Peer response is an increasingly common practice in second language (L2) writing classes. There are a number of reasons for having students engage in peer response, including providing students with an audience other than the teacher, encouraging students to work together during the writing process, and helping students edit and revise their text (see Ferris 2003). While such reasons seem acknowledged, teachers are beginning to explore how students actually evaluate their peers’ papers (Connor & Asenavage 1994;; Paulus 1999; Tsui S Ng 2000; Zhang 1995). One of the most important questions regarding peer response centers on what students actually do when asked to review a peer’s writing. Several researchers have examined what personae or “stances” students took (e.g., Lockhart S Ng 1995; Mangelsdorf S Schlumberger 1992;Min 2008), namely, how students orient themselves toward and position themselves vis-a-vis peers and the texts under review.

However, studies focusing on English as a second language (ESL) student peer response stances tended to describe (e.g., prescriptive, interpretive, or collaborative) rather than to explain the observed students’ stances. It is only recently that Zhu and Mitchell (2012) examine ESL students’ peer response activities by linking the students’ peer response stances to their otive/objects for participating in peer response. Their study attempts to advance peer response research by introducing a new line of inquiry informed by Leont’ev’s activity theory and to “address the why of the observed stances, thus examining student participation in peer response tasks as activity” (Zhu & Mitchell 2012, 366).

Seen as a part and extension of sociocultural theory, activity theory represents an intriguing framework for understanding human behaviors in sociocultural context (Lantolf & Thorne 2006). Activity theory, which grew out of the work of Vygotsky and Leont’ev, subsequently expanded by Engestörm, locates individual actions in larger cultural activity systems. Basically, Engeström (1999) developed and formalized what is often left implicit, underdeveloped, or ignored in Leont’ev’s individual model of activity. Moving towards a better accounting of the collective nature of activity, Engeström expanded Leont’ev’s concepts graphically to include rules, community, and division of labor. The three elements are included because human activity is always influenced by them (Kim 2009).

This study, guided by Engeström’s (1999, 2001) activity theory which owes its theoretical lineage to sociocultural theory (SCT), explores how roles (peer feedback givers and receivers) and tasks are distributed among English as a foreign language (EFL) students who engage in peer response. More specifically, as an extension of previous research of focusing on “stances” ESL students adopt, I investigate whether different roles in peer response groups make a difference in the nature of peer response and identify what underlays the different roles in peer group interaction. In addition, I examine whether different roles to the peer response create tensions and contradictions in peer response and how these created conflicts lead to changes in peer response activity system. By reading stories narrated by two EFL students who participated in peer response, I will interpret each story through the lens of Engeström’s theory of activity.

Activity theory is a unified account of Vygotsky’s original proposals on the nature and development of human behavior. Specifically, it addresses the implications of his claim that human behavior results from the integration of socially and culturally constructed forms of mediation into human activity (Lantolf 2000). Vygotsky began with the idea of mediation and constructed a basic triangle “to represent his reconceptualization of stimulus and response” (Swain et al. 2011, 98), as in Figure 1. This triangular representation of mediated action is Vygotsky’s attempt to explain human development that does not rely on the simple stimulusresponse process advocated by the growing behaviorist movement of his time. Vygotsky’s work focuses on the mediated learning of individuals. This is easily illustrated as follows.

Two approaches in current activity theory-based research can be considered the most well known and influential: the one developed by Leont’ev (1978, 1981) and the other proposed by Engeström (1987,1999). These two approaches provide two different views on the object of activity. For Leont’ev (1978), the object of activity is predominantly the object of individual activity, its “true motive” (62). And all activities are considered by Leont’ev to be social, including those that are not carried out collectively. Even if people work alone, their work is determined by social and cultural practices, tools, values, and so forth. In other words, activities can be either individual or collective in respect to their form, but they are always social.

