For what purpose? Swedish superintendents ’ perspectives of professional development for principals

Purpose – The purpose of this article is to expand knowledge on Swedish principals ’ professional development (PD) from the perspectives of superintendents. In particular, the article analyzes how superintendents understand and organize PD for principals. Design/methodology/approach – Empirical data are derived from a strategic sample of ten ( n 5 10) superintendents.Transcribedinterviewswereanalyzedintwosteps.ThefirststepwascarriedoutinductivelytoidentifyprominentaspectsofPDforprincipals.Inthesecondstep,thedetectedthemesandcategorieswere analyzedmoredeductivelythroughthetheoreticallensoflearninginorganizations. Findings – The analysis revealed thatthe purpose of PD for principalsand the principal leadership that must be nurtured from the perspective of superintendents spans a scale, from knowing what is already required to critically examining and exploring the unknown. In addition, the understanding of learning stretches from an individual enterprise to a collective activity. However, noteworthy differences between the superintendents were detected and organized into three ideal types. Research limitations/implications – Despite a profound research design and a careful selection of superintendents,thesamplesetssomelimitsbecauseofthepluralitywithinthedecentralizedSwedishschoolsystem. Practical implications – The results can support strategies from superintendents, principals and educational authorities to build infrastructures that foster PD at different levels of school systems. Originality/value – This article offers a novel perspective by analyzing principals ’ PD from the perspectives of superintendents.

and Martin, 2013; Lomos et al., 2011). Building a learning organization requires supporting structures and cultures (Senge, 2006). In addition, it is necessary to encourage and nurture individual and mutual learning (Katz and Ain Dack, 2013).
Professional learning takes many forms and features, depending on role, individual preferences and organizational prerequisites. Professional development (PD) can be seen as synonymous with professional learning or as that part of professional learning that is offered or organized together with others. Even if most PD for principals is connected to novice principals and their professional training, we argue that in such a complex environment as schools, with linked governance on several hierarchical levels, principals need more formalized learning together with others. PD is, from that perspective, the responsibility not only of the individual; it is also a governance strategy to improve student achievement and support the effective performance of leaders in their assignments. Accordingly, superintendents play a crucial role in providing high-quality PD for principals (Campbell et al., 2017;Goldring et al., 2012). However, earlier research has shown that PD for superintendents becomes a personal concern characterized by a lack of supportive infrastructures (see Liljenberg et al., 2022). These findings indicate that if PD for principals rests on similar premises, the personal understanding of learning, development and effective principalship will affect what PD superintendents offer their principals. This implies that content and form are more connected to the local education authority (LEA)'s interpretation on necessary PD for principals rather than based of principals' and local schools' needs. To make it even more complex, the requirements for PD to support learning and be effective vary. Research is focused on individuals' ability and motivation and on context and design (e.g. Burch and Spillane, 2004;Rainey, 2014, 2019;Netolicky, 2016;Zepeda et al., 2014). Hence, individual and organizational learning must be supported by a variety of activities and prerequisites, ranging from the local, spontaneous and occasional to the planned, research based and systematic (Erts as and Irgens, 2021).
Thus, this article rests on the assumption that there is a close relationship between superintendents' understanding of the concept of learning, the learning activities in which they take part, and the PD they offer principals. Accordingly, in this article, we strive to expand the knowledge base of how Swedish superintendents, as leaders at the LEA level, understand and organize PD for principals through the lens of learning in organizations. We define PD for principals as the activities superintendents and deputy superintendents [1] design or grant to generate learning to enhance professional knowledge, competence and attitudes that will improve the leadership and performance of principals in local schools. Taking this stand, we extracted information from existing PD practices for principals in nine LEAs of various Swedish municipalities through interviews with superintendents and deputy superintendents. Two research questions guided our work.
(1) What is the purpose of PD for principals from the perspective of superintendents?
(2) What principal leadership is nurtured through the PD?
The article is structured as follows. First, the roles, relations and responsibilities of LEAs, superintendents and principals in the Swedish context and in relation to previous research are outlined. Thereafter, PD for principals is considered through the theoretical lens of learning in organizations. The following section details the methods and data analysis. In the subsequent section, the results are presented. The article ends with a discussion, some conclusions and directions for further research.

