Teacher leadership : a survey analysis of KwaZulu-Natal teachers ’ perceptions

The notion of teacher leadership is implicit in official documentation in the South African education system post 1994, which emphasises a move towards a more shared and participatory approach to the practice of leadership and management in schools. The concept of teacher leadership is embedded in a distributed leadership theoretical framing which emphasises that leadership need not be located only in the position of the principal but can be stretched over a range of people who work at different levels in a school. We report on a study in which the perceptions of teachers’ on their understanding and experiences of teacher leadership were explored. The study adopted a survey approach and utilised closed questionnaires to gather data from 1,055 post level-one teachers across a range of schools of diverse contexts in KwaZulu-Natal. We found that while teachers supported the notion of shared leadership and believed they were equipped to lead, their leadership was largely restricted to their classrooms. There was some evidence of teacher leadership amongst teacher colleagues in certain curricular and extra-curricular activities. However, teacher leadership in relation to school-wide and community issues was almost non-existent. We signal two problematics regarding the leadership of school teachers and consider the implication of these for the distribution of leadership, and therefore change, in schools.


Introduction
A major step forward in the South African education system post 1994 has been its move, at a policy level, towards more participation and collaboration in the practice of school leadership and management (see for example the South African Schools' Act, 1996 and the Task Team Report on Education Management Development, 1996).Thus the field of Education Leadership and Management in South Africa, determined by the Department of Education, stresses "participative, 'democratic' management, collegiality, collaboration, schools as open systems and learning organisations, and, importantly, sitebased management" ( Van der Mescht, 2008:14).At the heart of this democratising process, is the decentralisation of decision-making in schools (Department of Education, 1996).This policy shift towards a more democratic and participatory decision-making process in schools offers the possibility of and opens up the space for the emergence of teacher leadership.
In its simplest form, teacher leadership can be described as a model of leadership in which teaching staff at various levels within the organisation have the opportunity to lead (Harris & Lambert, 2003).The main idea underpinning this view is that leadership is not individual or positional but instead is a group process in which a range of people can participate.Teacher leadership has as its core "a focus on improving learning and is a model of leadership premised on the principles of professional collaboration, development and growth" (Harris & Lambert, 2003:43).Thus central to the concept of teacher leadership is the notion of the expert teacher who continually works to elevate her own teaching.This view is endorsed by Zimpher who describes how teacher leadership is "an outgrowth of expert practice and expert knowledge" (1988:54).As a result of this expertise, the teacher develops the confidence to lead colleagues as she attempts new initiatives in the pursuit of excellence in the practice of teaching and learning.Thus it can be seen that the role of mentor (Anderson & Lucasse Shannon, 1988;Gehrke, 1988) is a core function of a teacher leader.Further roles include but are not limited to reflective practitioner (Day & Harris, 2002), coach (Joyce & Showers, 1982), professional developer (Zimpher, 1988), action researcher (Ash & Persall, 2000) and decision-maker (Griffin, 1995;Muijs & Harris, 2003).
The literature on teacher leadership, particularly within a democratic distributed leadership framing (Gunter, 2005), suggests that it can be used as a strategy to democratise schools.In line with this thinking, Hart (1995) suggests that teacher leadership is often promoted to nurture a more democratic, communal or communitarian social system for schools and schooling.This is because, as Harris and Muijs argue, teacher leadership is primarily concerned with "enhanced leadership roles and decision-making powers to teachers " (2005:16).In relation to the leadership of teachers and their participation in decision-making, Barth comments on research which suggests that "the greater the participation in decision-making, the greater the productivity, job satisfaction, and organisational commitment " (1990:130).Against this backdrop, the study reported on in this article was designed as a large scale survey which aimed to investigate teachers' perceptions and experiences of teacher leadership within the context of their schools.

Literature review
This literature review endeavours to obtain clarity on the concepts of leadership and management and their relationship to each other.It does this in order to underscore some of the central tenets of teacher leadership and, in so doing, develops distributed leadership as the theoretical framing of the study.

