Fostering collaborative teacher learning: A typology of school leadership

Previous research indicates that supportive school leadership is a key condition of collaborative teacher learning. The purpose of this study was to develop a typology of how school leaders foster collaborative teacher learning. We adopted an integrative perspective on leadership by examining both learning-centred leadership and distributed leadership practices that are supportive of collaborative teacher learning. Data were gathered by means of inter-views with ten school leaders and a questionnaire that was completed by 39 teachers from six secondary schools in the Netherlands. The aim of the interviews was to identify to what extent school leaders applied learning-centred leadership and distributed leadership practices. The questionnaire measured teachers' perceptions of the role of school leaders in teacher learning. As an outcome of this study, we constructed a typology that provides insights into how school leaders foster collaborative teacher learning. Four types of school leaders were distinguished: (a) integrators of teacher learning, (b) facilitators of teacher learning, (c) managers of teacher learning, and (d) managers of

The purpose of this study was to develop a typology of how school leaders foster collaborative teacher learning.
We adopted an integrative perspective on leadership by examining both learning-centred leadership and distributed leadership practices that are supportive of collaborative teacher learning.Data were gathered by means of interviews with ten school leaders and a questionnaire that was completed by 39 teachers from six secondary schools in the Netherlands.The aim of the interviews was to identify to what extent school leaders applied learning-centred leadership and distributed leadership practices.The questionnaire measured teachers' perceptions of the role of school leaders in teacher learning.As an outcome of this study, we constructed a typology that provides insights into how school leaders foster collaborative teacher learning.Four types of school leaders were distinguished: (a) integrators of teacher learning, (b) facilitators of teacher learning, (c) managers of teacher learning, and (d) managers of daily school practice.Our findings suggest that integration of learning-centred leadership and distributed leadership practices can help school leaders to support collaborative teacher learning.

| INTRODUC TI ON
Internationally, policy makers and scholars have increasingly recognized the significance of school leadership for teacher learning (OECD, 2019;Stoll & Kools, 2017;Vangrieken, Dochy, Raes & Kindt, 2015).Collaborative teacher learning, in which research knowledge and teachers' knowledge and experience are connected, is considered to have a particularly positive impact on teacher professional development, school development, and the professional learning climate in schools (Hubers, Schildkamp, Poortman, & Pieters, 2017;Stoll & Kools, 2017).In order to promote teacher learning, school leaders should articulate and communicate a clear vision of teacher learning, and facilitate, monitor and reward it.Yet, research shows, that leadership practices that support teacher learning are scarce (Van Schaik, Volman, Admiraal, & Schenke, 2019).The findings of most studies are based on teacher reports of their experience of school leadership; little is known about what leadership practices school leaders apply to foster collaborative teacher learning.
The focus of this study was on how school leaders foster collaborative teacher learning, particularly in teacher learning groups.This was investigated with a mixed-method study in ten secondary schools in the Netherlands.Two contemporary perspectives on school leadership were used as a framework for analysing leadership practices: (a) learning-centred leadership (Hallinger, Piyaman, & Viseshsiri, 2017;Liu, Hallinger, & Feng, 2016), and (b) distributed leadership (Bouwmans et al., 2019;Gronn, 2002;Harris, 2007;Heck & Hallinger, 2009;Spillane, 2006).Although both are assumed to stimulate collaborative teacher learning (Admiraal et al., 2016) they represent differing perspectives on leadership: learning-centred leadership focuses on formally assigned school leadership, whereas distributed leadership focuses on informal school leadership.Gronn (2009) introduced the concept of hybrid leadership to indicate the role formal leaders can play in distributing leadership.However, although formal leaders' influence on informal leadership is often suggested, little is known about which leadership practices are used in a hybrid leadership style (Crawford, 2012;Tian, Risku, & Collin, 2016).The current study comprises an examination of leadership practices that are supportive of collaborative teacher learning from an integrative perspective.We use the term "integrative leadership perspective" to denote the combination of a focus on learning-centred leadership practices and distributed leadership practices.An example of a learning-centred leadership practice is managing a learning program (Hallinger et al., 2017;Liu et al., 2016).An example of a distributed leadership practice is opening the boundaries of leadership (Bouwmans et al., 2019;Gronn, 2002;Harris, 2007;Spillane, 2006).

