Schools, Teachers, and Curriculum Change: a Balancing Act?

Educational change is a fact of life for teachers across the world, as schools are subjected to constant and ubiquitous pressures to innovate. And yet many school practices remain remarkably persistent in the face of such innovation. This paradox of innovation without change is perplexing for policymakers and practitioners alike. This paper investigates the gap between policy and practice, between innovation and the changes in social practices that occur in response to such innovation. It draws upon empirical data from two case studies in Scotland – schools responding to new curriculum policy – exploring contrasting approaches to the management of innovation. One is a laissez faire approach, and the other a more directive managerial strategy. Through an analytical separation of culture, structure, and agency, derived from the social theory of Margaret Archer, the paper sheds light on the social processes that accompanied innovation in these two settings demonstrating how teacher culture and differing management styles impact upon externally initiated policy.


Introduction
Change is a mantra for the modern age. Education systems worldwide mirror this overall trend, with the last fifteen years being characterized by what Levin (1998) has described as an epidemic of change. As a result, practitioners' work has intensified, paperwork and bureaucracy have increased, and teachers have felt increasingly disempowered and professionally marginalized (Ball, 2008). And, yet, despite this flurry of reform activity, what Elmore (2004) has described as the core of schooling has remained, in the view of many commentators, relatively stable (e.g., Cuban, 1988;Sarason, 1990;Tyack & Cuban, 1995).
Moreover, clarion calls for change in schooling have been accompanied by apparently contradictory discourses. On the one hand, teachers have been systematically positioned as barriers to the change via discourses of derision (Ball, 1990) and rhetorics of excellence (Fore, 1998) and subject to heavy duty quality assurance regimes (Goodson, 2003). Such a view tends to construct curriculum change as a matter of the simple implementation of teacher proof curricula. Much research over the years (e.g., Bowe, Ball, & Gold, 1992;Helsby, 1999;Skilbeck, 1998) suggests that this is a deeply problematic course of action based upon dubious assumptions about human activity.
On the other hand, teachers have become recently positioned in policy as agents of change, as the standard bearers of professional models of transformational change (e.g., Scottish Executive Education Department, 2006). This view demonstrates increased awareness that policies mutate as 4 they migrate from setting to setting, as they are mediated (Osborn et al., 1997) by professionals in differing ways that reflect their skills and prior experiences, their values and attitudes towards the policy in question, the contingencies of the setting into which the policy is to be introduced, and the social interactions that accompany this translation from policy to practice. Supovitz (2008) describes this process as iterative refraction. Inherently, such translation is an uncertain process that cannot be accurately predicted in advance. It thus poses problems for politicians who inevitably work on short term electoral cycles being naturally concerned about issues such as value for money, and local policymakers (for example local authority personnel and school managers) whose job it is to interpret national curriculum policy and who are accountable for its outcomes. Translation is also subject to agential factors such as teacher capacity and will (Spillane, 1999) to engage with the policy in question, and these are in turn affected by contextual issuesfor example access to cultural and material resources and the existence of social structures that might restrict or enable engagementthat are often beyond the control of those seeking to enact policy. A significant contextual factor in recent years, both within the prescriptive and more permissive approaches to policy described above, has been the use of outcomes steering backed by rigorous inspection regimes and the quantitative use of attainment data. Arguably, such methods have done more to erode teacher autonomy (Biesta, 2004;Reeves, 2008) than has any recourse to prescriptive inputs. Moreover, it is increasingly clear that such methods are subject to unintended consequences, which can be as unhelpful as the allegedly low standards that they seek to address. Such consequences include pressures on teachers to teach to the test, consequently narrowing the curriculum (e.g., Miller, Edwards, & Priestley, in press2010;Torrance, 1997).
This paper explores the processes through which iterative refraction (Supovitz, 2008) occurs examining the impact of different forms of policy management on the enactment of the policy in question. The context is the school-based development of models of provision for social studies in Scotland following the development of national curriculum policy, policy that arguably exerted pressure on schools to develop integrated teaching of the social subjects (geography, history, and modern studies) in the early years of secondary schooling. The paper utilizes a theoretical approach derived from the social theory of Margaret Archer (1988Archer ( , 1995Archer ( , 2000 drawing upon empirical data from two contrasting case studies set in Scottish secondary schools. The paper first locates the research in its national context before setting out the theoretical model that guided the analysis of the data. Finally, after providing an overview of the key features of each case, the paper engages in an extended analysis of the data.