Within Engeström’s approach the unit of analysis is defined as “objectoriented, collective, and culturally mediated human activity, or activity system” (Engeström & Miettinen 1999, 9). Therefore, individuals, according to Engeström (1999)), can only carry out actions within a larger- scale collective activity system. Leont’v (1978), 1981) views activity as a mediated “subject—object” interaction. Leont’ev’s theory emphasizes the centrality of motive for understanding human behaviors and focuses more on the motives of individuals and the connectedness of motives and behaviors. In contrast, Engeström’s approach focuses more on human behaviors in collective activity systems comprising subject, artifact, object, community, rules, and division of labor, and underscores the attainment of an outcome (Kaptelinin 2005).

Engeström expands the triangle as he attempts to make all of these interactive elements explicit (see Figure 2). In order to reach an outcome, it is necessary to produce certain objects (e.g. experiences, knowledge, and learning). Human activity is mediated by artifacts (e.g. tools, documents, computers, language, etc.). Engeström includes community as a component of activity because individuals live in communities and participate in human activity in the context of those communities. The addition of rules is intended to reflect the fact that rules specify and regulate work procedures and interactions among community participants, who are represented as a collective subject. A collective subject constituted within a community implies, in turn, a division of labor (roles) - the way tasks are divided horizontally, and the way roles and hierarchies are structured among members of community. In fact, Engeström’s framework for activity theory is useful for understanding how a wide range factors work together to impact an activity.

Activity systems are not static but dynamic. All of a system’s elements reciprocally and dynamically influence each other so that the system is continually adjusting, adapting, and changing (Nelson & Kim 2001). For Engeström, activity systems undergo constant change, partly because they intersect and interact with other activity systems. In Engeström’s view, change often occurs as the result of so-called contradictions (which are not necessarily negative) within the activity system itself. These contradictions reflect dialectical processes at work among components in Engeström’s expanded triangle. It is the dialectical processes among components that, in the end, cause, direct, and characterize the nature of the changes that the activity system undergoes. For Engeström, “such change not only affects the nature of the structural relations within the activity system but also implies (potentially, at least) positive development within the system, which Engeström often characterizes as a change in trajectory” (Witte 2005, 139).

The data I wish to consider is first-person narratives elicited from two EFL college students. Indeed, stories elicited from the learners are a legitimate source of data in SLA which one cannot attain in the more traditional approach. Although narratives have been marginalized by the social science as legitimate data, as Pavlenko and Lantolf (2000) argue, “first person accounts in the form of personal narratives provide a much richer source of data than do third-person distal observations” (157).

Two students took a course entitled “EFL Writing” which was offered at the College of Education in a large university in Seoul, Korea. This course was taught by the researcher. Among the topics addressed was teacher and peer responses to writing. Students were trained to read and respond to other students’ papers by viewing sample papers which came from ESL composition classes at the university level. Meanwhile, students had the opportunity to write an essay and to engage in peer response. I put students together in groups of four, each with a first draft in hand, and then had each student read the other’s paper and react to the strengths and weaknesses of the paper.

Although peer responses could contain feedback on both content and language, the instructor asked students to focus exclusively on response to content. Feedback given by the peer was written feedback. The students were to use English in their responses. Based on peer review comments, students were required to write their second draft and to submit the revised section in the following week. At the end of the semester, I had invited four students to craft reflective narratives describing their own experience on peer response. They voluntarily participated in the study. They are all prospective teachers of English. I stated that what I was interested in is a brief text in which they address anything regarding peer feedback activity.

I use Won’s and Choi’s (both pseudonyms) stories as a heuristic - “[a] heuristic is a device or a method that allows one to proceed fruitfully in finding information” (Swain et al. 2011, viii). In other words, based on Engeström’s (1987, 1999) activity theory framework, I will seek understanding of phenomena (peer response in EFL writing classes) via stories. Since “[s]tories and sociocultural theory (SCT) are natural partners” (Taniguchi 2009, as cited in Swain et al. 2011, vii), in the following section, the stories lead and activity theory serve (see Swain et al. 2011, vii).

1The stories are from my own data from previous research (Huh & Lee 2011). For the previous research, I invited a total of four EFL students. However, for the analysis from an activity theory perspective, I focus here on just two students. I am also convinced that “[s]tories are rich source of data and can yield new insights each time they are revisited” (Swain et al. 2011, 2). The story has been included as written; no changes have been made.

