The Cross-School Teacher Team as a Site for Learning

ABSTRACT In policy and in initiatives from the Swedish National Agency for Education for enhancing professional development, there is currently a strong emphasis on teachers´ collaborative professional development. As previous research suggests that teachers may need to engage in various types of collaboration for learning, extended knowledge on different teacher teams for collaboration is needed. In this study, a Cross-School Teacher Team (CSTT) of teachers who shared pedagogical interests but worked in different municipalities was followed through observations, interviews, text-collections of logbook-reflections and digital communication. To examine what constitutes the CSTT as a site for learning and how this is shaped by as well as shaping the enabling and constraining arrangements, the empirical material was analysed through the lens of the theory of practice architectures. The results show how a CSTT can provide a complement to local teams for work integrated learning through diverse perspectives, a shared focus on specific and professional issues and a safe space for sharing and reflection. Internal and external transparency worked as a catalyst both for processes of reflection and transformation of local arrangements and practices. Knowledge of the practice architectures of the CSTT offer tools to elaborate upon alternative or complementary spaces for collaboration


Introduction
The importance of teachers' continuous learning and professional development (PD) can hardly be underestimated. Previous research suggests that teachers may need to engage in various types of collaboration for learning (Kemmis, Wilkinsson, Edward-Growes et al, 2014;Levine and Marcus, 2010). Therefore, extended knowledge of various forms of collaborative PD (CPD) is valuable in order to develop collaborative spaces and opportunities for teachers to interact to enable teachers' continuous PD. This study contributes to such knowledge by exploring the Cross-School Teacher Team (CSTT) as a way to develop teaching. In this team teachers with a shared pedagogical interest but located in different parts of the country interacted and collaborated in physical meetings as well as online. The CSTT hereby represents a different form of CPD than peer learning in the local teams Swedish teachers are usually organised in.
The overall aim of this study is to develop an understanding of how the CSTT works as a site for PD. To do so, a CSTT of teachers participating in the larger research project DILS (Digital Arenas in Literacy Practices in Early Primary School) and the work within this group was followed for more than two years. Building on the theory of practice architectures (Kemmis, Wilkinsson, Edward-Growes et al, 2014), the study is neither limited to individuals nor organisations but deals with both by addressing issues of the interrelations between these. The CSTT is scrutinised from a notion of practices as connected to other practices, shaped not only by the participants' knowledge and actions but by conditions external to them (Kemmis & Grootenboer, 2008: Kemmis, Edwards-Groves, Wilkinson & Hardy, 2012. The study is structured by the following questions: What constitutes the CSTT practice as a site for professional development? How is this shaped by as well as shaping the enabling and constraining arrangements? Hereby the study addresses the gap identified in the recent research review on teacher collaboration by Vangreiken, Dochy, Raes et.al (2015) who noted a lack of empirical research on the characteristics of teacher teams. By investigating CSTT this study provides insights into alternative and ICT-supported CPD and thereby adding to previous research on school-based teacher teams. An analysis of what happens within the CSTT as well as how this interplays with structure or arrangements, provides knowledge of essential aspects to consider in developing collaborative spaces and opportunities for teachers to interact.