Integration and the social subjects curriculum in Scotland
This next section explains the context within which the research is located. The paper focuses on the ways in which teachers and managers in the case study schools responded to new curriculum guidance. The guidance in question, Environmental Studies 5-14 (Learning and Teaching Scotland, 2000), revised existing guidelines and arguably increased pressure for the integrated teaching of the social subjects (geography, history, and modern 6 studies) as opposed to separate provision of each in the first two years of secondary schooling (S1 and S2). While specific to Scotland, the case studies are of wider interest in that they provide good examples of teacher mediation of curricular guidance in the face of pressures to implement a controversial form of provision; as such they touch upon issues that are common to many current attempts to reform the curriculum.
The integration/subject debate has long been a source of controversy in Scotland despite the supremacy of subjects within the secondary school curriculum. This may be construed as a battle of paradigms. Primary education has a tradition of thematic teaching, with its roots in the 1965 (Scottish Education Department, 1965), whereas secondary education is firmly rooted in the teaching of traditional subjects. A second dimension of the debate is specific to the social subjects. The emergence of social studies (Gleeson & Whitty, 1976;Hill, 1994;Wesley & Wronski, 1973) as the predominant approach to teaching the humanities and social sciences in the early secondary school is a worldwide trend (Benavot, 2004;Wong, 1991), with which Scotland and the rest of the UK are largely out of step. There was considerable interest in such approaches in the UK during the 1970s and 1980s (e.g., Hargreaves, 1982;King, 1986), but these often tended to be associated with low ability pupils (Whitty, 1992), and largely dissipated in the face of pressures from Her Majesty's Inspectorate Education (HMIE) and national curricular developments (Ross, 1995). Nevertheless, despite the weak nature in the UK of these ideas about social studies, they existed as an alternative to the separate social subjects, and as a result, the idea of interdisciplinary provision 7 in this area has tended to resurface periodically when national debates about curriculum were taking place.

Primary Education in Scotland Memorandum
In Scotland, the Munn Report (Scottish Education Department, 1977) clearly identified the problems inherent in a traditional subjects curriculum, namely fragmentation and poor coverage of cross curricular issues. However, the report eventually fell into line with the predominant view in secondary schools reaffirming the "Hirstian subject-based curriculum with a nod in the direction of cross-curricular courses, but only for the less able" (Boyd, 1997, p. 60). These debates were to re-emerge in the late 1980s and early 1990s when the 5-14 curriculum was developed. The use of terminology such as people and place, people in the past, and people in society to replace the traditional subject names geography, history, and modern studies may be seen as giving an explicit message to schools about provision, reinforced suggesting an alternativeby their framing as a single coherent set of strands -, the social subjects -, with common enquiry skills descriptors, instead of discrete subjects.
This message was reinforced in the 2000 guidelines (Learning and Teaching Scotland, 2000) by the extension of an additional set of generic skills (developing informed attitudes) from people in society to the other strands. Simultaneously, schools were coming under pressure from HMIE (e.g., 1992, 1999, 2000) to reduce the amount of contact that young people had with different teachers, albeit through a different mechanism, that of short term subject rotations. An HMIE report identified that "it is not unusual for first year pupils to be taught by between 13 and 16 different teachers each week" automatically lead to integrated provision, it has increased pressures for the social subjects to be administered jointly. It is in this context that this study was conceived.