| THE ROLE OF LE ADER S HIP IN TE ACHER LE ARNING
Across Europe, we see an emphasis on collaborative and continuous teacher learning.The professional development of teachers is increasingly taking place in teacher learning groups (Stoll, Bolam, McMahon, Wallace, & Thomas, 2006;Van Schaik, Volman, Admiraal, & Schenke, 2019;Vrieling, Van den Beemt, & De Laat, 2016).
Our study has for this reason focused on collaborative teacher learning in teacher learning groups.Teacher learning groups are considered to contribute to educational quality, and to a culture of professional learning (Stoll & Kools, 2017;Vangrieken, Dochy, Raes, & Kyndt, 2015).In teacher learning groups, teachers learn by co-constructing knowledge; a process in which they make use of a variety of sources: a variety of classroom experiences, research literature, expert knowledge, and their own research (Hargreaves & O'Connor, 2017;Van Schaik et al., 2019).The development of a professional school culture is an important approach for promoting teacher learning (Stoll & Kools, 2017).Nevertheless, school leadership that supports, stimulates, and facilitates teacher learning, has been found to be a key condition for collaborative teacher learning (Stoll & Kools, 2017).

| Learning-centred school leadership
Learning-centred leadership is conceptualized by Hallinger et al. (2017) and Liu et al. (2016) as intentional efforts to inspire, guide, direct, support and participate in teacher learning with the goal of increasing professional knowledge and promoting school effectiveness.Empirical studies, on how to support schools as professional organizations, indicate that school leadership practices should be directed at teacher learning (Robinson, Lloyd, & Rowe, 2008;Stoll & Kools, 2017).
In learning-centred leadership, leadership refers to the behaviour or leadership practices of individual school leaders.Liu et al. (2016) distinguished four underlying dimensions of leadership practices in learning-centred leadership.The first dimension, building a learning vision, refers to the extent to which school leaders use leadership practices for setting, clarifying, communicating, demonstrating, and providing a vision of teacher learning.
Secondly, modelling refers to the extent to which school leaders display enthusiasm for learning in their own behaviour, and a willingness to share personal learning with teachers.Thirdly, providing learning support addresses the role school leaders play in creating an environment in which resources and opportunities for learning are made available, and in which teacher learning is respected, rewarded, supported and encouraged.Fourthly, managing the learning program refers to the extent to which school leaders organise, manage and monitor teacher learning.
In the learning-centred leadership perspective, a formal leadership position is assumed; leadership is performed by a school leader (Gronn, 2002).

| Distributed leadership
Distributed leadership builds on the idea that people are naturally inclined to influence their environment in a positive and proactive way (Deci & Ryan, 2000).The focus is on engaging expertise and leadership wherever it exists within the organization, rather than seeking this only through formal positions or roles (Harris, 2007).
Distributed leadership is characterized by four practices.
Firstly, leadership as a group characteristic highlights that leadership is the purview of a group or network of interacting individuals, instead of that of a formally assigned school leader, such as a principal or team leader (Woods, Bennett, Harvey, & Wise, 2004).Leadership is conceptualized as the interactions between leaders, followers, and the situation in which leadership is required (Spillane, 2006).Each situation might require new expertise, and therefore the positions of leaders and followers can change, depending on the situation.Secondly, opening the boundaries of leadership refers to the characteristic that both people in formal and in informal positions can take on leadership initiatives and responsibilities (Spillane, 2006;Woods et al., 2004).Leadership practices are shared by various members of the school organisation, such as teachers, who can be empowered to assume leadership roles (Admiraal et al., 2016).Gronn (2002) refers to this as the "leadership-plus" aspect of distributed leadership.
The third distributed leadership practice is that individuals or groups assume leadership roles or contribute to leadership on the basis of their expertise, affinity, or experience in a specific situation or at a particular time.
Consequently, distributed leadership includes a dynamic perspective on leadership; depending on the situation, members alternately claim, acquire or are assigned leadership positions (DeRue & Ashford, 2010).
Lastly, the fourth distributed leadership practice is aligning leadership initiatives, combining them into joint activities with shared goals, and connecting them to school development.This is what Woods et al. (2004) refer to as "conjoint activity".The latter means that members work, learn or develop together, striving towards joint goals, sharing meanings and ambitions.Because of this distributed leadership practice, Bush (2011) refers to distributed leadership as "a collegial perspective on leadership".Similarly, Heck and Hallinger (2009) refer to it as a form of collaboration practiced by school leaders and teachers in leading school development.

