In this chapter, I will first discuss the findings of this study. Next, I will summarize the findings and indicate the study limitations, pedagogical implications, and suggestions for further study.
Discussion
The findings reported in Chapter 4 can be further discussed under the four themes:
synchronous CMC as social milieu, patterns of interaction and regulation during on-line revision, mediated scaffolding during on-line peer revision, and non-scaffolding assistance.
Synchronous CMC as social milieu
According to Vygotsky (1962, 1978), the locus of learning is not exclusively within the individual’s mind, but rather extends outside the learner, specifically within the interaction with other individuals. In other words, human higher mental development, in a sociocultural view, is tightly dependent upon the social milieu within which learning may take place. In this study, this specific social learning context was MSN Messenger– one of Microsoft’s on-line instant messaging programs, via which the peers exchanged critical comments in order to help each other revise. However, it should be noticed that this synchronous messaging program in this study was not considered merely to be an electronic communication channel as demonstrated in most existing studies, but rather to be an
alternative technology-enhanced social context where authentic dyadic interactions could be co-constructed by peers via instant communication. To further realize this specific
synchronous CMC context, its emerging types of on-line interaction and its social impacts, whether positive or negative, will be discussed in detail in the following sections.
Types of on-line interaction in the synchronous CMC context
In this synchronous CMC context, the emerging interaction types were similar to those in face-to-face revision environments (e.g., de Guerrero & Villamil, 1994). Specifically, the participants, in such synchronous on-line revision sessions, stayed on-task for most of the recorded interactions, with only a few episodes dedicated to talking about the task rules or straying to unrelated topics (see Table 4.1). Also, the on-task episodes themselves were found to be highly varied in nature (see Table 4.2). That is, the participants not only engaged in interactive peer revisions as expected, but also made self-revision with non-interactive attitudes.
Nevertheless, unlike those in face-to-face revisions, the on-line reviewers and reviewees were noted to seek help through the Internet, due to their scarce chances of receiving
assistance or guidance from a “real” teacher or tutor. Such actions of resorting to “virtual”
experts not only gave rise to a new CMC facilitated reader/writer interaction
(“Reader/Writer/Expert Interactive Revisions”), but also distinguished this on-line peer revising environment from the traditional face-to-face classroom context documented in the existing literature. Most importantly, it was also discovered that even within the interactive peer revisions, the patterns of idea exchange were diversified in nature. Generally speaking, such different patterns of idea exchange may be consistent with the three categories of social behaviors proposed by Villamil and de Guerrero (1996), namely, relinquishing/appropriating, respect for authorship/lack of respect for authorship, and struggle for authorial
control/maintaining authorial control. These diversified social behaviors revealed in this study may result in the participants’ the implicit, dynamic role awareness, which, on one hand, urged the student readers to actively initiate and maintain interactional control, and, on the other hand, left the student writers greatly dependent on their partners. Based on this investigation of a wide variety of on-task episodes as well as interactive peer revisions, the
results seem to confirm the assertions from previous studies about the complexity of peer interaction during revision (Gere & Abbott, 1985; Gere & Stevens, 1985; Nelson & Murphy, 1993; Warshauer, 1992), and seem to also identify the following positive and negative impacts of synchronous CMC towards the dyadic interaction.
Social impacts of synchronous CMC
In this study, the positive social impacts of synchronous CMC towards the dyadic interactions were the peers’ increasing independence and self-confidence. As shown in the participants’ instant conversation logs (see Excerpt 4.5 and 4.6), the peers, since having no real teachers or tutors to counsel, had only to find ways out by themselves. In other words, the CMC as the mediation for “virtual” experts, or more precisely for the source of
scaffolding, promoted the spirit of autonomy by allowing the participants to search for information on their own and fostered peer participation by imbuing them with more self-confidence to doubt, to negotiate, and to express ideas. It was exactly such a desire to express and defend oneself that triggered further dyadic interactions and prolonged the on-line discussion. These results, in this light, corresponded to Lotman’s (1988) statement that on-line discussion serves the role of “thinking device” (p. 36) and is thus of great significance for collaborative construction of knowledge. Moreover, they are also consonant with most of the findings documented in the literature that synchronous CMC is indeed empowered to foster either the quantity of language production (Chun, 1994; Kern, 1995;
Warschauer, 1996) or the equality of peer participation (Chun, 1994; Gonzalez-Bueno, 1998;
Kern 1995).
