Title Investigating mediations in student activities in an Englishimmersion context in mainland China

This study aims to investigate the role of mediation in the learning process from a sociocultural perspective, activity theory in particular. This study was carried out in a primary English immersion school within the CCUEI Programs in Mainland China. Data were collected mainly through observations and interviews, which were then supplemented by field notes. Techniques of discourse analysis were used to analyze the data. The recorded peer talk presents an emic picture of what occurred during children’s interactions. Research findings show that mediation includes task, teacher talk, peer talk, subject content, semiotic artifacts and other semiotic tools. Peer talk as well as linguistic resources constitutes an important resource aside from teacher talk, where the subject content serves as an important mediational tool in the activity and the content-based immersion programs, while offering students opportunities of using the language in meaningful interactions. The results offer insights to both students and teachers about language learning and about pedagogy and professional education respectively.


Introduction
Mediation has long been of interest to researchers.Vygotsky (1978) took human action as the unit of analysis, and stated that the human mind is mediated.He identified two mediational tools that people employ in the development of higher cognition: physical tools and symbolic tools among which language plays a major role.My study aims to investigate whether there are other mediational tools students employ in their learning activities to develop their English spoken language proficiency in Mainland China.
In Mainland China English language teaching has experienced great changes.Recently a new English language curriculum has stipulated English as a compulsory subject starting from at least grade three in primary schools.Primary school English teachers face many challenges including limited English language proficiency (Yu, 2001;Zou, 1998), a lack of systematic in-service training, and few active researchers at this level (Cortazzi & Jin, 2001;Hu, 2002aHu, , 2002b)).The China-Canada-United States Collaborative English Immersion (CCUEI) Program, the first and the most influential immersion program in Mainland China, serves the purpose of improving English language teaching at the preschool and primary levels.The typical characteristic of English immersion programs is the teaching of the English language, subject content, and culture through the medium of English without the use of the students' first language (Francis & Reyhner, 2002;Reyhner, 1998).The aim of this project was to increase children's confidence in English language learning, to enhance their English language proficiency, to ensure the development of the students' content-based learning to a satisfactory level and to better the students' understanding of Western culture while fully developing the students' L1 (Chinese) and maintaining their knowledge of Chinese culture and identity (Qiang, 2000;Qiang & Zhao, 2000).
Many studies have been carried out on CCUEI (Fang, Wang, & Siegel, 2001;Pei, 1998;Qiang, 2000;Qiang & Zhao, 2001;Zhang & Pei, 2005;Zhao & Qiang, 2002), and nearly all of them focus on the outcome-oriented effectiveness of immersion.But there are very few research reports on what is occurring in the language learning process, and how the learners appropriate the English language.This research aims to address the following questions: 1) What are the mediations students employ in peer learning processes?
2) How do the mediations function in peer learning?

Mediation and Activity
According to Lantolf and Thorne (2006), mediation refers to "the process through which humans deploy culturally constructed artifacts, concepts, and activities to regulate (i.e. gain voluntary control over and transform) the material world or their own and each other's social and mental activities" (p.79).This means that the human mind is mediated through social interactions by semiotic tools, especially language, and mediation is the setting up of connection of the self to the outside world (Vygotsky, 1978).Human beings use symbolic tools (not only physical tools) to mediate and regulate their relationship with others and themselves, and change the nature of these relationships.Mediation results in children's internalization of language, knowledge and culture, which is "the reconstruction on the inner, psychological plane of the socially mediated external forms of goal-directed activity" and "anchored in the social activity" (Lantolf, 2000, p.13).
Tasks and activities have long been the interest of researchers engaged in second language teaching and learning.Tasks and activities serve as means to elicit language use from learners which are employed by both teachers and researchers (Ellis, 2003).According to sociocultural theories, activities are collective and social in nature (Vygotsky, 1978;Leont'ev, 1981;Engestrom, 1987, Tsui andLaw, 2007), and activities are not only cognition-based, but also include cognitive, executive, evaluative, and emotional aspects (Bedny & Meister, 1997).From a sociocultural perspective, activity is not a single action but a unit of actions embedded in the activity system that are collective and social in nature (Vogotsky, 1978;Leont'ev, 1981;Luria, 1974), which is addressed by the activity theory and illustrated by the complex model of activity (Engestrom, 1987(Engestrom, ,1999)).Activity Theory has been further developed by Engestrom (1987) who proposes three more components of an activity system: the community, rules and the division of labor (see Figure 1. the complex activity system).
(Figure 1: The complex activity system is here)

