GENDER, SPEECH STYLES

AND THE ASSESSMENT OF DISCUSSION


INTRODUCTION

The subject of this research is the linguistic activity of discussion, and the criteria used in the assessment of discussion. I am specifically interested in the applicability of a feminist model of conversation to the assessment of discussion in Scottish Secondary Education, and I will be considering the extent to which discussion assessment in schools is influenced by gender-specific conversational norms.

The introduction of talk assessment into 16+ all-abilities public examinations in Scotland, England and Wales is a relatively recent pedagogic development, and there has been interest amongst linguists in monitoring its progress, with the intention of making available to educationalists any knowledge about language developed in linguistics which may assist with teaching and assessing talk (eg. Brown and Yule 1985, Jenkins and Cheshire 1990, Swann et al.1992). The incorporation of linguistic knowledge into talk assessment has been limited, and it is notable that recent work within feminist linguistics, drawing attention to gender-based differences in conversational style, has not been taken into account as yet by the examining bodies. Given the evidence for gender differences in conversational styles, there is clearly a need for investigative studies into the implications of these differences for assessed discussions in the classroom, and this is the area in which my research is situated. Recent work by Jenkins and Cheshire (1991, 1992) also addresses these issues, and I place my work, and my results, in the context of theirs.

In this thesis, I aim to test the applicability of a model of co-operative / competitive talk, as defined by Jennifer Coates in 'Gossip revisited' (1989a) and other publications, to the assessment of discussion. The questions I set out to address in this study are as follows:

  1. 1. Is there a correlation between speech style and grade?
  2. 2. Is there a bias present in the Scottish Examination Board's Grade Related Criteria towards competitive talk, and/or insufficient acknowledgment of co-operative talk?
  3. 3. Is the classification of pupils' behaviour in class discussion as competitive or co- operative feasible?
  4. 4. Do boys and girls differ systematically in their use of competitive and co- operative linguistic forms in classroom discussion groups?
  5. 5. Does it appear that girls are penalised in the assessment of discussion?
  6. 6. Does the gender of the teacher appear to influence their assessment behaviour?
  7. 7. Do teachers experience conflict between the oracy skills which they would like to reward the pupils for, and those skills the Grade Related Criteria allow them to reward pupils for?
  8. 8. Is the gender of the teacher a factor in any sense of conflict identified in the point above?

My linguistic data consists of classroom discussions which took place in Scottish secondary school English lessons. These were recorded on video on behalf of the S.E.B. as part of their material for training secondary school teachers in the assessment of discussion at Standard Grade. I have concentrated on the discussions shown in the film Young People Talking, commissioned by the S.E.B. and produced by Jordanhill College. This film was part of a package which also included the S.E.B.'s analysis of the discussions featured and the levels/grades which they estimated the speakers shown in the film to have achieved. I am very grateful to the S.E.B. for making this material available to me.

Models of how conversation works are being continually developed and refined, and feminist research has contributed important insights to this field of knowledge. Work in this area has provided useful overviews and critiques of mainstream linguistics. Many of the assumptions made by non- feminist research into gendered language use have been questioned by feminists working in this area, who have pointed out the androcentrism behind the aspects of language use selected for investigation, the groups whose language use was investigated, the methods of data collection used, and the explanations proposed for the differences found (see Coates and Cameron 1989:13-26). The questions raised in the work of Coates and Cameron and other feminist linguists, concerning the assumptions which have been made in linguistics, are potentially a great contribution to the vitality of research within the field.

The model of discussion used by the S.E.B., as I will demonstrate below in section 2.4, does not take account of research which suggests there are gender differences in conversational style, although other linguistic models, such as those of the systemic linguist M.A.K. Halliday (1975, 1978), and those of educationalists such as James Britton (1970) and Douglas Barnes (1971), have been significant influences. It was necessary to extend the existing linguistic theoretical models in order to establish a theoretical and methodological framework substantial enough to cope with my data. To do this, I worked on the model proposed and developed by Coates, with reference to other current linguistic research. By transcribing and analysing substantial amounts of naturally occurring data, and interviewing teachers, I have tested the relevance of Coates's work to classroom assessment of talk. The aims of this study therefore are twofold. On the theoretical linguistic axis, I hope my work provides an extension of the co-operative/competitive model of conversation. On the education axis, I hope my application of this model to the data has drawn attention to important, and previously ignored, aspects of talk assessment.

I use the terms 'co-operative' and 'competitive' throughout this study. The terms are used as Coates (1986) uses them, to refer to specific linguistic forms. In the work of Coates and others, these features have been found to occur in the conversational interaction of women and men to different extents. Studies in this area have found that men in most situations use co-operative forms less frequently than women, and competitive forms more often than women. The features of discussion which are associated with these two styles will be described in detail in Chapter One. However, I will comment here on two problems arising from my selection of the terms 'co-operative' and 'competitive'.