Roles and responsibilities in Swedish LEA organizations
In the decentralized and marked-adopted Swedish school system, LEAs are accountable for school results, school improvements and administrative obligations, as regulated in the JEA Education Act and the national curricula for the various forms of school (see, e.g. R€ onnberg et al., 2019). The prevailing directives clarify that all LEAs are required to have a superintendent to ensure compliance with the Education Act (SFS, 2010(SFS, :800, 2010. However, no formal directives exist on which competencies superintendents must have. Therefore, LEAs appoint their superintendents and deputy superintendents and determine what competencies they should have. As superintendents prepare proposals for the decisions of the local board, it grants them important influence over the direction of school improvement and PD (Liljenberg, 2021;Moos et al., 2016;St ahlkrantz and Rapp, 2020).
Regarding principals, the national directives stipulate that the principal is responsible for the school unit, leading and coordinating pedagogical work. In addition, the principal has a special responsibility for school improvement. Most often, superintendents appoint their principals. To be appointed as a principal, one must have "pedagogical insight through education and experience" (SFS, 2010(SFS, :800, 2010. Consequently, there are no requirements for previous teacher education, even if most principals have a background as teachers (Sigridursdottir et al., 2023 forthcoming in press). A minimum of one semester of full-time university studies in pedagogy is required, but previous experience does not necessarily have to be obtained from the school system. However, all newly appointed principals must attend the National School Leadership Training Program (NSLTP) within one year after employment. The Education Act also states that LEAs are responsible for offering principals PD. Beyond the obligatory program, there are no formal regulations about content or approaches to learning for PD for principals. Therefore, the Swedish National Agency of Education offers, individually or in cooperation with universities, a variety of courses, online seminaries and improvement projects. Superintendents tend to give principals considerable autonomy (Adolfsson and Alvunger, 2020). However, before applying, principals need a superintendent's approval to participate.
As previously stated, planning PD for principals becomes an issue for superintendents and deputy superintendents when they consider national reforms and LEA requirements in combination with what is seen as deficits or improvement areas in the local school organizations. Until now, there has been little knowledge about PD for principals. However, there is a growing body of research on the central LEA level giving us relatively extensive insight into the sometimes challenging relationship between superintendents and principals (Addi-Raccah, 2015;Honig and Rainey, 2019;H akansson and Adolfsson, 2022). In addition, international research stresses the importance of middle-tier intermediaries being learningoriented and providing job-embedded PD for all staff if school improvement initiatives are to produce results and eventually close gaps in achievement between diverse groups of students (e.g. Leithwood and Azah, 2017). Nordholm's (2016) work also indicates that Swedish LEAs and their superintendents need support to interpret educational reforms innovatively to support principals and teachers in the local schools.
To conclude, even if research has revealed significant details about LEAs' and superintendents' important work, essential questions remain about whether schools are to develop into learning organizations and whether improvement efforts are to yield results. Due to the increasing heterogeneity among LEAs and local schools, it becomes even more important to focus on the aims and content of PD for principals. Specifically, few details are provided on how superintendents understand PD for principals, its purpose and the consequences for principals and schools. To this end, we strive in this article to offer a contribution from a Swedish perspective.

Professional development in light of learning organizations and individuals
As an educational leader at the LEA level, a superintendent needs to govern and create an environment that supports and encourages independent and knowledgeable principals. In Sweden, state governance has in recent decades emphasized a decentralized school model combined with increased accountability and state control . Given these developments, there is also an emphasis on cooperation and PD to handle complex issues. Therefore, the need for schools to act and develop into learning organizations has also increased. A learning organization needs systems thinking, personal mastery, mutual mental models, a shared vision and team learning (Senge, 2006). Learning together becomes important to handle a constantly changing context. PD for principals is important to create and uphold a learning organization of high organizational quality. In this regard, mature organizations gain and use insights in relation to their missions (Arnold and Wade, 2017). Moreover, PD for principals can be crucial in helping them choose to stay in their positions. However, what principals and LEAs consider important to learn as well as how this learning should be organized is more ambiguous. At all levels, educational leaders require knowledge on a broad range of issues related to administration and the schools' core-teaching and learning for all students. Educational leaders also need knowledge on how to build organizations and how leadership and management can support improvement and highquality education. This means that PD needs to support individual and team learning (Ellstr€ om, 2001;Senge, 2006).