Research Methodology
This research involved one English immersion primary school which was within the CCUEI Programs in Guangdong Province, where English immersion has been adopted since the establishment of the school in 2004.In September 2004, the school began to recruit students, and the English immersion program was introduced to the 5 classes of Grade 1.At present, the school has 21 English immersion classes with about 800 immersion students.Nearly all the classes in Grade 1, 2, 3, 4 and 5 are immersion classes.The time spent on immersion instruction occupies 20 percent of all the instructions given to the students.English is taught through social science, living science and English language arts.

Participants
The selection of the classes, the teachers and the students for this research was undertaken based on the following criteria: the teachers were willing to participate in the research; the classrooms were recommended by the principal with the consideration of warm and positive classroom atmosphere; and the students were selected based on a wide range of language ability such as higher, intermediate and lower proficiency levels in English.
This represented a large variation among the sample students.Class 4 of Grade 4 in this school was one among the first classes to start the immersion program in September 2004.Six students from this class were chosen as research participants.Most of these students were around 10 years old.The selection criteria aimed to ensure that gender balance and a range of language proficiency were represented.Selection interviews of the candidates recommended by the teacher were conducted in English to ensure that students of high, intermediate and lower proficiency levels were represented.

Data collection
The data of this case study came mainly from classroom observations and interviews.In the classroom observations, audio/video recordings and field notes were used to keep a record of in-class lessons, the students' after-class activities, and the teacher's behavior.Physical artifacts were also collected as a supplementary means.Digital recorders were attached to these 6 students in the class.These recordings were transcribed by the researcher.These methods of data collection offer the opportunity to obtain rich and naturally occurring thick data, which can grant an emic picture of this study.

Data analysis
Engestrom's (1999) framework of activity system formed the basis of the data analysis.The students in the class governed by class rules together with the teacher's instruction, guidance and assistance adopted various mediations in order to improve their language proficiency in their learning activities.For the analysis, techniques of spoken discourse analysis were used to analyze the transcriptions of the data.Mediations emerging from the data were then systematically categorized.

Results and Findings
The focus of this study is on the mediations in the students activities.In this regard, a detailed analysis has been done on the mediations from the peer resources, peer talk in particular in the activity of talking about living things.

Peer talk as mediation
The following excerpts illustrate how peer talk functions as important mediations in the activity from different aspects: 1) peer prompting and waiting, 2) reciprocity of peer assistance, 3) active engagement and non-verbal expressions, 4) active involvement, self-correction and voice volume, 5) peer talk expansion and creativity, 6) coinage and private speech, 7) resistance and inquiry, 8) comments and social function of private speech, and 9)code-switching, turn-taking and thinking.Owing to the limited space where I cannot show all the analysis on the talk of the three pairs, only excerpts from the talk were extracted and shown here: Excerpts 1 to 2 are from the talk by Pair W and N; Excerpts 3 to 5 are from the talk by Group Y, C and X; and Excepts 6 to 9 are from the talk by Pair L and H.

Introduction about the activity
As this was an English immersion school, living science was one the content subjects for the immersion.In the science class, the teacher introduced to the students the living things and the non-living things, as well as the criteria of how to judge the living things and non-living things.After the science class, the teacher assigned the students to talk about living and non-living things in the evening self-study class (as this was a boarding school), which included 8 pictures (a butterfly, a robot, the Milky Way, an electronic dog, an orange, a flower, a rock and a sea creature) in the textbook on the theme of "growth and change".Four features were identified in living things for students to judge them by: 1) can it grow?2) can it produce babies?3) can it produce or get nutrition?4) can it react to the stimulus around it?In this excerpt, W and N are trying to reason why a butterfly is a living thing.In Line 5, when N hesitates and pauses as he can not formulate and articulate the reason, W prompts in line 6 in a low voice with the two words needed "it can" in N's utterance.N picks up the prompted word "it" in line 7, but gets stuck there and still has difficulty to continue.W prompts again in line 8 with the word "can".N at this moment picks up the prompt and articulates the reason in line 9 with a pause of 6 seconds.From this episode, we can see that when N has some difficulty in the reasoning, W assists him with prompts patiently.When N comes to pick up the prompts in the end and gives the reason, W gives enough time to wait for N to talk.Peer prompting and waiting serve as mediation in their activity.

