Enriching Literature Reviews with Computer-Assisted Research Mining. Case: Profiling Group Support Systems Research

Johanna Bragge1, Sami Relander1, Anne Sunikka1 and Petri Mannonen2

1Helsinki School of Economics, 2Helsinki University of Technology

1firstname.lastname@hse.fi; 2firstname.lastname@soberit.hut.fi

1. Introduction

The literature review is a key concept within the scientific process of publishing academic research

articles. Here, we assume familiarity with the issue, which according to Hart [1, p. 13] is simply about

selecting “available documents on the topic written from a particular standpoint to fulfill certain aims or express certain views on the nature of the topic and how it is to be investigated, and the effective

evaluation of these documents in relation to the research being proposed.” In addition, to augment this

definition, Webster and Watson [2] state that a good, high-quality literature review is complete and focuses on concepts, not on authors. The literature review has many functions, perhaps the most notable of which is that it positions the study in question within an existing stream or body of

literature and demonstrates that the author is familiar with the most important prior works related to his or her field. Additionally, it shows that the author can synthesize and link existing works in a meaningful way to yield new insights and open research gaps. The literature review is thus a standard section in every research article, and its length is determined on the purpose of the article: from one chapter to whole stateof- the-art review articles with extended length. The former ones are by far the most common, especially in information systems research, which is a relatively young and multidisciplinary field, thus adding complexities to assembling thorough reviews.

With the advent of various online journal databases, the amount of literature at the fingertips of researchers has exploded. As a result, the classic literature review has had to adapt to accommodate these changes. No longer is the task about retrieving selected hard copies from the library, but more about mastering various search engines and indexes that categorize papers.

Recently, the literature review concept has received increased attention due to these technological changes and developments. Insightful text mining and information visualization tools are being developed to help the researcher in profiling, mapping and visualizing knowledge domains [3, 4]. Some of these tools are designed for fielded research abstracts imported from scientific databases, and some tools transform free form text into data that can be analyzed for information extraction [see e.g. 5 for a list].

Building on from the current state of affairs regarding the literature review and the issues it

presents, we have a twofold objective. Firstly, we seek to argue for an extension to the traditional notion of a literature review into the research profiling approach as initially presented by Porter and his colleagues [4]. This is an interesting new direction with potential to overcome at least some of the main challenges and pressures we face regarding the literature review and its place within research. The emphasis is not so much on a bibliometric perspective (statistics related to the production, distribution and usage of documents [6]), but more on the viewpoint of how to actively use

research profiling to uncover research gaps and/or new, emergent scientific domains by focusing on the content of prior articles. This should aid researchers to position the literature review away from a mere passive tool to an active means of generating and refining new ideas and concepts on which to conduct research.

Proceedings of the 40th Annual Hawaii International Conference on System Sciences (HICSS'07)

0-7695-2755-8/07 $20.00 © 2007

USING SYNCHRONOUS ONLINE PEER RESPONSE GROUPS IN EFL WRITING: REVISION-RELATED DISCOURSE

Mei-Ya Liang                  
National Central University

In recent years, synchronous online peer response groups have been increasingly used in English as a foreign language (EFL) writing. This article describes a study of synchronous online interaction among three small peer groups in a Taiwanese undergraduate EFL writing class. An environmental analysis of students’ online discourse in two writing tasks showed that meaning negotiation, error correction, and technical actions seldom occurred and that social talk, task management, and content discussion predominated the chat. Further analysis indicates that relationships among different types of online interaction and their connections with subsequent writing and revision are complex and depend on group makeup and dynamics. Findings suggest that such complex activity may not guarantee revision. Writing instructors may need to proactively model, scaffold and support revision-related online discourse if it is to be of benefit.

INTRODUCTION
As a learner-centered process approach to second language (L2) writing, peer response has been widely adopted and studied since the 1990s (Hyland & Hyland, 2006). The dialogic nature of peer response seems to foster multiple support systems (Hyland, 2000) and communicative behaviors (Villamil & de Guerrero, 1996). L2 research has shown that peer response can increase chances for meaning negotiation and language practice (Lockhart & Ng, 1995; Mendonca & Johnson, 1994), encourage collaborative reading and writing (Tsui & Ng, 2000), and promote writing revisions (Berg, 1999; Mendonca & Johnson, 1994; Min, 2006, 2008; Stanley, 1992). These interactive practices appear to draw upon and enhance interactional and writing skills.
Recently, online peer response has also been used as an alternative to face-to-face (F2F) communication. Online peer response that blends spoken, written, and electronic communication can promote student motivation, participation, and collaboration (Warschauer, 1996, 2002), an awareness of audiences (Ware, 2004), a critical analysis of linguistic, negotiation, and writing features through the use of printouts (DiGiovanni & Nagaswami, 2001), as well as frequent use of peer ideas in revisions (Hewett, 2000; Tuzi, 2004). Applying electronic technologies in L2 writing classes, research has set out to explore such issues as effective uses of synchronous online peer responses and revisions (Hansen & Liu, 2005). While training procedures for improving revision-related discourse have been proposed in composition studies (e.g., Berg, 1999; DiGiovanni & Nagaswami, 2001; Hansen & Liu, 2005; Min, 2006, 2008; Rollinson, 2005; Stanley, 1992), relatively few studies have provided adequate frameworks for describing the nature of L2 interaction in synchronous online peer response groups. To further this research, it is necessary to understand how L2 learners interact in synchronous computer-mediated communication (SCMC).
This study explores L2 students’ synchronous online peer response in an EFL writing class. Research on revision-related L2 discourse in SCMC is reviewed as are the ways that revision-related online discourse might facilitate L2 writing and revision processes. A coding scheme (Liang, 2008) is employed in analyzing peer discussions about their writing. The author concludes with suggested strategies for supporting and facilitating synchronous online peer response groups.
Mei-Ya Liang Synchronous Online Peer Response Groups
Language Learning & Technology 46

LITERATURE REVIEW
L2 Interaction in Face-to-Face and Online Contexts
Interactive processes of L2 peer response have been characterized from two perspectives: (a) meaning negotiation and (b) collaborative learning. Meaning negotiation refers to interactive feedback (e.g., confirmation checks, comprehension checks, clarification requests) that deals with unclear messages. Varonis and Gass (1985) proposed that “the types of linguistic activities that occur in NNS-NNS conversations differ from those in other types of discourse, particularly with respect to the negotiation of meaning when there has been an actual or potential breakdown” (p. 71). During negotiation, L2 learners notice linguistic features and modify messages (Pica, 1996). Long (1996) also suggested that speaking and writing partners in an L2 environment not only provide comprehensible input, but also facilitate learner output through meaning negotiation and error correction. Foster and Ohta (2005) further found that, in L2 peer interaction, modified output in the forms of self-correction and supportive talk was more common than other-correction and meaning negotiation.
Following Vygotsky, sociocultural theorists, such as Donato (1994, 2000) and Swain (2000), have highlighted peer assistance and mutual scaffolding in collaborative dialogues. Van Lier (1996, 2000) has also specified the importance of contingency in collaborative discourse through the concept of “intersubjectivity”—that is, a shared social context for interaction where “participants are jointly focused on the activity and its goals, and they draw each other’s attention into a common direction” (van Lier, 1996, p.161). He argues for the ecological value of learners’ meaningful actions in social spaces of interaction (van Lier, 2000). De Guerrero and Villamil (2000), for example, had pairs of students role-play as writer and reviewer, helping to give them a sense of personal investment in the peer revision task. Learners may even use an L2 to establish and maintain social relationships while discussing content. By considering social contexts, sociocultural perspectives enrich our understanding of L2 learning, including speaking, writing, and collaborative dialogues.
Text-based SCMC brings with it instant messaging, which leads to similar discourse functions and negotiation sequences to F2F communication (e.g., Blake, 2000; Smith, 2003; Sotillo, 2000). On the other hand, there are some differences: (a) technical actions in various forms of keyboard strokes (e.g., emoticons and punctuations) replace nonverbal cues (e.g., gestures and facial expressions) and paralinguistic features (e.g., pitch, volume, and intonation) (Negretti, 1999); (b) delayed responses to messages allow L2 learners to see and correct errors (Lee, 2001); and (c) lack of turn-taking provokes L2 students’ extensive use of communicative strategies for discourse management (Chun, 1994; Sotillo, 2000). Fitze (2006) designed an experiment to compare F2F and written electronic whole-class discussion in two intact classes and found that advanced L2 students utilized a wider variety of vocabulary and communicative strategies (e.g., clarification requests, dis/agreement statements, social formulations, topic managements) in online discussion.
Several L2 studies have focused on communicative features and discourse functions unique to the temporal and spatial context established by SCMC. Analyzing chat discourse in Webchat between English nonnative speakers and native speakers, Negretti (1999) reported that the participants often used explicit and economical strategies in order to manage procedures and tasks, maintain social cohesion, and show awareness of chat features. Using the WebCT chat environment, Darhower (2002) also found that non-native learners and their teachers created a sense of intersubjective communication by means of teasing, joking, and off-topic discussion, as well as accepting, rejecting, and explaining ideas within conversations. While meaning negotiation is the essence of L2 interaction in both F2F and online conversations, researchers have also observed that L2 learners use other interactional strategies, such as technical actions, social formulations, error corrections, and discourse management more frequently than they do in F2F discussions.
Mei-Ya Liang Synchronous Online Peer Response Groups
Language Learning & Technology 47
Revision-Related Discourse in SCMC
L2 researchers have developed coding schemes to explore revision-related discourse in synchronous online peer response. Adapting Mendonca and Johnson’s (1994) descriptive categories, DiGiovanni and Nagaswami (2001) examined L2 students’ revision-related discourse based on four major categories—questions, explanations, restatements, and suggestions—in both online and F2F oral settings. They found that the number of negotiations was higher in F2F but that the proportion of agreement or disagreement with ideas or with the organization of ideas was higher in synchronous online peer discussion. DiGiovanni and Nagaswami’s (2001) framework, though not exhaustive, provided some key features of negotiating meaning when discussing writing. Taking Halliday’s (1994) functional-semantic view of dialogue as a basis, Jones, Garralda, Li, and Lock (2006) divided revision-related discourse into two move types— initiating moves (i.e., offer, directive, statement, and question) and responding moves (i.e., clarification, confirmation, acceptance, rejection, acknowledgement). They found that tutees asked more questions and made more statements in online interaction than in F2F peer tutoring. Specifically, they asked questions to elicit information and evaluation rather than provide explanations.
Jones et al. (2006) further investigated the area of online peer response. They discovered that the EFL students in a first year writing class were more likely to discuss textual issues (e.g., grammar, vocabulary, and style) in F2F sessions, but chose to focus on global concerns (e.g., content, organization, topic, and thesis) and relational communication in synchronous online sessions. Jones et al.’s finding contradicts Liu and Sadler’s (2003), in which computer-enhanced groups tended to focus on local revisions, whereas the F2F group covered both local and global revisions. These L2 writing studies suggest different functions of peer comments in the two modes of communication. While DiGiovanni and Nagaswami (2001) and Jones et al. (2006) have put emphasis on interactional dynamics that facilitate collaborative responses, another line of research has turned the focus to revision and has questioned the quality of peer response in SCMC. For example, Braine (2001) and Liu and Sadler (2003) have noted disjointed discourse and off-task messages in simultaneous online composing.
A few investigators have attempted to develop a framework specific to CMC contexts to discuss types of revision-related discourse and areas of revision resulting from synchronous online conferences. Using an electronic whiteboard in synchronous online conferencing, Hewett (2006) examined the types of communicative utterances (e.g., content, form, process, context, and phatic) in her first year English classes. The results showed that half of the talk was for interpersonal connections, interaction facilitation, and workspace discussion. Regarding students’ revision-related discourse, 25% focused on content and context of writing, 62% on writing processes and problems (e.g., thesis, supporting ideas, organization), and 13% on formal concerns (e.g., grammar, mechanisms, citation practices). Comparing revision-related discourse and revision changes, she further found that most of the online interactions could be connected with writing and revision. Although none of the students’ essay drafts had formal connections related to the synchronous conferences, Hewett’s study has shown that synchronous online conferences could result in new writing practices and revision changes in an L1 writing context. Relationships among types of interaction and their connections to revision, however, remain unclear in L2 contexts.
Drawing upon studies on both meaning negotiation and collaborative learning in L2 contexts and in SCMC, Liang (2008) proposed a framework which outlines six major types of synchronous online interaction to explore L2 peer groups’ engagement in a summary writing and revision task. They are (a) meaning negotiation, (b) content discussion, (c) error correction, (d) task management, (e) social talk, and (f) technical action. In the 2008 study, the patterns observed across the six groups showed that the total percentage of turns for meaning negotiation, error correction, task management, and technical action was very low and that two-thirds of the turns were spent on social talk and content discussion. This framework seems to be adequate to reveal the relative contributions among different types of interaction. The current study investigates peer response in SCMC through which writing and revision resulting from revision-related discourse or other important interactional processes in SCMC can be traced and considered.
Mei-Ya Liang Synchronous Online Peer Response Groups
Language Learning & Technology 48

