It was our own experience as students and teachers that originally persuaded us that small groups of students, working together, can advance their learning in ways not available when a teacher talks to a whole class. It has never been our purpose to present small group work as a panacea, however; for teachers organizing their studentsā learning it is one option in the repertoire, an option with its own considerable strengths, as we shall show. Quite young students can and will take joint responsibility for the progress of their learning and gain considerably from so doing.
It was this conviction that led us some years ago to set up a study of small group talk in two schools. One of us had already carried out some study of group talk (Barnes 1976), and we wanted to look more systematically at what happens when children in schools are given a task to discuss. Our report on that research was published (Barnes and Todd 1977) exactly as we had sent it to the Social Studies Research Council which had funded our work. Since then it has been read and used by many teachers on both sides of the Atlantic. It has been recommended to teachers by the National Oracy Project in the United Kingdom (UK), by members of the whole language network in the United States (US), and by Project Talk in Ontario, Canada. Nevertheless, we have always been aware that when we first wrote we had not addressed ourselves specifically to teachers, so that the first edition contained details of the research that some teachers might not find very relevant to their concerns.
The present volume sets out to remedy this. We have retained substantial elements of the original materials but have recast them so that teachersā concerns are addressed directly. Our central task has been to speak to teachers and to make explicit what light our research throws on learning in classrooms. Moreover, we have included a chapter of direct advice about how to manage small group work in classrooms. In the final chapter we have supported this advice with a discussion of how meaning is generated in conversation generally, and how it is generated particularly in discussions directed towards learning. Since the first edition much has been published which has led us to reconsider and clarify what we wrote before.
In this new edition we quote as extensively as before from the tape recordings we made and use the same framework in analyzing them that helped us to understand them in the first place. We have recast our discussion of the progressive construction of understanding during group talk and the part played in this by studentsā questions, and we also include a chapter (No. 5) which uses our experience during the research to offer advice to teachers who wish to record and study group talk in their own classes or those of colleagues. We hope that teachers will find this useful whether they are informally attempting to understand their studentsā learning or preparing a thesis for an academic qualification.
Changing Perspectives and Practices
Our interest in childrenās talking was far from unique, of course. There have for many decades been liberal-minded educators who wished to give students more responsibility for controlling the pace and direction of their learning. Some teachers and theorists in the United States (Anderson 1959; Thelen 1960, for example) had long maintained that small group discussion was an appropriate way of achieving this. What was new in the sixties and seventies was that cheap recording equipment made it possible to investigate in detail what students said, and thus move toward an informed view of what the talk was contributing to their understanding of the topics discussed. English teachers in London (Britton 1969; Martin et al. 1976) transcribed studentsā talk to show that it contributes to learning. They set out to demonstrate how in discussion students could collaborate to make meanings, could reshape their thoughts and feelings through talk. It became possible also to ask how teachers influenced talk both directly and indirectly, and to explore other ways in which the talk and therefore the learning was influenced by the social context in which it took place. Beyond these were questions about the kinds of talk encouraged or inhibited by these influences and how the value of these kinds of talk as means of learning might be established. It was with concerns such as these that we set out on our research in 1973. We hoped that our work would initiate a tradition of research that would soon move on beyond our thinking. That this has not happened in the way we expected is one of the reasons for this new version of the book.
This is not to undervalue some important initiatives that were taking place in the broad area of language and learning. In the United Kingdom, the Bullock Report (DES 1975) gave an official statement of the importance of group talk in learning, and in 1989 spoken language became for the first time a compulsory part of the UKās National Curriculum for all students. However, in the late eighties this was modified by an official insistence on testing oral skills and on the teaching of Standard English. (There were even ministerial statements which implied disapproval of small group teaching methods.) Nevertheless the British government had funded from 1987 until 1993 a National Oracy Project which took as one of its main tasks to encourage teachers of all subjects to make use of small group talk with students of any age. (Oracy was used by Wilkinson et al. 1965 and Wilkinson et al. 1990 to emphasizeāin parallel with literacyāthe importance of spoken language.) The National Oracy Project chose to work through small clusters of schools, organized and partly funded as local projects, and this placed the development work in the hands of thousands of teachers who voluntarily gave up their time to take part.