Differentiating education leadership and education management
In her book, The good high school, Sara Lawrence Lightfoot maintains that the literature tends to agree that "an essential ingredient of good schools is strong, consistent and inspired leadership " (1983:323).Here leadership is understood as the process which brings about change in the organisation and which "mobilizes members to think, believe, and behave in a manner that satisfies emerging organisational needs, not simply their individual needs or wants or the status quo" (Donaldson, 2006:7).In other words, Donaldson continues, "leadership helps the school adapt to its changing function in society " (2006:8).This central focus of change in the practice of leadership stands in contrast to the practice of management which, according to much of the literature, presents the purpose of management as being to ensure the stability, preservation and maintenance of the organisation (Astin & Astin, 2000).While we fully support the view of Spillane, Halverson and Diamond that leadership is "critical to innovation in schools" (2004:1), we believe further that the two processes of leadership and management complement each other and both are needed for an organisation to prosper (Kotter, 1990).
In addition, we argue that one cannot talk about education leadership without talking about issues of power.Power is central to leadership and becomes visible in "the way people are positioned in schools, where people are positioned and who does the positioning" (Grant, 2010:57).This positioning tells us much about the distribution, or otherwise, of power and authority in schools.In other words, as Gunter so aptly puts it, "educational leadership meets the issue of power head on" (2005:45).It is to a discussion on the distribution of leadership that we now turn.

Distributed leadership
Traditionally, research on education leadership has been premised on a singular view of leadership and has focused on the difference principals make to schools and, in so doing, has reinforced the assumption that "school leadership is synonymous with the principal" (Spillane et al., 2004:4).In direct contrast, more recent theories work from the premise that leadership need not be located only in the principal of a school but should be "stretched over multiple leaders" (Spillane, 2006:15), including teachers.In line with this thinking, Harris and Muijs contend that "both senior managers and teachers (our emphasis) have to function as leaders and decision makers and try to bring about fundamental changes " (2005:133).
One such alternative theory has been referred to as distributed leadership which, at its best, can be conceptualised as "an emergent property of a group or network of individuals in which group members pool their expertise" (Gronn, 2000:324).It requires the "re-distribution of power" (Grant, 2010:57) and "the capacity to relinquish, so that the latent, creative powers of teachers can be released" (Barth, 1988:640).For Spillane, the distributed leadership perspective foregrounds leadership practice which is "constructed in the interactions between leaders, followers and their situations " (2006:26).Shared decision-making within this distributed practice requires a school context and culture which is supportive, collaborative (Lieberman, Saxl & Miles, 1988) and collegial (Muijs & Harris, 2003) and which opens up a safe space for life-long learning and continuing professional development (Katzenmeyer & Moller, 2001).This necessitates a range of leaders, whether teachers or SMT members, who are emotionally intelligent (Goleman, Boyatzis & McKee, 2002;Graven, 2004), purposeful (Donaldson, 2006), courageous (Grant, 2006), and who are willing to take considered risks (Lieberman, Saxl & Miles, 1988) in the quest for school improvement.
The principal as leader of leaders At this juncture, it must be noted that inviting teachers into the practice of leadership does not make the role of the school principal redundant.On the contrary, the principal's role remains central to the practice of leadership and becomes one of holding "the pieces of the organisation together in a productive relationship" (Harris & Muijs, 2005:28).However, it is important to remember that school leadership and management takes place within the legal framework of South Africa's education system and, within this framework, school principals are ultimately accountable because they possess statutory delegated authority (Mbatha, Grobler & Loock, 2006).By virtue of their positional power, principals cannot abdicate their accountability but must instead become the 'leader of leaders' (Ash & Persall, 2000;Harris & Lambert, 2003) in their schools.As 'leader of leaders', it is their responsibility to build a school culture premised on trust (Lieberman, Saxl & Miles, 1988;Grant, 2006) and mutual learning which will facilitate the distribution of leadership.
However, as MacBeath (2005) warns, building a culture of trust in the face of accountability pressures implies a risk.The risk is that trust may be placed in people who do not honour that trust and, in so doing, the principal is left accountable for the task poorly done.But the alternative of working in a culture of mistrust is equally unappealing because "without mutual trust, relationships and respect are compromised and mistrust exerts a corrosive influence" (MacBeath, 2005:353).Thus the development of mutual trust is a non-negotiable in the practice of leadership and the challenge for school leaders, whether they are teachers or SMT members, is to find areas of expertise in colleagues because trust and respect are earned through expertise.The problematic for school principals is this: How does the principal balance the important issue of developing trust with the equally important issue of accountability?In response to this problematic, Van der Mescht and Tyala suggest that the principal's role becomes "a balancing act " (2008:227) where the principal is challenged to determine, in line with legislation, what practices can be distributed and how the distribution will happen.This is dependent, to a large extent, on the unique culture and context of each school and we turn now to this discussion.

