute to the language development of their higher counterparts.
Given the extensive research was done to learners with different proficiency level, there is lack of data that confirm the characteristics of learners with a similar proven to be successful in drawing learners' attention to form-focused and lexical choices (Storch, 2007).The task provided in the study was adapted from first year preparation exam and practice of Roma Tre University (First year preparation and practice).It required participants to correct one error appeared on each numbered line.The errors included in the task were 10 grammar, two word order, one vocabulary, and two spelling errors.Afterwards, the data were collected after the pair finished completing the editing task for analysis.The analysis used Language Related Episodes (LREs), which will be explained further in data analysis, to investigate accuracy over grammar and lexis during the completion of the given task.

Participants
Two Indonesian learners of English at intermediate level participated in the study.The participants were graduated from English department at one of universities in Banda Aceh, Indonesia.However, the learners are still learning English in order to improve their TOEFL score at one of private English courses.They are currently working as administrative staffs and have part-time jobs teaching general English at elementary level.The learners have similar Institutional Testing Program (ITP) TOEFL scores from 490 to 500.This means that their English L2 proficiency levels were intermediate levels, which were fairly homogenous (Murakawa, 1997, as cited in Watanabe and Swain, 2007).The learners were chosen because of their close relationship with the researcher as well as their willingness to participate in the study.

The Context
The study was conducted within the context of the teaching English department of a university in Banda Aceh, Indonesia.The learners attended only one session to complete the given editing task.The session was held on Sunday to avoid the participants getting distracted by the activities that they normally do during weekdays.

Data Analysis
In analysing the characteristics of the pair talk that learners produced, the talk was recorded during the editing task completion and transcribed after that.The transcribed data was analysed by using the same approach in Storch's (2007) study.
It examined the nature of the pair talks by using analysis of Language Related Episodes (LREs).LREs occurs when 2 nd language learners 'talk about the language they are producing, question their language use, or correct themselves or others,' (Swain and Lapkin, 1998, p. 326).Furthermore, LREs are divided into three categories (Swain, 1998, as cited in Storch, 2007).Firstly, Form-focused LRE (F-LRE) was where learners focused on grammatical form.Secondly, Lexical LRE (L-LRE) was where learners paid attention on word meanings, word choices, and prepositions.
Finally, Mechanical LRE (M-LRE) was where learners focused on punctuation, spelling, and punctuation  Excerpt 1: F-LRE (modal auxiliary verb) Excerpt 1 is an example of F-LRE where the learners discuss about modal auxiliary verb.First, Olin (O) reads aloud sentence number 2 in the given the text and questions the meaning of word 'frank'.However, Lely (L) shifts Olin's attention to another part of sentence and makes suggestion of what supposed to be the correct answer is.Later, Olin agrees and also provides Lely with the alternative answer (Line 24-25).This process is referred to as 'collective scaffolding' by Donato (1994) where learners internalize the new-shared knowledge. Excerpt 2: L-LRE (word meaning) From the three excerpts above, it can be analysed that the episodes are the interactive process.The interactive process is defined when two learners are involved in the decision-making process.Non-interactive process, meanwhile, is another process where only one participant makes decision during the talk (Storch, 2007).
In addition, LREs were also coded the outcome of the talk as correct/acceptable, incorrect/unacceptable, or unresolved (Leeser, 2004)

Implications for teaching
Given the small sample size (2) and one-attempt test, it is difficult to generalize findings, to say that peer interaction can promote learners to reach more correct decisions over grammatical items in the given task.Thus, to know whether the English language learners get benefit from working in pairs particularly on writing and form-focused task, future study should include larger participants to examine the nature of the pair talks produced by different proficiencies as well as to investigate whether peer interaction can assist learners in developing their cognitive learning potential at the university in Banda Aceh.Further, the future study should also interview learners whether they benefit from working in pair after giving the editing task.
It should also be noted that working in pair during writing and form-focused task is seldom conducted at the university.Thus, the study investigating pair work might be very useful for L2 teachers in Banda Aceh design group learners effectively to improve learners' English proficiency.
level especially towards intermediate-intermediate level learners.METHOD Research Questions While Storch (2007) examined the nature of pair talks on the editing task without focusing on proficiency levels, this study focused on the characteristics of pair talks with the same proficiency level (intermediate-intermediate levels).Thus, it formulated two research questions as follows; 1.What are the characteristics of the pair talk at intermediate level when completing the editing task? 2. Can working in pair help intermediate learners make correctly grammatical decisions when solving the editing task?Data Collection This study was conducted online via Skype where participants were recorded during completing the editing task (see Appendix A).The editing task has been Risa Fitria Englisia Vol. 4, No. 1, NOVEMBER 2016 | 45 24. O: But I must to be frank, what does frank mean?At first I was… 25.L: Don't think about that.Must right?Must is no 'to'.After must, no 'to'.26.O: Oh yeah yeah…but I have to, right?27.L: I have to, okay…have have.28.O: Oh yeah…But I have to, okay.29.L: to be frank, right?30.O: right.
. Three excerpts above are examples of correct/acceptable LREs.Excerpt 4 is an example of incorrect LRE where Olin reads out sentence number 3 and Lely suggests that word order is the error in the given sentence.In line 32 of the transcript, Lely suggests that the possible error in the sentence is word order but she seems unsure about her answer.From line 33 to line 40, both of the learners try out new possibilities by trial and error.However, in line 41, Olin suggests Lely that the problem is not word order but prepositional phrase 'interested in.' Lely, in this matter, agrees with Olin and fails to convince Olin that black short hair should be used in the given sentence