THIS STUDY
As part of a larger research project that explores online interaction impacts on EFL university students’ L2 development, this study focuses on the use of synchronous online peer response groups in EFL writing. Specifically, this study asks two questions:
1. What are the different types of interaction in synchronous online L2 discourse?
2. How does synchronous online peer revision-related discourse facilitate subsequent writing and revision?

Course Context
This study involved a sophomore EFL writing course during the fall semester of 2008 at a major university in Taiwan. The course was open to English majors and minors. The course took a process writing approach and focused on expository writing. The goal was for students to help each other write clear and well-reasoned prose. Students used a variety of print and electronic resources to discuss ideas, compose multiple drafts, and edit and revise texts. The class met 3 hours per week for 18 weeks. Course grades were determined by class participation and several writing assignments: an issue paper, annotations and peer comments, exam essays, reflective essays, a book review, and a research paper and presentation. Each assignment was evaluated based on (a) title and thesis, (b) main and supporting ideas, (c) organization and style, (d) word choice and grammar, and (e) editing and revision.
In the class meetings, through both print and multimedia writing prompts (e.g., videos, audio clips, films, artworks, advertisements, etc), the instructor led the class to practice peer response strategies in their discussion about how to improve writing. The class discussions on revision-related strategies were organized around the following topics: (a) reading-writing connections, (b) quoting, paraphrasing, and summarizing, (c) text structures in expository writing, (d) coherence and cohesion, and (e) weak reasoning and errors of logic. Students also participated in synchronous online peer response sessions in a computer lab during the class time as part of the class requirement. They were placed in three groups. The online sessions offered students opportunities to discuss and review peer drafts at different stages of the writing process:
• Prewriting and Drafting. There was one 2-hour online session in the second week. Students brainstormed ideas for their issue papers. During the process, students collected all their resources to compose topic sentences and thesis statements as well as to outline paragraphs in their essays. After students posted issue papers in their blogs, the instructor left comments on each student’s post.
• Revising and Editing. There were four 2-hour online sessions on the 3rd, 6th, 9th, and 13th
• Presentation and Submission. There were two 2-hour sessions on the 15 weeks. Before class meetings, students posted their drafts of annotations in their personal blogs. In class, they shared ideas in drafts by asking questions or by making suggestions and comments about unclear ideas. After each synchronous online peer session, students reviewed the transcripts, left comments on their peers’ blogs, and posted their own revised drafts.
th and 17th
Participants weeks. The first was a collaborative writing of book review task, in which students reviewed writing strategies and discussed what to write. The second was a research paper presentation task. Students posted their drafts before class. During class, they reviewed ideas and evaluated writing. After that, students compiled and submitted their work for grading.
The participants for this study were 12 university students (10 females and 2 males) who took Sophomore Composition in the department of English. According to the results of a pre-course questionnaire, three
Mei-Ya Liang Synchronous Online Peer Response Groups
Language Learning & Technology 49
students had studied English for 8 years and the other 9 students had had more than 9 years of previous English study. Based on their self reports, six students had passed the Intermediate-level GEPT (i.e., a general English proficiency test developed in 1999, commissioned by the Ministry of Education in Taiwan, see Roever and Pan, 2008), and three had achieved the High-Intermediate level. The other three students did not specify or take any GEPT, but their English level was approximately TOEFL CBT 193- 250.
At the beginning of the semester, students were also asked to specify which aspects of their writing in English needed further improvement. The results from the questionnaire show that most of the students wished to improve writing skills in global idea development and organization, such as having a clear main idea in the paragraph, developing additional support for main points, writing a good concluding paragraph, and having smooth connections and transitions between sentences.

Revision-Related Discourse Training Procedures
The students were asked to form three groups of four. Both Group 1 and Group 2 were composed of English majors, whereas Group 3 consisted of 2 English majors, 1 Chinese major, and 1 French major. The English majors had taken the author’s “Freshman English” course in the fall semester of 2007. In the course, students also formed small groups and discussed electronic and print texts in both F2F meetings and online interaction. Given example sentences, students in small groups had practiced group collaborative skills by negotiating meaning (e.g., “What does this word/idea mean?” “Are you saying ...?” “Do you mean that…”), discussing content (e.g., “I don’t understand because…” “A/Another reason for this might be...” “It’s not the point. The main point is …”) and managing tasks (e.g., “What might be important here...?” “How is this task related to...?”). Students were also asked to correct errors and manage tasks as well as to clarify and comment on messages by using a checklist:
• Is the title or topic attractive?
• Are the words in the essay appropriate?
• Does the essay cover important points of an issue?
• Is the essay written in the writer’s own words?
• Does the essay include mistakes in spelling and grammar?
• Does the essay include multimedia aids that serve clear purposes or make the stories more interesting?
In the writing class, students received additional training procedures in the writing process. At the prewriting and drafting stage, the instructor adopted Rollinson’s (2005) suggestions: (a) explaining the potential benefits of peer feedback, (b) discussing with students the purposes of peer responses and the role of the reviewer in reader-writer dialogues, and (c) modeling comments and coaching synchronous peer response on sample paragraphs in class. Two F2F teacher-student conferences were held in the 7th and 14th
• Why are you writing this paper? weeks to review and discuss students’ writing drafts and peer revision. After each conference, students posted reflections on their blogs. As students progressed toward the later stages of editing, revisions, and presentations, the instructor summarized important writing and revision strategies that had been discussed during class time. To guide revision and possibly facilitate synchronous online peer revision-related discourse, the following list of questions was also posted on the class blog:
• Do the title and keywords forecast an issue or an attitude toward the subject?
• Do the library and Web sources provide sound reasons and proper evidence for the issue?
Mei-Ya Liang Synchronous Online Peer Response Groups
Language Learning & Technology 50
• Do you share experiences, emotions, or cultures to support your opinions?
• Do you organize the ideas with a smooth connection and transition between sentences?
• Do you quote some words, phrases, or sentences or paraphrase them?
• Do you adopt your peer’s opinions?
• Do you polish your language and check grammar (e.g., verb tense, subject-verb agreement, word form, run-on sentence, informal register, etc.)?
• Do you find weak reasoning and errors of logic (e.g., irrelevant, oversimplification, overgeneralization, false analogy, undefined terms, etc.)?
• Do you change your question or problem during the course of study?