| An integrative leadership perspective
The formal learning-centred leadership perspective and the informal distributed leadership perspective seem to be contrasting perspectives on school leadership.However, some authors suggested that formal and informal leadership can co-exist and interact (Bolden, 2011;Bouwmans et al., 2019;Tian et al., 2016).Moreover, in an extensively cited study on claims of successful school leadership by Leithwood, Harris, and Hopkins (2019), the authors argued that distributed leadership practices are part of what successful formally assigned school leaders do.
In a recent update of their article, the authors asserted that findings from a considerable body of research support the validity of this claim (Leithwood et al., 2019).MacBeath (2005) distinguishes four ways in which formal leaders can create opportunities for distributed leadership.Firstly, in formal leadership, initiatives and activities are distributed by delegating influence and responsibility by formalising teacher leadership tasks.Secondly, pragmatic distribution refers to ad hoc influence and responsibility, which is delegated by school leaders as a response to increased demands.Thirdly, strategic distribution refers to new members with specific expertise or affinity taking on specific leadership tasks.Fourthly, incremental distribution, refers to the transition of leadership practices from formal leadership to informal teacher leadership.According to Bouwmans et al. (2019), incremental distribution implies that formal leaders give increased responsibilities and leadership to teachers based on their ability when teachers are willing to take on these responsibilities.Each of these distributed leadership practices can support teacher learning.
The aim of the present study was to develop a typology of school leadership that integrates distributed leadership practices with learning-centred leadership practices, resulting in an integrative leadership perspective.The study was guided by the following question: Which types of leadership for fostering collaborative teacher learning in schools can be distinguished?

| ME THOD
In order to answer the research question, we conducted a mixed-method study.Qualitative data was gathered with in-depth interviews of ten school leaders from secondary schools; the aim was to identify to what extent school leaders applied learning-centred leadership and distributed leadership practices.Quantitative data for supporting our analysis was gathered with an online questionnaire completed by 39 teachers from the same schools as the school leaders; the aim was to measure teacher perceptions of the role of school leaders in teacher learning.

| Participants
Ten participants with leadership responsibilities at six secondary schools in the Netherlands participated in the study: five principals, four assistant principals and two team leaders.Dutch secondary schools are usually structured with leadership on two levels: principals and assistant principals on the first level; team leaders and department leaders on the second level.We refer to participants at both levels as school leaders.In the Netherlands, school leaders have a high degree of autonomy to make decisions about the quality assurance of teaching and learning, innovations in school, to hire teachers and take responsibility for the professional development of teachers.Teachers have a high degree of autonomy regarding their teaching and professional development.Secondary schools in the Netherlands include preparatory vocational tracks (age 12 to 16) that prepare students for secondary vocational education, and general education tracks (age 12 to 18) that prepare students for higher education such as higher vocational education and university education.
Interview participants were school leaders highly engaged in knowledge co-construction; interviewees were suggested by teachers in teacher learning groups in respective school.Three were women and seven were men.
Five had more than ten years of experience in being a school leader, two had more than five years of experience, and two had less than three years of experience.Also, 38 teachers participated.An overview of the participants and a general description of the teacher learning groups is provided in Table 1 below.

