MICRO-LEVEL TECHNIQUES OF POWER IN THE CLASSROOM PRODUCTION OF CLASS, RACE, GENDER AND OTHER RELATIONS

Jennifer M Gore

Faculty of Education

The University of Newcastle

University Drive

Callaghan, NSW

Australia. 2308

vejmg@cc.newcastle.edu.au

DRAFT ONLY: Please do not quote or reproduce without the permission of the author

Paper prepared for presentation at the Annual Conference of the Australian Association for Research in Education, Adelaide, November 30-December 3, 1998

In this paper, I draw upon data collected in two pedagogical sites to explore how social differentiation operates at the classroom level. So-called "hidden curriculum" research has documented the production or reproduction of class, race, and gender dynamics in schools and/or classrooms, typically using a methodology which used one (sometimes more) of these aspects of social difference as a foundation. More recent analyses continue this approach of a primary focus on one major dynamic, despite consideration of intersections with other dynamics, demonstrating how it operates or has operated in schools (e.g., Kamler, et al, 1994; Kenway and Willis, 1997; Middleton, 1998). For instance, Epstein and Johnson (1998) in a recent publication select sexuality as their primary concern, as conveyed in the title of the book Schooling Sexualities, despite detailed investigations of other dynamics, especially race. Quite rightly, on the basis of their theoretical and empirical work, they state that "sexuality is present throughout the formal and informal curriculum" (p. 195).

While I would not want to devalue such important work or to dispute such a statement, I think it is important to consider the theoretical and practical implications of focussing on a single dynamic, such as sexuality. Race and class and gender and ethnicity, and potentially a range of other dynamics, are also present throughout formal and informal curricula. How can we understand the intersection of all these structural dynamics, the way they interact or compete with or complement each other? It is interesting that despite the insights of poststructuralism and the embrace by many who are concerned with identity politics of the notion of multiple or hybrid subjectivities, much of the empirical work still focuses on, or at least commences from, singular dynamics. Of course, we have to start somewhere and we can't do it all at once. But are radical educators (for a lack of a better term) making slower progress than they might because of the approaches taken to documenting and interpreting the schooling experience of non-dominant social groups? Moreover, what are educators to do who are concerned about minimising the most harmful effects of these dynamics? How can "most harmful" even be determined? And how are they to deal with the complexities of classrooms in which multiple dynamics circulate?

As a preliminary answer to these questions, I argue, following Foucault, that we can start from a micro-level focus on power and build up to an identification of structural dynamics involved or constituted rather than commencing from the structural concerns. There is a danger when starting from a single (or even more than one) social dynamic that everything will be seen as contributing to that dynamic. The approach I have adopted attempts to account more fully for the enormous complexity of schooling and social life in general.

In so doing, this paper responds directly to those "critical" educational researchers who make rather broad claims about the pervasiveness of gender, or class, or race. It also responds to those who emphasise interactionist approaches which claim to be "a useful corrective to crude empirical claims that are sometimes made about what happens in schools" and to "modify any extreme notions we might hold of the monolithic behaviour, beliefs and attitudes of individuals that some subcultural studies might lead us toward" (Woods, 1990, p.140).

Methodological Issues

"There is always more going on than we know or remember [or can see] . . . in classrooms" (Ellsworth, 1997, p. 65).

While the full study involved observations and interviews in four different sites, in this paper I focus only on the observational data from the two institutional sites. This decision was based on a particular concern to illuminate institutionalised practices which may relate to the construction of structural social dynamics. The two sites were a high school in which a Year 8 and a Year 11 class were observed for two school terms, and a university teacher education program in which a first year cohort was studied for one semester. The PE classes shared the one female teacher. The TE program involved a team of three lecturers and around 100 students who each week attended a lecture, a talk from a guest speaker, and a tutorial, and who also participated in a related four-week practicum during the semester. Both groups were highly homogeneous in terms of race and ethnicity. The PE site was low in socio-economic status, the TE site more diverse and higher in SES terms. There were more females than males in the TE site, with roughly equal numbers in the PE site. I did not have access to the sexuality of participants.