Data Collection and Analysis
Through MSN Messenger (tw.msn.com), the EFL university students participated in synchronous online peer response group activities. On their personal blogs, they posted their writing assignments. Data were collected from the online chat sessions, student blogs, and two corresponding writing assignments: the book review and research paper tasks. The following directions were listed in the course syllabus:
• Book review (Group work). Provide a brief description of the main ideas and provide an appraisal of the strengths and weaknesses of the book. Your review should include your group members’ opinions about the book. It should range from 500-750 words.
• Research Paper. Use your annotations to support your points of view by the evidence. Provide the reasons why you choose this issue as the focus of your paper. Write questions on the issue and contrast two or more competing perspectives. Write an 8-10 page research paper (2000-2500 words) with supporting citations in APA format.
During the online sessions, the instructor was present in the computer lab to assist with computer problems and remind students of the task requirements. Otherwise students engaged in group discussions and made progress with the two assignments without instructor participation.
Following Liang’s (2008) coding scheme, the author first identified participants’ discourse types. The taxonomy of revision-related discourse includes four categories: (a) meaning negotiation, (b) content discussion, (c) error correction, and (d) task management. Additional codes were included to accommodate the (e) social talk, and (f) technical action that were not directly related to student writing. Definitions and examples for each coding category are presented in Appendix B. The units of analysis to describe L2 learners’ co-constructed online discourses are “turns.” Using the taxonomy, the author and a trained research assistant coded chat data independently and then reviewed all cases of disagreement and resolved differences together. Almost all of the turns included only one type of interaction per turn, but a few turns (less than 10 cases in this study) that included two types of interaction were counted as two turns. χ2 Goodness of Fit Tests were then performed in order to check for differences across discourse types and groups.
The process of investigating the connections between the chat and subsequent writing and revision was recursive and iterative. For the book review task, students’ collaborative texts were examined by the research assistant and the author. We highlighted the words and phrases that were used in the revision-related discourse in each of the three groups’ transcripts. For the research paper, students’ revision changes between drafts were also highlighted and then their chat transcripts were read to look for signals of relationships between revision-related online discourse and individual students’ revised work (for samples, see Appendices C and D).
Mei-Ya Liang Synchronous Online Peer Response Groups
Language Learning & Technology 51

CONCLUSION

Using peer response groups in EFL writing enables students to collaboratively brainstorm, share, and review texts. Synchronous Web technology adds a valuable tool for facilitating and recording the dynamics of group interaction. On the other hand, there are possible reasons why synchronous online peer response groups might be fun, but not very effective. The revision-related and non-revision-related discourse taxonomy used in this study can give an overall view of online interaction from which instructors can proceed to make task implementations more consistent and the results correspondingly stronger. For example, writing and revision processes can be improved by asking students to concentrate on revision-related discourse and to play down non-revision-related discourse. However, the results of this study have also shown that the relationships between revision-related discourse and discourse-related revision are not straightforward. If we consider the ecological nature of online interaction, we will encourage students to make meaningful use of overall online interactional features in the collaborative process for better composition and revision.
To maximize learner-centered, collaborative opportunities for L2 learning, writing, and communication, training procedures and support systems should be employed according to group interaction and task performance along with students’ progress in the writing process. Modeling peer response strategies at the beginning of the course can prepare students at different L2 proficiency levels for online negotiation and discussion. Training should focus on connecting students’ prior experiences with current writing pedagogy. During the revising and editing process, teachers might draw attention to student variation and group interaction. Chat transcripts and selected episodes can be reviewed to help less experienced students learn diverse ways of interaction and locate revision-related discourse. At the presentation and submission stage, while students might feel under pressure as writing tasks are developed into final products, they still need to demonstrate a range of communicative strategies in order to accomplish various learning goals. To revise their writing after the peer response session, these L2 writers can continue the process of negotiating with various resources and audiences either inside or outside the writing class. In short, revision-related online discourse in small-group synchronous writing tasks can provide potentially useful pedagogical insights and tools for the teaching of writing.
Mei-Ya Liang Synchronous Online Peer Response Groups
Language Learning & Technology 58

ABOUT THE AUTHOR
Mei-Ya Liang is an associate professor in the Department of English at National Central University, Taiwan. Her current research focuses on the use of technology in the teaching of English.
E-mail: mliang@cc.ncu.edu.tw

Critical Thinking and Transferability: A Review of the Literature

By

Gwendolyn Reece

April 9, 2002

Since the 1960s, concern that American students may not capable of transferring
the skills they have gained from their education to the practical problems of life has
troubled educators. Of greatest concern is whether students have mastered “critical
thinking” or “higher order thinking skills” and can apply them outside of school
curricula. These concerns have given rise to the “critical thinking movement.”
To demonstrate that the movement is successful, it must prove that its efforts not
only increase the critical thinking of students in school, but that students can transfer
critical thinking to novel situations, including those encountered in daily life. The
primary purpose of this review is to ascertain if there is compelling evidence that efforts
to teach critical thinking have had this result.
What became apparent in the process of this review, however, was that several
subsidiary problems must first be answered before the problem of evaluating the
effectiveness of critical thinking transfer can be approached. The first of these problems
is whether the movement has a common theme or definition of “critical thinking.”

Second is the question, does “critical thinking” encompass “creative thinking” or is it
antithetical to it. The third problem might be formulated thus: is “critical thinking”
generalizable or is it tied to subject matter? The fourth problem is whether adequate
evaluative measures of critical thinking are available to measure the effectiveness of
efforts to teach critical thinking. Answering these prior question is essential before
inquiring whether there is compelling evidence that teaching critical thinking results in a
transfer of skills or dispositions that students can use in other arenas. This line of inquiry
supplies the structure for this review of the relevant literature.

The scope of this review is limited. Most critical thinking literature provides
program and instructional technique description. This material is out of scope for this
review except as it bears directly upon the question concerning subject-dependence in
relation to critical thinking. Furthermore, although this review addresses the works of
most seminal thinkers in the critical thinking movement, constraints and limited access to
information means that some major figures, such as Harvey Siegel, have not been
included. Finally, although philosophical literature on this subject abounds, evaluative
studies using either qualitative or quantitative methods to measure the effectiveness of
whole programs are comparatively scarce. I have included relevant examples of these
studies, yet it can be said at the outset that the dearth such studies needs to be redressed
by the research community.

The Common Theme of the Critical Thinking Movement
The first step in understanding the Critical Thinking Movement is to uncover the
essential characteristics of critical thought and examine the commonality of agendas for
the Critical Thinking Movement. Proponents of the critical thinking movement posit
numerous reasons for teaching critical thinking. A common reason is a reflection of the
shift in economic patterns away from an industrial society into arenas in which laborers
must solve complex problems (Bloom, 1956; Reich, 1992; Paul, 1993; Nickerson, 1987).
Another reason frequently proposed is that critical thinking skills are necessary for
effective citizenship in a democracy, for example, in selecting leaders and being a juror
(Ennis, 1987; Paul, 1993; Nickerson, 1987). Paul (1993) and Nickerson (1987) also call
attention to the capacity of human beings for self-delusion and note that irrational human
behavior causes great suffering in the world. They see critical thinking the antidote.

Finally, both these thinkers uphold the notion that thinking is a significant part of being
human; therefore, mastery of critical thinking is a necessary for being a fully developed
human being.
Another premise of the proponents of the critical thinking movement is that
critical thinking does not always unfold naturally as a part of growth. Furthermore,
critical thinking is not effectively taught in traditional school settings that rely heavily
upon rote memorization and didactic teaching methods (Kennedy, 1991; Paul, 1993;
Schrag, 1988; Nickerson, 1987). Therefore, leaders of the movement have developed
numerous programs to teach critical thinking.
The common theme of the critical thinking movement is that critical thinking
skills involve the ability to make reasonable decisions in complex situations, such as
those found in a rapidly changing and complex society. The movement emphasizes
“knowing how” more than “knowing that” (Roland, 1961). Furthermore, helping
individuals gain these abilities, requires a self- conscious attempt on the part of educators
to address the cultivation of critical thinking by utilizing methods other than simply rote
memorization and didactic instruction.

What is Critical Thinking?
The unity of the movement disintegrates once the question “what do you mean by
critical thinking?” is asked. There is a significant divergence of opinion about what
constitutes critical thinking.
Some scholars identify critical thinking with the mastery of specific skill sets and
provide schematics or taxonomies to express their inter-relationships. A Committee of
College and University Examiners created one of the early taxonomies. (Bloom, 1956).

Bloom and his colleagues identified six major classes of cognitive skills: Knowledge (by
which they mean recall); Comprehension; Application; Analysis; Synthesis; and
Evaluation. One reason for this construction is that the lower skills are required in order
for the higher skills to be used. Comprehension requires Knowledge, or recall.
Therefore, critical thinking, in Bloom’s view, is gaining mastery of these skill sets and
selecting the appropriate techniques when encountering a novel situation.
The primary strength of Bloom’s taxonomy is that it is logical and hierarchical,
guiding the educator in a process leading from the most simple to the most complex form
of cognitive skills. It is also comparatively easy to evaluate the mastery of these skills
because they link to particular behaviors (Bloom, 1956, p. 12). Bloom supplies
numerous evaluative techniques linked to the taxonomy.
There are, however, disadvantages with Bloom’s taxonomy. Historically, many
teachers have used it as a “cookbook” without demonstrating critical thinking skills
themselves (Paul, 1974, pp. 375-383). Paul also criticizes Bloom for overemphasizing
recall and for insisting on neutrality. Paul believes that critical thinking should be used to
reach substantial value judgments. Finally, Paul conceives Bloom’s taxonomy as
neglecting the dialectical dynamic of critical thinking. With regard to his first point, Paul
overstates his case; misuse of the taxonomy does not invalidate the design itself. The
emphasis given to Knowledge or recall is a more controversial, relating to the question
whether or not critical thinking is subject-dependent. I do believe that Paul is correct in
criticizing Bloom’s view that critical thinking is value neutral, since real life decisions are
never value neutral; but again, that does not invalidate the structure of his taxonomy. The
neglect of the dialectical process in critical thinking, however, is a substantial criticism
that does seem borne out in the construction of the taxonomy, which is designed to flow
from simple to complex.
Another general criticism of defining critical thinking as being comprised of a set
of skills is that critical thought also requires particular dispositions or habits to use those
skills. Dispositions, unlike skills, cannot be taught ; they can only be cultivated through
such activities as modeling. Many proponents of critical thinking classify both abilities
and dispositions that are necessary in critical thought.
In light of these criticisms, Ennis (1987) revised his taxonomy to incorporate
dispositions as well as specific abilities, thus defining critical thinking as a combination
of the two. He drew his list of abilities from the field of logic and these abilities are often
taught in “informal logic” courses. The abilities are comparatively easy to evaluate
(Ennis, 1961). In fact, Ennis authored the Cornell Critical Thinking Test, which is a
popular evaluative tool that measures critical thinking according to his taxonomy of
abilities (Royalty, 1995). Dispositions are more difficult to measure.
Similar to Ennis, Paul (1993) created a scheme that addresses both abilities and
dispositions, but Paul tended to stress the activities of the thinker more than the thought
itself. He defined critical thinking as a “unique kind of purposeful thinking in which the
thinker systematically and habitually imposes criteria and intellectual standards upon the
thinking; taking charge of the construction of thinking; guiding the construction of
thinking according to the standards; and assessing the effectiveness of the thinking
according to the purpose, the criteria, and the standards” (Paul, 1987, p. 21). In this way,
Paul accounts for the dispositions of the thinker as well as requiring that the thinker
master a certain skill set in identifying and using standards and criteria. His standards
and criteria seem, like Ennis, to be primarily drawn from the field of logic.
At the far end of the spectrum are thinkers who identify critical thinking with the
virtue of thoughtfulness (Schrag, 1988). The cultivation thoughtfulness leads individuals
to engage in purposeful and deliberate thinking (Schrag, 1988, p. 7). Schrag defines a
virtue as mediating between two contrary inclinations, in this case, between the
inclination towards impulsiveness and the inclination towards rigidity (Schrag, 1988, p.
14). If we ascribe to Schrag’s notion that critical thinking is a virtue, then it is highly
transferable. However, it is also exceedingly difficult to evaluate whether or not
someone possesses a virtue. Although not explicitly stated, Schrag seems still to hold
that deliberate thinking should be in alignment with the laws of logic, although his
emphasis is on character building rather than skill- mastery. However, an important
distinction is that, unlike skills, virtues cannot be taught ; they can only be fostered. To
accept Schrag’s proposition would entail that more attention be given to creating
environments and situations that cultivate virtues rather than teaching them. In this he
seems to be influenced by John Dewey’s ideas of reflective thinking (Dewey, 1933).