Learning as a concept can be based on the understanding of various ontological and epistemological concepts, ranging from cognitive-based perspectives, for which prior experience and second-hand information create a fixed body of knowledge, to perspectives building on a constructivist understanding focusing on complexity and problem-based learning, including a critical view. Single-loop and double-loop learning by Argyris and Sch€ on (1978) is a well-known version of this distinction, which was followed by related distinctions other researchers had made (Senge, 2006;Ellstr€ om, 2001). Regarding PD, this means that knowledge building can be ambiguous and can stretch between increased knowledge, confirming and extending current knowledge and interpretation and sensemaking to understand and critically examine reality. Ellstr€ om (2001,2011) used the concepts of adaptive learning and developmental learning to distinguish between the two perspectives. Adaptive learning refers to the mastery of given tasks and situations or the formation of competencies to improve routines or to handle frequently occurring problems in organizations. In contrast, developmental learning refers to situations in which individuals or groups begin to question and explore established working procedures, norms and ways of defining problems in organizations, but they act to develop new ways of handling complex situations and duties. Therefore, combining learning intended to solve current problems and to provide positive organizational affects with learning intended to foster critical thinking, altered perspectives and independent decision-making can be challenging. Although cognitive-based and constructivist-based perspectives are often presented as dichotomies, these perspectives overlap on a sliding scale. Ellstr€ om (2001), for example, argued that adaptive and developmental learning are complementary because professional work includes predefined tasks and routines to follow as well as complex problems and questioning of routines to develop new ways of working.
Furthermore, Targama (2007, 2013) put forward an interpretative perspective on individuals' and organizations' development, suggesting that individuals' attributes at work, such as their knowledge and competence, are strongly bound to their understanding of work. Therefore, the way individuals understand a specific work task governs their response. This means that individuals who hold the same position can understand their work differently. They often agree on what to do, but how they act differs. The variation in understanding and how they realize their work is connected to competence. Those who are assessed as having lower competence have problems describing what colleagues who hold higher competence do to perform better, but the person with higher competence can describe the difference. This indicates that assessment and interpretations are related to competence, which can hinder people from identifying necessary knowledge to enhance their or others' professionalism (ibid). JEA A work novice must rely on existing norms and social practices and acquire new perspectives (Fenwick, 2003). However, acquiring new experiences that are not problematized can be counterproductive and may strengthen inherent cultures, structures and methods. On the other hand, if the organization is not ready, a more open, reflective and critical stance can cause confusion and counteract implementations. Trying out new perspectives and methods can take a long time and may initially have a small or even negative impact (Lewitt and March, 1988). To be caught in an adaptive mode of learning tends to be more problematic for individuals and organizations.
To conclude, superintendents can have various notions and understandings of PD for principals based on how they understand principals' work and how they understand and manage their own work as well as their own learning experiences. In this article, we apply this take-off point in combination with the concepts of adaptive and developmental learning as well as individual and collective learning as a lens for assessing the purpose of PD for principals and nurturing leadership.

Method and analysis
As previously noted, empirical data are derived from a strategic sample of Swedish superintendents in a variety of municipalities (see Table 1). Sweden contains 290 local municipalities with between 2,000 and 900,000 inhabitants (SKL, 2016). Three municipalities have more than 300,000 inhabitants, with more than 200,000 inhabitants in the inner cities. The average municipality in Sweden has approximately 40,000 inhabitants. In addition, the municipalities are classified based on their locations: municipalities with major cities (1a) or close to major cities (1b); municipalities with large cities (2a) or close to large cities (2b); and municipalities with smaller cities (3a), close to smaller cities (3b) and rural municipalities (3c). In the current sample, nine municipalities were strategically selected with between 2,000 and 135,000 inhabitants. These municipalities' locations provide a broad representation of Sweden. Moreover, this sample also provides heterogeneity in terms of the LEAs' role and function. Somewhat specific to Sweden, the Swedish school boards at the local municipal level consist of appointed politicians representing the political parties of the municipality council. Accordingly, LEAs and their superintendents are obligated to collaborate with and implement local political intentions and decisions. However, given the fact that there are 290 municipalities in Sweden, there is also variety in local school governance and organization, which is not seldom linked to municipal size and location (Johansson et al., 2016). From former work (Nordholm, 2016), we also know that LEAs have the potential to provide innovative middle-tier translations, but this job can look very different, e.g. depending on whether LEAs and superintendents have access to administrative support, professional colleagues and networks or are forced to undertake this work mainly on their own (Liljenberg et al., 2022 We also know from former work (Nordholm et al., 2021) that there is variation in terms of culture between municipalities, which, for example, manifests in the degree of autonomy that LEAs and their superintendents give principals in their leadership. Accordingly, the sample offers a valuable heterogeneity from additional perspectives. The article builds on data from nine semi-structured digital interviews conducted with five superintendents and five deputy superintendents. A request for participation was sent by e-mail to the respondents. In one municipality, a deputy superintendent wanted to be interviewed with a colleague, which we approved to avoid losing their participation. All superintendents and deputy superintendents (men and women) were selected based on municipality size, the responsibility of various school forms and the number of employees. The participating superintendents and deputy superintendents had 2-13 years of experience in their current positions. To increase knowledge about how superintendents understand and organize PD for principals, we posed open questions to the interviewees [2] about their PD and PD for principals. Each interview lasted approximately 60 min and was transcribed verbatim afterward.

