ENHANCING COLLABORATION THROUGH CHAT IN ESP:
A CONVERSATION ANALYSIS

by Dafne González & Leticia Esteves
Universidad Simón Bolívar
Caracas, Venezuela
dygonza @ usb.ve   lesteves @ usb.ve

Abstract

Text and/or voice chat rooms are ideal spaces for interaction, collaboration and negotiation of meaning. However, they have not been fully exploited in language teaching. This study shows the results of a conversation analysis of 24 transcriptions of text chat sessions where architecture students were carrying out collaborative group work. The researchers wanted to discover the discourse patterns and conversational strategies used by the students in this online context to study their possible repercussions in English learning. The results suggest that the students were building the kind of discourse that is thought to lead to language learning.

Introduction

            Have you thought about the benefits chats can offer our English teachers and language learners? Would you like to know how to include chats as one of your tools to share information with your students, have them work collaboratively and interact with one another, help them engage in real tasks, promote negotiation of meaning,  focus on their own learning, and practice the target language?  In this article, we will relate our experience using chats in our English for Architecture class at the Simón Bolívar University (USB) in Caracas, Venezuela, using data collected for the doctoral dissertation of one of the co-authors (González, 2004).

The aim was to incorporate an online unit in the last level of the English for Architecture courses at the USB with the intention of studying the ramifications of the use of different web tools in the learning of English. We decided to incorporate chats as part of the unit because they allow for group work, resemble regular face-to-face (f2f) conversations (González, 2003), and give students freedom to express themselves at their own pace. In addition, conversations can be automatically saved in the form of transcripts for further analysis in class and for research purposes.

 During the implementation period, we noticed the online chat sessions were fundamental in the accomplishment of the course objectives, and the students also reported this in their self-evaluations and unit evaluation. Therefore, in this article, we will concentrate on the analysis of the chatlogs.

Context of the study

The online unit was designed for 56 students of architecture in their third year of studies at the university and their last English course (Level 6). Their level of English varied between lower and upper intermediate. One of the units of this last course is dedicated to the topic of Modernism and that made us think a focus on modernism in Valencia, Spain might interest our students. The aim of the unit was to describe the characteristics of Modern Architecture in Valencia, and the final task was to write an essay on the characteristics of the modernist architecture found in that city.

In this blended course (50% in the f2f classroom and 50% online), there was an online moderator in Spain, while two other teachers were the f2f facilitators in Venezuela.

The training of the f2f teachers was itself an example of e-learning, considering that it was delivered through chat and e-mail. Using these tools, the f2f teachers learned about the different programs and applications needed to run streaming videos, carry out chats, send voice messages, write summaries, share images and descriptions, and write journals and essays in online environments.

At the same time, we booked a computer lab with Internet access for our classes, and designed the 7-week online unit. The class activities, materials, and assessment procedures were very carefully planned by the e-instructor and later revised by the f2f teachers in Venezuela.  A Yahoo Group (YG) was created to be the asynchronous communication center, and a web site (http://www.geocities.com/dygonza/unitindex.html) was designed as a portal.

Two international colleagues, from Argentina and Denmark respectively, were invited to observe our online classes. They joined our Yahoo Group, participated in some of the chat sessions in the YG, and had access to all the teacher-created material as well as to the students’ work.

Implementation

The first activity in our online unit was a diagnostic survey to detect the students’ skills in the use of web tools. The results showed that many students did not have Internet access and had no experience in the use of web tools for academic purposes. Fortunately, our students were able to overcome the technological barrier through clear instruction and systematic training in the computer lab.

In this preparatory week, the students got acquainted with the Yahoo Group and the procedures for the online unit, and expressed their expectations in a journal entry.  The students also watched a video about Modernism in Barcelona, Spain, to activate their previous knowledge on the topic. (They already knew about Gaudi, the major representative of Modernism in Spain, from their Theory of Architecture content course). Pre-viewing and post-viewing interactive exercises and activities were designed using Hot Potatoes to introduce new vocabulary and structures presented in the video.