Critical vs. Creative Thinking
A significant criticism of all of these theories about critical thinking is that,
drawing on the laws of logic, they focus almost all attention on teaching students how to
evaluate propositions. This approach may help individuals decide what to do or believe,
but it does not address what Nickerson (1987) calls “thinking at goals” or “thinking about
value systems.” Nickerson is referring to an even higher level of critical thinking, not
merely evaluating propositions, but thinking about what one’s goals should be. Such
critical thinking also has a creative aspect that helps generate new propositions.
For all of the above thinkers, creative thinking is not in conflict with critical
thinking, even if it is not part of it. Using Dewey (1933), Schrag (1988) argues that
thinking is an action that creates experience. Therefore, any thinking is creative and adds
to the individual’s repertoire of experience that will generate new growth.
Some scholars, however, believe that critical thinking and creative thinking are
different cognitive skills (Belenky, 1986; Walters, 1990). Part of their concern is that the
emphasis on critical thinking privileges certain epistemologies at the cost of others.
Walters (1990) places special value on imagination and intuition but is silent on the
subject of whether or not they can be taught.
This position is most clearly articulated by the gleeful maverick Edward DeBono
(1971). DeBono’s main thesis is that so-called “vertical thinking” cannot construct new
hypotheses but can only evaluate propositions. He creates a dichotomy between “vertical
thinking” and “lateral thinking,” which might be called “creative thinking.” Lateral
thinking involves creating whole new ways of looking at things, instead of evaluating old
ways of looking at things. Additionally, DeBono claims that lateral thinking can be
taught.
The creative/critical thinking dichotomy is problematic because new, creative
insight is clearly essential to good thinking, yet how it occurs remains shrouded in
mystery. Furthermore, creative thinking is clearly not tied to specific skill sets, nor is it
readily identified by the possession of certain dispositions. Although creativity may be
recognized, it is difficult to evaluate.

Is Critical Thinking Subject-Dependent?
In addition to the critique of the critical thinking movement privileging rational
epistemologies at the expense of creative thinking, another essential debate within the
movement is whether general critical thinking skills that are not subject dependent even
exist. McPeck (1981), for example, argues that since thinking is always thinking about
something, then any kind of thinking is dependent upon the subject being thought about.
Furthermore, he claims that different subjects or domains have different epistemologies,
the idea of critical thinking having different meanings from subject to subject. “[S]kills,
like critical thinking in general, are parasitic upon detailed knowledge of, and experience
in, parent fields and problem areas” (McPeck, 1981, p. 10). He redefines critical thinking
to mean reflective epistemology that is “the analysis of good reasons for belief,
understanding the various kinds of reason involves understanding complex meanings of
field-dependent concepts and evidence” (McPeck, 1981, p. 24).
The ramifications of the proposition for subject dependence are enormous. If
true, it undermines the claim that there are universal intellectual standards of reasonable
thought. Moreover, it means that critical thinking can only be taught from within a
particular subject, invalidating the many “critical thinking courses” that currently exist.
Finally, the claim of subject specificity utterly negates all claims to the general
transferability of critical thinking skills.
Paul (1993) and Ennis (1989) criticize McPeck for reifying the concept of
“subject,” which he uses interchangeably with the academic disciplines. Although one
cannot think about “everything in general” but must think about a topic, that is not the
same as thinking within a discipline, they argue.

Four possible overarching schemes for teaching critical thinking emerge from this
debate. The “general” approach involves teaching generalized critical thinking skills in a
critical thinking course. The “infusion” approach requires self-consciously teaching
critical thinking skills from within a subject course. The “immersion” method assumes
that students will gain the subject-specific critical thinking skills through taking the
subject course. The “mixed- model” approach combines a general course with either an
“infusion” or “immersion” approach (Ennis, 1989). Neither the “mixed- model” nor the
“infusion” approaches rule out the possibility of generalizable critical thinking skills.
Wilen (1995) emphasizes the role of metacognition in critical thinking and advocates an
infusion approach in which the expert models critical thinking in light of their subject
matter through thinking aloud. He endorses this teaching method not because critical
thinking is restricted to subject matter but because such teaching is efficacious. Only
through evaluating and comparing the general and the infusion methods will it be made
apparent whether critical thinking skills are generalizable or subject-dependent.

Evaluation of Critical Thinking
Careful evaluation of critical thinking is vital, not only to answer the question of
the generalizability, but also to assess program strengths and weaknesses and to redress
the weaknesses. Baron (1987) identifies four dimensions in the evaluation of programs
teaching critical thinking skills. The first axis is Formative-Summative. The Formative
evaluation has to do with improving the program, while the Summative evaluates the
effectiveness of the program. The second is Product-Process. The Product evaluation
focuses on what the students produce while the Process is concerned with the workings of
the critical thinking instruction and thinking of the students. A third axis is Qualitative
Quantitative. Both Qualitative and Quantitative studies are meant to capture the
experiences of the people in the program. Finally, Experimental-Quasi- Experimental is
another way of testing program effectiveness. Baron (1987) also suggests testing for
sustained effect, transfer, side effects and metacognition.
A number of tests measure critical thinking skills, (Baron, 1987; Ennis 1961,
1989), but measuring dispositions and “virtues” is more difficult. A danger is that the
ideals of critical thinking dispositions and Schrag’s idea of the virtue of thoughtfulness
(Schrag, 1988) could be overlooked in application because evaluating skills is easier.
Halpern (1993) highlights a number of challenges in evaluating the effectiveness
of critical thinking instruction. Pre-tests and post-tests convey some information, but
reveal nothing about retention. Also, cognitive skills improve with practice, so
presumably the real effects of critical thinking instruction would take some time to
become apparent. Long-term gains are difficult to assess by testing students long after
they have taken a critical thinking course, because it is difficult to measure whether any
improvement from their baseline resulted from the course or from normal maturation and
skills gained in other course. This problem of assessing program effectiveness is further
complicated by the fact that virtually all campuses that offer critical thinking courses
require them for all students; there is no control group. One possibility is to create a
control group from students at a similar university.

There are also significant criticisms about the critical thinking instruments in
addition to the fact that they only test skills and neglect habits, creativity and the “virtue
of thoughtfulness.” McPeck (1981) argues that they are indistinguishable from IQ tests.
One (Royalty, 1995) has been to test students using both IQ tests as well as critical
thinking instruments. However, even if an immediate post-test does reveal information
and is distinguishable from an IQ test, the program cannot be deemed successful until it
can demonstrate transferability of the skills by the student into a new context.

Transferability of Critical Thinking
As previously mentioned, if critical thinking is subject-dependent, critical
thinking should not be transferable across domains. If critical thinking is not
transferable, then most of the reasons for teaching it are invalidated. However, other
impediments may prevent the transfer of critical thinking skills. Perkins (1987)
enumerates three stages of critical thinking development: acquisition, making it
automatic, and transfer. Most critical thinking programs focus on acquisition, but
without the other two steps, Perkins argues, critical thinking tends to remain within the
context of the course. Perkins postulates two types of transfer. “High-road transfer” is
the intentional transfer of a “frame” or critical thinking tactic from one learning context
to another context. Instructors should create exercises that help students achieve this type
of transfer because it typically does not occur automatically. “Low-road transfer” occurs
more spontaneously. By “low-road transfer” he means the phenomenon of perceiving
similarity in a new circumstance and applying the “frame.” The primary significance of
Perkin’s thesis is that without instruction in transfer, students will be less likely to be able
to apply critical thinking skills to novel situations.