My own study proceeded from a Foucauldian understanding of power as operating from the ground up. Hence, I examined the operation of power, including the specific techniques of power and the direction of any exercises of power, prior to assessing any effects in terms of class, race, gender, or other such categories. Two sets of codes were used in relation to the data which pertain to this paper. The first set of codes was designed to categorise the techniques of power employed by teachers or students (Gore, 1995) while the second set coded for the explicit presence of the social dynamics which are addressed here. The techniques of power were defined as follows:

Surveillance: Supervising, closely observing, watching, threatening to watch, avoiding being watched

Normalisation: Invoking, requiring, setting or conforming to a standard, defining the normal

Exclusion: Tracing the limits that will define difference, boundary, zone, defining the pathological

Distribution: Dividing into parts, arranging, ranking bodies in space

Classification: Differentiating individuals and/or groups from one another

Individualisation: Giving individual character to, specifying an individual

Totalisation: Giving collective character to, specifying a collectivity/total, will to conform

Regulation: Controlling by rule, subject to restrictions; adapt to requirements; act of invoking a rule, including sanction, reward, punishment

Figure 1. Coding categories for the techniques of power

The categories of social differentiation were defined as follows. In this study, 'sexuality' was used to code matters which pertained to hetero-, homo-, and bi-sexuality, "queer" identity, and any other acknowledged form of identity based in one's sexuality. 'Gender' was used wherever issues of masculinity and femininity, or male and female, arose. 'Race' pertained to issues of Aboriginality, "whiteness" or other racialized categories. 'Ethnicity' referred to cultural and national identities. 'Class' referred to socioeconomic issues and groupings. Each segment of relevant fieldnote text was coded for one or more of these social dynamics.

Given the interconnectedness of, and theoretical debates about, naming these categories, specific limits had to be placed on the usage of each. I acknowledge the huge problems of definition, of the tendency to use binaries, and of the difficulties of observation (none of which can be addressed within the scope of this paper). My goal in this work is theory generation, not claims to "truth". I have used the data collected to develop propositions which may be tested in subsequent work and which may be worth feeding into existing theoretical and methodological debates.

When coding the data for one or more social dynamic, I was looking for explicit statements or actions in these classes which pertained to any group or topic which might be identified within the category. Despite the important theoretical and empirical work that demonstrates the countless indirect ways in which such dynamics are implicated in schools and classrooms, I am nonetheless focussing here on direct instances when classroom processes or discourses were about class, race, gender, etc. Certainly, these codes would have been used much more frequently if I had included indirect constructions of the categories of social differentiation outlined above, but as I shall elaborate shortly, it is interesting how few classroom interactions were directly about the social dynamics which are my focus here.

For this paper, I linked my earlier analysis of techniques of power (such as surveillance, normalisation and classification) (see Gore, 1996) with the more traditional radical educational concerns with structural dynamics of power-class, race, gender, ethnicity, and sexuality.

Some of the key processes adopted for the analyses which have informed this paper are:

1. Noting the prevalence of each dimension in these classes. How often was the data coded for each social dynamic?

2. Mapping which techniques of power coincided with those instances when social dynamics were coded and noting the patterns across race, class, gender, sexuality, ethnicity.

3. Summarising how the techniques of power connected with each category of social difference to construct a more contextualised account of the production of each.

4. Examining the interrelationship of techniques of power and social dynamics (e.g., were they context bound such that race was rarely raised in classes with all white students?)

5. Identifying site differences between and within the physical education and teacher education sites.

Findings

The prevalence of the social dynamics

It is important to note at the outset that for the vast majority of instances in which these techniques of power were exercised, traditional social dynamics (as defined above) were not implicated in any way. This means that most exercises of power in these classrooms were not apparently about class, race, gender and other relations. Instead, the majority of exercises of power contributed more directly to the production of other power relations, to constructions of self, and/or to constructions of knowledge, centring on, for instance, what it means to be a student or certain kind of student in a particular class or institution. Indeed, my research has led me to see these three elements -relations, self, and knowledge-as fundamental to, and constitutive of, pedagogy.

It is also worth noting when segments of the data were coded for social dynamics, by far the greatest number of codes were for gender rather than any of the other categories of social differentiation. That is, in these two sites, 190 segments of gender were found, with around 30 segments each of sexuality, ethnicity, and race, and less than 20 segments of class (see Table 1).

PE8

PE11

TE

Total

gender

104

39

47

190

race

2

0

27

29

ethnicity

4

0

27

31

sexuality

8

18

5

31

class

1

5

13

19 Table 1. Number of segments within each site coded for specific social dynamics

Differences between the sites pose some important questions such as: Why does sexuality feature more in the school setting than in the teacher education setting? Why is class least prevalent overall? Why is gender so prevalent? Why do race and ethnicity emerge so frequently in the TE site compared with the school site? I will provide some preliminary answers to such questions later in the paper. First, I provide a qualitative account of the data which was coded for each of the social dynamics.