Here, I can build two triangles from Won’s/Choi’s perspectives. In the first triangle (Action 1), Won and Choi responded to their peers’ writing in this particular peer response. Won and Choi are the subjects or agents. Their immediate goal is to respond effectively to their peers’ writing. Included among mediational tools are computer, knowledge of writing conventions, response strategies, and above all language. Most crucially, the action 1 is mediated by Won’s/Choi’s use of written English. In the second triangle (Action 2), Won and Choi are also the agents. Their goal is to revise their own writings. Won and Choi are to make more textbased (global) changes after receiving peer feedback. Their actions are mediated by their use of written English as well as their peer commentaries. “The actions and goals are specific to the individuals, the time and the place” (Swain et al. 2011, 99). So, Won and Choi took the action 1 of peer response to read each other’s paper for the purpose of providing feedback, and they took action 2 to make global changes after receiving peer feedback.

As noted earlier, the terms at the base of the triangle, rules, community, and division of labor, make up what Engeström (1999) refers to as the “social basis” of the activity system. The social basis situates the activity in a broader context that allows us to account for the influences that shape the activity. Thus, in this peer response activity, Won’s and Choi’s actions towards objects are mediated by four interrelated factors: mediating artifacts; rules; community; and division of labor (see Figure 2). Engeström’s perspective picks out the rules of peer response activities which Won, Choi and their peers explicitly agreed on at the beginning of this peer response activity. In order to respond effectively to the content of a peer’s writing, Won, Choi, and their peers should follow four suggestions2: 1) write personalized comments; 2) provide guidance or direction when necessary; 3) make text-specific comments; and 4) balance positive and negative comments.

The division of labor for peer feedback activity is the shared responsibilities among peers in groups. A division of labor exists between Won/Choi and her peers in this peer feedback activity. Won’s (Choi’s) job as the peer reviewer is to give more-effective feedback. Perhaps, the immediate goal of Won/Choi in this peer response activity may respond to peers’ writing in a timely, supportive and knowledgable way. As a feedback receiver, Won/Choi can revise and rewrite based on peer comments. Her peers also have the same jobs. Thus, Won and Choi participate in “the reader role” as well as “the writer role.” The roles - drafter, peer reviewer, reviser, and responder—are shaped by rules, which specify and regulate peer response procedures and interactions among peers. Although this particular peer response action immediately involves only Won, Choi and peers, they are all part of an EFL classroom community that engages in the same actions.

All the goals of an action are stated at the beginning of this peer response activity. However, many goals or objects emerge in the course of peer response in addition to the ones articulated at the beginning. For example, Won’s narrative indicates that her participation in peer response is primarily driven by an object “[not to] fall behind other classmates or at least [not to] reveal [her] shortcoming.” This object, in turn, is related to a need “to be clear in the writing.” Won emphasizes her purpose of making writing clear for her reader to understand, stating that “I especially paid attention to choosing the right words to express my opinion because I know that one important aspect of measuring competence in language is vocabulary.” Won said:

In Won’s action, it is possible to see evidence in the following narrative of yet other emergent goals, setting up a goal perhaps for herself. The experience of being a feedback giver made her give more consideration to her own potential readers. Won comes to realize that responding to peers’ writings is a learning process that will help her identify her own weaknesses in writing (Villamil & de Guerrero 1996). Further, the importance pointed out here is that, for Won, responding to peers’ writing builds the critical skills needed to analyze and to revise her own writing. This is in line with Rollinson (2005) who states that peer feedback also trains students to be a critical reader on their own writing. Won’s role as a feedback giver during peer response can be described as responding and cooperative. Won played a much more prominent role in peer response as a feedback provider. Viewed in this light, Won seems to primarily participate in peer response as a feedback giver. A simple yet powerful illustration was voiced by Won:

Choi is a student who demonstrates an absolute degree of trust in her peers by saying that “the comments can be very rich, much richer than those of the teacher’s who usually rushes through all the compositions in a short time.” Choi tried to interpret whatever her peers had done for her very positively and utilized peer feedback in constructing revision of her text. Choi actively checked peers’ comprehension of her text, responded to peers’ questions, and elicited peers’ feedback on her writing. The most detailed account comes from Choi’s comments:

Choi then emphasized the benefits of peer response for her as the writer (Mangelsdorf 1992). Her role for participating in peer response is to obtain peer feedback for revision. Guided by this role, Choi came to view peer response as a feedback-oriented activity with the writer by regarding the recipient of the feedback as the primary beneficiary (Zhu & Mitchell 2012). In particular, her “reader-centered, active, and eliciting” (Zhu & Mitchell 2012, 369) stance in her writer role is very well illustrated as follows:

From an activity theory perspective, the division of labor in this peer response activity—who gives feedback and who receives feedback—is shaped by Choi. Choi primarily participated in peer response as a feedback receiver. The labor is divided between the participants in groups of four—Choi is responsible for peer feedback receiver and three peers for givers. In this sense, as illustrated in Figure 2, the subject and object are mediated by division of labour (or roles). This is most explicitly evidenced in the words of Choi:

Although she talked a lot in a positive way about peer response, Choi highlighted, for the most part, the benefits of peer response for the writer as opposed to the reader and participated in the writer role primarily through responding to peers’ comments. This suggests that, for Choi, peer response is primarily an activity serving the needs of the writer. Besides, Choi seems to be willing to let her peers take the lead. She commented, “I only revised the parts that was pointed. I didn’t make a correction in the introduction part because I thought no one pointed out something . . . I followed only their comments and I didn’t think critically about my writing.” By no means is it easy to draw a tentative conclusion of whether or not Choi took advantages of peer comments to complete the tasks more quickly and with less effort. Unfortunately, the more feedback the peers offered, the less responsibility Choi took. Clues to this are provided in the following story.

Within the classroom community, students might disagree about how labor should be divided or how valuable various positions within that division are, causing tensions within the activity system. During peer response, Won experienced the tensions because she was reluctant to rely on her nonnative peers for feedback. Won thought that her peers, due to their limitation of English proficiency, are simply not good at giving her helpful feedback. Besides, Won’s peers were less tolerant of unclear writing and tended to convey more directly their confusion when they failed to understand Won’s writing. It would appear that such feedback may not well motivate Won to engage in text-based revision. After examining peers’ comments on her paper, Won castigated peer-responders:

Yet, oddly enough, MIN only gave highly judgemental and negative feedback. MIN seems to “establish her role as trouble-shooter” (Lockhart & Ng 1995, 620). Won strongly argued that she saw very little benefit in this kind of feedback. In the end, she reacted defensively to MIN’s comments, saying that “she wasn’t more competent in English than me, but still gave instructions in a harsh manner.” Won expressed discomfort with her peer’s acting like teachers:

The double arrows in the reciprocal relationships in Figure 2 imply that elements may potentially oppose one another and in such cases tensions and contradictions may occur. Engeström sees the tensions and contradictions that can occur within and between activity systems “as possible sites for potential change or learning” (Swain et al. 2011, 105). Change produces advances and improvements as well as complications and challenges that need to be addressed and resolved by participants within activity systems. We can see tensions and contradictions within Won’s activity system as well.

For example, Won, due to her peer’s limitation of English proficiency, called into question of rule of this activity. As noted earlier, Won felt uncomfortable with an authoritative reader like MIN. These role relationship caused conflicts within the activity system. In particular, although writers working with authoritative readers seem to reinforce this characteristic by assuming a passive role and relying on the reader to set the agenda for the discussion (Lockhart & Ng 1995), Won made less use of peer comments in making revisions. Forced by contradictions, Won changed her participation in peer response to “a practice-oriented activity geared toward the reader” (Zhu & Mitchell 2012, 378) and adopted a reader role. Won, as an agent in this action, positioned herself as a feedback givers. Won herself has the responsibility to choose the direction and focus of the peer response activity. Clearly, for Won, peer response is primarily an activity beneficial to the reader, although she may have also seen some value in peer feedback for other writers. Moreover, responding to peers’ writing should be a learning process for Won (Bates, Lane, & Lange, 1993). Tensions and contradictions are not necessarily to be avoided, however, because they are the driving force of transformation within an activity system (Engeström, 1987). In this view, the following story gives tantalizing hints of how contradictions could lead Won to expand into new learning activities:

Won’s and Choi’s stories highlight the ways in which a writer’s division of labour may change. From an activity theory perspective, Won and Choi and the object of peer response activity are mediated by the roles Won and Choi choose in the classroom community. Even when Won and Choi assumed the same roles, different roles occurred depending on the way the interaction between peers was constructed. In this strand, all of a system’s elements reciprocally and dynamically influence each other so that the system is continually adjusting, adapting, and changing. This reminds us that “the action is not an isolated one but one of many that are part of a dynamic, socially constructed way of being and acting in the world” (Swain et al. 2011, 102). In sum, as a human activity, peer response activity cannot be fully understood or explained without observing the sociocultural contexts in which it takes place.

2Four suggestions are from Writing clearly: Responding to ESL compositions by Bates, Lane, and Lange (1993).

This study investigates student participation in peer response by linking students’ roles with the division of labour among the students in peer response groups. Foregrounded in this study are the students’ different roles in the same peer response activity. A division of labor exists between Won/Choi and their peers - the way tasks are divided up and the way roles are structured. Yet Won and Choi adopted rather divergent roles when participating in peer response activity and carried out qualitatively different peer response activities (Zhu & Mitchell 2012). It is obvious here that the distribution of their roles in carrying out this particular peer response was shaped by Won’ and Choi’s perception about the validity of their peers’ responses. For example, Won expressed concern about her peers’ competency to evaluate her writing. Consequently, Won performed peer response activity in pursuit of the benefits for the reader. Namely, responding to peers’ writing builds the critical skills needed to analyze and revise her own writing. And this conception of the value of giving peer feedback propelled her to adopt a reader role oriented toward critical analysis of peer writing.

In contrast, Choi is a student who demonstrates an absolute degree of trust in her peers. Most critically, she is very much guided by her peers. Choi did not seem to have a complete grasp of task goals and was unable to undertake revision on her own initiative. However, she could achieve a certain degree of control over the task as she let herself be guided by her peers (de Guerrero & Villamil 1994). That is the reason why Choi primarily participated in peer response as a feedback receiver. The labor is divided between the peers in groups of four — Choi is responsible for peer feedback receiver and three peers for givers. In Choi’s peer response activity, the division of labour is divided vertically with respect to asymmetrical relationship. In short, Choi would hold her peers’ feedback in higher regard than Won would.

Here, I find activity theory useful in articulating Won’s plight. Interpreting Won’s case in activity theoretical terms allows us to see a major “contradiction” (Engeström 1999a,2001) between Won and her peers. This contradiction occurs because Won was reluctant to rely on her nonnative peers for feedback. Won thought that her peers, due to their limitation of English proficiency, are simply not good at giving her helpful feedback. It would appear that her peers’ comments may not well motivate Won to write better revision. However, as Engeström (2001) suggests, a contradiction can be used as a springboard for growth. Presumably, such conflict has the potential for learning and change to take place. In Won’s case, a resolution of such contradiction might include for Won changing her peer response to “a practice-oriented activity geared toward the reader” (Zhu & Mitchell 2012, 378) and adopting a reader role in which she applied concepts learned about writing to the analysis and evaluation of peers’ writing. Won, as an agent in this action, positioned herself as a feedback giver. Won herself has the responsibility to choose the direction and focus of the peer response activity.

In this regard, neither Won nor Choi performed the peer response activity exactly in accordance with the purposes presented by the teacher at the beginning of the peer response activity. Thus, their task performance could be viewed as partial failure to carry out the teacher assigned task (Zhu & Mitchell 2012). Although teachers set instructional goals for peer response activities, Won and Choi did not simply follow these goals but rather their adopted or preferred roles in groups of four determined the way how peer response was carried out. “It is [Won’ and Choi’s] definition of the situation” (Storch 2004, 473), demonstrating how peer response will be carried out. From both Won and Choi, we learn “how these individuals-in-context interact with their peers in soliciting and providing quality feedback” (Nelson & Carson 2006, 54).