The importance of context
Researchers tend to agree that school context is central to an understanding of teacher leadership (see for example Katzenmeyer & Moller, 2001;Harris, 2004;Spillane et al., 2004).This is because, as Spillane et al. argue: "leadership practice cannot be extracted from its socio-cultural context " (2004:22).In other words, teacher leadership is likely to vary depending on the his-torical, cultural and institutional settings in which it is situated.This is particularly pertinent in South Africa where the apartheid legacy with its dysfunctional schooling system has "not simply disappeared with the replacement of the apartheid government with a new government" (Christie, 1998:284).Instead, this inheritance "continues to be determinative in shaping and accounting for the character of current social behaviour in the country, including the performance of children in schools" (Soudien, 2007:183).It is imperative therefore, as Grant argues, that "any understanding of teacher leadership from a South African perspective be expansive and sufficiently flexible to accommodate these vastly differing school contexts" (2010:243).In so doing, we need to heed the warning of Smylie that "it may be difficult to develop teacher leadership to its full potential without also developing its contexts" (1995:6).
In this regard, South African research findings into teacher leadership are varied as determined by the culture and context of the schools in which the research was done.For example, Singh's (2007) study found that although there was evidence of teacher leadership in her two case study schools, access to leadership opportunities was based on the seniority, expertise and teaching experience of teachers and determined by the SMT.Ntuzela's qualitative study uncovered that in his two case study schools, the enactment of teacher leadership differed considerably and he concludes that "the different contexts of the two schools resulted in their different understandings of teacher leadership" (2008:83).In the historically disadvantaged school, a nonparticipatory culture "prevented teachers from taking initiative as leaders in their school and this led to teacher demotivation" (Ntuzela, 2008:78).In direct contrast, a culture of collegiality and shared decision-making was the hallmark of the historically advantaged school, resulting in the enactment of teacher leadership within and extended beyond the classroom within what Gunter (2005) refers to as a dispersed distributed leadership characterisation.In her auto-ethnographic study, Pillay (2008) found that the role of the principal was crucial in developing a school culture conducive to the practice of distributed and teacher leadership and she argues that "an environment must be created where teachers are nurtured and developed so that they will be able to meet the challenges of an ever changing educational system and keep abreast with the changes" (Pillay, 2008:122).
In concluding this section, this article works from the premise that South African schools require the characteristics of flexibility and adaptability as they respond to the ever changing South African educational system as well as their own local needs.We argue that teacher leadership, as but one manifestation of a distributed approach to leadership can be a powerful tool in bringing about school change.Teacher leaders are agents of change and this agency should be nurtured and tapped so that teachers learn to lead new initiatives and challenge the existing status quo in schools in the pursuit of excellence in teaching and learning.To do this, teachers require support from the principal as 'leader of leaders' and through continuing professional development initiatives, both inside and outside the school.

Research design and methodology Aim and research questions
In this study we set out to explore teachers' understanding and experiences of teacher leadership.We report on findings of a large scale survey in conveniently selected primary and secondary schools in KwaZulu-Natal.The following specific research questions underpinned the study: 1.What are teachers' perceptions about leadership in schools? 2. To what extent is teacher leadership happening in schools and what roles do teachers take up? 3. What are teachers' perceptions of the leadership context and culture in their schools?