The research
The research took the form of a naturalistic inquiry employing qualitative methods. The natural settings were two clusters of social subjects departments within their wider school settings. The research can be broken down into three distinct phases as follows: 1. The collection of data about school social studies provision by questionnaire. This was a pre-research phase in some respects, as it enabled the development of subsequent methodology and the selection of case studies. 9 2. Initial site visits to each of the case study schools to explore school and departmental culture and teacher attitudes towards teaching and learning and curriculum change.
3. Follow-up site visits later in the school year to explore the issues that arose at the planning and enactment stages of the 5-14 social subjects guidelines.
In the latter two phases, data were generated through observation of teaching and the day-to-day life of the school departments, semi-structured interviews with teachers and managers, and analysis of relevant documentation. During the first set of visits, I interviewed the specialist teachers of geography, history, and modern studies once. Each teacher was observed teaching on at least one occasion before the interview; this provided a common frame of reference for the subsequent interview, as well as enabled the identification of critical incidents in teaching which could be discussed to explore underpinning decision making processes. The purpose of these interviews was to build a picture of the beliefs and practices of the teachers within the faculty. During the first site visits, I also interviewed members of the senior management team to explore aspects of school policy, philosophy, and management. The second site visit to each school had a different purpose, specifically to explore attitudes towards the developments that had occurred in the school following the publication of the revised curriculum guidelines. During this visit, each teacher was interviewed for a second time. Pseudonyms have been used in all research outputs to protect respondents.
Initial analysis consisted of open interpretive coding of the interview data guided by the conceptual framework outlined in figure one, with codes generated in terms of the initial loose categories of culture, structure, and agency in line with Archer's (1988Archer's ( , 1985Archer's ( , 2000 social theory. Coding was initially undertaken using the smallest case study; the codes were then applied to the remainder of the interview data and emerging further themes elaborated in the light of complementary and contradictory cases. In the latter cases, careful consideration was given to discovering whether these were due to particular or idiosyncratic factors of the case in question, or whether they could be examples of generalizable findings that were simply absent (or not apparent) in analysis of the initial case. If merited, the new codes were applied to the data, and the other case studies were reexamined for instances of the issue in question. This progressive focusing (Parlett & Hamilton, 1972, as cited in Stark & Torrance, 2005) allowed me to determine if there were grounds for inference through comparison of case studies.
I then undertook a more detailed analysis of the themes emerging from the data drawing more explicitly upon the Archer model. This analysis sought to further develop the analytical separation of the data by posing generic questions derived from the following categories.
 Questions concerned with culture. These largely relate to the kinds of knowledge that inform everyday practice and shape teacher values.   The next section of the paper further develops this conceptual framework, known as morphogenesis/morphostasis (M/M). M/M explicitly addresses issues of social reproduction and transformation, offering both an epistemological framework for understanding social change, through concepts such as emergence, and a set of methodological tools for those seeking to investigate change contextspractitioners and policymakers as well as academic researchers. The final sections of the paper explore the research data in the light of this theoretical model.

A conceptual model for analyzing continuity and change
The contradictions described in the first section of this paper occur in part, in my view, because of confused notions that surround the concept of change. Policy is frequently envisaged as something monolithic, change is often seen as a linear process of implementation, and success consequently viewed in terms of fidelity to the original policy. As suggested in the previous section, the process is considerably more complex than this. I do not seek to revisit this critique which has been the subject of much writing (e.g., Cuban  (1988) conflates the two, consequently obscuring the processes that occur when social interaction leads to cultural, structural, and individual elaboration.
Archer's notion of analytical dualism (or separation) allows us to disentangle the various aspects that contribute to the unfolding of a given social situation enabling us to make judgments about the relative causative weight of culture, structure, and agency.
For example, M/M may allow us to analyze how ideologies relating to accountability (cultural forms) enable the persistence of power that is an emergent property of certain roles and systems in schools (social structures), how these roles serve to perpetuate and develop the ideology in question, how the existence of the ideology and the roles and systems serves to enable and constrain teacher agency, and how certain teachers are able to bring to bear particular experiences and values in acting within these constraints. The diagram below (Figure 2) illustrates how a social system might be represented in this way.

Figure 2. Social reproduction and transformation
Viewed within such a conceptual framework, policies thus become cultural forms that migrate from setting to setting, encountering and combining with existing cultural forms to provide the cultural resources that underpin human activity. An important point to note is that policies, as official discourses or cultural forms, will inevitably have to compete with other less official cultural forms (such as local traditions) as well as with other, possibly contradictory, official discourses such as local policy. It is important to reiterate also that, at each level where ideas are mediated, they will be inevitably filtered by a range of different contextual considerations, personal values, and interpretationthe process described by Supovitz (2008) as iterative refraction.
Such a view of innovation and change has considerable implications for both policymakers and practitioners. It highlights the need for the building of professional capacity to engage with policy, boosting the ability of teachers to respond creatively from a wide repertoire for maneuver (Biesta & Tedder, 2007) to the problems faced when engaging with innovation. A related issue concerns the way in which we conceptualize capacity. Too often, it is seen as an individual quality, as personal capacity or ability. While this is clearly an element of capacity, it only conveys part of the picture. The concept of capacity is elicited more thoroughly by Biesta and Tedder's (2007) ecological view of agency. Such a view suggests that that, even if actors have some kind of capacities, whether they can achieve agency depends on the interaction of the both capacities and the ecological conditions (Biesta & Tedder, 2007)), and that individuals who exercise considerable agency in one setting might be disempowered in another. Thus it is necessary that all involved in the process of change become aware of structural factors that might enable or constrain agency, including school systems, the demands of attainment and quality assurance, relationships and power structures within schools, artifacts available, and the physical layout of buildings. It is also necessary when seeking to understand the dynamics of any change context to take account of cultural considerations, particularly values and attitudes that might militate for or against agency and the access that practitioners have to alternative cultural resources. In this vein, one should be cognizant of the important role of coherent policy in stimulating and facilitating innovation (Lieberman & Miller, 1999).