Some Studies of Critical Thinking Program Effectiveness and Transfer
Although most works dealing with critical thinking tend to be abstract and
philosophical, there have been some notable attempts to evaluate the effectiveness of
critical thinking instruction. One of the most impressive was conducted in Venezuela
(Hernstein, et.al. 1986). Four hundred seventh graders took a yearlong critical thinking
course in which they were taught fifty-six lessons. Four tests were given before, during,
and after the course. The same tests were administered to a control group. These tests
were supplemented with a design and an oral argument post-test of randomly selected
individuals from the experimental and control group. The results of one objective test
were marginal, but the others were significant. The higher performance on the General
Abilities Test and a Target Abilities Test developed for the program was especially
notable. The two post-tests also revealed higher achievement in the experimental group.
The tests measured critical thinking as skills. This study seems to strongly indicate
transferability. The objective tests were cross-discipline and the oral arguments and
design exercises cross domains.
Royalty (1995) created a study in an attempt to prove the generalizability thesis of
critical thinking, and therefore, its transfer. He acknowledges that the lack of agreement
about what constitutes critical thinking creates problems in its measurement. He
conducted two studies to see whether or not critical thinking skills could be applied to
novel domains. To measure this, he attempted to identify areas that did not require
specialized subject knowledge in order to test them.
Royalty’s first study used the Cornell Critical Thinking Test, which was created
by Ennis and reflects his theory of critical thinking. He also administered an IQ test, the
Belief in the Paranormal Scale, and The Paranormal Experience Scale. His results
showed a strong correlation between IQ and critical thinking but that neither critical
thinking nor IQ accounted for the variance in belief in the paranormal. There was,
however, a strong correlation between The Paranormal Experience Scale and the Belief
in the Paranormal Scale. He explained his conclusion by postulating that belief in the
paranormal and metaphysical speculation may rely upon other “ways of knowing” that
are not part of critical thinking.
I believe that his first study was flawed because of some of the researcher’s
assumptions. “Although the relationship between paranormal beliefs and experiences
may represent logical consistency, it would represent critical thinking only if one
discounted the importance of the content truth of the premise” (Royalty, 1995). Clearly,
the researcher has already identified belief in the paranormal as illogical. He seems to be
subscribing to a materialist ontology and a logical positivist epistemology. If Dewey
(1933) is correct, then it is no surprise that experience of the paranormal is positively
correlated with belief in the paranormal. It seems to me that his first study was flawed in
its conception.
Royalty’s (1995) second study, however, is more promising. He administered the
Cornell Critical Thinking Test, an IQ test, and a Statistical Reasoning test. Both IQ and
the critical thinking accounted for variability, but critical thinking accounted for a unique
portion of the variability. Therefore, the test indicates that critical thinking skills can be
applied to novel subject areas and, therefore, transfer.
In an attempt to understand transfer, Lehman and Nisbett (1990) examined three
types of reasoning: verbal, statistical-methodological, and conditioned reasoning. They
tested University of Michigan undergraduates during their first term in the first year, and
then tested the same undergraduates during their second term in their fourth year. This
study was looking primarily at the immersion model, not testing students who had formal
critical thinking instruction. The students were given three tests, one for each reasoning
type, that included both academic and “real life” questions. The first year showed no
distinction based upon discipline major. After four years, there was no statistical
significance shown for verbal reasoning. Those majoring in the Social Sciences or
Psychology made significant gains in the statistical- methodological reasoning, while
those majoring in the Natural Sciences or Humanities made marginal gains. The Natural
Science and Humanities majors made significant gains in conditional reasoning, but the
Social Science and Psychology majors did not improve. This study seems to support the
conclusion that different types of reasoning are taught in different fields. Still, it does not
rule out the possibility of general critical thinking skills. The improvement on the “real
life” sections of the tests indicates that the skills were transferable, although different,
depending upon the major in which the student was immersed.
In opposition to these studies, Hendricks (2001) studied critical thinking
instruction looking at the distinction between traditional schooling that decontextualizes
knowledge vs. situated learning. Two hundred and twenty seventh graders were studied,
assigned randomly to experimental or control. The students were then taught about
causality, an important component of critical thinking. They were given a “transfer task”
to complete two weeks after the instruction and another six weeks after the instruction.
Interviews were conducted after six weeks. Transfer was very poor for both groups of
students. Hendricks suggests, in agreement with Perkins (1987), that more direct transfer
training was probably needed.

Conclusions and Needs for Additional Research
At this point, the evidence is mixed concerning the transferability of critical
thinking skills and, therefore, the usefulness of critical thinking instruction. The
confounded state of research into the transferability of critical thinking skills stems from
fundamental disagreement about what is meant by critical thinking. All of the evaluation
instruments test critical thinking as skills and do not test for dispositions nor for
thoughtfulness. They are unable to account neither for “wrong” answers that might be
reached through critical thought nor for “right” answers that might be reached through
“test wiseness.”
The evidence also demonstrates the limitations of philosophy as a tool to explore
these problems. Although philosophy is important to articulate goals and aims, all of the
models have logical consistency and cannot be discredited using philosophical methods.
I am most persuaded by the philosophy that defines critical thinking as requiring both
skills and dispositions. I do, also, acknowledge that different fields privilege different
epistemologies, but I do not think that this discredits the existence of general critical
thinking skills and am wary about reifying subject domains.
Clearly additional research is needed. Generally emphasis should shift from the
philosophy of critical thinking to the evaluation of critical thinking and critical thinking
programs. Detailed comparison between IQ test results and the Critical Thinking
instruments is needed in order to ascertain if they are actually measuring different
phenomena. Ways of studying critical thinking dispositions or thoughtfulness need to be
created to ascertain whether or not these characteristics have been inculcated through
critical thinking instruction. More quantitative and qualitative studies are needed to
amass sufficient data for macroanalysis. The infusion model and the general model of
critical thinking instruction should be subjected to tests and compared in order to
ascertain whether general critical thinking skills exist or whether they are subject
dependent. The impact of “transfer instruction” needs to be tested through more studies
of transferability, including longitudinal studies and studies that measure the application
of critical thinking to novel situations and new domains. Furthermore, the interaction
between creative and critical thinking requires more exploration, including whether
creative thinking, imagination and intuition can be taught. Of course, formative
evaluative work to gauge the effectiveness of various methods of critical thinking
instruction is still important.

Nature of Discussions in a Foreign Language Literature Class

Author

Dogan Yuksel

Kocaeli University, Turkey

Bio
Dr. Dogan Yuksel, currently an Assistant Professor at Kocaeli University's Department of Foreign Languages, teaches courses on ELT Research and Methodology in the English Education program. He received his PhD from Florida State University's Multilingual Multicultural Education program in 2007. His research interests include classroom discourse, literary discussions, and Vygotskyan Sociocultural Theory. His email address is doganyuksel@yahoo.com
Abstract
Use of literature in foreign language teaching has been viewed as one of the traditional ways of teaching for centuries. However, the research about the use of literature is quite limited compared to the popularity of its use. In this study, I analyzed the discursive structure of classroom talk both academically and socially to identify the nature of literary discussions. A foreign language literature class in Turkey was observed for one semester and qualitative analyses were conducted following a discourse analysis method. The findings of the study revealed mixed results. Possible reasons of the findings were explained with educational implications.

Key Words: Use of literature, classroom discourse, teacher questions, EFL

Introduction
Use of literature in foreign and second language teaching is taken for granted for many teachers and researchers. However, how literature is studied and nature of classroom discourse in second/foreign language literature classrooms have not been examined thoroughly (Donato & Brooks, 2004; Kim, 2004; Mantero, 2001). Studies that have investigated what is happening in literature classes are very few in number. As it is often articulated, previous research on the use of literature in second and foreign language teaching has focused mostly on the reading comprehension of cultural texts (Donato & Brooks, 2004), and very few studies have examined the verbal interactions of second/foreign language learners in literature discussions (Kim, 2004). A review of previous studies and books that are devoted to the relationship between literature and language learning demonstrate that the study of literature seems to entail mostly receptive skills: for the most part reading and literacy development, and only sometimes writing or speaking (Akyel & Yalcin, 1990; Carter & Long, 1991; Chen, 2006; Cho & Krashen 1994; Ghosn, 2002; Lao & Krashen, 2000; Lazar, 1993, 1994; Liaw, 2001; McKay, 1982; Parkinson & Thomas, 2000; Shanahan; 1997; Widdowson, 1984).

Use of Literature in Foreign Language Teaching
Most of the studies that examined the relationship between literature and language teaching have articulated four benefits of literature: (1) literature helps developing linguistic knowledge both on usage and use level (Lazar, 1994; McKay, 1982; Parkinson & Thomas, 2000; Widdowson, 1984), (2) literature may enhance students’ motivation (Akyel & Yalcin, 1990; Ghosn, 2002; Lazar, 1993; McKay, 1982; Parkinson & Thomas, 2000), (3) literature has the potential to increase learners’ understanding of the target culture (Akyel & Yalcin, 1990; Ghosn, 2002; Lazar, 1993, 1994; McKay, 1982; Parkinson & Thomas 2000; Shanahan; 1997), (4) literature may help develop skills of cognitive and critical thinking (Ghosn, 2002, Lazar, 1993; Parkinson & Thomas, 2000). Most of these benefits are based on the experiences of teachers and researchers who had substantial background in the teaching of literature; however, none of them is supported by research that is coming from real classroom settings.
Motivated from the lack of research on the nature of discussions in literature classrooms, this study examined the nature of literary discussions in an advanced level English literature class in a Turkish EFL (English as a Foreign Language) setting. The study focused on the discursive features of a foreign language literature course offered at a Turkish university. The discursive structure of classroom talk was analyzed both academically and socially to identify the nature of the literary discussions. Students’ perceptions of the class and on-going discourse, and their views about their roles were investigated as well. In other words, both academic and social aspects of the classroom discourse were addressed to increase the understanding of ‘what-is-going-on’ in an advanced level foreign language literature class.
By conducting this study, the researcher aimed to contribute to the literature by providing naturalistic, uncontrolled data about the nature of literary discussions in an advanced level English literature class at a Turkish university. In other words, by analyzing the nature of whole group discussions in a college level foreign language literature class, this study provided empirical data regarding the nature of discussions in a literature class, which in turn might offer some insights about the validity of the claim that the study of literature in the collegiate curriculum is useful in developing foreign language proficiency.
The discursive structure of the classroom talk, i.e., how the discussion evolves, may help us see the major significant points in a discussion, namely how the discussion is constructed, who starts it, who finalizes it and who contributes to it. Characteristics of discussions may further shed some light on the perception of knowledge and schooling by the teachers and students. Many previous studies demonstrate that most of the schooling in many settings have been built on the traditional IRE (Initiation- Response- Evaluation) routine, which may indicate that, among other things, the teacher have the ultimate control in the classroom (Gutierrez, 1994; Lemke, 1990; Nystrand, 1997). On the other hand, open-ended discussions are less prescribed, including less teacher control, and may be a sign of teachers’ interest in students’ ideas and comments (Nystrand, 1997). The significance of the nature of discussions has been acknowledged by many researchers who deal with classroom discourse in different fields of education (Cazden, 2001; Gutierrez, 1994; Lemke, 1990; Nystrand, 1997; Walsh, 2006; Wells, 1999b).
The following research questions guided this study:

1- What is the nature of literary discussions in an advanced level literature course in a Turkish EFL setting based on the analyses of the discursive structure of the classroom talk?
2- What do students think about this specific literature class and their roles in this class?