Research orientation and design
This study was underpinned by an interpretive research orientation.We were interested in "understanding the meaning people have constructed, that is, how they make sense of their world, and the experiences they have in the world" (Merriam, 1998:6).Specifically we were interested to describe and understand the meaning South African teachers gave to the concept of teacher leader and to explore the practices they associated with teacher leadership in their schools.However, we were aware that a range of qualitative case studies were already completed on the topic (see for example Grant, 2006;Rajagopaul, 2007;Singh, 2007;Khumalo, 2008;Ntuzela, 2008) and, while these yielded rich descriptions, they were limited in terms of breadth.We elected therefore to introduce a quantitative dimension to the existing research on teacher leadership.
The study was designed as a large-scale survey and called for numerical data and descriptive statistics to ascertain teachers' understanding and experiences of teacher leadership in the sample schools.Its attractiveness was in its "ability to make statements which are supported by large data banks and its ability to establish the degree of confidence which can be placed in a set of findings" (Cohen, Manion & Morrison, 2007:207).To this end, questionnaires were employed because of their ability to gather information from a large population in one or several locations using pencil and paper without necessarily making personal contact with the respondents (Bless & Achola, 1990).Furthermore, questionnaires lend themselves to logical and organised data entry and analysis (Anderson & Arsenault, 1998).Thus a self-administered, structured, closed questionnaire was deemed an appropriate instrument to gather data in this study.

Research schools and participants
This study included a total of 81 schools across three districts in KwaZulu-Natal, selected because of their accessibility to the team of researchers.The majority of schools (70%) were located in the Umgungundlovu District, 24% were situated in the Umzinyathi District while a mere 6% could be found in the far northern district of Umkhanyakude.To give a sense of the socio-economic status of these schools, 37% of the sample charged schools fees of between R1 and R500 per annum while a further 20% charged fees of between R501 and R1000.Of the sample, a total of 18% were no-fee schools.Thus, three quarters of the schools in the study serviced poor communities.In direct contrast, the remaining one quarter of schools charged fees of more than R1000 with 16% of this one quarter charging school fees in excess of R5000.Thus it can be seen that a minority of schools in the study were wellresourced schools servicing the more privileged communities.
Of the 81 schools in the study, 54% were primary schools, 39% were secondary and 7% combined.Of the school sample, 72% had a learner enrolment of more than 600.In other words, they were fairly large schools.In addition, the majority of schools (41%) had between 20-28 educators on the staff while a further 28% had between 29-37 educators.It can be said therefore that the schools, in the main, had a large teaching staff.
The participants in the study included 1,055 post level-one teachers from across the 81 schools, selected purposively and on the basis of their availability and willingness to participate in the research.Thus a limitation of the study was that the sample was not representative of the whole population.However, the size and demographic spread of the data set reduced this limitation to a certain extent.
The majority of the teachers in the study were women (76%) and 83 % of the total were permanently appointed to their teaching posts.The bulk of the teachers (71%) were aged between 31 and 50 years of age.The group consisted mainly of qualified teachers (88%) with 23% holding a three-year teaching qualification, 49% holding a four-year teaching qualification and 15% holding a qualification of five or more year's duration.The teaching experience of the cohort was spread fairly evenly across the data set with the majority (36% ) having 16 or more years experience, 18% having 11-15 years experience, 24% having 6-10 years experience and 21% in their first five years of the profession.
Thus the participants were fairly representative of KwaZulu-Natal teachers and were sufficiently qualified and experienced to reflect on their perceptions and experiences of teacher leadership in the schools in which they taught.

Ethical issues
Permission to undertake the study was granted from the principals of the 81 schools participating in the study and written consent was received from each of the 1,055 teachers who participated in the study.Participants were guaranteed anonymity and were also advised that they were free to withdraw from the study at any time, should the need arise.Ethical clearance to proceed with the research was received from the higher education institution under whose auspices the study was conducted.