The schools
16 The research was undertaken in two schools. These are quite different in character, being selected from questionnaire (phase one) returns which suggested interesting curriculum development in response to the 5-14 guidelines.

Hillview School
Hillview School is a small school situated in a fairly isolated, rural town, with a mixed socioeconomic profile. The town is not a poor locale, but contains some relatively disadvantaged families. Attainment in exams is high in comparison with national and Education Authority (EA) averages. The school is composite, with primary and secondary sections of roughly 200 students each. The secondary section comprises cohorts from years S1-4 (ages 11-16); pupils choosing to continue with post-compulsory schooling attend a larger secondary school in a neighboring town.
Geography, history, and modern studies were organized through a social subjects faculty, and one teacher, Sam, had a management overview of all three subjects, although he was only qualified to teach geography. This was a longstanding arrangement, somewhat unusual at the time of the research given the tendency for subject departments to be run entirely by promoted subject specialists. The subjects were taught in two adjacent classrooms with an adjoining workroom. Despite the administrative structure, the teaching of the constituent subjects was kept largely separate; geography was taught by Sam and history by Frank. These teachers shared the teaching of modern studies, although modern studies enjoyed only 50% of the time allocation given to geography and history over the course of S1 and S2.
According to the headteacher, Hillview School is "a small, community school, the ethos of which is people working together in teams" (Headteacher). This view was largely supported by the teachers interviewed during the research project who commented favorably upon the style of the senior management team. Such a community ethos, combined with the small size of the school, facilitated relationships, and extended to the day-to-day dealings between management and staff. The teachers in the faculty enjoyed a considerable amount of autonomy in their daily work, and it was evident from the data that they were trusted by senior management to get on with their work and achieve good results. Both departmental teachers and the headteacher reported extensive teamwork in the development of courses and resources and in the undertaking of administrative tasks.
The backgrounds and beliefs of the departmental teachers are worthy of consideration as these are significant in shaping their responses to pressures for change. Sam saw his work very much in terms of teaching geography, being particularly passionate about keeping the three social subjects separate. His colleague, Frank, was originally a teacher of modern studies, who primarily taught history after a reorganization. Again, and perhaps surprisingly given Frank's professional biography, there was a strongly espoused opposition to integrating the teaching of the social subjects; this case illustrates a teacher trained in one social subject, but primarily engaged in the teaching of a second for which he had less formal training and at the time of reorganization, less experience. Frank, likewise, grounded his opposition in the integrity of the subject specialism taught by specialist teachers. While confidence to teach unfamiliar subject matter was an issue for both teachers, they believed that fundamental, even essential differences exist between the subjects justifying their continued separation.
Nevertheless, it is clear from the data that these teachers were not opposed to innovation per se. There was considerable evidence of a thoughtful and pragmatic approach to day-to-day issues, a highly reflective approach to practice, and above all, a caring focus on the learning of the students in their classes. Regarding interdisciplinary approaches to teaching the social subjects, the picture is not as clear cut as some of the teachers' statements might imply. Administratively, there was a good deal of collaboration between the teachers especially in planning courses to avoid duplication.
In summary, teaching in this faculty was varied and focused on learning and on the needs of pupils. It was rooted in the demands of the subject up to a point, but interviews revealed a situation that was more complex than it seemed at first glance, and there was little evidence of the fragmented approach often associated with separate provision.