N: grow (correcting W)
In this excerpt, W and N are talking about whether the robot is a living thing or a non-living thing.In line 3 when N shows his difficulty with his stuttering filler "er-" and hesitation pause, W offers him prompts of "living thing" in line 4.In line 5, N picks up the prompts and completes his utterance with the question.In line 8, when W has some erroneous pronunciation with "grow", N corrects her and provides her with the correct pronunciation in line 9.This excerpt shows that students show their multi-competency in their learning, and reciprocity emerges in the student peer assistance.In this excerpt, the first part from line 1 to line 8 is mainly between Y and X, however, C actively involves himself in the talk in line 4 prompting X with "why" when he finds X stuttering with difficulty by the pause.X picks up in line 5 and Y explains the reason in line 6, where she shows difficulty in articulating the reason of "reproducing babies."X prompts her with the word, and Y picks up though confused with the words "produce" and "reproduce".In line 9, X projects the talk to C and asks him whether the robot is a living thing or not.C, influenced by the L1 in mixing that answer of negative and positive, gives the erroneous answer to the question with "No, it is".X corrects him and has the stress laid on the word "isn't" in order to direct C's attention to the form.C immediately picks up in line 12.In the last 3 lines C, Y and X are co-constructing the reasons.What is interesting here is that in line 15, Y corrects herself and finds the correct form to express in the talk.This excerpt shows us 1) that students are trying to involve the other party into the talk in the process, 2) that when peers correct their partners, they utilize voice volume to stress or emphasize and direct their attention consciously, and 3) that self-correction also occurs in the peer talk.

This is a flower
In the excerpt, the students, not following the instructed pattern of asking and answering the questions, are conducting a free talk about the pictures.Through the expansion, they are using what they have learned in the situations they can find in the utterance "Do you know their names?",and even creatively using the language by creating a new phrase "water animal" for the "jellyfish" in line 10 and 11. 12. H: Really?
In this excerpt, L and H are trying to manipulate the free language use by giving comments to what they see and expressing their affect through such the utterance as "Oh, it's beautiful."In line 7 when H hesitates to give the answer, L prompts with an answer in a low voice.H refuses to pick it up as he is not sure about the answer, so he just answers "I don't know", which surprises L, as this is unusual.When L rephrases in line 11 the utterance with "non-living thing", H gives his inquiry with "really?" in line 12, which shows that he is not taking in what he is not sure of.Therefore, the students' free language use begins to emerge in their talk, and agency also shows in their talk.In this excerpt, in line 3 and line 7, L and H use code-switching to express what they cannot express in English in order to maintain their talk.Similarly, in line 17, L seeks help to express "terrible" in English, and H tells her to code-switch in line 18.In line 19, L uses code-switching as H prompts. L prompts in line 8 in a low voice when she finds H has some difficulty in expressing himself, and H picks up the prompt immediately.The talk about the assignment appears to be over.However, H and L seize the chances to use English as much as possible, thus expand their talk.Then L uses "oh?oh!" as a transition in line 10 to signal the change of the theme, and she asks about the time in line 11, then they go on to talk about living things from line 14 to 24.Students not only prompt the language expression to use, but also prompt turns for each other.In line 23, L reminds H of the turn he should take.What is exciting in this talk is that in the last part from line 24 to 35, L and H are co-constructing what else is living thing and expanding their talk in authentic thinking from concrete concepts such as "trees, flowers" to abstract concepts such as grass and animal.In the end, they connect humans to what they have learned about living things and non-living things, which shows that they are applying the abstract concepts into practice, connecting the remote to their intimate situation.In sum all these excerpts show that peer talk and peer assistance form the most important source of mediation in the student activity.