From my data analysis, it soon became apparent that a speaker who was not using co- operative forms was not necessarily using competitive forms either; the absence of a co-operative form does not entail the use of a corresponding competitive form. Thus I needed to increase the scope of the analytic structure to include a category denoting the absence of co-operative forms, without implying speakers are actually using competitive forms. I applied the term 'non-co-operative' to this situation.

There is also an inherent problem in the use of the terms 'competitive' and 'co-operative', since these are not only descriptive, but also evaluative. The labels may seem to present co-operative talk as essentially 'good' and competitive talk as 'bad'. Whether the use of certain linguistic forms can be good or bad is not self-evident, and is an issue to which I will return in Chapter Six. For an account of the origins of the label 'co-operative' to describe the features associated with women's single-sex conversation style, see section 1.2. Despite the evaluative dimension of the terms, I have retained them since they are fundamental to the linguistic model I have taken from Coates.

In my analysis of the discussions in Chapter Four, I make use of both qualitative and quantitative analyses. Developments in linguistic theory have challenged the notion of a simple form/function correlation (see, for example, Cameron, McAlinden and O'Leary 1989, Coates 1987, Winefield et al. 1989). These studies have illustrated the problems of merely counting forms whose function depends on the context in which they appear. I have taken account of their findings by making my main approach close analyses of small samples from my data, and by using quantification when appropriate, in the light of a careful, context-sensitive assessment of the functions of the forms being counted.

I transcribed six discussions featured in the Young People Talking film, and therefore accompanied by the S.E.B.'s grading and assessment comments, and I also used one discussion which was not used in the film. I analysed these seven discussions for pupils' use of co- operative, non-co-operative, and competitive linguistic forms, in order to profile which pupils used them, to what extent, and in what contexts.

I took an audio-visual recording of two extracts from the discussions I had analysed into the English departments of local Glasgow schools, for teachers to assess one pupil from each discussion (this is also the practice used by the S.E.B. for training and moderation). One of these pupils was a very co-operative speaker in a co-operative discussion, while the other was much less co-operative speaker, in a discussion which was non-co-operative, and sometimes competitive. The aim of this activity was to acquire information on the criteria teachers used when they made assessments, and to discover how comfortable they were with applying the S.E.B.'s criteria to the extracts.


Chapter Contents

In Chapter One, I outline the theoretical foundations of my research, and describe in detail the linguistic features associated with co-operative and competitive styles of talk. In Chapter Two, the implementation of discussion assessment in Scottish schools is described. The history and process of the assessment of discussion at Standard Grade are reviewed and the terminology associated with assessment is introduced. I also discuss attitudes amongst staff, pupils, and educationalists towards discussion assessment, to establish the context in which discussion assessment takes place. The chapter concludes with a section reviewing the Scottish Examination Board's assessment criteria from the perspective of co-operative and competitive models of language. In this section I consider the evidence for the assessment system containing a possible bias towards competitive linguistic forms. Chapter Three covers my data collection methodology.

There are two chapters containing analyses of data. Chapter Four contains the analyses of the videoed classroom discussions. The linguistic model proposed in Chapter One is applied to transcripts of the data, to investigate the extent to which girls and boys can be said to be using co-operative or competitive discussion styles. Chapter Five presents teachers' assessments of, and comments on, two video-recorded sample discussions. This data is analysed to highlight the criteria teachers actually use when they assess. I look in particular at whether the gender of the teacher influences the criteria they use, and whether the use of co- operative linguistic forms appears to influence the grades a pupil is awarded.

In the concluding chapter of the thesis, Chapter Six, I will return to the questions listed above, and consider whether there is a correlation between the gender of a pupil and the style of talk they use in a classroom discussion, whether the use of either co-operative skills or competitive skills can be seen to be either rewarded or penalised in the assessment of classroom discussion, and whether the gender of a teacher is an important factor in the criteria that teacher uses to assess discussion skills. I will also review any issues arising from my theoretical framework, and any methodological considerations.

There are three appendices which contain the S.E.B.'s Grade Related Criteria for the assessment of discussion (Appendix I), the questionnaire used in the collection of data about teachers' assessment practices (Appendix II), and the full transcriptions of the seven classroom discussions which I analyse (Appendix III).

I am indebted to the Scottish Examination Board, the schools and classroom teachers for their help and support throughout my study. The members of staff I met were all interested in my work, and enthusiastic about their own. All were committed to improving opportunities for discussion and refining assessment methods. They were actively engaged in solving the constraints of teaching and assessing in their classrooms, and resolving any problems and inconsistencies in assessment procedure as they encountered them. I am aware that in addressing one area of talk assessment, my results will leave unanswered many of the uncertainties faced by pupils, teachers, and moderators. I have, however, attempted to present a perspective and a set of terms with which to discuss classroom discussion, and I hope this will be a further step towards a satisfactory model of language for teaching, assessment, and research purposes, and a useful addition to the debates surrounding oral assessment.


Back to Main Page