Gender

In the PE site, most instances of gender arose from the teacher using sex as a basis for the organisation of class activities or from observations of behaviour between and among male and female students (e.g., boys engaging in mock fighting, girls squealing, girls cuddling, different levels of engagement/passivity between boys and girls). There was also talk among the students about various aspects of femininity and masculinity (e.g., references to female body-builders, to roles of mothers and fathers) and instances of them organising themselves along sex lines (e.g., when the girls chose not to do a fat test in front of the boys). In the TE site, many of the gender instances were introduced by lecturers and guest speakers, some of these taking the form of lecturers attempting to get students to construct/reconstruct their gendered understandings. Some instances centred on students' opposition to the feminist perspectives introduced in the lectures, and/or discussions of sexism. From these data, gender was clearly a legitimate basis for class organisation, for naming individuals and groups. Similarly, aspects of feminist discourse have clearly been widely incorporated into the language of classrooms, such that kids were accusing each other of sexism (including boys accusing girls), and feminist perspectives were being debated. Interestingly, in an interview one Year 8 boy claimed that he used sexist arguments to get out of work: "I say one thing and it's a bit sexist so all the girls come back [at me], and then all the boys come back. All the boys do it so they can get out of work (laughs) . . . I'm pretty cunning when it comes to things for getting out of work." This explanation in itself points to a more complex manifestation of gender relations than a traditional research approach might allow.

Race

The only race codes in the PE site occurred in the Year 8 setting when the teacher named Aboriginal Australians as the most Australian Australians and when a boy told an Aboriginal joke. In the TE site, all of the segments coded for race related to the formal curriculum, either by way of lecturers or guest speakers presenting race issues (such as Aboriginal deaths in custody, or cultural inclusivity in the curriculum), relaying racist experiences and discussing racism in Australia, describing characteristics of Aboriginal and non-Aboriginal people, employing racial categories, or by way of student discussions of these issues or role plays of related issues.

Ethnicity

A similar pattern was found for ethnicity. In the PE setting, the only segments coded for ethnicity occurred in the Year 8 class. Most of these related to the formal curriculum focus on the social construction of the family. The only "enactment" of ethnicity occurred when a student was asked by another about her ethnicity and she named herself as Macedonian. In the TE site, the majority of segments coded for ethnicity arose in relation to the formal curriculum. Many instances took the form personal experiences recounted by lecturers and guest speakers or of discussions of NESB and ESL students and multiculturalism. Other instances related to student discussions of the formal curriculum and role plays. A small number of segments related to researcher observations of different participation rates and interaction patterns which seemed to correspond with the ethnic backgrounds of students. For instance, the few students of Asian backgrounds in the tutorials did not speak often and when they did the other students did not seem to engage with their contributions.

Sexuality

The incidence of sexuality in the school site was in part a function of the curriculum content and in part a function of the extracurricular concerns and categories articulated by the school students. The Year 8s completed a unit on families in which the issue of gay families was raised. There was general agreement among students that a whole range of different family types is acceptable and that the sexuality of parents shouldn't matter. Interestingly, gay or lesbian parents were only discussed as couples while the possibility of a single homosexual parent was not acknowledged. There was one case when a boy used the word "faggot" as an insult directed at another boy. In the Year 11 class, direct references to sexuality were more widespread. They included references to other students as "poofters" or as sexy; discussion of male sexuality in relation to a man entering a beauty contest; discussion of sex/kissing/sexual acts; discussion of heterosexual and homosexual relationships and of actions that could be construed as gay or lesbian; and male objectification of women. Many of these instances occurred outside of the formal curriculum in the many unrelated conversations that students had while "working". The few instances of sexuality in the TE site all took the form of discussions of sexuality and included: appropriate dress for female teachers in order to minimise their sexuality; sexual assault; questioning of a guest lecturer's marital status (and, by implication, her sexuality) given her lecture on feminism which many students found confronting; and, the recounting of stories from prac teaching experience in which students (all female) had been sexualised by their students (one had been asked if she used condoms, another hugged inappropriately by a male student, another given the finger). From these data, it appears that sexuality was more overt in the school site. Even many of the teacher education examples were derived from experiences in schools. The greater access to students' informal discussions in the PE site may also explain its relative prevalence there.