Yet, in spite of the well established literature on distance collaboration (e.g., Cummins &
Sayers, 1995), the distance nature of synchronous CMC is to a certain extent found to hinder the revision processing. In the existing CMC literature, physical distancing is commonly considered to be the key factor that reduces the participants’ pressure of facing authority and
that could create a non-threatening learning environment. However, in the present study, such distant idea exchanges via CMC surprisingly became the factor that distracted the participants from the revision works. Specifically, with less anxiety to come up with immediate answers, the peers were more likely to retard the whole revision processing by engaging themselves in other activities that were unrelated to revision. Such distraction, as shown in Excerpt 5.1 below, might eventually result in various short replies, incoherent idea exchanges, the partners’ monologues, and finally the occurrence of non-interactive revisions.
Excerpt 5.1 (Revision Session 6) (Reader: Cindy; Writer: Peter)
1. Cindy: Look at this sentence…”I am infatuate with Kreisler’s music and most of his work.”
2. Cindy: infatuate is a verb.
3. Cindy: u should add “d”
4. Cindy: and “work” should add “s”
5. Cindy: Hello?????
6. Cindy: Is anybody there?
7. Cindy: Do you know what I am talking about?
8. Peter: Yes?
9. Peter: I see
Beyond the issue of pros and cons, one more social impact of synchronous CMC observed in this study is its electronic variety of language. As Herring (1996) indicated, the language of CMC is typed and hence like writing, but it is exchanged rapidly and thus like spoken conversation as well. Corresponding to Herring’s statement, the language the
participants used in this synchronous CMC context was neither spoken nor written in terms of the conventional sense of speaking or writing. More specifically, the textual function of language used here was more similar to written language in terms of the vocabulary use.
However, due to the purpose of reducing typing time, such an electronic written language also manifested various features resembling speaking, such as lack of capitalization (e.g., “i
think english is hard.”) and absence of punctuation (e.g., the omission of the question mark in
“Do you know how to revise it”). Moreover, on account of the absence of paralinguistic cues, the synchronous CMC language was also characteristic of abbreviations (e.g., “u” for you,
“ur” for your, and “thx” for thanks), emoticons (e.g., @@ for dizziness, =D for a smiling face, Orz for an embarrassed person kneeling and touching the forehead to the ground), and even onomatopoetic devices (e.g., “oh”). Also, it is obvious that the participants tended to utilize these alternative communicative devices to compensate the lack of aural or visual cues in the “text” chats. Excerpt 5.2 below exactly shows this CMC language lying between the two extremes of speaking and writing.
Excerpt 5.2 (Revision Session 15) (Reader: Wendy; Writer: Billy)
1. Wendy: it seems that u compare the two movies from the stories itself 2. Wendy: themselves =D
3. Billy: yes~
4. Wendy: maybe u need key words 5. Billy: haha…I think so…
6. Wendy for example, compare the “ending”
7. Billy: Oh thx~~ it is a good idea.~~
8. Billy: it makes others read easier.
9. Wendy: @@ yeahhaha 3Q
In addition, it is especially noted that some of the speech-like language in CMC was culturally shaped. For instance, the peers in this study were found to use such words as “3Q”
(thank you), “haha” (hehe), “88” (bye-bye) and “OKla” (okay) to present their emotions.
These words were partially or completely created with Chinese pronunciation and therefore displayed the specific cultural influence upon the participants. In a sociocultural stance, those Chinese-like pronunciations are culturally specific and thus reflected the essence of L1 as mediation. That is, L1 as one of the mediation tools originally created by humans would further help to organize the biologically specified brain into a higher, or culturally shaped,
mind through the integration of symbolic artifacts into thinking (Lantolf, 2000).
Regulation and patterns of interaction during on-line peer revision
In this study, different patterns of interaction and the social relationships that resulted from the participants’ cognitive stages of regulation emerging in the on-line peer revisions were also explored. Consistent with the study of de Guerrero and Villamil (1994), the participants’ particular behaviors might characterize each cognitive regulation. Generally speaking, self-regulated participants show greater independence in troublesource
identification and problem solving. Such independence is also reflected in their attitude of self-confidence in terms of the content, language uses, and the ways of initiating interactions and providing scaffolding. Other-regulated learners are otherwise more uncertain about their revising actions; they are unable to undertake revision successfully on their own and hence often show the need for peer assistance. Yet, other-regulated learners are also seen to display better grasp of goals and improve control over the task after being guided and assisted.