The use of chat, for collaborative learning, was carried out in weeks 2 to 5 through a jigsaw reading activity where each participant in a group had a piece of information unknown to the others and fundamental to the final task. Each group was made up of 5 students with different levels of English and different levels in their knowledge of architecture. In the Yahoo Group we had posted a folder for each group with five different illustrated descriptions of buildings in Valencia (for example, the one at http://www.geocities.com/dyg_usb/id3126/group_5a.html). Each group participant had to choose one building, and write a summary highlighting the most typical characteristics as well as the architectural elements found in each.

With their summaries, the students got together in a chat conference, to share and discuss the elements and features identified in each of their buildings. At the end of the chat, each group was to start a collaborative summary about the characteristics of the 5 buildings discussed. They used e-mail to circulate their summaries until they were satisfied with the results.

The next step was to form other groups and participate in a second chat activity. These second groups comprised one student from each of the previous groups.  Each member came to the new groups with all the data they had gathered in the collaborative summary. As a whole, each group had information on the 25 buildings selected for the unit, which had to be discussed to get at the general characteristics of Valencia Modernist architecture as expressed in those buildings.  An illustration of the group formation will probably make the process easier to understand: http://daf4.free.fr/wiaoc/groupformation.html .

Using the information obtained in the chat, and as an individual assignment, students had to write a final essay describing the characteristics of Modernism in Valencia. (See this example: http://daf4.free.fr/wiaoc/IndEssay-Sandra.doc ).

Week 6 was dedicated to overall assessment, since each activity had been evaluated during its implementation, with the use of different rubrics and checklists. Students filled in an online questionnaire to evaluate their learning gains in each of the completed activities: journals, chat, summaries, online consulting hours, etc. They also evaluated their participation in group work: http://dafnegonzalez.com/id3-124-05/content/teamwork-eval.htm. The whole unit was evaluated with a journal entry where students answered questions such as:

·         What did you like about the online unit?

·         What problems did you find?

·         Did you find enough help from the face-to-face and the online teachers? Explain.

·         Was this unit helpful to practice your English? Why?

·         What do you think about the content of the unit?

·         What else would you like to say about this unit?

Finally, the students completed a poll about the characteristics of Modernist architecture in Valencia in the Yahoo Group. It was a checklist where they had to select those characteristics present in the buildings they had been discussing. In a nutshell, the chats were the core tasks of the unit, but it is important to note that they were leading up to the final task and not just isolated activities.

The methodology

The research as a whole was a combination of Case Study (Kemmis & McTaggart, 1988) and Action Research (Burns, 1999). The case study was the unit with all of its components (students, teachers, materials, external observers, web tools, political, social and economical contextual variables, etc), and the action research was the implementation of the online unit per se.  It was carried out in a natural context: the classroom environment with intact groups.  It was collaboratively oriented since the design involved different people, and the research approach was mainly qualitative with some quantitative information in the form of percentages to reinforce the results of the qualitative analysis. 

The study included the analysis of all the activities completed during the unit, f2f teachers’ reports, the reports sent by the two external observers, and a conversational analysis of the chatlogs  The objectives of the chatlogs analysis were in the first place, to discover the patterns and conversational strategies used by students while chatting online; and secondly, to examine any possible effects on English language learning. In this article we will focus on the conversation analysis of the chatlogs.

Analyzing

The conversations in a chat session, where students meet to carry out a structured task, are not very different from regular f2f conversations which had originally been the target of Conversation Analysis (CA). Given the nature of the chat as a social interactive space, and since CA is a method used to analyze different discourse exchange systems (Schegloff et al., 2002), we thought it would be an appropriate method to analyze our transcripts.

CA describes the speech exchange system, the distribution of power, and the structure and sequence of the conversation.  There are different speech-exchange systems such as regular f2f conversations, interviews, business meetings, classroom speech, and others; and among those, each system has different structures for the organization of turn-taking while taking part in a conversation (Markee, 2000). Other authors state that ordinary conversation is considered the basic form of exchange since it provides space for equalitarian distribution of power among the speakers, and, according to Markee, this kind of situation would be the ideal context for language learning to take place, since this equalitarian context tends to promote negotiation of meaning which is a paramount element in language learning.  Markee adds that in the classroom we can only observe this happen in group work, so we thought the chat would be a suitable medium to observe how discourse was used by students in this environment, and if there had been the four kinds of negotiation of meaning described by Markee.  In other words, we wanted to study the discourse in this new virtual environment.