Data analysis
In the first step, the analysis was carried out inductively to identify prominent aspects of principals' PD from the perspective of superintendents. Specifically, the analysis strived to detect interview extracts and dictums detailing the purpose of PD for principals, the activities taking place and the leadership to be nurtured. After reading the transcribed interviews several times, basic categories and themes were abstracted from the material, compared and combined into organized themes (Attride-Stirling, 2001). A sliding scale on the approaches and strategies for PD activities was initially identified, ranging from ad hoc to strategic planning.
In the second step, a more theory-grounded analysis was undertaken based on the framework of learning organizations and individual learning. With the themes identified in the first step of the analysis, the second step elaborated on the concepts of adaptive and developmental learning, individual and collective learning, as well as perspective on competence and variation in the understanding of work (Ellstr€ om, 2001(Ellstr€ om, , 2011Targama, 2007, 2013). This step of the analysis also included several occasions for peer debriefing (Robson, 2002) in the research team to guard against researcher bias.
The concepts provided a valuable lens for assessing the PD the superintendents designed or granted and the purposes behind those decisions. The analysis also revealed three categories representing different understandings of PD for principals that emerged during the analysis. However, although three categories emerged from the analysis, they are ideal types (see Weber, 1977) that are analytically separate but overlap in practice.

Results
The results demonstrate that PD is necessary for individual careers as well as for meeting organizational and societal changes and expectations. Even if learning is expected of successful leaders, the overall results show that the support and requirements that principals receive from their superiors affect what they can learn. How superintendents understand their subordinates' roles and tasks and their own work and role as superintendents in turn affects what PD they find necessary to provide. In the results, the purpose of PD for principals spans on a scale from knowing what is already required (adaptive learning) to critically examining and exploring the unknown (developmental learning). In addition, the understanding of learning stretches from learning as an individual enterprise to an understanding of learning as a collective activity. In addition to the mandatory NSLTP, the infrastructure for principals' PD was underdeveloped and built on the individual superintendents' understandings rather than on a well-defined and structured practice. However, noteworthy differences between the superintendents were JEA detected and elaborated into three ideal types that represented important differences regarding the purpose of PD for principals and the principal leadership to be nurtured. The three ideal types are presented under separate subheadings below. The overall image, built up by the three ideal types, is further problematized in the discussion section.
Ideal type I-professional development as a remedy for individual shortcomings In the interviews, the superintendents who describe PD for principals as a remedy for individual shortcomings also express principalship as heavily dependent on personal traits. This influences how these superintendents think principals should act to be successful in their roles. In the quote below, this is demonstrated by Superintendent A, who expresses a strong belief in the individual leader: Superintendent A: I'm a bit into how you look at your own role if you are like the captain of the ship who points out a direction and gets people with you or if you are the one who gathers people and asks, "Now we have to think together.. . . Where are we going? I have tended to think that we should have the first one. Principals who have their own ideas and can get people to follow them, explain it so that people become enthusiastic and have a little charisma, a little authority that makes you be perceived as credible in this captain's role, which I perceive as more effective.
Superintendents often express the same ideas when they talk about their positions and the expectations that politicians and other decision-makers have of them. They talk about themselves as strategic experts with knowledge about the municipality as well as about rules and regulations and with the mission to "fix the school so it becomes cheaper and better". Consequently, they express no need for PD in any specific knowledge area. This also holds for how they arrange for PD for principals in the organization.