Class

The only instance where class was coded in the PE Year 8 site related to the teacher using a farting incident to "remind" students that not all children have the "privilege of learning how to do things". She talked of behaviours exhibited by people because they "don't know any better". In the Year 11 site, three of the coded segments related to student interaction. Given the few segments of data code with class, I provide more detail of these instances than for some of the other social dynamics. Madeleine was eating another student's food in class (eating in class was acceptable practice). Zac told her to bring her own food to school. She said "I live in poverty" and bit into Remus' roll. Later in the same lesson, Zac used the F word and so received a demerit point. He protested that the reason was because every time he put his hand in his bag Madeleine asked for food. The teacher told him to be assertive but Madeleine said again "I told him I live in poverty". In another incident, Madeleine remarks about another student that she "doesn't know why he is like that, he's got everything". Another girl responds by saying "maybe they've given him too much". Another segment related to the curriculum content (talking about different stress levels and life circumstances between businessmen and women and many women who are in lower paid full time employment and still taking on the bulk of housework and child rearing) and the final instance was when a girl said "You're poverty" to a boy. This particular insult, constructed from class-based categories, seemed quite unusual. In the TE site, class arose primarily when lecturers or guest speakers raised issues such as class considerations in curriculum development and public versus private schooling. In a couple of instances, students raised class in discussions of appropriate teacher dress, for instance when a student talked of being assigned to Neutral Bay for practicum and not being able to afford the appropriate style of dress.

The intersection of social dynamics

There were relatively few instances in which more than one social dynamic was simultaneously invoked or operating (see Table 2). As depicted in Table 2, which provides frequencies and row percentages, the intersection of social dynamics occurred infrequently (some such instances are evident in the examples provided above). For example, when gender was coded, race was also coded for only 0.5% of the 190 segments, class for 1.6%, sexuality for 2.1%, and ethnicity for 3.2%, and so on.

gender

race

class

sexuality

ethnicity

gender %

n

100

[190]

0.5

[1]

1.6

[3]

2.1

[4]

3.2

[6]

race %

n

6.9

[2]

100

[29]

0

[0]

3.4

[1]

10.3

[3]

class %

n

21.1

[4]

0

[0]

100

[19]

0

[0]

10.5

[2]

sexuality %

n

19.4

[6]

3.2

[1]

0

[0]

100

[31]

3.2

[1]

ethnicity %

n

22.6

[7]

12.9

[4]

3.2

[1]

3.2

[1]

100

[31]

Table 2. The coincidence of social dynamics, showing row percentages and [frequencies].

(Note: Minor discrepancies due to inter-rater instances).

Despite all the work on the intersection of differences and on multiple subjectivities, these data show that race, gender, class, etc. were most commonly operating in isolation.

The micro level focus of the analysis may in part account for this result. Nonetheless, the data show that in most instances, only one aspect of identity was used in the discursive structures and practices of these classrooms. This finding may also be a function of a need/attempt to avoid too much complexity or the reliance on discursive structures which centre on a single piece of information, often constructed around binary categories. Indeed, the continued reliance in the radical educational literature on a single dynamic as an organising feature of both empirical and theoretical work is indicative of similar discursive structures, of the inability to actually speak a complexity which moves beyond existing linguistic and political affiliations with existing categories.

Techniques used in the production of the social dynamics

The coincidence of segments coded for each social dynamic and for specific techniques of power is depicted in the frequency chart which follows (see Table 3). This table depicts the number of segments in each site coded for each of the social dynamics, as well as the number of segments for which specific techniques of power were also coded. For instance, in the PE site, gender was coded in 143 segments of data. When gender was coded, it coincided with codes for surveillance in 34 of those segments, for normalisation in 26 of those segments, and so on. The total of gender codes across the two sites is also presented with the total number of times each technique of power coincided with gender. The same applies for each social dynamic. A number of observations can be made from these data. In particular, I have examined the frequency of specific techniques, patterns in the relative frequency with which specific techniques were employed in relation to each social dynamic, and patterns by site. A summary of these analyses is provided below.

Overall frequency of specific techniques

regulation was the least used technique in relation to the set of social dynamics

surveillance also played a lesser role

* individualisation and exclusion were the most frequently employed techniques

in descending order of frequency, the techniques were individualisation, exclusion, totalisation, distribution, normalisation, classification, surveillance and regulation

Of interest here is the fact that these social dynamics were not addressed by way of rules or direct sanctions, but by the more indirect means of exclusion, normalisation, and classification. Such a finding is consistent with Foucault's account of disciplinary society.