Data Analysis

We checked that our transcripts met the requirements needed for CA (Heritage 1989): structure (given by the tasks assigned), context, and naturalness (this is the kind of task our students usually accomplish in the f2f classes). They also met the requirements of heuristic research (Seliger & Shohamy, 1989): data validity (saved in the form of transcripts that would be available and accessible to the researchers at any moment), and. reliability since they are used in their original form with no editions or changes, and we used all the transcripts for our analysis.

To analyze the transcripts, we first read all the chatlogs to find common patterns and differences in the structure of the group work carried out by the students.  Then, we analyzed the sequence of the transactions, negotiation of meaning, turn taking, topic initiation and termination, and other aspects that we found could be relevant for the study.

At the same time, we were paying attention to how our findings in this new media could be related to regular f2f conversations regarding the distribution of power, which is ultimately the aim of conversation analysis.

Aspects highlighted by the analysis:

General structure and sequence of the interaction

• Most of the typical principles of a speech exchange system of an ordinary conversation were used. The students used strategies to overcome the restrictions imposed by the medium.
• The structure of the interaction is characterized by defined sequences; i.e. an opening of social actions (welcome and greetings), work logistics, and discussion of the content topic, sprinkled by social and technological threads, or logistics of the work, but always returning to the topic in discussion. Finally, the farewells are direct, and, on very few occasions, students anticipate they are leaving the chat (pre-farewells), for example: “I need to go”.

Power and autonomy

• There was an equal distribution of power among the participants. We think this was so because they needed the information that others had in order to complete their task. When one student was not contributing, the others asked him or her to share the information needed.
• Students exhibited autonomy in solving the technological and academic problems they confronted. The last resource was the teacher.
• When the teachers intervened in the conversation without having been asked for help, they were ignored. The students let them know directly or indirectly that they had everything under control, and in some cases the teachers were informed that their intervention was interrupting the thread of the discussion.

Turn taking distribution

• The turn-taking was free. The students did not have a sequence nor a predetermined extension.
• Turns were not pre-assigned, so strategies of assignment of turns were used. Students indicated the name of the person to whom a comment or question was directed; otherwise, the comment or question went to the complete group, so that anyone could take the floor.
• The distribution of turns was equitably distributed, and organized through a natural process of passing the turn to others, or individuals taking their turns as they thought necessary.

• No student participated more than the others, and the teacher-student rate of talk shows that students held the power of the conversation. For a total of 4,798 turns, students took 88.04% (4,224 turns), while teachers participated only 11.96% (574 turns; see the table and pie chart at http://dafnegonzalez.com/pics/turns-table.jpg and http://dafnegonzalez.com/pics/turns-pie.jpg).

Coherence and Cohesion

• Different strategies were used to establish and maintain the coherence and cohesion of the discourse. (Extracts taken from the chatlogs can be found at http://daf4.free.fr//wiaoc/extracts-chatlogs.html).
• There was the presence of the regular structure of an ordinary conversation in adjacency pairs (Sacks, 1995), but not with the regularity that would have been observed between two people in an f2f conversation. The nature of parallel and not serial interaction observed in chats often prevents adjacency pairs from being observed from one turn to another.
• Participants made use of different paralinguistic strategies to overcome the restrictions of the medium. We saw the use of uppercase letters, punctuation signs, onomatopoeias, icons and emoticons to express diverse feelings and moods, as well as to stress or emphasize their discourse.

Negotiation of meaning

• Students negotiated meaning regarding: language (unknown words), technology (how to go about the use of the tools), content (architectural content) and logistics (the processes to carry out the tasks). 
• Markee (2000) mentions 4 levels of negotiation of meaning, depending on how many turns away from the problem the repair occurs.  First position repair (a self correction), is not essential for language learning, according to Markee.  Second position repair is where the communication problem is solved in the turn immediately after it was presented. Third position repair is resolved in the third turn of a repair sequence, fourth position repair in the fourth, and so on.  Markee says that fourth position repair is almost never found in classroom situations. In contrast, we found all 4 levels of negotiation of meaning to be present in our transcripts, those of second and third position being the most frequent.