Comparison among the three pairs/groups
Aside from the mediation address above, the dynamics and interaction between the elements of the activity is to be explained through the comparison made among the three pairs, based on the components of the complex activity system: object, subject, mediation, rules, community, division of labor, outcome, peer relations in the peer talk and the strategies used respectively (see Table 1).
(Table 1 is here) The students used a variety of strategies in their talk.Self-repetition and pauses were used to gain time for thinking and phrasing what they wanted to express.Peer correction and peer prompting scaffolded their talk where the low volume of their voice showed their efforts to minimize the asymmetry in their role relations.Code-switching was used when they could not express themselves in English.Self-repetition and self-correction were used in their talk together with peer correction, peer prompting, and giggling as a reminder.X and Y were dominating the talk without giving C any chance to participate in their talk.C listened to them carefully and skillfully used the word "yes" with the rising tone to initiate his participation in the activity.In H and L's talk, correction and peer prompting were also used in their talk to give each other scaffolding in the talk where the volume of voice was also lowered to show respect to their partners.H used coinage of the pronunciation for the new word to help him memorize the word.Code-switching occurred when L and H found they could not express themselves in English.Besides code-switching, tone-switching was used by L, for example she used "en" with the rising tone to show her surprise and disagreement, just as in L1; she used "hen" to show her irritation.Lastly, L used "oh?oh!" to mark as a sign poster for the ending of one picture and shifting to another picture.

Categories of mediations
Besides the mediations coming from the peer resources, peer talk in particular, other mediations are evidenced in this study, though no detailed analysis can be given here because of the limit of space.In this study, results show that in the activities the mediations include the following categories: teacher resources, peer resources, e realia, content, semiotic artifacts and other semiotic artifacts.Teacher resources include teacher's talk such as instruction, questioning, elaborating and prompting, teacher's tone and volume of voice; teacher's gestures; teacher's eye contact; teacher's presence and teacher's silence.Peer resources involve peer talk such as tutoring, questioning, elaborating, collaborating and prompting; peers' tone and volume of voice; peer play such as kidding, chanting and play with words; and peer gestures and actions.Semiotic artifacts contain texts, textbooks and the language used.Other semiotic artifacts comprise of worksheets, PPT, pictures and maps.Subject content plays a very important role in mediation as the present study involves an English immersion context.In addition, genre "is likely to affect the linguistic form" (Ellis, 2003, p.92), which enable students to accomplish their activities.The type of mediations can be summarized as follows (see table 2): (Table 2 is here)

Theoretical framework revisited: representations in this study
The variety of mediations function as the most important part in the activity where Engestrom's model of activity system with its seven correlated components of object, mediation, rules, community, division of labor, object and outcome serves as the analytical framework.The subject refers to the peer students.When the students participate in the activities, the mediation refers to the psycholinguistic and other semiotic tools, as described above.Since the current study examines peer talk, the subjects in the activity system involved are peers either in pairs or in groups.Rules refer to the relations, norms, conventions and expectations in the community.Peer role relation is a very important aspect of the rules.Community in the current study refers to the pair or the group, the class and the school, as the study pertains to a boarding school and students stay at the school most of the time except during weekends.Classroom is the place where students have their class studies and do their self-studies in the evenings.Division of labor incorporates how the students divide their work in the activity.Object refers to the task-driven/related goals of achievements: the language, the content, and the culture behind the language.Outcome refers to participants' goal-directed achievements, the participants' discourse, and the affective satisfaction of collaborating with peers.Tasks are designed with the goal of eliciting the language in use (Ellis, 2003) and peer talk in the current study is selected as an indicator of the outcome.Peer talk, the indicator of outcome serves as a type of mediation at the same time.Figure 2 illustrates the representations which grounded from the data in the current study in Engestrom's (1999) framework.
(Figure 2 is here)

Dynamic relations among the components of activity
The relations among these components can be described as follows: Peer students, who are regulated by the rules in the community, divide their works, and deploy the devices of mediation to try to achieve their goals in the object.The goals refer to acquiring English language proficiency, mastering the content, and learning the culture embedded in the content subjects.Through the division of labors and through the process of their effort, their goals will be transformed into the outcome, which include the goal-directed achievement, or the process of the experience of achieving the goals, and the affective dissatisfaction or satisfaction of collaborating with peers, and the peer talk.As the nature of activity is situated, dynamic, dialectical and internally correlated, these arrows show the situatedness, the dynamical and relational nature of activity, which means that the change of one component may cause changes in the others.