The NSLTP, organized by universities, is mandatory for all novice principals in their first principal positions, and superintendents must send their novice principals to the program. However, the superintendents who speak of PD for principals as being related to individual shortcomings express no thoughts about the need for PD for experienced principals or how novice principals who attend the program can contribute with new insights to experienced principals or to the organization as a whole. Rather, they talk about the program as something that "has to be done" and that, when completed, gives principals more time for their "real work": Superintendent A: National School Leadership Training Program lasts for three years, and it takes about six months or a year before you understand the expectations and feel comfortable . . .. But then, when you have finished, you are a little tired of . . . you are a little tired of working extra hours for such a long period, so we have not had any major activities aimed at principals' professional development. However, we have not said no to anyone, I think, that wants to do something themselves.
Thinking of PD for principals as a remedy for shortcomings also implies that when they give examples of PD activities that may become relevant to the organization, the superintendents talk about activities that can make the principals stronger leaders. For example, the superintendents have noticed that principals have difficulties "dealing with problematic employees" and "troublesome parents". To become better at handling this type of challenge, the superintendents believe that powerful leadership training can give principals the necessary tools for and greater insight into their own personalities as leaders. From the superintendents' perspectives, principals can receive significant support from the LEA level, and if they fail in their work despite this, they do so due to their own personalities. Given that some superintendents believe that PD cannot remedy all shortcomings, they sometimes conclude that principals who encounter difficulties in their work, regardless of previous experience, simply do not have what it takes to handle the principal position. In these cases, the superintendents try to convince the principals that the best thing for them and the organization would be for them to leave their position: Superintendent B: I have made her understand this, but it is so ingrained in her personality that it is not something she can relearn. She's like that. And then she understood . . .. She has said to me like this: "Thanks to those conversations we had, now I have understood that I should not work as a principal anymore." She has worked as principal for twelve years, only two years in my organization, and she says, "I understand that I have to do something else.
Ideal type II-professional development as a tool for school improvement When these superintendents talk about their PD, some of them emphasize the importance of colleagues and forums in which they can meet and discuss the things that they struggle with in their work. Similarly, some of them argue that PD for principals must be arranged for principals as a group. If PD is to be of any use, it must impact the local schools in connection with the aims and visions at the LEA level. Therefore, it has to build on jointly identified needs and be arranged so that the principals can learn together and from each other. This implies that working with PD for principals becomes an important part of implementing LEA's aims and in that sense a governance tool. One superintendent described this as a process. First, "What do we have to improve in our organization?" Second, "We must work better with digitalization." Third, "What professional development do principals need, then?" The strong connection between school improvement and PD for principals can also be related to the Educational Act (SFS, 2010:800, 2010, 2 cap. 9x), which states that Swedish principals "must especially work to improve the education". When needs for improvement extend beyond the superintendents' competence, they build collaboration with external partners. However, in larger municipalities, the superintendents try to take advantage of their knowledge in the entire LEA organization. More senior principals who have improved their schools in various areas are given the possibility to share their knowledge but also develop as leaders taking responsibility for their colleagues' PD, as in the school-age educare network Superintendent E1 mentions: Superintendent E1: We have formed networks for different school forms; for example, we have a network for principals who lead school-age educare. In the network, we work together to systematically develop and lead school-age educare. It becomes a professional development activity for the principals that they plan together with us for their colleagues.
In municipalities with several novice principals, these superintendents also find arranging hands-on PD activities for principals important. As Swedish principals attend training in parallel with their first assignments, novice principals have many tasks to complete although they have not received any training for them. According to these superintendents, helping novice principals get a good start in their new positions is an important part of their work. Without this support, they stress that school improvement in the organization is at risk of losing momentum: Superintendent F: I help them in the craft being a principal entails, and I help them design these "year wheels" for systematic quality work. So, you can say that I hold ongoing internal training in systematic quality work where I go side by side with them.