Riverside High School
Riverside High School is a medium-sized, denominational school serving a large suburban and rural hinterland and mixed in terms of socioeconomic status. Attainment was roughly in line with national and EA averages at the time of the research. The school contains around 1,100 pupils, and provision covers the full secondary range from S1-S6.
Prior to the issue of revised 5-14 guidelines for the social subjects (Learning and Teaching Scotland, 2000), social subjects provision had been mainly organized across S1 and S2 via the mechanism of separate subjects running in rotation. The exception to this general situation was a first year pupil induction unit for the three social subjects utilizing an interdisciplinary approach based around the organizing theme of Europe. This managementled innovation predated the revised 5-14 curriculum, being originally conceived as a phased transition from the primary to secondary stage within this curricular area. Geography, history, and modern studies had traditionally been associated with one another through being situated geographically in the same area of the school with a common staff base; the integrated first year unit further cemented these links. However, the departments retained strong individual identities prior to 2000.
Following the revised guidelines and the McCrone Agreement, the departments were brought together more formally through the institution of a new humanities faculty structure. In S1, the school developed a form of provision that was termed "integrated" by the teachers. This consisted of the existing interdisciplinary European unit followed by a modular, multidisciplinary approach (one teacher, three subjects) for the remainder of the first year. In S2, the social subjects continued to be taught separately in rotation by specialist staff. While some of the staff included experienced teachers with a long history in the school and middle management experience, most of the teachers were recent external appointees, often newly qualified. Despite an enmity on the behalf of some teachers towards senior managers and their management practices, the teachers shared a view of a faculty characterized by good social relations. However, in professional terms, the situation was more fragmented. Curricular issues still tended to be discussed in the former subject departments despite the fact that this form of organization had been supplanted by the faculty. A major theme emerging from the data was an apparent balkanization (Hargreaves, 1994) of the school in departmental terms. Comments from the teachers pointed to an organizational fragmentation within the school, a lack of familiarity with the workings of other departments, and a general sense of isolation.  Angela was perhaps typical of this phenomenon. On the subject of the multidisciplinary provision adopted at Riverside, she was unequivocal in her view that specialists should teach specialist subjects and that they should not teach outside their specialism. Significantly, such attitudes were also expressed by Jim, the faculty leader, who was one of a series of appointments specifically made to encourage "very whole school" thinking (Steve, Deputy Headteacher). Jim showed a distinct lack of enthusiasm for teaching outside the boundaries of his own subject, going beyond the negative rationale of lack of interest and enthusiasm, positing essential differences between the three subjects.
Another teacher, Sue, justified her adherence to her subject in terms of the need to maintain academic standards. Sue indicated that an integrated approach was adequate for first year pupils. However, her view was that it was not satisfactory for more advanced study as she saw expertise as being contained within the boundaries of the subject in question. Thus, Sue was quite happy about the prospect of teaching unfamiliar history syllabi at an advanced level because she saw herself as a historian, but was considerably more reticent about tackling equally unfamiliar, but less advanced geography content.
Because the majority of teachers did not really believe in the approach adopted even where they were prepared to go along with the innovation, planning and course development were thus given low priority. In this faculty, there was little evidence of an active grass roots approach to foster such provision; despite the clearly cordial social relations between the majority of 2 3 the staff, program planning did not readily make such links, and connectedness seemed to largely depend upon the member of staff teaching any particular class.

Analysis
This final section of the paper provides an analysis of the data using the M/M conceptual framework. This analysis applies the categories of culture, structure, and agency in turn, before drawing together the strands in conclusion.

I first examine how consistencies and conflicts between existing
cultures and new cultural forms play out through social interaction amongst teachers and managers in the case study schools. The data powerfully suggest that the subject was seen as being the raison d'être of teaching by the majority of teachers in both schools. This aspect of teacher beliefs seems to be the most overtly powerful factor in affecting the form that practice took and the teachers' responses to externally initiated reform. Subjects were often seen in essentialist and content-driven terms; indeed, there was little evidence of awareness amongst teachers that school subjects may be viewed as socially constructed entities (Beane, 1997) rather than as absolute categories.
The formative role of academic and professional training was mentioned by several of the teachers at both schools. The reinforcing effect of professional dialogue is also significant. At Riverside, as has been noted, such dialogue tended to occur between subject specialists, rather than 2 4 occurring more widely within the faculty. Thus, the subject both provides a frame for professional dialogue (and indeed for the establishment of structure to maintain this) and a collective tradition within which teachers define themselves.
This is not simply a cognitive affair. Balkin (1998) suggests that affective processes play a major role in the acceptance of new ideas. In the case of schools, these include issues of emotional attachment to the subject (e.g., Hansen & Olson, 1996;Siskin, 1994) and fear of failure (linked to professional confidence) in moving beyond the familiar (e.g., Lang et al., 1999). This was certainly apparent in both schools where issues of confidence and motivation were evident.