Discussion
This study shows results consistent from the previous studies that students use language and physical tools such as realia as mediational tools in their peer learning processes.The study also expands the previous literature on mediation by discovering that in the student peer activities, mediation covers a wide range of areas from teacher's talk to peer talk, content, genre, semiotic artifacts and other semiotic tools, and that mediation includes task, teacher talk, peer talk, subject contents and other semiotic tools.Peer talk demonstrates multifaceted functions as mediation, which together with linguistic resources constitutes an important resource aside from teacher talk, where subject content serves as an important mediational tool in the activity and the content-based immersion programs, offering students opportunities of language use in meaningful interaction.
This study investigates the students' peer learning process through the lens of sociocultural theories, which offers insights into the immersion practice.It investigates mediational tools which students employed in their activities from an emic perspective, which provides a better understanding of the English immersion programs in China.It enriches practices of the classroom interaction among learners.The findings generated from this study of the English immersion programs present an in-depth analysis of the spoken English language learning process and offer valuable insights for pedagogy and teacher professional education.
Some implications of pedagogy and teacher education can be drawn.First, when designing tasks, teachers should be able to predict and have clear expectations of the dynamics of activities, as students conduct the activities based their own interpretation of the goals, the rules, the mediation and the setting, which results in different outcomes.Second, peer talk function as a most important mediation in students' English language learning, which, as an indicator, can reveal what is happening in the process of performing the task, and equip teachers to explain the reasons for the students' achievement.This can provide teachers with a clear and more complete picture of students' performance in the activities.Third, the fact that the change in one component within the activity system leads to different activities may give insights into the teacher education that the feasibility of variation of activities can be a way to satisfy the needs of different students.
The limitation of the study is that it investigated only the student spoken language, i.e., their peer talk in the activities.
Further research can be done on the aspects of student written language discourse in their L2 proficiency development.

Conclusion
Mediations include the multiple resources such as teacher resources, peer resources, physical tools and semiotic tools.Peer talk functions as a most important source of mediation, where students demonstrate the deployment of communicative strategies such as prompts, and scaffolding help.Subject content serves as sources of mediation in the activity and the content-based immersion programs offer students opportunities of language use in meaningful interaction.The discourse mode students use to accomplish the task such as 'object description" , narrative, expository, argumentational.

Figure 1 .Figure 2 .
Figure 1.The Complex Activity System (Engestrom et al., 1999, p. 31) ni shuo ta hui sheng zhang de {You say it can grow} (prompting in a low voice).In this excerpt L and H are talking about whether orange is living thing or not.In line 4 and 5, private speech occurs when L tries to find out what she wants to express.H prompts in line 6 in L1 in a low voice, but L cannot pick up the prompt as she does not know how to say "grow".L asks H for help in line 7, and H prompts in line 8 with a wrong word "gray".However, though the prompted word is not correct, it reminds L of the correct word.In line 10 L finds the correct form and corrects H immediately.In line 11 H repeats after L and tries in line 12 to internalize the word by building a connection between this new word and his previous knowledge of pronunciation: he coined the sound in mandarin as a strategy to learn the language.
4.1.8Resistance and inquiry Excerpt 7:Talk about Milky Way by Pair L and H L= Student L H= Student H 1. H: Is // this a 2. L: // What is this.3. H: It's Milk Way. 4. L: Oh, it's beautiful.5. H: Yes, it is.6. L: Is this the living thing?7. H: en: 8. L: No, it isn't (prompting in a very low voice).9. H: I don't know.10.L: en? (surprisingly) 11.Oh he is non-living thing.

Table 1 .
Comparison among the three pairs

Table 2 .
Types of mediation