Expressing the idea of PD as a tool for school improvement, the superintendents also argue that having novice principals attending the national training program is a benefit for the entire organization, for them and for the more experienced principals. When novice principals attend the program, they can share new knowledge in the meetings together with the superintendent and the other principals. When they finish the program, the superintendents arrange a seminar during which they can present the improvement work that they have conducted in their schools and written about in their final exams. As all principals in the organization attend the seminar, this becomes an opportunity for PD for them all, but Superintendent J also mentioned an opportunity for the exchange of experiences and hence for team learning: JEA Superintendent J: Since I believe so strongly in co-learning and collegial learning and actually being able to be good role models for each other, I have, after they finished their principal degree, set aside time for all of us so that we can listen to them and ask wise questions, all to enable an exchange of experiences. I say that we will be Sweden's best municipality because we are so up-to-date.. . . I think we are building a bank [of knowledge] together here.
Moreover, as school improvement is a constant process, PD, according to these superintendents, must be developed in parallel. As these superintendents stress the training program as basic training, they have recurring meetings with all principals to come to common conclusions about the next steps and ways to top their PD in accordance with the ongoing improvement processes in the schools. This indicates that the superintendents use PD to create team learning and togetherness, which are considered important features of organizational learning. This in turn was viewed as a way to strengthen principals' work to improve teaching and learning in their local schools.
Ideal type III-professional development as a path to individual and organizational awareness and critical thinking Giving time to read research and discuss results with colleagues are, for these superintendents, a necessity for becoming proficient in their roles. However, staying updated professionally does not aim solely to respond to identified needs in the organization. Rather, for these superintendents, it aims to develop a critical mindset that can challenge well-established assumptions about "how things are and should be" and thus contribute to better-balanced decision-making in the organization. Having these ideas about their own PD affects how the superintendents arrange for PD for principals and what they believe it should support, as Superintendent C expressed: Superintendent C: I think it applies both to me and the principals. You must have pretty good judgment and some kind of moral compass . . .. You need to have the ability to put yourself in someone else's situation and think about what it's like to be a parent if you go to this meeting, and so on. And you need to think about it before you are inviting to the meeting.. . . If you train an organization to think like that, the first meeting gets much better, and you don't have to correct so many things afterwards.
To steer PD in the direction they desire and to promote awareness and critical thinking, the superintendents apply various strategies. At first, they arrange for regular meetings during which they meet the principals and encourage them to help each other with clarity when they express their individual opinions rather than well-grounded arguments or when they request decisions to follow to avoid taking a stand and thus miss out on opportunities for learning. Encouraging principals to learn from each other also requires that superintendents are aware of how they act to avoid counteracting what they want to promote: Superintendent H: It comes a lot to the principals, many ideas about what they should do and how they should do it. To be able to talk about what's behind different methods, the ideas about learning, in relation to that, I think it is important for me to be updated and keep those thoughts alive all the time to be able to steer, sift, ask the right questions . . . but also help them develop their own awareness. I cannot tell them that. It will not happen. It's about getting those conversations started so that they can form those arguments together, create that understanding.
Second, these superintendents encourage principals to continue their university studies toward master's degrees. To make this possible, the superintendents try to facilitate principals' work situations and give them time for writing when needed. They also work out strategies for using principals' knowledge when they take their exams. Given that Swedish principals have few career advancement options beyond becoming superintendents, they try to create alternative assignments that can make more experienced principals want to stay in the organization, for example, making them responsible for various improvement areas or as mentors for other principals. These assignments are intended to keep the principals in the organization and to raise the general level of critical thinking in decisions in the organization.
Third, these superintendents try to create partnerships with researchers who can put theoretical perspectives on many of the things that principals want to improve in their organizations, contribute to critical discussions and provide support when improvements are necessary. Recurrently, these superintendents also stressed the importance of PD for principals being research based and continuous. Overall, they consider principals' knowledge and professional judgment crucial for the development of the Swedish school, as Superintendent G demonstrated: Superintendent G: I think a lot about the fact that we need a more well-educated force of both principals and superintendents. I'm thinking so. It's a tickling thought that you should have master's degree before you can become a principal. I think that would have given a completely different conversation about bildung and education.. . . You cannot think that the national training program is enough. No, it cannot be like that. You are just in the beginning of your new position. That's really what I think. Continuous professional development is a legal requirement. If you are to keep your position, it must be included, but also that the threshold for the principal's position must be higher. Yes, in the long run, I think it's a necessity.
Consequently, the purpose of PD for principals from these superintendents' perspective stretches beyond the current situation and addresses the individual, the local school and the future of the Swedish school.