• Students made very little use of the mother tongue. It was used only to translate a term, or to call the attention of a participant who was not, according to the group, following the norms that they had implicitly set.

Content, language and technology

• The topic under discussion and the task to carry out were the central axes of the interactions. Participants concentrated on the architectural content, and English was the means to do it. They only concentrated on the language when there was a communication breakdown, and they immediately looked for a solution. Grammatical errors or misspellings were not taken into consideration in most of the cases. The students’ interest in the message was paramount; thus, they behaved like authentic audiences as characterized by the literature in language acquisition.
• When students considered one topic had been sufficiently discussed, they introduced a new one, as opposed to in an f2f classroom setting where the teacher is the one who introduces new topics.
• The chat as a technological support was the most appropriate medium to carry out online collaborative group work because of the immediacy of the feedback which allowed for negotiation of meaning to take place in a natural way and for the ease with which transcripts could be later analyzed without losing any detail of the transactions.

In general, we can say that students collaboratively built up their vocabulary and their knowledge of architecture as if trying to solve a big puzzle in which each student had a piece needed to complete the whole and the chat was the space where the pieces were laid. Students showed in all the transcripts that they spontaneously learned how to interact using the discourse in this new medium.

Conclusions

            Through this CA of the chatlogs, we observed how our students took advantage of the structure of the conversations as a resource that allowed them to describe, discuss and understand the characteristics of Valencia modernist architecture using the target language in a chat.

            We realized how, with little teacher participation, students could start a sound discussion and complete their tasks solving each problem they found through the use of the target language and technology. The students final compositions evidenced how group discussions through chat contributed to vocabulary building and the acquisition of new knowledge. 

We interpret the results of this study to suggest that student interaction based on the performance of small-group tasks through chat sessions is quite close to the open organization and characteristics of regular f2f conversations. The students’ previous knowledge was activated and negotiation of meaning took place throughout our data.  Our observations could be of great interest not only to those specialized in language teaching but also to researchers interested in the development of theories of learning since the data analyzed presupposes the construction of the kind of balanced discourse that is thought to promote language learning.

References

Burns, A. (1999). Collaborative Action Research for English Language Teachers. Cambridge: Cambridge University Press.

González, D. (2003). Teaching and learning through chat: A taxonomy of educational chat for EFL/ESL. Teaching English with Technology,  3 (4), October 2003. Retrieved May 1, 2006 from: http://www.iatefl.org.pl/call/j_review15.htm.

González, D. (2004). Dilemas de la evaluación del aprendizaje de Inglés con Propósitos Específicos a través de soportes electrónicos: Estudio de un caso. Unpublished doctoral dissertation. Universidad de Valencia, Spain.

Heritage, J. (1989). Current developments in Conversation Analysis.  In D. Roger & P. Bull (Eds.), Conversation: An Interdisciplinary Perspective, pp.21-47. Avon: Multilingual Matters.

Kemmis, S. &  McTaggart, R. (Eds.) (1988). The Action Research Planner. Geelong, Victoria: Deakin University Press.

Markee, N. (2000). Conversation Analysis. Mahwah, NJ: Lawrence Erlbaum Associate Publishers.

Sacks, H. (1995).  Lectures on Conversation. Oxford: Blackwell Publishing.

Schegloff, E.A., Koshick, I, Jacoby, S., & Olsher, D. (2002). Conversational Analysis and Applied Linguistics. Annual Review of Applied Linguistics. 22: 3-31.

Seliger, H. W., & Shohamy, E. (1989). Second Language Research Methods. Oxford: Oxford University Press.

Editor’s notes:

This presentation was made as a regular session at theWebheads in Action Online Convergence on November 18, 2005. 

• The URL for the presentation is here: http://daf4.free.fr/wiaoc/chatpres.html.
• The session took place in the Elluminate presentation room at Learning Times.  A recording was made and can be heard at http://home.learningtimes.net/learningtimes?go=1042112.