Conversely, Wubbells and
Poppleton (1999) found that enthusiasm for change is likely to contribute to the success of an initiative and that this is more likely when teachers identify strongly with the change, perhaps having played a role in its inception. This was especially the case for Jock at Riverside High School; his positive prior experience of integrated humanities led to his support for the management initiative even where he strongly disapproved of the ways in which it was enacted.
A major difficulty concerns shared understanding of the concepts. The term, integration, is one that is subject to multiple meanings, and teachers' often incomplete and/or misconstrued notions of integration inevitably affected their responses to it. The research suggests two aspects to this. The first factor is the tendency to see integration as an organizational problematic; thus, curricular fragmentation is dealt with via the establishment of 2 5 organizational frameworks rather than as a pedagogical issue to be addressed through social and classroom practice. For example, at Riverside, the senior management team seemed to genuinely believe that putting into place an overarching inter/multidisciplinary structure would ensure that integration occurred. However, my data suggest otherwise. At Hillview, the teachers expressed open hostility to any form of interdisciplinary organization, even though close examination of teaching suggested that, despite this, pedagogic integration was an important strand of practice across the faculty through shared approaches (Fogarty, 1991). In both cases, the term, integration, was explicitly associated with putting one teacher in charge of three subjects rather than the sort of pedagogic link-making that was implied within 5-14.
This process provides a good example of iterative refraction (Supovitz, 2008). For instance, at Riverside High School, the notion of integration was mediated via social interaction at two levels. First, the notion of integrated teaching was watered down at the level of senior management with the potentially controversial interdisciplinary variety becoming the less contentious multidisciplinary version, more in tune with prevailing notions of subjectcenteredness (with the exception of the year one European Union unit, which was viewed in terms of transition from primary school). The second level was its operationalization at the faculty level. Such an approach focused on establishing externally visible frameworks rather than addressing the underlying practices that support and maintain those frameworks.
Consequently, there was little evidence amongst teachers of serious engagement with the notion of integration which did not seriously contradict was not explicitly recognized as such. In both schools, the new cultural form was "lethally mutated" (Spillane, 2002, p. 378) as it was filtered through existing beliefs and values either at the level of senior management or in its enactment by classroom practitioners. I conclude here that the key issue is teacher identity rooted in the subject. Integration creates uncertainty and tension amongst teachers, foregrounding "notions of otherness" and heralding an "invasion of others' territories" (Hansen & Olson, 1996, p. 676).

Horizontal relationships
The small size of Hillview School exerted an effect on the organizational structures of the department. There is no economy of scale in a school of this size; therefore, teachers tend to take on multiple roles. Thus, the school was unusual in establishing a social subjects' faculty pre-McCrone with a single principal teacher or faculty head in charge of all three subjects.
The small size of the department also meant that two teachers taught three subjects between them; both had a specialist main subject and then shared the teaching of modern studies between them. This had the effect of bringing the two teachers together professionally for at least the third subject. The geography of the school was also significant in this respect. The location of the department (two adjoining classrooms with a connecting workroom) had the effect of bringing the teachers together, at least in social terms, and potentially facilitated closer working in terms of provision and pedagogy; as Cuban (1984) reminds us, the form that these take is often a practical response to the spatial characteristics of the school.
The boundaries between the social subjects departments had been further eroded through the redeployment of staff in the early 1990s that had led Frank (originally a modern studies teacher) to become the sole teacher of 2 8 history, and the appointment of a coordinator of all three subjects. Such trends were further reinforced by close working, joint planning, and the requirement for both teachers to teach a third subject, modern studies. It is clear that, while the subject disciplines of history and geography continued to form a major part of the identity of the two teachers, this identity was perhaps not as clearly rooted here in terms of the subject as it might be elsewhere, and dialogue necessarily took on a degree of interdisciplinarity.
The situation was considerably different at Riverside, partly as a result of scale and the organizational and relational complexity that this engendered.
Differences included a lower incidence of professional dialogue between colleagues from different subjects (despite the new need for people to discuss teaching beyond the boundaries of their own subjects), the changed dynamics caused by the influx of new (especially younger) staff, and the balkanization of departments within the school and apparent lack of opportunities for whole staff dialogue except when mediated by the senior management through formal structures (e.g., committees).
Substantive professional dialogue about teaching (pedagogy and provision) tended to focus on the subject taking place within the subject subsets. In this large faculty, the situation mirrored Siskin's (1994) findings with the existence of strong professional subcultures based on the teaching of subjects. For example, in geography, professional dialogue about teaching tended to coalesce around Martin, the former (and still unofficial) principal teacher. Where intra-faculty professional dialogue existed, it tended to revolve around themes such as pupil behavior rather than substantive issues of 2 9 pedagogy and provision. In this school, the structures associated with subjects provided clear boundaries that the faculty lacked. Similarly, subjects provided role models for younger staff -Martin in the case of geography, and Jim in history and modern studies -despite the senior management's attempts to promote integrated structures and in spite of Jim's official role as the leader of geography. The structures that emerged from these roles and relationships, along with subject identity, constituted the departmental immune system that acted as filters to maintain existing practice and resist the influx of invading cultural forms.