To visualize how the ideal types relate to the theoretical concepts of learning (Ellstr€ om, 2001, 2011) we constructed a matrix (see Figure 1). In the matrix, each superintendent participating in the study is represented by a dot.
In addition, Targama's (2007, 2013) perspective on competence and variation in understanding of work in organizations showed that the superintendents positioned in ideal type I emphasized the individual principal learning to handle issues and processes in an advanced, decided and ideal way. From this perspective, the purpose of PD is to give principals a certain kind of knowledge based on the best or the correct practice. Moreover, PD in ideal type I aims to form a principal role that follows what others indicate is correct rather than principals undertaking their own initiatives. This understanding of work also indicates that individual abilities and knowledge are important, and a lack of a specific competence could be the reason a principal needs a certain PD. As the superintendents focus on current practice, there is little time for meta reflection or a critical perspective that The dots represent the superintendents participating in the study JEA questions aims and structures. Experience is mostly exchanged when people give advice and solve problems. The intention is that new knowledge shall immediately be used in practice.
Superintendents positioned in ideal type III, on the other hand, see PD as opportunities for principals to critically review their practice and to create new understanding by sharing and reflecting together. Based on this idea, critical and theoretical aspects become evident because they can open new perspectives. Joint learning helps build a common understanding of what is important and can be an opportunity to understand and execute leadership in various ways. Thus, these superintendents emphasize developmental and collective learning. This view of learning helps form principals who are more independent and prepared to solve unknown challenges in schools. Learning and building competence become an aim in itself.
Ideal type II is situated between the two other types. Here, the commonality is the mutual issues in the local organization and the principal's role is to bring about change. Therefore, superintendents positioned in ideal type II try to balance adaptive and developmental learning although a predominance of adaptive learning was identified. Often, the superintendents direct the content of PD for principals toward an area that is important for the whole LEA organization and that helps them achieve a common vision that includes an opportunity to learn theory and practice, depending on content and form. Experience and practical examples help spread ideas and encourage joint learning. In PD with this purpose, finding a level that suits each principal and school can be difficult. It is also risky to remain on the problem-solving level, where more complex issues are left aside. There is also a risk that the learning focuses on issues on which all can agree and therefore are not challenging or relevant enough for all.

Discussion
This article strived to expand the knowledge base of PD for principals by focusing on how Swedish superintendents, as leaders at the LEA level, understand and organize PD for principals through the lens of learning in organizations. The analysis made clear that from the superintendent's perspective, PD for principals was based on various points of departure. Even if there were initial similarities, the superintendents' arguments rested on various understandings of learning, ranging from adaptive to developmental (Ellstr€ om, 2001(Ellstr€ om, , 2011. PD for principals also reflected superintendents' experiences of PD and their views of principals' leadership, ranging from an individual enterprise to a collective activity. However, it is important to remember that besides the superintendents' experience of PD, differences in size and structure between LEA organizations add to the variation. Even with a limited sample, we could see that PD for principals existed on a scale, making each organization's provision of PD for principals unique. However, based on the noteworthy differences that emerged in the analysis, we constructed the three ideal types to make visible the differences in purpose of PD for principals and the principal leadership nurtured in accordance with the study's research questions. The detected ideal types show various aims and priorities in learning, even if they all strive to meet principals' individual and organizational development. For the superintendents in ideal type I, it is hard to understand the necessity of critical perspectives and theoretical models that are used in ideal type III. This becomes a potential limitation because having the ability to combine theoretical understanding with practice can enhance collective capacity building (Erts as and Irgens, 2021). These superintendents also see learning more as an advantage for the individual principal than for creating a mutual understanding in the principal group. In accordance with Targama (2007, 2013), the understanding of what necessary competencies are might differ among those who initiate PD for principals because they do not have experience of how critical thinking and theoretical perspectives can add knowledge. This, in turn, can explain the heterogeneity among principals regarding what is most important in their roles.