| Data collection
The aim of the interviews was to identify school leaders' use of learning-centred leadership and distributed leadership practices.Semi-structured interview guides were used, to ensure a comprehensive account of leadership practices.Open-ended questions on how leaders manage, facilitate, and stimulate teacher learning were asked, such as "In what way do you keep yourself informed about the progress in teacher learning groups?" and "Do you participate in a teacher learning group yourself?"School leaders received the complete interview guides prior to interviews.All the interviews were digitally recorded and transcribed.
Data on teacher perceptions of the role of school leaders in teacher learning were gathered using a brief questionnaire.The latter included two items directly related to the role of school leaders and their leadership practices: "My school leader regularly asks questions about the progress in teacher learning groups" and "My school leader shows appreciation for my participation in teacher learning groups."Furthermore, four items concerned the role of school leaders indirectly, such as "I have sufficient professional development time annually for teacher learning group meetings and teacher learning group -related activities."All six items were rated on a five-point Likert scale, and the teachers were asked to what extent the items applied to knowledge co-construction in teacher learning groups (1 = not true, 5 = completely true).

| Data analysis
First, a matrix was set up to structure the information from school leader interviews; each school leader was defined as a case and was assigned a row in the matrix.Learning-centred leadership practices were placed in columns, along with distributed leadership practices.Descriptions of indicators of learning-centred leadership and distributed leadership practices are presented in Table 2. Interview responses that fitted these descriptions were entered into the cells per school leader per leadership practice.Second, based on the information that was entered into the cells, a description of characteristics was made for each learning-centred leadership practice and

Leadership practices Descriptions
Learning centred leadership: Managing learning in such a way that when members assume leadership and learn or develop together, they pool their initiatives and expertise, and the outcome is a product, knowledge or energy that is greater than the sum of their individual actions.Source: Authors.distributed leadership practice per school leader.Third, all school leaders were given a score for each learningcentred leadership and distributed leadership practice on the basis of these descriptions.School leaders' use of learning-centred leadership and distributed leadership practices were scored using a five-point Likert scale, as presented in Table 3 (1 = the school leader does not apply the practice at all, 5 = the school leader fully applies the practice).Fourth, a manual cluster analysis of the scores was used for identifying meaningful patterns in leadership practices.School leaders with comparable scores on learning-centred leadership and distributed leadership practices were paired.This resulted in a typology of leadership for fostering collaborative teacher learning.The four types of leadership identified were given a label based on characteristics.The analyses were performed by the first author, and both the process and outcomes were checked and discussed in the team of authors in order to guarantee the reliability of the interpretations.Lastly, the data from the teacher questionnaire was analysed using descriptive statistics, such as mean and standard deviation per type of school leadership (see Table 4).These descriptive were used to compare the types with regard to how teachers perceived the role of school leaders in teacher learning.The analyses were performed by the first author and both process and outcomes of the analyses were checked by the fourth author and discussed in the author team in order to enhance the reliability of the findings.

| A T YP OLOGY OF SCHOOL LE ADER S HIP
Four types of leadership aimed at fostering collaborative teacher learning were found: (a) integration of teacher learning, (b), management of teacher learning, (c) facilitation of teacher learning, and (d) management of daily school practice.Table 3 contains a summary of the four types of school leaders, with their mean scores on the eight indicators of learning-centred leadership and distributed leadership practices.A summary of learning-centred leadership and distributed leadership practices per type of school leadership is provided in Table 5.

| Type 1: Integrators of teacher learning
Three school leaders (A, H, and I) were characterized as Type 1 school leaders because of their average high scores on both learning-centred leadership and distributed leadership (see Table 3).The same school leaders received the highest scores from their teachers on their engagement with the teacher learning groups (see Table 4).Type 1 school leaders have a clear vision of collaborative teacher learning and integrate this vision in the organisation of the school.
The vision of school leader I is typical.According to I, collaborative learning through peer support, and sharing knowledge and experiences with colleagues, is part of the professional culture she wants to establish.Likewise, school leader H's focus on collaborative learning is illustrated by the mandatory participation of his teachers in teacher learning groups, where internal expertise is utilized as much as possible.Also, these school leaders communicate and articulate their vision on teacher learning for their schools.
All three school leaders have in common that they managed to translate their learning vision into learning support and a variety of opportunities for learning in practice, such as regular joint sessions to learn from expert teachers, and teacher learning groups.For example, school leader A has provided her teachers with extra time, so they are able to meet more regularly for peer-coaching.
Type 1 school leaders generally participate in the teacher learning groups themselves and are therefore explicitly involved in the learning programs.This gives them the opportunity to model values of openness and risk-taking through their own learning behaviour.School leader I participates in one of the teacher learning groups; in interviews, she used the term "we" instead of "they" when discussing teacher learning at her school