Vertical relationships
Moving beyond the faculty, one can see that the formal organization of the two schools at the time of the research was fairly similar. Both had a hierarchical organizational structure, with faculty organization nominally bringing subject departments together as a greater whole. However, the superficial similarities in terms of organizational structure belied significant differences between the two schools. At Hillview, relationships between staff at various tiers of this organizational structure were informal and largely cordial. There was a good deal of vertical interaction in the school, and a mutual view of good relations between staff and senior management. This was combined with a "hands off" approach to management in respect of the social subjects.
However, this is only half of the story. The exercise of power (or a lack of it) can enable or constrain social practice. In the case of Hillview, there was some evidence that a lack of central impetus limited the potential for 3 0 engagement with new ideas. In the case of the teachers in the Hillview social subjects faculty, one can see the positive effects of autonomy (creative mediation of innovation, for instance) combined with more negative effects (a lack of serious engagement with cultural alternatives). The data suggest that the rejection of many of the more formal aspects of integration was at least partially rooted in such a lack of engagement with, and understanding of, the concept of inter-disciplinarity. The hands-off approach of the senior management in the case of these particular teachers, especially its apparent lack of a major role as a mediator of curriculum policy at a departmental level, may be a reason for this phenomenon.
At Riverside, social practice was more overtly influenced by senior management direction, sometimes in a negative or muscular fashion (Smyth & Shacklock, 1998). The move to a faculty organizational model ostensibly simplified the hierarchy within the school. However, the situation was more complicated in the case of the social subjects due to the continued role of the former principal teacher of geography and the lack of interest of the new faculty head in this role. It is probably more accurate to describe the reorganization as narrowing the gap between middle and senior management and distancing the role from classroom practice and the subject; Jim's contact time was reduced heavily compared with the former principal teacher role, and his management responsibility increased to cover several subjects.
The relationship between senior management and staff was a key issue in influencing the development of practice in the faculty. The data suggest that different groups within the school were treated differently depending on the degree to which they supported the management agenda and whether their "faces fit." Arguably, such a management style acted as an impediment to fundamental change, encouraging teacher compliance and risk avoidance strategies rather than genuine engagement with innovation (Blase, 1998;Doyle & Ponder, 1977).

The wider school
Another area of significance concerns the relationships amongst staff across the wider school. The interview and observation data reveal the multidimensional nature of staff relationships at Hillview, where staff across the school had regular opportunities to come together, both socially and professionally. The structural characteristics of the school provided an environment where cross-school interaction was facilitated. The situation at Riverside was very different, typical of that described by Bernstein (1975) for schools organized around a collection code conception of curricular provision.
In line with Bernstein's observations, managers at Riverside tended to have both horizontal and vertical relationships. On the other hand, more junior teachers were more likely to have vertical relationships confined to the school hierarchy, horizontal relationships being largely restricted to their subject or faculty groupings. Where inter-faculty professional dialogue was evident, membership of groups was allocated by senior managers and agendas seemed to be fairly closely circumscribed. Bernstein states that this type of organization can encourage gossip, conspiracy theories, and distrust, issues that were certainly in evidence at Riverside High School.

Agency: Human activity in its social contexts 2
The next section of the paper examines agency, the third dimension of the triad. Agency can be defined in terms of the capacity of individuals to act reflexively within the possibilities bounded by their social and material environments to effect changes to their conditions or to reproduce them. In such a view, agency is past oriented (Emirbayer & Mische, 1998) in terms of the cultural software (Balkin, 1998) that individuals acquire from past experience, but projected to the future and rooted in the possibilities of the present.
I will focus here briefly on how agency is enhanced and/or constrained by its social context of cultural and structures, showing how these impact on individuals to form cultural software or a mosaic of personal efficacy and selfconfidence (Lang et al.,1999), "expertise and practical knowledge" (Wallace & Kang, 2004, p. 936), teacher beliefs (Wallace & Kang, 2004), and will or motivation (Spillane, 1999). This past orientation, in tandem with the material and structural possibilities of the present and the reflexive ability of humans to plan for the future, provides the opportunity or space for maneuver that makes agency possible. Agency can be negative, something that can destroy or distort a worthwhile enterprise (Osborn et al., 1997). However, agency can also be a desirable attribute leading to protective or creative mediation of innovation (Osborn et al., 1997), local adaptation of change (Cuban, 1998), and social benefit. Positive agency can also be about opposing change, especially when such change can be construed as ill-conceived and/or harmful. Agency is, thus, also about reflexive and purposeful human activity that leads to the reproduction of social structures or cultural forms.