The questionnaire
As mentioned earlier, a structured closed questionnaire was adopted as the data collection tool in this study.The questionnaire was divided into three broad sections; biographical information of the teachers in the study (Section A), school information (Section B) as well as key information on teacher leadership (Section C).Section C was further divided into four sections: C1 was organised as a response to Research Question 1, C2 and C3 were organised as responses to Research Question 2 while C3 was organised as a response to Research Question 3. Sections A and B each consisted of five questions while Section C consisted of 46 questions (4, 16, 16 and 10, respectively).Responses to sections A and B of the questionnaire were gathered using nominal scales while Section C of the questionnaire adopted a five-point Likert rating scale to capture the data.The questionnaire fairly and comprehensively covered the domains of the three research questions in an effort towards content validity whilst ensuring that it remained contained to reduce the possibility of respondent fatigue.The questionnaire was designed with some questions repeated in different sections in different ways to determine if they would "yield the same result each time" (Babbie & Mouton, 1998:119) in an effort towards reliability.The survey questionnaire was originally developed for, piloted and implemented in 2007 (see Khumalo, 2008).Based on the learning and feedback from this 2007 study, it was adapted and used in a pilot in 2008.Following the pilot, it was further refined before implementation in this study, enhancing the construct validity of the instrument.
To avoid possible non-return of questionnaires, researchers communicated regularly with and personally visited the schools to follow up on and collect questionnaires within as short a time-frame as possible.

Data analysis
The data were analysed descriptively using the Statistical Package for Social Sciences (SPSS).The data from Section C2 and C3 were then further analysed using Grant's (2008) model of teacher leadership in relation to zones and roles (Figure 1).The findings were recorded and the data interpreted in the light of the available literature on teacher leadership, in an attempt at literature triangulation in the quest for reliability (Liebenberg & Roos, 2008:585).
A possible limitation to the study was that a team of 17 novice researchers was responsible for inputting the data and errors may have occurred during the process.However, the process was co-ordinated by the project leader and her team of colleagues and the final data set was cleansed and inconsistencies removed.
A further limitation of the study was that, for the majority of the teacher participants, English was a second language.This was particularly evident in section C3 of the questionnaire where the level of comprehension of the questionnaire was weak.To overcome this limitation, the returned questionnaires were checked for accuracy and completion and, following this process, it was agreed that responses to section C3 should be discarded.Leading across neighbouring schools in the community In presenting the findings, reference is made to the four zones and six roles of teacher leadership as described in Grant's (2008) model of teacher leadership and depicted in Figure 1.For ease of reading, we do not reference the model again in this section but simply refer to the various zones or roles as necessary.

Support for the notion of teacher leadership
Responses to the survey questionnaire revealed that the majority of the teachers in the study (71.7%) believed that school teachers were confident and capable of leading (Q12).From the responses, it emerged that only 7.9% of the teachers were of the opinion that it was only the School Management Team (SMT) who should make decisions in the school (Q11).Only 21.5% of the sample were of the view that people in positions of authority should always or often lead (Q13).This perception reinforces the view of the Task Team Report on Education Management Development (Department of Education 1996:27) which states that "management should not be seen as being the task of the few; it should be seen as an activity in which all members of educational organizations engage".At a level of rhetoric then, teachers in the study supported the notion of distributed leadership and believed that teachers could and should lead.
However, while teachers in the study supported the notion of teacher leadership, in practice their experience of leadership was largely restricted to their classrooms as the next section demonstrates.
Leading within the zone of the classroom Data from the survey confirmed the existence of teacher leadership in the zone of the classroom (zone 1 of the model).In this zone, the leadership work of teachers is to continually improve their teaching and learning (role 1) and, in so doing, develop expert power in the classroom.It was evident from the data presented that teachers were engaged in activities that promoted teaching in their classrooms.Three questions on the questionnaire were indicators of teacher leadership in this zone.Firstly, expert knowledge and expert practice are key indicators of leadership in the zone of the classroom and, in the study, 71.5% of the teachers asserted that they regularly (often or always) updated their knowledge on pedagogical developments in their learning area (Q29).Secondly, reflection on classroom practice is a pre-requisite for im-proved teaching and learning and, in the survey, 76.7% of the teachers stated that they often or always critically reflected on their classroom practice (Q16).Finally teacher initiative is a key indicator of teacher leadership.In the survey, 61.3 % of the teachers indicated that they often or always took the initiative without duties being formally assigned to them (Q15).
In summary, the statistics indicate that, according to the teachers in the study, their understanding and experiences of teacher leadership were strong in the zone of the classroom.In contrast, the statistics decreased considerably in relation to indicators of teacher leadership in zone two where teachers worked with other teachers outside the classroom in curricular and extracurricular activities.It is to the presentation of findings in zone two that we now turn.