TA B L E 4
Scores on how teachers perceived their school leaders' engagement with teacher learning groups (TLGs) per type of school leadership "I wanted to participate in the teacher learning group myself.As their school leader, however, I also discuss their personal development.I first asked the teachers if they agreed that I would participate in their learning.Consequently, I try to participate as a co-learner, and not as their school leader."(School leader I, when the interview was conducted) Furthermore, a characteristic that emerged for school leaders A, H, and I was that they persistently connected teacher professional learning to school development, and monitoring the process as well as the outcomes of the broader learning program.For example, school leader H offered a variety of opportunities for teacher learning, such as teacher learning groups and peer-coaching, that were in line with the goals for school development.School leader I explicitly stated that an important role for him was to connect teacher learning to school development.Referring to a teacher learning group on research he noted, "My role is to monitor whether the questions the research group is dealing with fit, or contribute to the vision for our school development."(School leader H, when the interview was conducted) In line with distributed leadership practices, all three school leaders provided space for teachers who were not formally school leaders to take on leadership initiatives, activities, and responsibilities.This was particularly evident for teachers who showed initiative in spreading their expertise and enthusiasm.For instance, school leader H mentioned that he worked for a shift in leadership initiatives: "Initially, I initiated and led the data team myself.Currently, the data team is led by teachers who alternately bring in their own leadership initiatives.So, my own role is changing, which is actually quite interesting and exciting."(School leader H, when the interview was conducted) The above quotation illustrates that leadership, considered as taking on responsibility, is certainly seen as a group concern.Furthermore, Type 1 school leaders award leadership of informal leaders; particularly within the context of teacher learning groups, or when the knowledge that is co-constructed in teacher learning groups is shared with other colleagues, or disseminated within the school for the purpose of school development.
Corresponding to the high scores on learning-centred leadership and distributed leadership, teachers evaluated the engagement of Type 1 school leaders with teacher learning groups very positively (see Table 4).Teachers reported that their school leaders regularly asked about the progress in their teacher learning groups, that they experienced a high degree of appreciation from their school leaders, and that their co-constructed knowledge aligned with the schools' broader vision on education.Furthermore, teachers indicated that they had sufficient time for professional development.