In the decentralized Swedish system characterized by large differences regarding municipality size and ability to provide PD for principals, the results show that the various understandings of what is considered necessary PD for principals can increase the differences in what is expected of principals as well as their prerequisites. In a fast-changing society in which principals have a heavy workload, it becomes even more important to pay attention to the opportunities and expectations superintendents put on PD for principals. To handle complex situations proactively and sustainably, the ability to combine theoretical perspectives with practical experience and knowledge is necessary. However, it is not enough to solve problems and share best practices. There is a risk when PD primarily targets individuals who have not acquired the necessary knowledge. As in ideal type I, the focus is on how to acquire ideal solutions. There is a risk that the learning stalls at confirmation and adaptive learning (Ellstr€ om, 2001(Ellstr€ om, , 2011. Each new situation is seen as specific, and there is too little time to see patterns and find new and alternative perspectives. Therefore, superficial knowledge dominates at the expense of depth, with the risk that principals stagnate in their learning. They become rooted in their current knowledge levels and rely on others' perspectives and suggestions. Moreover, with the limited time, more long-term perspectives become downgraded in favor of issues that can provide faster results (Lewitt and March, 1988).
Viewing PD as educational governance makes investigating what kind of knowledge building is necessary even more important (Sigridursdottir et al., 2023 forthcoming in press). There is a difference if the learning is directed toward practice or generalized learning. Even if both dimensions are important, the first one often creates expectations regarding immediate affects and changes whereas learning as a process broadens understanding and therefore can render a slower phase that is more sustainable. This also applies to various processes in which the PD is initiated and occurs. If the provision occurs at the end of a decision process, the aim might be solely how to learn and understand what the new reform is based on and how to execute the reform rather than a discussion about why the learning is important, what the alternatives are, and how the principal and school can meet their goals.
Moreover, the results show that it is common for PD to connect locally with approved improvement projects. Most superintendents confirm that they use what national agencies and universities offer. Overall, it seems that PD is primarily based on what is externally offered and on the school system's current needs. This implies that principals depend on their superintendents and on national offers to build a common knowledge base and to develop as professionals. If most superintendents belong to ideal type I, where the focus is on supporting knowledge that is based on best practices and implementing already approved models and practices, principals risk becoming dependent on others and less able to make decisions based on their organizational context. However, creating opportunities to combine theory and practice, as in ideal type III, can strengthen independence and the possibility of novel practice.
Most LEAs devote time and resources to PD. In our interviews, there are limited signs of a more strategic governance perspective. Instead, the superintendents' experiences and understanding of learning, combined with current national initiatives and trends, influence what is offered. In the long term, this can impact what is considered important in the principal's professional role, a situation that can undermine the value of theoretical competence and individual interpretation (Sandberg and Targama, 2013). At some point, this will also restrict principals' professional knowledge.

Conclusion
PD for principals has in recent years been emphasized because it is considered crucial in handling a constantly changing environment. This visualizes the importance of making conscious decisions about how to arrange and support PD for principals in school organizations. PD provides not only new insights but also a governance tool with various aims and views of what is important to learn, which affect expectations regarding principals' JEA work and ability. The categorization into ideal types shows variation between adaptive learning and development learning as well as learning directed to individual actors and principals as a group. Increased mobility and high expectations regarding improvement require conscious and systematic work to provide adequate PD simultaneously because our data indicate that PD for principals relates to current suggestions and superintendents' individual experiences, which can make the principals obedient implementers rather than independent decision-makers. From a policy and practice perspective, the findings highlight the importance of consciousness regarding the aims of principals PD to choose the right content and working forms. The theoretical contribution of the paper is to expand the knowledge on how various forms of PD has underlaying aims and understanding of principals' ability and role.
Regarding advantages and limitations, it is initially important to emphasize that this study of PD for principals is conducted from the perspective of superintendents. Although Swedish superintendents are responsible for providing principals with adequate PD, principals can take their own initiative, especially if their ambitions for learning exceed those of their superiors. Therefore, it is necessary to examine PD for principals also from principals' perspectives. Moreover, as professional knowledge develops over time, it is reasonable to believe that principals with several years in the profession have needs different from those of newly appointed principals. By determining what senior principals value in their PD, researchers may find aspects associated with PD that make principals stay in their current positions. Based on the extensive principal turnover, this would be valuable knowledge. In addition, we based this paper on a limited sample of respondents; context matters in leadership, and larger studies can reveal a more comprehensive picture. However, this and future studies can contribute to more strategic PD and support superintendents' and LEAs' strategies in building infrastructures to foster more relevant PD at various school system levels.
Notes 1. In large municipalities, superintendents often have deputy superintendents with certain responsibilities, e.g. preschool, compulsory school, and upper secondary school. 2. In the following text, all interviewees are referred to as superintendents.