Unlike the self- and other-regulated participants, the object-regulated peers seem to be totally controlled by the rudimentary drafts and thus failed to engage in any constructive dialogue with their partners. Those aforementioned behaviors, in actuality, thed light on the nature of regulatory levels. That is, self-regulation suggests high self-assurance, leadership, and great willingness to share. Other-regulation suggests hesitancy, the need for help, and the potential progress under scaffolding. Object-regulation indicates the naive self-satisfaction, the learner distraction, the absence of dialogical interactions, and even the self-abandoning (de Guerrero
& Villamil, 1994).
Moreover, through the observation, I also found that the participants’ regulatory levels were never fixed, but were rather dynamic. More specifically, the participants’ cognitive stages of regulation were seen to fluctuate according to their shifting awareness and attitudes towards the roles. As similarly reflected in the diversified social behaviors (see Excerpt 4.3
and 4.4), the ways how the participants assumed their own roles would to a certain extent influence their cognitive stages of regulation. That is, the student readers, who tended to assume greater responsibility to pick over possible troublesources, were mostly self-regulated, while the writers, who tended to sit back waiting for guidance and directions, were therefore other-regulated or even object-regulated. In this sense, once the roles are shifted, the
regulatory levels would also be altered.
Most importantly, it is found that different patterns of social relationships resulted from varied combinations of the peers’ cognitive stages of regulation. As shown in Table 4.5, asymmetrical social relationships predominated in on-line interactive revisions. Moreover, within the asymmetrical relationships, the OTR/SER category was the most common. This prevalence of asymmetrical OTR/SER interactions, on one hand, once again suggests the potentially unbalanced power relationships between the readers and the writers due to their different, or even nearly opposite, role awareness. On the other hand, it also successfully recapitulates the Vygotskian idea of ZPD by creating a mutual activity frame in which the OTR participants carry out revision actions in conjunction with the assistance of the SER ones. Moreover, such a symbolic mutual activity frame, in this study, was found to be implemented in two distinct manners: collaborative and authoritative (see Table 4.7). The collaborative interventions were constructed by the self-regulated peers helping their
less-regulated partners to understand the changes proposed with softened critical comments, and, therefore, were characteristic of dialogic on-line speech as well as the camaraderie and empathy. However, on the contrary, the authoritative interventions were constructed under the absence of negotiated process and thus were featured with monologues of the self-regulated peers and simple acceptance on the part of less-regulated partners.
Finally, it is edifying to note the significance of the external assistance in the participants’ growth of cognitive regulation. As observed, different patterns of dyadic interaction form a continuum where at the optimal extreme the two self-regulated peers
identify and solve troublesources independently through a highly negotiated process, whereas, at the other extreme, the two objected-regulated peers are both stuck by troubles and fail to conduct any constructive communication. Yet, such symmetrical social relationships between two OBR and two OTR participants seem to be more ineffective, due to their limited abilities to scaffold each other. Hence, along this continuum, the key factor that pushes the
participants to move from the basis (OBR/OBR), through the midpoint (OTR/OTR), and to the top (SER/SER) would be the behavior of resorting to “outside help” (de Guerrero &
Villamil, 1994, p. 492). From the data, this “outside help” could be the assistance offered by the partners or the “virtual” sources, such as magazines, on-line dictionaries, translation programs, or any informative websites. With these scaffolding devices, the participants can gradually gain the impetus to expand the ZPDs and move forward to a higher degree of regulation. To further realize the relationships between the participants’ regulatory levels and scaffolding received, the issue of mediated scaffolding will be further discussed in the following section.
Mediated scaffolding during on-line peer revision
As mentioned, the traditional line of sociocultural theorists contend that human cognitive development is a result of social interaction in which a less skilled individual can extend his/her current capabilities with the scaffolding provided by a more experienced or skilled individual. In other words, the expert has traditionally been believed to be the major social mediation via which the novice could move from the actual developmental level to the potential level of development in the ZPD. This belief could be well reflected in Donato’s (1994) contention that social interaction is a mediation via which the novice can be drawn into, and operates within, the expert’s strategic processes and thus result in individual cognitive development.