| Type 2: Managers of teacher learning
School leaders F and G were categorized as Type 2, because of their high scores on learning-centred leadership, and moderate scores on distributed leadership.In line with learning-centred leadership, like Type 1 school leaders, they had a clear vision of the importance of teacher learning in their schools, with a specific focus on collaborative learning.Type 2 school leaders differed from Type 1 school leaders because of their lower scores for distributed leadership practices.A main characteristic was that in their leadership approach, leadership remained the purview of formally recognised leaders; leadership was not approached as a group effort.
Type 2 school leaders articulated their vision mainly by the hierarchical lines of the school organisation, for example by communicating through team leaders.School leader F mentioned that it was part of the role and responsibility of his team leaders to disseminate his vision in their teams.Furthermore, like Type 1 school leaders, Type 2 school leaders expressed their vision by making time and resources available for teacher learning and learning support.This is illustrated by the following remark from school leader G: "I have increased our budget for teacher learning by 50%, with the main objective of stimulating learning and development among teachers.We use this budget for a combination of joint thematic school development efforts, with specific attention paid to peer and collaborative learning in teacher learning groups, and individual developments when desired.My approach here is that whenever a teacher wishes to learn or develop professionally, we think along and cooperate in providing time and resources."(School leader G, when the interview was conducted) Both of the second type of school leaders participated in the learning program.In contrast to Type 1 school leaders, however, they both held a leading, guiding or coordinating position, for instance by leading a teacher learning group.The following remark by school leader G characterizes their participation: "I see that especially teacher learning groups with active leadership meet regularly and are successful in terms of outcomes.The teacher learning groups that are implicitly led meet irregularly and have less impact, in my opinion.Facilitation and active management are very important at our school.I manage two teacher learning groups myself.These teacher learning groups are successful because I invest a lot of time in actively managing these groups."(School leader G, when the interview was conducted) In line with distributed leadership practices, school leaders F and G both provided space for teachers to assume leadership responsibilities based on expertise and affinity, mainly in the context of teacher learning groups.This is comparable with school leader Type 1. Yet, a difference was that the main responsibility was assumed by leaders in formal positions; teacher responsibility and leadership was characterised as contributing to responsibility and leadership, or being supportive of the formal leaders.School leader F illustrated this with the following remark: "Formal leaders, such as my team leaders, have time allotted for creating a learning vision and working on educational developments that align with that vision.Of course, this can be achieved in cooperation with other teachers, yet only to a certain extent.In my opinion, [responsibility for] spreading the vision really belongs to the team leader; he or she should be the team's visionary.I see teachers who cooperate in creating a vision as co-visionaries.An example of such a co-visionary is a former team leader and now a teacher and member of one of the other teams.Without any doubt, he contributes to the development of a new vision for that team.Yet, he cannot take over leadership responsibilities from his current team leader."(School leader F, when the interview was conducted) The average of scores assigned by teachers to school leaders F (M = 3.50) and G (M = 3.63) were similar.However, the two school leaders scored differently on the individual variables (see Table 1).School leader G was assigned higher scores on variables reflecting his personal involvement with teachers, such as asking for progress and expressing appreciation for their learning and development.By contrast, school leader F was assigned higher scores for more organisational variables, such as facilitation and time allocation.

| Type 3: Facilitators of teacher learning
We characterized school leaders B, C and D as Type 3 school leaders, because of their moderate scores on learning-centred leadership and high scores on distributed leadership.One of the main characteristics is their strong focus on organising, structuring, and facilitating teacher learning, without being part of the learning program themselves.What stands out is their low average score on modelling.Although, during the interview, all three school leaders were enthusiastic about the opportunities for the teacher learning they provided, none of them indicated that they were learners themselves.
Type 3 school leaders focus on creating organisational conditions for teacher learning, such as time allocation, assigning teachers to specific teacher learning groups, providing easy access to support, and connecting professional development to school development.In this, their aim was to structure teacher learning.As with school leader Type 1 and Type 2, they showed a clear vision of the importance of teacher learning, and how this should be organised.
According to school leaders B and C, their vision for learning was communicated and articulated in respective school.How their vision was communicated reflected a strong organisational focus.The way in which school leader C explained how she communicates her vision is illustrative: "We defined our vision and policy, which was subsequently communicated to the teachers via the participation council.After the response of the participation council, we fine-tuned our vision, after which we just have to get started."(School leader C, when the interview was conducted) Type 3 school leaders have in common a limited monitoring of the learning program.School leader C, for example, says that she does not know whether what is learned is actually applied in the lessons; she thinks she should do more regular classroom visits in all her departments to check this.
With regard to distributed leadership practices within the structure and organisation of teacher learning, all three school leaders suggested that they provided space for teacher leadership initiatives for other than formally recognised leaders, based on expertise and affinity.Characteristic was that these leadership initiatives often resulted in formal leadership positions, that involved coordinating or guiding tasks, such as leading teacher learning groups, supportive teams, subject experts, or coaches.Assigning leadership in the sense of responsibility to teacher learning groups, or individual teachers, was a characteristic of the organisationally focused third leadership style.In the interview with school leader D, he indicated that he had started to create opportunities for leadership initiatives by teachers other than formally recognized leaders.All three school leaders indicated that an important role they played was to manage the various initiatives and developments in such a way that they contributed to a joint educational development in respective school.For example, school leader D mentioned that it was his responsibility to ensure that the findings from teacher-researchers at his school were further distributed and aligned with the school vision.
School leaders B and C received overall an average score from teachers, with higher scores on structure variables, such as providing time and facilitating teacher learning; both received lower scores for personal involvement variables, such as expressing appreciation and asking about progress (see Table 4).These scores underline the organisational focus on structure and facilitation, although the latter can be improved, according to the teachers.The overall scores for school leader D were lower.