Both within the project and elsewhere, many UK teachers have enthusiastically adopted group methods and been delighted with the results. The publications of the National Oracy Project, for example, have been filled with articles by teachers in which they quote with delight transcribed extracts from their studentsā classroom talk. On the other side of the Atlantic, the same has been true of the publications of the Oracy Project in Peel County, Ontario, Canada. Teachers found that students of all ages took part enthusiastically in group talk and that they often displayed unexpected understanding, knowledge, and ability to think for themselves.
In Canada, Brubaker, Payne, and Rickett (1990) published a collection of practical and theoretical articles from different countries which shows how widespread the interest in small group learning had become. In the United States, as a result of the widespread whole language movement, many elementary school teachers have become more aware that development in literacy cannot be separated from spoken language and are therefore experimenting with various formats for talk with very young children, including small groups (Short and Pierce 1990; Pierce, Gilles, and Barnes 1993). Two volumes, partly concerned with group talk, were published as the culmination of the Talk Project of Peel County, Ontario (Booth and Thornley-Hall 1991a; 1991b). It appears likely, however, that in North America educators take it for granted that small group talk is for young children and irrelevant to learning in high schools. This is not a view that we share. In Australia, Reid, Forrestal, and Cook (1989) had taken off from the earlier work of Martin and others in London to produce a systematic approach to group work addressed to classroom teachers, which includes a set of categories for recognizing and planning appropriate stages within group learning. This work in Western Australia, like the other projects we have referred to, involved groups of teachers in developing methods of working with groups in their own classrooms.
The spread of small group methods did not take place without the expression of doubts of its efficacy, however. Boydell (1975) and Galton et al. (1980) had reported that most of the talk in groups sitting together in primary classrooms in Britain was not work-related. Bennett and his associates (1984) carried out a general study of the learning tasks given to six- and seven-year-olds, and in the course of it found that younger children sitting in groups spoke to one another primarily (27 percent) to share information and seldom (8 per-cent) to explain. It seems likely that these depressing results came from classrooms where the students had been given insectionidual tasks and had not received any encouragement to work together or been helped to understand the value of collaborative talk.
In the United Kingdom teachers have not only used small group methods but have undertaken many small-scale studies. Some have been primarily concerned to investigate effective learning from group discussion in subjects such as English literature (Wilson 1976; Dewhirst and Wade 1984), mathematics (Webb 1980), and geography (Baldwin 1976). Others have addressed specific issues: Jones (1988) investigated inter alia the effect of studentsā conceptions of the nature of learning upon their participation in group talk; Phillips (1985) devised a set of functional categories for kinds of talk; and Moss (1989) studied how girls deal with advertising clichĆ©s. By the end of the National Oracy Project, some members were investigating oracy more analytically: Des-Fountain and Howe described the kinds of talk that should be valued for learning and Corden identified a range of roles for teachers and the effects of these on group talk (Norman 1992).
At the same time larger scale programs of research into what was called cooperative learning were being carried out by teams of researchers in the United States and Israel (Sharan et al. 1984; Yager, Johnson and Johnson 1985; Sharan 1990; Slavin 1990). Galton and Williamson (1992) refer to a survey by Johnson and others (1981) which reported that nearly two-thirds of studies of collaborative methods had reported superior learning from small group work in comparison with standard teacher-class teaching. (Some of the North American work seems to have been limited by an unspoken assumption that the benefits of small group work arise from peer instruction.) Summarizing the outcomes of later work, Sharan (1990, 288) wrote: āThe potentially positive effects of cooperative learning on studentsā achievement have been documented many times in the . . . research literature on cooperative learning.ā These researchers set up precisely defined versions of āgroup workā and trained teachers to put them into practice. Sharan et al. (1984) compared class teaching with two ways of working in groups, Group Investigation, which emphasized collaborative inquiry and shared responsibility, and STAD (Students Teams and Academic Divisions) which rewarded groups on the basis of the conflated improvement in the test performance of insectionidual members. They were able to show, for example, that collaborative group discussions made possible the most effective academic learning when tested with high-level interpretive questions about literature. This was complemented by the observation that group work is less effective than class teaching when tested by low-level questions requiring merely reproduction of content. Lazarovitz and Karsenty (1990), using pre- and post-tests in the teaching of photosynthesis, concluded that high school students who had worked in groups reached scores superior or equivalent to those who were taught in class and that they gained greater self-esteem and satisfaction in learning.