However, on account of the unique nature of this study, here I only examined peer
interaction, or more specifically peers as mediation, in on-line reviewing sessions. As previously indicated, Vygotsky’s theoretical framework has recently been expanded and employed by a number of L2 researchers to investigate peer response activities during group work in second language writing classrooms (de Guerrero & Villamil, 1994; Donato, 1994; ; DiCamilla & Anton, 1997; Lantolf & Appel, 1994; Villamil & de Guerrero, 1996; Villamil &
de Guerrero, 1998). Those positive results reveal the fact that “the speakers are at the same time individually novices and collectively experts, sources of new orientations for each other, and guides through this complex linguistic problem solving” (Donato, 1994, p. 46). In other words, expertise can also be collaboratively constructed via the positive dialogic mediation in which the peers may share the goals of working out a linguistically-based solution to a
problem.
In this sense, it has been assumed that the peer mediation could be saliently embodied in the form of scaffolding behaviors. The investigation of a wide variety of peer scaffolding activities undertaken by the participants in revision identifies the previous characterizations of scaffolding mechanisms (Bruner, 1978; Villamil & de Guerrero, 1996, 2000; Wood, et al., 1976) to be a complex process involving a myriad of recursive behaviors. Indeed, as
demonstrated in various reviewing sessions of this study, peer revision emerged as a
collaborative, communicative experience in which the peers had not only to read and write, but also to learn various complex skills such as assessing, composing, copying, and even persuading. Through such collaborative dyadic interactions, I not only observed the six traditionally recognized scaffolding behaviors(“Recruitment”, “Reduction in Degree of Freedom”, “Direction Maintenance”, “Marking Critical Features”, “Frustration Control”,
“Demonstration”), but also identified one new scaffolding type, “Sharing”, which is unique to the optimal SER/SER peer interaction. These findings above showed that learners were capable of providing mutual support in ways analogous to expert scaffolding documented in the literature (Bruner, 1978; Lidz, 1991; Wood, et al., 1976).
Also, the findings, as the expansion of the second research question, showed that different patterns of scaffolding behaviors would be given according to the peers’ different levels of cognitive regulation. This phenomenon was especially preeminent when one of the peers in a dyad was highly self-regulated and the other one was in his/her lower level of regulation (see Excerpt 4.12 and 4.13). This result, on one hand, is consonant with the finding of Ohta (2000) that learners might be sensitive to each other’s ZPD and therefore would not provide scaffolding randomly. On the other hand, it implies the ways how peer feedback is negotiated in their ZPD in terms of the innate regulatory hierarchy (see Excerpt 4.12 and 4.13). That is, language learning as a kind of human higher mental activities may greatly rely on the mediation provided by other individuals, who co-construct their ZPD in consort with the learner dialogically. In this sense, feedback as regulation would become relevant and could, therefore, be appropriated by learners to modify their inter-language systems (Aljaafreh & Lantolf, 1994).
Similarly, different natures of scaffolding assistance would also be offered according to different interaction interventions. As shown in Excerpt 4.7 and 4.8, the OTR/SER interaction presented two subtypes of peer interventions, that is, authoritative and collaborative. Through the observation of the scaffolding assistance provided in these two interventions, the close relationships between the scaffolding types and the interventions were identified. Specifically, the authoritative student reviewers tended to apply more extrinsic types of assistance, such as
“Demonstrating”, to direct the revision process and results. On the contrary, the collaborative reviewers tended to use more intrinsic scaffolding, such as “Reduction in Degree of
Freedom” and “Direction Maintenance”, to help simplify the revision tasks and to keep their partners in pursuit of the final goals. This phenomenon might be further reinforced by the feedback of the reviewees. That is, the relinquishment of the reviewees in authoritative interventions might even reinforce the occurrence of extrinsic assistance, while great participation in collaborative interventions might eventually stimulate the occurrence of
intrinsic scaffolding. This finding once again confirms the strong correlation between the
mediated scaffolding and the received feedback, as mentioned in the previous paragraphs.
The last significant phenomenon about scaffolding as mediation is the participants’ L1.
As shown in the peer interactions, the participants were observed to use Chinese, their L1, as the tool of mediation to make meaning of text (e.g., Excerpt 4.10), help retrieve words in the L2 (e.g., Excerpt 3.1), explore and expand ideas (e.g., Excerpt 4.11), guide the action through the task (e.g., Excerpt 3.2), and keep conversation going (e.g., Excerpt 4.13). As Excerpt 5.3
As shown in the peer interactions, the participants were observed to use Chinese, their L1, as the tool of mediation to make meaning of text (e.g., Excerpt 4.10), help retrieve words in the L2 (e.g., Excerpt 3.1), explore and expand ideas (e.g., Excerpt 4.11), guide the action through the task (e.g., Excerpt 3.2), and keep conversation going (e.g., Excerpt 4.13). As Excerpt 5.3