3
The case study data provide many examples of the exercise of agency. These

Opposition to school policy
A second example of agency in opposition is provided by the case of Jock at Riverside High School. In this case, his exercise of agency had detrimental effects on his career, leading to him being overlooked for a promotion and becoming relatively marginalized in the school decision-making processes. Jock was a reflective teacher and a reflexive individual. He was 3 5 able to stand back from his job and to acknowledge his own strong views on provision while seeing the value of alternative approaches. He was also able to critique the school's approach to a policy which he broadly supported.
Jock's cultural software was influenced by his background in another education system and by his varied experience of teaching and management.
He espoused strong values in respect of education and was a proponent of a pupil-centered approach to learning in contrast with many of his colleagues.  (Doyle & Ponder, 1977). Jock acted in line with his deeply held values, when he challenged senior managers. This is an example of culture and structure being in contradiction. In such a scenario, agency resolves this tension; this is not just a case of someone blindly operating in response to societal pressures but genuine agency is exercised as a reflexive response to contradictions within the individual's immediate social world. In 6 this case, space for maneuver was limited but this teacher was still able to exercise agency as an internal critic of school policy and practice and through his own teaching approaches (despite pressure from the school's management).
The two examples illustrate how agency is tailored by societal pressures but can also act in opposition to those pressures. As Archer's (1995) social model would predict, agency is affected but not determined by the pressures exerted by social structures and cultural forms. Human reflexivity comes into play to exercise choices from a repertoire of possible decisions. Indeed, such agency is necessary given the contradictions that are faced in daily life.

Conclusions: From social interaction to structural and cultural elaboration
The discussion in this paper illustrates how the interplay of culture, structure, and agency through social interaction leads to social reproduction and transformation or, in Archer's (1995) terms, cultural and structural elaboration. This analysis of the interplay of culture, social structure, and agency casts considerable light on the processes of curriculum construction and educational change at the various strata of the management hierarchies in the two schools. The analysis of the data suggests that while there are significant contextual differences between the two schools, there are also substantial overlaps and similarities. These tend to lie at a generic level, suggesting that while a definitive blueprint for the management of externally initiated change is not a possibility (given the complexity of contextual factors), 7 there is a possibility of developing a generic model for guiding those who seek to enact such change locally. I suggest that M/M provides a good basistool making tools (Balkin, 1998)for the development of such a model as it offers both a conceptual frame and a methodology for research.
At Hillview, the influx of ideas about integration did not change any of the structural conditions within which teaching is organized. Formal systems, such as the faculty organization and the various roles within the school had been established as a result of the contingencies of managing a small school rather than in response to changed policy imperatives and/or new ideas or policies about provision. Structural reproduction occurred, and there is no evidence that such ideas impacted on this process. Nevertheless, there is some evidence that they have had a greater influence in terms of cultural elaboration. Integration has been considered by the teachers, and while one can argue that this engagement is based upon a simplified, incomplete, or erroneous understanding of this cultural form, it is clear that the new ideas have been largely rejected. However, regular social interaction between the teachers in the faculty has led to a limited form of integration which can be termed shared (Fogarty, 1991). Here, we see some cultural elaboration as a result of the active engagement with a set of ideas by interested and reflexive practitioners. The role of social structures was important here. The lack of economy of scale in a small school meant that standard departmental organization had been supplanted by a faculty system that brought together teachers from different subjects to discuss pedagogy and provision on a regular basis. The existence of this system, combined with the emergent properties of the relationship between the two teachers, led to high quality This research suggests that strong leadership is important in the promotion and sustaining of change, being a source of both impetus and support. At Riverside, managers provided impetus, but support was more sporadic. At Hillview, support was evident, but impetus less so. This contributed, to some extent, to regular and proactive, but unsustained and unembedded change at Riverside (with a focus on systems), and slow reactive but sustainable change at Hillview (with a focus on practice).
This combines with the existence of participation, professional trust, and autonomy. At Hillview, where all these features were clearly evident, teachers engaged clearly with changes, which were in effect their changes being conceived, internalized, and practiced by them. At Riverside, where there were lower levels of all three factors, change tended to be accompanied by strategic compliance and was so weakly internalized by some teachers that they were able to undertake largely incompatible practices with little apparent cognitive dissonance (for example the pragmatic partial continuation of the former departmental structure alongside new faculty organization).
This brings us back to the discussion of agency. I suggest that meaningful engagement with innovation will be facilitated if those charged with enacting new policy are able to enhance the agency of the change agents, namely senior managers and teachers in schools and local policymakers in education authorities. In the case of managers, such agency is about knowing when not to act as well as being decisive in action. It is a question of exercising agency at this level that does not deny or repress the agency of classroom teachers as their agency is key to the success of innovation, especially in terms of its enactment in practice. It is a case of accepting that the exercise of teacher agency may result in change outcomes 4 0 that depart from the original plan, and that this is acceptable as long as any such decisions have been well informed and thoughtfully taken.

Acknowledgement
I wish to offer my thanks to Professor Julie Allan for her continued interest in this work and for her valuable insights about the paper.