Methodological Framework
The research questions influenced the choice of the methodological framework, the social interactionist perspective (Green, 1983; Green & Wallat, 1981; Mehan, 1979, 1998) that provided a comprehensive outlook about the classroom discourse. A social interactionist perspective to teaching focuses on the discourse and interaction in the classroom, and uses a discourse analysis framework to analyze the classroom interaction. In his seminal book Learning Lessons, published in 1979, Mehan pointed out the significance of studying interaction in the classroom context by stating “because educational facts are constituted in interaction, we need to study interaction in educational contexts… in order to understand the nature of schooling” (p. 6). After more than 25 years, as Wells (2005) states, many people who study classroom learning and teaching today agree that “the nature of the interaction that takes place in class is one of the most significant influences on the quality of student learning” (p. 1).
From this perspective, a second or foreign language classroom is not only an academic environment, but also it is a social context (Cazden, 2001). Classroom language is analyzed as it pertains to two different functions in classroom life: (a) the communication of propositional information, which is also called as referential, cognitive or ideational function, and (b) the establishment and maintenance of social relationship and identity that refers to the social and affective features of language (Cazden, 2001).

Data Collection Procedures
Following Harklau (2005), the present study can be defined as ethnographic research that involves a case study, which is quite common in qualitative research (Ellis & Barkhuizen, 2005). Natural and systematic audio and video recordings of the classroom sessions for nine weeks (a whole semester), field notes taken as a non-participant observer during this time, and semi-structured interviews with the instructor, students, and the administrators of the program constituted the primary data sources. Discourse analysis was used as a major research tool to examine the nature of literary discussions. The main aim of using multiple data collection techniques was to get various perspectives of the same classroom phenomenon to increase the credibility of the study. Another concern of the study was to get the insider’s perspective and understanding of the classroom interaction.

Data Analysis

Discourse analysis technique was used while analyzing the data. The following steps were followed prior to data analysis: (i) transcription of the audio and video recordings, (ii) data indexing, (iii) data reduction, and (iv) data coding. Data transcription process involved the transcription of the video and audio-recordings in their entirety. Data indexing was performed by describing the context in which data occurred, in other words, where and when the activity took place in the study. Each teacher-fronted whole group discussion was indexed by providing the context of when and where it took place. The description of the topic, pseudonyms for each participant, and major activities of the lesson in each episode were included in this session as well. After indexing the relevant parts, data was reduced into smaller and manageable chunks of information (episodes) according to the research questions. Finally, the data was coded using a top-down, deductive coding procedure where the analyst pinpointed a set of codes before engagement with data began. The four constructs were already identified by utilizing the findings of the previous literature and theoretical framework.

Interrater Reliability

After I identified 69 episodes of teacher-fronted text-based discussion, and coded them according to the research constructs, two external reviewers were asked to re-code some parts of the data to increase the reliability of the findings. External raters were familiar with classroom discourse research and each coded 10% of the data, which was randomly selected. For the part of the study that was discussed in this paper, external raters coded the level of learner utterances. I prepared a training manual that included the definitions of each level of learner utterances together with at least two examples. After the final codings, there was 98% interrater reliability between the researcher and the first rater, and 97% interrater reliability between the researcher and the second rater.
Context of the Research

The research was conducted in the Drama Analysis and Teaching class offered at the sixth semester of English Education program at a major Turkish University, during the spring semester of 2006, which started on February 27th and ended on June 16th. The class met regularly on Tuesdays between 8:30 A.M. and 11:30 A.M. It had three hours of class with two breaks for ten minutes after each hour. The participants of this study were advanced level English Education majors attending a Turkish public university. I was particularly interested in advanced level learners, because they had adequate English proficiency and the necessary background in literature to carry out the classroom discussions in the target language. The instructor of the course, Dr. Anne, held a PhD in English Literature and had been teaching the same course for more than 10 years. The background survey indicated that the participants ranged in age from 20 to 22, and they had been studying English for 5 to 12 years. Pseudonyms were used to protect the identities of the participants of the study.
During videorecordings, I positioned the camcorder I used to the right corner of the classroom, and kept it at the same place during the whole semester. I sat next to the camcorder, and remained silent as much as I could. My main concern was being as invisible as possible, and I did it as best as I could, even though I needed to talk in the classroom from time to time.

Figure 1. Layout of the classroom.

Findings
Discursive Structure of Classroom Talk

During nine weeks of recordings, I identified 69 whole group teacher fronted text-based discussions. The teacher fronted text-based discussions, i.e., the episodes, ranged between one minute and ten seconds and twenty-six minutes and twelve in duration. In each of these episodes, the focus was the text that had been read, and the teacher asked at least two questions. After each text-based discussion was identified, they were coded based on (a) initiation move of each episode, (b) major patterns of the moves (e.g., initiation, response, evaluation), (c) level of the learner utterances (i.e., word, phrase, sentence, or discourse), and (d) finalization move of each episode.
The average number of episodes per week was 7.6. The first week of the recordings had fourteen episodes, and eighth week had only four. Time spent on discussions in each week varied between 47 and 104 minutes. The average time spent on discussion was 77 minutes. Week 3 had the shortest time with 47 minutes, on the other hand in Week 5, 104 minutes of class time was spent on discussions. The distribution of the episodes in each week is provided in Table 1.

Conclusion
The findings of this study, among other things, reiterated the lack of speaking and practice opportunities that literature classes may provide to foreign language learners. This problem, if taken together with the exploration of new ways about using literary texts to create pedagogically effective environments for the development of proficiency (Carter, 2007), might be a good start to think about specific features of literature and literature teaching that might have some influence on the development of second language skills such as speaking.
As it is widely acknowledged, speaking is a major problem in foreign language contexts and learners do not find enough opportunities to develop their speaking skills. Teachers and students often feel frustrated because of the limited opportunities of “speak[ing], read[ing] and writ[ing] meaningfully in English in a learning situation in which there is little of substance worth talking about” (Handscombe, 1994, p. 334). The findings of this study demonstrated that literature classes in the foreign language curriculum would be a venue to provide opportunities for the meaningful use of language in a context while most other interactions were based on some in-class mechanical drills, repetition, and memorization of chunks. However, in order to do that, foreign language literature teachers should be aware of the tool they have and develop the opportunity for discourse accordingly, which remind us, again, of the suggestions of Gibbons (2006) about the mediation role of teachers in the foreign language teaching classrooms.

Notes
1. Detailed analysis of the teacher questions in the Drama Analysis and Teaching class was conducted as well. However, because teacher questions were not the specific focus of this study, further discussion of the topic is not included. You can refer to Yuksel & Yu (2008) for further exploration of this aspect of the classroom discourse.

Acknowledgments
I would like to thank Miao Yu, Dr. Deborah Hasson and Dr. Frank B. Brooks for their support and feedback during the preparation of this paper. I am also grateful to the instructor and students of the course for their cooperation and understanding during data collection.

References
Akyel, A., & Yalcin, E. (1990). Literature in the EFL class: A study of goal
achievement incongruence. ELT Journal, 44(3), 174-180.

Berg, C. (1990). Conversation activities based on literary readings. The French Review, 63 (4).

Carter, R. (2007). Literature and language teaching 1986-2006: A review. International Journal of Applied Linguistics, 17, 3-13.

Carter, R., & Long M. (1991). Teaching literature. London: Longman.

Cazden, C. (2001). Classroom discourse: The language of teaching and learning. Portsmouth, NH: Heinemann.

Chen, Y.M. (2006). Using children’s literature for reading and writing stories. Asian EFL Journal, 8(4).

Cho, K.S., & Krashen, S. (1994). Acquisition of vocabulary from the Sweet Valley Kids series. Journal of Reading 37, 662-667.

Donato, R., & Brooks F. B. (2004). Literary discussions and advanced speaking functions: Researching the (dis)connection. Foreign Language Annals, 37(2).

Ellis, R., & Barkhuizen, G. (2005). Analyzing learner language. New York: Oxford University Press.

Gibbons, P. (2006). Bridging discourses in the ESL classroom: Students, teachers and researchers. London ; New York: Continuum.

Ghosn, I. (2002). Four good reasons to use literature in primary school ELT. ELT Journal, 56(2), 172-179.

Green, J. L. (1983). Exploring classroom discourse: Linguistic perspectives on
teaching-learning processes. Educational Psychologist, 18, 180-99.

Green, J. L., & Wallat, C. (1981). Mapping instructional conversations: A sociolinguistic ethnography. In J. L. Green, & C. Wallat (Eds.), Ethnography and language in educational settings. Norwood: Ablex.

Gutierrez, K. D. (1994). How talk, context, and script shape contexts for learning: A cross-case comparison of journal sharing. Linguistics and Education, 5, 335-65.

Handscombe, J. (1994). Putting it all together. In F. Genesee (Ed.), Educating second language children: The whole child, the whole curriculum, the whole community (pp 331-356). New York: Cambridge University Press.

Harklau, L. (2005). Ethnography and ethnographic research on second language
teaching and learning. In E. Hinkel (Ed.), Handbook of research in second language learning and teaching (pp. 179-194). Mahwah, NJ: Lawrence Erlbaum.

Kim, M. (2004). Literature discussions in adult L2 learning. Language and Education, 18(2), 145-166.

Kramsch, C. (1985). Literary texts in the classroom: A discourse perspective. The Modern Language Journal 69(4), 356-366.

Lao, C. Y., & Krashen, S. (2000). The impact of popular literature study on literacy development in EFL: More evidence for the power of reading. System 28, 261- 270.

Lazar, G. (1993). Literature and language teaching. Cambridge: Cambridge University Press.

Lazar, G. (1994). Using literature at lower levels. ELT Journal, 48(2), 115-124.

Lemke, J. L. (1990). Talking science: Language, learning, and values. Norwood, NJ: Ablex.

Liaw, M. L. (2001). Exploring literary responses in an EFL classroom. Foreign Language Annals, 34, 35-45.

Martin, L. (1993). Breaking the sounds of silence: Promoting discussion of literary texts in intermediate courses. The French Review, 66(4), 549-61.

Mantero, M. (2001). The reasons we speak: A sociocultural discourse analysis of text-based talk in a university level foreign language classroom. Unpublished doctoral dissertation, The Florida State University.