Leadership of teacher colleagues in curricular and extra-curricular activities
In this zone the teacher leader is likely to be involved in the provision of curriculum knowledge (role 2), managing in-service training and providing assistance to other educators (role 3) and finally, participating in the performance evaluation of other educators (role 4).Five questions on the questionnaire were indicators of teacher leadership in this zone.
A mere 19.2% of the teachers in this study claimed to often or always provide in-service training (role 3) to their colleagues (Q19), whilst 31.2% of the teachers said they sometimes provided in-service training to assist other educators.The data also revealed that 32.2% of the teachers often or always led outside the classroom by providing curriculum development knowledge to their colleagues (role 2, Q20).Despite performance evaluation of peers being an integral aspect of the Integrated Quality Management System (IQMS) in which all South African schools are compelled to engage, only 38.4% of teachers often or always participated in the performance evaluation of their colleagues (role 4, Q22).The two areas in which teachers were most actively engaged in this second zone were related to role 2 and involved the planning of extra-mural activities in their schools (47.7% often or always, Q24) and in the selection of textbooks and instructional materials for their grade or learning area (71.6% often or always, Q23).
These findings point to a restricted form of teacher leadership (Harris & Muijs, 2005) within this zone and emphasise management functions at the expense of leadership processes, as defined in an earlier section of this article.Opportunities for authentic leadership and teacher empowerment through team work, peer support and collaboration in relation to curriculum issues were the exception rather than the norm.This minimalist embracing of a collaborative practice suggests that teacher leadership was understood more as an isolated activity and, as a consequence, its transformatory power was not fully tapped in this second zone.Furthermore, this raises a concern about the possibility of authentic leadership beyond the zone of the classroom because, as Harris and Lambert argue, "collaboration is at the heart of teacher leadership, as it is premised on change that is undertaken collectively" (2003: 44).

Leading outside the classroom in whole schools development
This third zone of the model comprises two roles of a teacher leader, the one involving teacher participation in school level decision making (role 6) while the other involves the teacher in organising and leading reviews of school practice (role 5).Five questions on the questionnaire were indicators of teacher leadership in this zone.
For Muijs and Harris (2003) involvement in decision-making is a key indicator of the strength of teacher leadership.In our study, the data revealed that teachers were seldom fully involved in in-school decision making with only 30.5% responding often or always (Q18).The role which enjoyed the highest level of involvement by teachers within the zone of the whole school (67.3%) related to the setting of standards for pupil behaviour in the school (role 6, Q26).However, only 27.2% of educators often or always organized and led reviews of the school year plan (Q17) whilst a mere 14.1% of teachers often or always set the duty roster for their colleagues (Q30).In other words, teachers in the study were not often involved in school-wide decision-making processes and when teachers were involved, this was usually restricted and took the form described by Harris and Muijs of "individual or collective consultation with the senior management team" (2005:90).
Another finding which demonstrated that teachers were not adequately empowered as leaders in this zone was their failure to engage in designing staff development programmes.The study revealed that a massive 65.6% of even the most seasoned teachers (51+ age group), seldom or never participated in designing staff development programmes for their school (Q27).This finding is illustrated in Table 1.Only 11.8% of all the participants in the study were often or always involved in this role.According to Harris and Muijs (2005:126) one of the key problems in developing teacher leadership is that "staff lack confidence and in some cases leadership skills to perform the roles and responsibilities".In cases like these where "teachers are expected to move into leadership roles, they must be provided with meaningful professional development experiences, in both formal and informal settings" (Katzenmeyer & Moller, 2001:53).

Leading across neighbouring schools in the community
Roles in this fourth zone are associated with, firstly, providing curriculum development knowledge across schools (role 2) and, secondly, leading inservice education and assisting other teachers across schools (role 3).Three questions on the questionnaire were indicators of teacher leadership in this zone.
Only 15.9% of the teachers in the study often or always provided curriculum development knowledge to teachers in other schools (role 2, Q21).The data also revealed that 25.4% of the teachers co-ordinated aspects of extramural activities beyond their school (role 2, Q25).Of the teachers, 22.5% often or always co-ordinated cluster meetings for their learning areas within their districts (Q28).This demonstrated limited involvement of teachers in providing curriculum development knowledge to teachers in other schools (role 2).We can conclude, on the basis of these statistics, that teacher leadership within zone 4 was not a common practice for teachers in our study.This finding concurs with the case study research of Rajagopaul (2007) and the survey research of Khumalo ( 2008) that teacher leadership was not especially evident in zone 4.