| Type 4: Managers of daily school practice
School leaders E and J were categorised as school leader Type 4, because of their lowest scores, both for learningcentred leadership and distributed leadership.Also, these school leaders received the lowest scores from teachers.Although both school leaders indicated a vision of teacher learning, it seemed to remain largely their own vision.The vision was either not connected (school leader J) or was only implicitly connected (school leader E) with further leadership practices aimed at uptake of their vision.Both underscored the daily management of the school, not a learning program, as illustrated by the following remarks from school leader J: "I would like to focus more on pedagogical aspects and educational developments in my department.
Therefore, a relocation of tasks needs to be considered; for instance, taking away the care of daily student affairs.
I also notice that I have little facilitation time to offer to my teachers to stimulate or intensify collaborative learning.It is now considered a standard part of teachers' annual tasks and responsibilities.Yet, other developments or responsibilities are being prioritized."(School leader J, when the interview was conducted) A second characteristic was the absence of distributed leadership practices.Both school leaders were assigned low scores on all four aspects of distributed leadership (see Table 3).They approached leadership almost exclusively as a characteristic of a formally recognised leader.The only space for leadership initiatives seemed to be within teacher learning groups, in which teachers could contribute to leadership based on their expertise or affinity.For instance, members of a teacher research group discussed topics with school leader E annually.However, formal leaders decided whether these topics would be used and how.
Teacher scores for the fourth type of leader were comparatively low (see Table 4).In particular, scores were low for variables that reflected interaction between school leaders and teachers about the knowledge constructed in teacher learning groups, such as "I share the co-constructed knowledge in conversations with my school leader."

| D ISCUSS I ON AND CON CLUS I ON
We analysed leadership practices for fostering collaborative teacher learning among school leaders in ten secondary schools in the Netherlands.We have developed a typology of school leadership that integrates leadership practices that draw on learning-centred leadership theory (Hallinger et al., 2017;Liu et al., 2016), and leadership practices that draw on distributed leadership theory (Bouwmans et al., 2019;Spillane, 2006).Four types of school leaders were distinguished: integrators of teacher learning, facilitators of teacher learning, managers of teacher learning, and managers of daily school practice.
In general, integrators of learning demonstrated a strength in using both learning-centred leadership practices and distributed leadership practices.The combination of these two kinds of leadership practices reflects what Leithwood et al. (2019) claim to be leadership practices of successful school leaders.
Managers of learning showed a combination of strong learning-centred leadership practices and moderate distributed leadership practices.The difference between them and integrators of learning, however, was in the use of distributed leadership practices.Although managers of learning expressed support for teacher leadership initiatives, those initiatives were mainly directed by formally recognised leaders.In line with Leithwood et al. (2019), we propose that the effectiveness of leaders in this category would benefit from an increase in the use of distributed leadership practices.We think this typology could be useful in further research and could also be practically applied to supporting school leaders who want to improve their leadership practices for fostering collaborative teacher learning.