Conclusions like these have not gone unchallenged and have often been considerably qualified in the course of other studies. For example, Bennett and Cass (1989) warned that groups tended to accept inappropriate explanations in order to maintain consensus and found that there was most interaction in groups that contained one high-attainer with two low-attainers. They also discovered that it was practical tasks that led to more interaction in groups of younger students, but that it was often more open topics (as, for example, talking about stories) that led to what they called abstract talk. They also suggested that students need more immediate rewards during the early life of a group, perhaps through short, structured tasks, but that eventually intrinsic motivation is likely to become more important through the influence of group loyalty.
The Israel/United States research tradition has been mainly concerned with testing the outcomes of cooperative learning, paying little or no attention to what goes on in the discussions of a working group. It is possible to read whole collections of papers reporting these researches without seeing a single transcript of studentsā talk. For this reason their work has not contributed significantly to our understanding of what happens in group discussion and why it is effective for learning. Although their publications propose precise methods for setting up group work (which are discussed in Chapter 4), they do not help teachers decide how they should intervene or what aspects of group talk they should encourage.
In many of the publications there has been a useful stress on the need for teachers to foster and even to teach the skills needed by pupils for successful group work (Kagan 1985). Bennett and Dunne (1992) suggest many exercises that can be used as preliminaries, though they also insist that the learning can be done in the course of substantive tasks. Students need to understand the unspoken ground rules of group work and must be helped to recognize the quality of their own contributions (Biott 1987). Excessive dependence on the teacherās approval has been identified as a major stumbling block, and this is related to fear of failure. Students should be helped to develop self-evaluating skills so that they can monitor their own performance. There has been surprisingly little interest in how the norms of teacher-student relationships in a school are likely to influence the studentsā view of the roles open to them when set into groups, though Galton and Williamson (1992) point out when discussing teachersā instructions to students about group work that āsuch instructions must be embedded within class norms that encourage cooperation.ā
Some investigators have set out to define normal stages in the work of a group during a given task (Tann 1981; Reid et al. 1989), a new and useful area of inquiry. Few, however, have considered the place of group work within the range of activities that makes up a lesson or a sequence of lessons (Harlen 1985; Howe 1988). Bennett and Dunne (1992) note in passing how often teachers fail to see to it that there is time for reporting back to the rest of the class, which would provide opportunities for greater explicitness and for the teacher to join in a wider discussion.
Surprisingly there has been little interest shown in the question of whether all kinds of talk are equally useful for learning, beyond some investigation of high or low cognitive levels. Wells (1989) feared students might be trapped in the everyday thinking they brought from their homes and proposed that teachers encourage what he called literate thinking, that is, critical and analytic discussion framed in the explicit terms appropriate to some kinds of writing.
An important change of attitudes and practices has occurred during these years among teachers of foreign languages, who have realized that since their purpose is to enable students to talk and write in real contexts, they are more likely to learn to do so if they have opportunities to use language in ways that simulate these contexts (Widdowson 1978). There is now an extensive specialist literature on communicative approaches to language teaching (Day 1986, for example). Activities recommended for group work include role play, discussion of the subject matter of printed texts, and the collaborative telling and writing of stories. This is an area of the curriculum where group talk has earned an undisputed status.
Thus in 1994 we address our writing to a different situation from that we envisaged nearly twenty years ago. Many books of advice to teachers about how to manage small groups are now available (Howe 1988; Jones 1988; Reid et al. 1989; Galton and Williamson 1992; Bennett and Dunne 1992), as well as two packs of in-service training materials published for the National Oracy Project (Baddeley 1992; Kemeny 1993). We have good reason to know that across the world there are many teachers who are using small group methods in their classes. No longer are we addressing the unconverted alone. However, although teachers are successfully using small group methods which imply a social constructivist view of learning, it seems likely that it is still a minority who could give an account of how the learning takes place or describe the influences that shape studentsā participation. Although an agenda of practical issues has emerged, no new theoretical framework has been developed on which future research might call, so we have become aware that elements in our 1977 study still have value. In what follows we have aimed to retain those elements of the original work that have not been replaced by subsequent studies while setting them in a new framework that may be more helpful to teachers.