McKay, S. (1982). Literature in the ESL classroom. TESOL Quarterly, 16(4), 529-36.

Mehan, H. (1979). Learning lessons: social organization in the classroom. Cambridge, MA: Harvard University Press.

Nystrand, M. (1997). Opening dialogue: Understanding the dynamics of language and learning in the English classroom. New York: Teachers College Press (with A. Gamoran, R Kachur, & C. Prendergast).

Parkinson, B., & Thomas H. R. (2000). Teaching literature in a second language. Edinburgh: Edinburgh University Press.

Shanahan, D. (1997). Articulating the relationship between language, literature, and culture: Toward a new agenda for foreign language teaching and research. The Modern Language Journal, 81, 164-74.

Skidmore, D. (2000). From pedagogical dialogue to dialogical pedagogy. Language and Education, 14(4), 283-296.

Slimani, A. (1989). The role of topicalization in classroom language learning. System, 17, 223–234.

Troudi, S. (1994). The nature of speaking opportunities in an English as a second language class. Unpublished doctoral dissertation, The Florida State University.

Walsh, S. (2006). Investigating classroom discourse. London and New York: Routledge.

Wells, G. (1999a). Dialogic inquiry: Toward sociocultural practice and theory of education. Cambridge, UK: Cambridge University Press.

Wells, G. (1999b). Language and education: Reconceptualizing education as dialogue. Annual Review of Applied Linguistics, 19, 135-155.

Wells, G. (2005). Toward dialogue in the classroom. Unpublished manuscript, Retrieved May 17, 2005, from http://education.ucsc.edu/faculty/gwells/Files/PapersFolder/-ClassroomDialogue.html

Widdowson, H. G. (1984). Stylistics and the teaching of literature. Essex: Longman.

Yuksel, D., & Yu. (2008). Inside the classroom: Student and teacher questions in a foreign language literature class. TESL Reporter, 41(1), 12-32.

Teaching Classical Literature Classically

by Andrew Kern

The classical purpose for teaching literature is the same as the classical purpose for teaching anything: to cultivate wisdom and virtue so that the student is better able to know and enjoy God. Classical literature exposes the student to models of virtue. It also places demands on his intellect, thus developing his intellectual virtues. As the intellectual virtues are developed, the student’s capacity to know and understand facts, ideas, relationships, and persons is nurtured. Furthermore, classical literature cultivates what James Taylor and John Senior have called Poetic Knowledge.

Poetic knowledge is the type of knowledge children gain when they read, e.g. The Chronicles of Narnia or fairytales. It is an intuitive knowledge of the nature of things, but the reader may not always–or even often–be directly aware of what he is learning or how he is being nourished by it. Poetic knowledge removes the reader from the realm of pragmatism (love of power) and lifts him to the realm of genuine ideals (love of virtue). By presenting models of virtues, training the intellectual virtues, and nourishing poetic knowledge, classical literature cultivates wisdom and virtue in the student.

Poetry cannot, as Matthew Arnold fondly hoped, take the place of religion… What it can do–when fully understood and articulated-is to restore human feeling. It can become a discipline for that ordering of the soul that the Greeks called “Paideia.” Dr. Louise Cowan

Because of the classical purpose for teaching literature, classical parents and educators are careful in their selection of what they and their students read. Young children should spend little time on books used to teach decoding letters (i.e. phonetics). Their minds are far beyond what their decoding can reveal to them. Therefore, it is imperative that parents and teachers read to children at a level beyond the child’s decoding skills. Children need to read or have read to them Bible stories, fairytales, folk tales, fables, mythology, legends, nursery rhymes, and good poetry.

Poetry will appeal to children sensually long before it speaks to them intellectually. Fine. Let them hear the music of Shakespeare even if they don’t understand the words. As soon as possible children should be hearing or reading Homer, Virgil, Ovid, Dante, the King James version of the Bible (for literary, not theological, reasons), Shakespeare, Chaucer, and other genuine classics. If these books seem too hard, slow down. But don’t renounce the pleasure of the music for the obscurity of the words. Rather, take your time. Don’t discard the bard for trivial tunes. By all means, use good books to prepare for the great books, but make sure the good books are good.

Which brings me back to my second point: be careful in your selection. Keeping in mind the purpose described above, select books that present models of the virtues to imitate. This does include virtuous characters who overcome evil, but it also includes virtuous writers who overcome the evil of sloth, self-indulgence, or greed to produce excellent art. Read texts that are beautiful and noble. Prefer stories that have been refined by time (Homer, fairytales, and fables are classic examples of such texts). Present tales that fill children’s imaginations with metaphors: fairies, elves, hobbits, wizards, knights, damsels in distress, saviors, princesses, beasts, and transformations. Children interpret life through these metaphors (so do adults), so when the metaphors are unreliable (like many modern pictures) or absent, they interpret life amiss.

Classical educators and students also read literature differently. To begin with, the classical student reads texts rather than merely reading about them. It is the text and the ideas contained in the text that the student focuses on. He reads intensively and repeatedly. He explores the form as well as the matter of the text. He criticizes everything he reads in light of the single dominating law of propriety–are the form, language, plot, characters, settings, motifs, etc. appropriate to the idea expressed? This is simply another way of asking whether the artist was successful. The noblest ideas need the noblest expression. When students contemplate such noble ideas in a noble form, they are ennobled.

To achieve this, classical educators differ in their approach to teaching literature. They use the didactic and Socratic approaches described in last issue’s article. Note that both approaches train the mind to think about ideas, one through contemplating models (didactic) and the other through exploring one’s opinions on the subject of the given text (Socratic).

By clarifying your purpose, deliberating over your materials, reading contemplatively, and teaching intentionally, you enable literature to fulfill its function as what Charles Grosvenor Osgood once called “means of grace.” Feed those souls.

Bridging the Language-Literature Gap: Introducing Literature Electronically to the Undergraduate Language Student

Mary Ann Lyman-Hager
San Diego State University

Abstract:
Foreign language teaching practices have emphasized oral language instruction at the expense of written language instruction. In addition, many foreign language and literature departments have made a division between language studies and literary studies, a division in which the number of students who pursue literary studies are far less numerous than those who pursue language studies. Some educators now believe that it is time to reemphasize reading and literature in the profession. Computer-enhanced reading instruction holds considerable promise as a means to reintroduce "culturally dense" texts into the curriculum, perhaps earlier than originally possible in traditional approaches to teaching. Finally, the tracking capabilities of reading programs allow researchers to collect data that can shed light on students' use of reading strategies.

INTRODUCTION: PARADOXES IN FOREIGN LANGUAGE TEACHING

Language teaching practices in the United States are in many ways paradoxical and contradictory.¹ Since the 1940s, language teaching in the United States has increasingly emphasized oral production and has distanced itself from grammar-translation methods that had dominated the profession in prior decades. Yet some, feeling we have strayed too far from literacy in our nearly exclusive emphasis on oral production skills, have called for a reemphasis on reading and writing. Multimedia and Internet tools for learning languages are becoming commonplace, but their use in the foreign language curriculum, especially in less commonly taught languages, is far from widespread. Finally, the National Standards movement, ushered in by the organizers from the American Council for the Teaching of Foreign Languages (ACTFL), calls for a holistic, interdisciplinary approach in language teaching. This approach would blend culture and content together in a rich learner-centered environment, yet few teachers are prepared to collaborate with others outside the discipline in creating these new approaches. Even within the discipline of foreign languages, the gap between language teaching and literature continues to widen.

Kramsch (1985) has referred to the transition between language and literacy study and has warned of the bimodal split of language departments into language versus literature factions. Student reluctance to study literature has been observed in North America by Muyskens (1993), who has written that "… while undergraduate language enrollments seem to be increasing, fewer students now choose to study second language literatures." This phenomena is not limited to North America, however. In Australia, Leopold (1985) has also noted that "as soon as a choice was available, students appeared to choose almost anything except literature."

L2 READING COMPREHENSION AND THE LITERACY GAP

Schulz (1982) has also referred to the major differences between undergraduate courses in language and those in literature. Most lower-level courses tend to emphasize communicative, linguistic-based tasks or daily "hearthstone" culture, balancing somewhat the four skills of reading, writing, listening, and speaking. These courses tend to emphasize oral language, while upper-level literature courses focus more intensely on reading and interpreting literary passages, often reflecting "high culture." Over 15 years ago, Schulz (1982) referred to the difficulties many students have in transitioning to advanced foreign language courses as the "literacy gap" and described the mismatch between students' abilities and course content. Suddenly, instructors expect a transition from the stage of painful word-by-word decoding of contrived written dialogs and narrations dealing with simple everyday events to comprehension of relatively lengthy literary texts containing highly abstract vocabulary, complex syntactical patterns, and sophisticated style and content which even an educated native speaker often cannot read without effort. (p. 43)

Intermediate-level students are required to read portions of authentic texts by the third semester of French undergraduate study, usually short passages of French literature. Unfortunately, this exposure to authentic texts comes as a shock to many students, who have been used to primarily oral-based class instruction. They often cannot understand the cultural frame of reference underlying the text because of its dissimilarity to their own (Bernhardt, 1990, 1991; Swaffar et al., 1991). So, in addition to the problems mentioned by Schulz, literary texts may have a high frequency of familiar words employed in unusual ways to create desired stylistic effects, and the cultural contexts described in target language literature tend to be unfamiliar to nonnative readers.

FOREIGN LANGUAGE READING COMPREHENSION

From reading research dating from the 1980s (Hudson, 1982; Carrell, 1983, 1984; Johnson, 1982), a general consensus exists that L2 reading comprehension results from the interaction of readers' use of bottom-up and top-down reading strategies. Many L1 studies also stress the importance of background, or world, knowledge enriching or guiding readers' understanding of specific text-bound features (Stanovich, 1980; Rumelhart, 1975, 1977).

Second language acquisition theorists view reading as a good source of comprehensible input (Krashen, 1984, 1991; Terrell, 1986; and Winitz, 1981). Because of the solitary nature of this skill, it is difficult to ascertain what psycholinguistic behaviors are occurring in the individual reader (Brumfit, 1981). Research on L1 reading (Zvetina, 1987) has been the primary focus of reading research to date, which has relied upon videotaped observations of readers, researchers' notations of readers' behaviors, and postreading interviews with learners. The computer offers a more efficient, less obtrusive manner of data gathering and, when used in conjunction with traditional data-gathering techniques, can yield new insights into foreign language reading comprehension.