Discussion of findings
In summary, it is evident from the data presented so far that glimpses of teacher leadership were apparent across all four zones in the study but the degree of teacher leadership varied dramatically from zone to zone.Irrespective of age, gender and qualification, the majority of teachers in the study supported the notion of teacher leadership and believed that all teachers had the potential to lead.In relation to zone 1, the majority of teachers were of the view that they critically reflected on their teaching with the purpose of continuously improving their classroom practice.In contrast, the take-up of teacher leadership in the other three zones dropped dramatically.In relation to curricular activities in zone 2, teachers were mainly involved in the selection of materials and text books for their grade or learning area.Furthermore, they were involved in extra-curricular activities such as sport.However, approximately half of the teachers seldom or never provided curriculum development knowledge to their colleagues, they did not lead in-service education nor did they participate in peer performance evaluation.Teachers defined themselves as leaders within zone 3 primarily in relation to their participation in schoollevel decision-making on the issue of learner discipline.There was little further evidence of teacher leadership in relation to other school decisionmaking contexts and teachers did not seem to be involved in reviews of school practice.Furthermore, there was little teacher leadership evident in zone 4 beyond some involvement in learning area cluster meetings and involvement in extra-mural activities, indicating a restricted form of teacher leadership (Harris & Muijs, 2005).Given this restricted understanding of teacher leadership by the teachers in the survey, one can speculate on the reasons for the findings.Two thoughts come to mind in this regard.The first is the impact of school context on teacher leadership in these KZN schools and the second involves the role of principal as 'leader of leaders' in the practice of teacher leadership.These two thoughts each constitute a problematic for us and it is to this discussion that we now turn.
The impact of school context on teachers' understanding of teacher leadership: the first problematic The first problematic can be described in the following way.Data from the first sections of the survey confirmed a restricted, as opposed to an emergent or successful, form of teacher leadership as the norm in the study schools with teacher leadership being understood as a fairly isolated activity.In direct contradiction, one of the findings in the final section of the survey indicated that, for 79.9% of the teachers, team work was often or always encouraged in their schools (Q55) and only a mere 1.6% indicated that they had never experienced teamwork.This constituted the first problematic in the study.How can schools operate as teams when indicators of participation and collaboration were nominal in the study?
In response to this question, we work from the premise that context is crucial to the take-up (or not) of teacher leadership in a school.As discussed earlier, three-quarters of the schools in the study were inadequately resourced schools servicing poor communities.Thus, the take-up of teacher leadership in these schools should be understood "against a backdrop of a fledgling democracy emerging from an apartheid history whilst still carrying the legacy of poverty and inequality" (Grant, 2006:522).As a direct consequence of this history of control and inequality, the legacy of patriarchal and hierarchical relations within the education system generally and within schools particularly, still prevails, despite national and local policies to the contrary.Thus, while the teachers in the study referred to team work in their schools, the evidence from earlier data indicated that the key indicator of authentic teamwork (participation in decision-making processes) was minimal in many of the schools.This is perhaps because in some of the study schools "the cultural and structural changes required to support teacher leadership have not been put in place" (Harris & Muijs, 2005:116).While democratic management and governance structures, through legislation, existed in the survey schools, the data suggests that the culture of some of the schools remained largely unchanged.While there was the appearance of team work, in practice leadership and decision-making was not distributed and power remained centralised with the SMT.Thus, the findings of this study confirm the view of Sayed that the goals of democracy have remained largely at the level of rhetoric and ignored the "realities on the ground" (Sayed, 2004:252).
The role of the principal as leader of leaders: the second problematic The role of the SMT in the take-up (or not) of teacher leadership in this study constitutes the second problematic.Data confirmed that, in about half of the survey schools, the SMT was perceived as an impediment to teacher leader-ship because SMT members did not distribute leadership but instead autocratically controlled the leadership practice.Data revealed that 43.9% of teachers indicated that the SMT seldom or never trusted their ability to lead (Q47).For many of the teachers, while acknowledging varying degrees of trust exhibited by the SMT, felt that they were not fully acknowledged as leaders.Some 45.3% of the teachers believed that the SMT seldom or never valued their opinion (Q50).Furthermore, for 44.2% of the teachers, their SMT often or always took the important decisions (Q52).This perceived lack of confidence of the SMT in the ability of teachers to lead was a barrier to teacher leadership for just under half of the teachers in the study.This finding of the SMT as a barrier to teacher leadership in the survey schools confirms both the findings of the original survey study (Khumalo, 2008) and the qualitative studies of Rajagopaul (2007), Singh (2007) and Ntuzela (2008).
To begin to understand why many of the SMTs in the survey were perceived as barriers to teacher leadership, it is important to return to the view of MacBeath ( 2005) who warns of the risk of building a culture of trust in the face of accountability pressures.The data led us to believe that the SMTs felt the full weight of their accountability and, consequently, were unwilling to redistribute power to teachers in case the task was inadequately performed.This signals a form of authorised distributed leadership (Gunter, 2005) where work was distributed only at the discretion of the SMT within a hierarchical system of relations.Here power remained at the organisational level and teachers were not encouraged to take initiative and work in innovative ways but instead were required only to follow directives.This tension of 'holding on' rather than 'letting go' (MacBeath, 2005;Van der Mescht & Tyala, 2008) for fear of losing control goes some way to help us understand why emergent teacher leadership (Gronn, 2000) was not forthcoming in the study.
But how does one go about developing a culture of trust in a school whilst still holding in focus the issue of accountability?MacBeath (2005) suggests that initially one ought to tread cautiously, working within the existing culture and history of the school and only later on should one become more strategic.This strategy includes identifying expertise and then supporting and developing potential teacher leaders through dialogue and staff development with the aim of building both confidence and mastery.This requires 'emotional maturity' on the part of the SMT by which we mean the "insight to know when you do not know, the confidence to admit this, and the ability to access the necessary information (or experience) and support from the broader professional community" (Graven, 2004:207).