Patterns in frequency of techniques for specific social dynamics

(from most to least frequent; a slash indicates equal frequency)

for gender ind, dist, tot, exc, norm, sur, class, reg

for race class, exc, norm, tot, ind, dist, sur, reg

for ethnicity ind, class, exc, norm, tot, dist, reg, sur

for sexuality exc, ind, norm/ class, tot, dist, sur, reg

for class norm, exc/ ind, class/ tot, dist/ sur, reg

These patterns give more weight to the overall frequencies. In particular, it is interesting to note that distribution, surveillance, and regulation group as the least prevalent techniques in relation to each social dynamic except gender. Distribution clearly played a much more central part in the enactment and production of classroom gender relations than for any other dynamic. Exclusion and normalisation also lie in close proximity to each other for each social dynamic, highlighting their connection (to be addressed more fully later). Classification was important in relation to race and ethnicity. Exclusion and individualisation were important in sexuality.

Patterns for the two sites

(not shown on the table without additional calculation)

PE tot, dist, ind, exc, sur, norm, class, reg

TE norm, exc, class/ ind, tot, dist, sur, reg

This pattern is not inconsistent with that obtained in an analysis of the total use of these techniques in the two sites (see Gore, 1996). That is, it appears that the techniques which dominate a site tend to hold whether or not social dynamics are implicated in the use of the techniques. In earlier work, I argued that radical pedagogy had paid too little attention to the actual practices of pedagogy, focussing instead on "larger" concerns with societal transformation around class, race, gender and other categories of social differentiation. What these data suggest is that the patterns of power associated with pedagogical practice are attributable much more to institutional location than to the particular radical or otherwise concerns of the educators. Nonetheless, despite the dominance of the institutional location, there were clear differences in terms of the pedagogical approach associated with specific social dynamics. I need to do more analysis of this point in order to make a stronger claim. I will, however, explore implications in the final section of this paper.

A Three Dimensional Mapping of Results

Correspondence analysis takes the data plotted in the frequency table provided above (Table 3) and calculates correspondences between the variables represented by the rows and columns of that table (see Figure 1). It is a technique for "describing and establishing relationships in qualitative data. . . . [It] has as its principal characteristic an exchange of the roles of variables and observations and seeks to represent them in the same space" (Henry, 1988, p. 621). In so doing, correspondence analysis identifies the primary axes along which variation occurred. With the data I am using, two axes were identified (there can be more), that is, these axes depict almost 95% of the inertia. In correspondence analysis, the analyst interprets what the dimensions are that describe the axes. I have interpreted the correspondence analysis as follows.

Figure 1. Correspondence analysis results: Techniques of power and categories of social differentiation

1. Dimension one (the x axis) can be understood when examined in conjunction with the other analyses presented here, as representing a distinction between actually doing or enacting classroom gender relations on the one hand, and, on the other hand, talking about social differences. Both of these processes clearly contribute to the production of social group differences, but gender clearly operated very differently in these classrooms than did the other dynamics. In another sense the distinction may be viewed as one which is about the production of classroom power relations and constructions of self rather than about the production of knowledge.

In classrooms it is both legitimate and common to name, to "essentialise", and to enact pedagogical practices around, the categories "boys" and "girls", while it is not legitimate to explicitly name or organise pedagogy around other groups such as "poor kids" and "rich kids" or "Black kids" and "white kids"(except perhaps in affirmative action initiatives). Hence the correspondence between gender, distribution, and totalisation is indicative of the institutional practising of gender relations in the classrooms observed and helps explain its prevalence. Such use of gender is not restricted to teachers. Students also freely use/d gender as a category of social organisation.

As with my earlier analysis (Gore, 1996), the techniques of normalisation, exclusion, and classification are more about the production of knowledge than the techniques of distribution, totalisation and surveillance (which I have called the more "corporeal" techniques) which link more directly with the production of bodies and relations between people.

2. The second dimension (represented on the y axis) distinguishes in particular between individualisation and classification. It is particularly interesting to note that class and sexuality cluster with individualisation while race and ethnicity cluster with classification. In these classrooms, issues of class and sexuality were commonly raised in relation to particular individuals (either in or beyond the classroom itself). There might be something about the middle-class institutions of schooling and higher education which, when coupled with discourses of "schooling for all", explain why class is not very prevalent as an explicit bases for differentiation, either among students or in discussions of the broader population. Similarly, it is not surprising given institutional attempts to de-sexualise schools (Epstein and Johnson, 1998) that sexuality "broke free" mainly in relation to specific individuals and circumstances. Race and ethnicity, on the other hand, were raised in reference to groups (Aborigines, NESB students, etc.) and issues (racism , multiculturalism, etc.) presented as part of the formal curriculum by lecturers and guest speakers. Along this dimension, gender falls right in the centre, indicating both the classifying and individualising techniques used in the production of gender relations and gendered understandings.