| Limitations and further research
The set-up of the study constitutes a limitation in that it does not allow us to generalise the findings.First, it was a small-scale qualitative study, using data from a limited number of participants.Second, data were collected in the particular context of the Dutch education system.The Dutch school system is characterised by school autonomy, resulting in the use of a diversity of pedagogical concepts, different types of school boards and management approaches.It is possible that different leadership practices are used in other contexts.The typology of school leaders we have developed could, as a next step, be validated with a large and more varied (preferably international) sample.A second limitation of the study is that we only focussed on the influence of school leadership on teacher learning, whereas there are other influences, such as professional school culture (Stoll & Kools, 2017), that are known to be relevant for collaborative teacher learning.Further research could take such influences into account.A third limitation is that we focused on collaborative teacher learning, whereas teacher learning is a multi-level study object that also entails questions concerning individual teacher learning.We focused on collaborative teacher learning as this has been found to have a positive impact on teacher professional development, school development, and the professional learning climate in schools.However, questions concerning individual teacher learning within teacher learning groups, or concerning how our integrative approach would affect individual teacher learning, are relevant to investigate in future research.A final limitation concerns the composition of our sample, which consisted of school leaders who operate at different levels of the school organisation ranging from principals to assistant principals and team leaders.Although all the school leaders were responsibile for teacher learning, their differing responsibilities might, for example, explain variations in the perceived distance between leaders and teachers.Further research could focus on a single level of leadership or could compare the different levels with each other.Fourthly, our aim was to describe and categorise leadership practices.Further research could explore the effects of these leadership practices on collaborative teacher learning, and their connection to school development.

| Implications
Our findings align with current ideas for leading schools as professional learning communities in Europe (Stoll & Kools, 2017).Yet, according to the Training and Learning International Survey (TALIS) (OECD, 2019), school leaders spend only 16% of their time on educational leadership, which means their role can be strengthened.The results from TALIS call for professional standards for educational school leadership.In the Netherlands, requirements for school leadership are set by the Ministry of Education and professional organisations for school leaders (Andersen & Krüger, 2013;VO-raad, 2014).These include, among other, teacher professional development as an ongoing process and strengthening professional learning cultures in schools.School leaders are increasingly expected to be responsible for achieving coherence between education goals and teacher learning.In education and staff policies, their focus should be on creating a culture in which teachers learn collaboratively how to improve the quality of learning processes.
The added value of this article is that our perspective, that integrates learning-centred leadership and distributed leadership, provides a fruitful lens to empirically study collaborative teacher learning and to highlight the importance of distributed leadership and its relationship to learning-centred leadership.Our typology of school leadership provides insight into how school leaders can foster collaborative teacher learning.However, according to TALIS 2018, sufficient time and support is necessary for school leaders to successfully foster a culture of collaborative teacher learning.
. The following remark by her is illustrative: TA B L E 3 Means for learning-centred leadership practices and distributed leadership practices, per type of school leadership enacted a combination of moderate learning-centred leadership practices and strong distributed leadership practices.They focused on organising, structuring, and facilitating teacher learning.Their distributed leadership practices align with what MacBeath (2005) refers to as formal distribution, because facilitators of learning tend to delegate influence and responsibility by more or less formalising teacher leadership tasks.Their learning-centred leadership practices were rated moderate, as their participation in the learning program and monitoring of what was learned and applied was limited.Drawing on our findings and the concept of successful school leadership inLeithwood et al. (2019), we suggest that facilitators of learning can be more successful if they become more active in participating and monitoring teacher learning.Accompanying teachers in their learning would offer opportunities for modelling participation in continued learning and could positively influence teachers' appraisal of the school leaders' personal involvement.Lastly, managers of daily school practice scored low on learning-centred leadership practices and distributed leadership practices.Whereas the other three types of school leaders could be characterised as more or less learning-centred and distributing, managers of daily school practice were limited in both areas.Therefore, managers of daily school practice were not successful in influencing teacher learning and school development.The results of our study substantiate claims regarding the usefulness of an integrative perspective on school leadership.A single learning-centred leadership perspective would have meant that we had missed distributed leadership practices that contribute to teacher learning, such as assigning leadership based on expertise and affinity, and aligning leadership initiatives to school development.Conversely, a single distributed leadership perspective would have meant ignoring learning-centred leadership practices such as building a learning vision and modelling.Using both perspectives enabled us to distinguish four ways in which school leaders foster collaborative teacher learning, (a) as integrators of learning, (b) managers of learning, (c) facilitators of learning, and (d) managers of daily school practice.
Participants and general description of teacher learning groups TA B L E 1

centred leadership practices Distributed leadership practices Level Core elements Level Core elements
Source: Authors.