Conclusion
Based on the survey data, the study confirmed that teacher leadership was supported at a level of rhetoric, most likely because the concept can be justified because of its 'representational power' (Harris & Spillane, 2008) and its leaning towards democratic ideals in schools.However, in practice the leadership of teachers was mainly restricted to the classroom.This restricted take-up of teacher leadership was attributed to two problematics: the impact of school context and the critical role of the principal as 'leader of leaders'.The prevalence of these problematics in the findings suggest that teacher leadership is not yet institutionalised in the majority of these survey schools and the full potential of teacher leadership as a tool for school change has not been harnessed.
At a methodological level, the findings of this survey study using quantitative methods confirmed the findings of the smaller qualitative studies, regarding the restricted nature of teacher leadership in many KwaZulu-Natal schools.This corroboration affords a level of confidence in the findings of the study and leads us to consider some of the implications for practice.Firstly, we argue for a reinsertion of leadership practices into schools and for vigorous debates about how leadership (in relation to management) should be reconceptualised so that the change agent role of educators comes to the fore.Secondly, we contend that teaching programmes, at the undergraduate and the postgraduate levels, should introduce the concept of teacher leadership where its advantages and disadvantages are debated.Finally, we acknowledge that schools themselves are the best places to learn about teacher leadership and this calls for courageous school leaders, whether SMT members or teachers, who are unafraid to take risks and who use their initiative and work collaboratively to inspire further leadership as people work towards the shared school vision.If school leaders take up this call, then the possibility exists that the existing 'sleeping giant' (Katzenmeyer & Moller, 2001) of teacher leadership will be awakened.
Farhana Kajee is an acting HOD at a high school and has eight years teaching experience.Her research focuses on the role of induction and mentoring of teachers by teacher leaders.
Ronnie Moodley is an HOD at a primary school and has 21 years teaching experience.His research focuses on decision making within the practice of leadership.
Sharila Somaroo is an educator at a secondary school and has 21 years teaching experience.The focus of her research is decision making within the practice of leadership.

Table 1
C ross-tabulation: Age with design of staff developm ent program m es