3. Of further interest is the location of normalisation and exclusion, first, in close proximity to each other, and, second, roughly equidistant from each of the social dynamics. As the frequency data indicate, these techniques were frequently employed in relation to each of the power dynamics. The pathological is often named in the process of establishing a norm (Canguilhem, 1991). Given the formal curricular emphasis on topics like feminism and multiculturalism, particularly in the TE site, the important role of normalisation and exclusion in encouraging students and student teachers to challenge some of their own beliefs and some societal practices is not surprising.

Conclusion

In short, the various analyses conducted here, in conjunction with the theoretical framework, have produced the following propositions about the classroom production of class, race, gender, and other relations.

1. The social dynamics of class, race, gender, etc. do not all function and are not all produced in the same way. Different techniques of power appear to prevail in relation to specific categories. The context in which these data were produced may well be a factor in the specific findings, but nonetheless, it seems reasonable to conclude that such differences as race, class, and gender function differently.

2. The "place" of gender in social organisation and hence in classrooms here (and, I suspect, in many other places) is quite different to the other dynamics, both in terms of its prevalence and in terms of its enactment.

3. Race and ethnicity correspond most clearly with classifying techniques of power.

4. Sexuality and class correspond most clearly with individualising techniques of power.

Implications

Educators and educational theorists may need to be more cognisant of the specific techniques by which social dynamics are addressed and produced in classrooms. In particular these findings raise questions about both the productive and repressive effects of taking gender so much for granted, of speaking about race and ethnicity in categorical and essentialist terms, and of speaking about class and sexuality in individualistic terms. The political and pedagogical ramifications of the different ways in which these differences are "produced" may well be something that those committed to social change/transformation will want to address.

At another level, educators and theorists may need to be more cognisant also of how the production of student and teacher subjectivities cuts across the categories of difference which have been the traditional concern of so-called radical educators. If the preponderance of exercises of power in classrooms are not about race, class, gender, and so on, but about other kinds of relationships between and among teachers and students, then transforming both the experiences of those in educational institutions and wider social dynamics may require greater attention to such fundamental power relations as those which shape us during the twelve of so years most people now spend in formal education.

I return, here, to earlier arguments I've made about what I call a "regime of pedagogy (Gore, 1993)." Pedagogy is much more than a relay for power relations external to itself. Pedagogy creates and constructs its own power relations, its own subjectivities, and its own discourses. In my view this way of thinking about pedagogy warrants much greater attention than it has been given in either radical or mainstream educational literature.

Acknowledgement

I wish to thank Jim Ladwig, Gavin Hazel, and Kellie Morrison, each of whom has provided various forms of invaluable support in the production of this paper. I also acknowledge the Australian Research Council for grants which enabled the study to be conducted.

References

Canguilhem, G. (1991). The Normal and the Pathological. New York: Zone Books.

Ellsworth, E. (1998). Teaching Positions: Difference, Pedagogy, and the Power of Address. New York: Teachers College Press.

Epstein, D. and Johnson, R. (1998). Schooling Sexualities. Buckingham: Open University Press.

Foucault, M.

Gore, J.M. (1995). On the continuity of power relations in pedagogy, International Studies in Sociology of Education, 5(2), pp. 165-188.

Gore, J.M. (1996). Understanding Power Relations in Pedagogy. Paper presented at the Annual Conference of the Australian Educational Research Association, Singapore.

Gore, J. M. (1998). On the Limits to Empowerment through Critical and Feminist Pedagogies. In D. Carlson and M. Apple (Eds). Critical Educational Theory in Unsettling Times. Minneapolis: University of Minnesota Press.

Henry, G. (1988). "Correspondence Analysis" in J. Keeves (Ed.). Educational Research, Methodology, and Measurement: An International Handbook, (pp. 621-624), Oxford: Pergamon Press.

Kamler, B., et al. (1994). Shaping Up Nicely: The Formation of Schoolgirls and Schoolboys in the First Month of School. Canberra: Australian Government Publishing Service.

Kenway, J. and Willis, S. (1997). Answering Back: Girls, Boys and Feminism in School. St. Leonards: Allen and Unwin.

Middleton, S. (1998). Disciplining Sexuality: Foucault, Life Histories, and Education. New York: Teachers College Press.

Woods, P. (1990). The Happiest Days? How Pupils Cope with School. London: The Falmer Press.

3