Based on data from undergraduate foreign language student reading research, Bernhardt (1990) has posited six components of effective reading comprehension which relate to vocabulary acquisition: (a) word recognition, (b) phonemic/graphemic decoding, (c) syntactic feature recognition, (d) intratextual perception, (e) prior knowledge, and (f) metacognition. Background information has been shown to be critical to forming a correct schema of a text (Bernhardt, 1990, 1991; Feldman, 1990; Hewitt, 1990; Swaffar et al., 1991). Reading material that is culturally weighted, that is, related to specific in-depth knowledge of the culture, is more difficult than reading material which is unweighted (Feldman, 1990; Hewitt, 1990). In other words, although students may not intuitively understand the importance of background knowledge and linguistic knowledge, they need more than simple dictionary definitions to understand foreign language texts.²

RESEARCH ON FOREIGN LANGUAGE GLOSSING

Glosses figure prominently in most L2 reading materials. Their role in comprehension has been investigated in several studies, usually with a focus on two central questions: (a) whether glossing of L2 texts enhances comprehension and (b) whether glossing inhibits fluent L2 reading. Johnson's (1982) study of ESL readers indicated that glossing may interfere with global comprehension by focusing on the meaning of individual words. Pak (1986) in ESL and Jacobs and Dufon (1990) in Spanish also found no significant differences in learners' passage comprehension with or without glosses. In contrast, Davis (1989) and Luo (1993) found, in two separate studies, that glosses did improve the comprehension of L2 readers of French. Their results may have diverged from previous studies because of differences in the difficulty level and the authenticity of the passages read. Johnson and Pak used texts especially written for L2 readers, and Jacobs and Dufon asked subjects to read fairly simple newspaper articles. In contrast, Davis and Luo used literary passages designed for the target language population (not modified for L2 readers) with many low-frequency vocabulary items. These studies indicate that glosses may enhance readers' comprehension, if the text contains a high incidence of unknown words.

READING RESEARCH

Reading is a so-called "passive" skill—focusing on learners' comprehension rather than their production. It is, therefore, difficult to measure reading directly without the aid of specialized techniques to track learners while they are engaged in the act of reading. Further, reading is the key to success in upper-division undergraduate language programs, yet it remains the least favored of the four skills among intermediate language students  (Sieloff-Magnan, 1995). What do we know about the nature of foreign language reading in a technological environment? How can instructors best introduce foreign language students to reading strategies for authentic texts of increasing length and complexity? Can computers assist instructors in introducing these strategies to students tackling difficult texts for the first time?

Computers have the ability to generate information about learners while they are engaged in the act of reading, recording "complex processes accurately and unobtrusively" (Goodfellow & Laurillard, 1994). The problem lies in interpreting the mass of data generated by the computer program and weaving it into a comprehensive picture of second language acquisition and reading strategies. Information about the efficacy of electronic textual glossing has been gleaned from tracking data generated by the computerized reading program Une Vie de boy (Davis & Lyman-Hager, 1994a, 1994b) which provides computerized annotations or glosses for the first 1,754 words of Ferdinand Oyono's 1956 Cameroonian novel Une Vie de boy. (See the relevant passages of the novel in Appendix A.) This small computer program has generated a number of research studies at The Pennsylvania State University, most recently a doctoral study by Hayden (1997), An earlier study by Lyman-Hager, Davis, Burnett, and Chennault (1993) used a comprehensive case study, recall protocols, vocabulary tests, pretreatment student questionnaires, and software evaluation sheets. These data collection measures were combined with tracking data collected electronically by the computer to help understand the reactions of individual students selected from a large pool of students who had used the software. The study illustrated how these students made use of informational glosses and how they interpreted the text as a result of working with the electronic materials. What is of interest here is the interaction between text and student.

The students were enrolled in third semester French at The Pennsylvania State University and exhibited at least intermediate-low level language skills (as described in the ACTFL Proficiency Guidelines). Their understanding of the underlying culture was closely interrelated to their reading proficiency, and their lack of vocabulary caused them to make incorrect assumptions about the meaning of the text. The Oyono text was situated in the (1930s) colonial period in the French-speaking country of Cameroon. The text describes the reaction of a young boy, Toundi, to his French colonizers. This first chapter describes his relations with his family and then with the Catholic priest who takes him in after he runs away from his family. The text is divided into the following three critical incidents:

1) a fight among some schoolboys (including the protagonist, Toundi) for sugar cubes distributed by the priest just before their tribal ritual of initiation into manhood; 2) the conflict between Toundi and his father over the sugar cube incident and his subsequent exclusion from a family dinner feast (a partially decomposed porcupine); and

3) Toundi's arrival at Father Gilbert's house and his mother's attempt to bring him back home.

Clearly, the selected passages were not very user friendly for the target population of the third-semester American students. Nothing in their background had prepared them to understand the initiation ritual and the cultural conflict with Catholicism. The family dinner, the fight over the porcupine left in the trap a bit too long, and colonial practices of the French in a far-away African country were quite foreign to most of the students in the study. Culturally dense, even inaccessible, the passages could have easily challenged learners even in the third or fourth year of language study. To make the text perceptually salient and understandable, the creators of the computer program, then faculty members in the Department of French at The Pennsylvania State University, packed the accompanying glosses with definitions, cultural background information, pronunciation aids, character and concept maps, and pictures. They attempted to operationalize Bernhardt's theories of reading into the computerized version of the Oyono text. In addition, prereading questions were created to relate students' backgrounds and childhood experiences to the backdrop of the text. An eight-page computerized introductory section to the text additionally presented an explicit, systematic discussion of strategies to be used by novice readers of French literature. This introduction also set the context for the Oyono text and outlined the use of the various types of glosses available.

The following case study, taken from Lyman-Hager and Burnett (1999), illustrates the power of combining data from multiple sources to shed light on individual learners' responses to a text in a foreign language and on the highly individualized manner in which learners process information from the text. Recall protocols, questionnaires, student academic background data, tracking data, and other sources of data provided the researchers with extensive descriptions of the language learners. In this case, the learner, Sara, caught the major idea of the text but missed many important details. Other profiles, also discussed in Lyman-Hager and Burnett, show the effects of textual interpretation by learners who exclusively used bottom up processing strategies and who spun their own versions of the story, often proceeding from the misperception of a single word.

CONCLUSIONS

Quite a few computerized foreign language reading programs are currently available. None of the programs available at the time of the study, however, provided the opportunity to: (a) follow a specific theoretical model of reading comprehension in a second language, (b) record students' behaviors while they are reading on-line, and (c) offer the requisite cultural background necessary to understand the text—here, a unique African perspective on Francophone culture.

The computer-enhanced version of the chapter by Oyono afforded learners a complete story, which has aesthetic and pedagogical merits over the fragmented excerpts usually found in literary passages annotated for language learners. Further, in order to be accessible to beginning readers, the text had to include extensive annotations, which, if found on the printed page, might interfere with the flow of "natural" reading. Computerized annotations can be hidden from the view of readers until needed. This type of hypertextual reading may offer students a greater variety of glosses than could be presented in the print version, due to the computer's ability to hide and reveal information at the request of the individual reader. It remains to be seen whether this type of glossing will allow learners to process and store textual information more efficiently and whether the software will help activate appropriate text processing strategies. A recent study by Chun and Plass (1996), as well as a doctoral study by Karp (personal communication, 1988), make use of tracking data and may offer partial answers to these and other questions.

Can the results of university-based research on individual readers' reactions to computerized second language reading have relevance to establishing guidelines or best practices for the creation of educational materials tailored to meet the needs of all learners? By examining the profiles of individual readers, can we describe strategies that may be useful for readers to adopt? A 1991 study by Noblitt and Bland suggested that valuable insights into individual learning strategies used by foreign language readers could be obtained by electronically tracking what they actually do when interacting with computerized instructional materials. The case study described in this article analyzes the results of data gathered by tracking the learner with the assistance of the very technology used in the presentation of the material.

Clearly, new approaches and new materials will be needed to equip the global electronic schoolhouse of the second millennium. Great hope has been placed on the Internet and computers in general as providers of authentic materials and as mediators of unfamiliar cultural and linguistic materials. Certainly the cost of individualizing instruction and of integrating computers into the curriculum should be considered along with the short- and long-term influence of computer use on learning (see Woolfolk, 1990). We now have enough computers in educational settings to determine how, under what conditions, and to what degree computers can affect student learning (Bozeman & House, 1988). Although some believe that comparing computer-based learning with traditional learning is similar to comparing apples and golf balls, the need for accountability is great (Pogrow, 1988).

Technology is called upon to play an important role in this restructured school, particularly in the area of "authentic performance assessment in foreign language education" (Nielson & Hoffman, 1996). Foreign language reading in particular has been an area thought to be well suited to technological applications (Blake, 1992). The new restructured school envisioned by writers of the National Standards calls for increased attention to cultural literacy and to learner-centered instruction. Two of the five standards (Cultures and Comparisons) refer specifically to understanding and appreciating cultural differences.³ Increasingly, second language researchers and computer scientists are recognizing the value of hypertextual glosses and electronic reading which lead the reader to a contextualized interpretation of an authentic text (see Landow, 1990). Glosses relating to culturally "dense" readings provide insights about specific texts and cultures, which individual language teachers may not be able to provide, thus bridging the gap between language and literature and between the reader and the text. More research studies are needed to determine which gloss types are the most useful for which students under which learning circumstances. Meanwhile, with the emergence of E-Books (rocket books) and other "modern" approaches, faculty will have to enter into alliances with colleagues in other disciplines to create more sophisticated and appropriate approaches to electronic reading. At the very least, we should collaborate with our colleagues in literary studies. By adopting a team approach to software development, we might impact the quality and diversity of the authentic foreign language readings and be able to introduce them earlier into the foreign language course sequence. By working with colleagues in computer science and instructional technology, we might be able to invent educational approaches involving multimedia glosses in order to present those readings to an even larger reading public, such as general audiences interested in foreign culture and travel.