Continuing professional development and collaborationa brief overview
In all professions, continuous learning throughout professional life is expected by the professionals themselves as well as the stakeholders. Teachers are no exceptions in this matter. Hargreaves (2003:24-25) argues that teachers in the knowledge society must learn to teach in ways they were not taught themselves, engaging in continuous professional learning and collaboration with colleagues. PD is here described as both an individual obligation and an institutional right (ibid). In Sweden, these obligations are even legally stipulated. Laws and directives regulating the Swedish educational system require educators to engage in organisational development, and principals must ensure opportunities for PD for all teachers (SFS 2010:800;Skolverket, 2011). In policy as well as in PD-initiatives from the Swedish National Agency for Education there is currently a strong emphasis on CPD which calls for extended research.
Teacher collaboration for exchanging ideas, discussion and feedback is generally considered to create a powerful environment for teachers' PD (Meirink, Meijer & Verloop, 2007;Lohman, 2005;Little, 2002). Muijs and Harris (2006) even argue that arrangements providing communicative spaces for teachers to learn from each other are necessary for teacher development in schools to take place. One way to enable teacher CPD is organising local teacher teams. Over the last decades, schools organising their staff into teacher teams have become well established in the Swedish schools (Blossing & Ekholm, 2008). Yet reports from The Swedish National Agency for Education show how teachers' work, in many respects, is isolated (Skolverket, 2013;2015). On the one hand, there are expectations at the policy level for teachers to collaborate, engage in reflective discussions and learn in the local teacher teams (Knutas, 2008;Ohlsson, 2004). On the other hand, organising teams does not in itself lead to collaboration and PD (Hargreaves, 1998). In a study of three local teacher teams, Ohlsson (2004) found that during meetings pedagogical issues were given considerably less attention than practical, organisational and administrative matters. Questions about the teachers' own actions were avoided and Ohlsson as well as Stedt (2013) highlights the fear of criticism from colleagues as a restraining factor for collaboration. The results of these studies show the importance of noticing relational aspects in researching teacher teams and CPD as is done in this study. Furthermore, the relational aspects are studied in a different context than the local team.
Collaboration has been put forward as essential to PD by enabling the exchange of experiences and reflective conversations (Stedt, 2013;Tjernberg, 2013). Yet collaboration is neither uncomplicated nor self-evident as a way to promote learning. Different types of collaboration generate diverse opportunities for learning according to Levine and Marcus (2010) who suggest that teachers may need to engage in various types of collaborative activities. While Levine and Marcus primarily discuss teacher teams within schools, other studies imply that interdisciplinary or cross-school constellations offer complementing opportunities for learning. For example, Kemmis et al (2014:135-136) describe how a strong whole-school approach may foster staff cohesion and practice development but concurrently constrain wider access to external contacts. Hereby teacher agency and PD becomes restricted for some teachers, who value access to diverse perspectives. Also Knutas (2008) problematizes local teacher teams showing how teachers' agency, as well as variation and diversity, can be restricted by organisation and management. The benefits of complementing local and mandatory teams with participation in voluntary teams are further highlighted by Meirink, Imats, Meijer et al (2010) in a study of interdisciplinary teams. The voluntary team provided diversity in its constellation, which enhanced learning. Moreover, these teams could focus on instructional innovation rather than practical and organisational issues (ibid.).
Previous research shows that alternative or complementary forms of collaboration, as the voluntary and cross-school collaboration presented in this article, are valuable for teachers. Further knowledge of various forms of collaborative practices is useful in order to enable and support teachers' PD. Yet, while different types of constellations as interdisciplinary teams (Meirink et al, 2010) have been studied, all studies reviewed above concern local teams within schools. The present study hereby adds to previous research by shedding light on the characteristics, including relational aspects, of a CSTT.
Moreover, reviews of research on teacher learning and collaboration show a need for more research with a holistic interest in complex systems (Opfer & Pedder, 2011) and more empirical research on how teachers learn in collaborative practices (Borko, 2004). The importance of a view not restricted to either the individual nor system, but connecting the social to the material, is highlighted both in complex theoretical (Fenwick;, sociomaterial perspectives (Fenwick, Jensen & Nerland, 2012;Mäkitalo, 2012) and in perspectives based in practice theory (Reich & Hager, 2014). In recent research on educational change and CPD-initiatives, such as action research, the theory of practice architectures (Kemmis & Grootenboer, 2008) has been used to grasp the complexities of educational practice (e.g, Aspfors, 2012;Aspfors, Pörn, Forsman el al, 2015;Kemmis et.al 2014;Langelotz, 2014). By studying not just practice itself but how practice and its outcomes are intertwined with and affected by existing conditions, these studies contribute to an understanding of how educational development is enabled and constrained.
From this theoretical standpoint, Aspfors et al (2015) highlight the researchers' role in CPD practices in the intersection between university and school. Their results show how the researcher continuously must negotiate with the practitioners as well as at an organisational level to support CPD activities. They conclude that further research is needed on how to open up and support communicative spaces for participants in CPD contexts (Aspfors et al, 2015:413). While the theory of practice architecture has been used in various studies on CPD, it has not previously been used for studies on teams interacting both in physical meetings and digital media. By both examining the characteristics of the CSTT and its enabling and constraining arrangements, the present study contributes to knowledge of how to support such a communicative space for CPD. As digital development has opened up possibilities for interaction that do not require geographical proximity, the specific context of the study reported here could provide fruitful insights with the purpose to develop and support CPD beyond the local.
Theoretical frame -Practice theory and practice architecture The theory of practice architecture developed by Kemmis and Grootenboer (2008) provides analytical tools to include attention to the holistic interest in complex systems highlighted as desirable in previous research. This theoretical perspective offers both focus on what happens within practices and pays attention to practices as connected to, and mutually affecting, external conditions and other practices (Kemmis, 2010). In this study, practice in all its complexity of discursive, material and relational aspects both within the practice in question or brought to the site is in focus, whereas other theories rather stress on either actions, language or systems (Schatzki, Knorr, Cetina & von Savigny, 2001;Raelin 2008;Reich & Hager, 2014).
According to the theory of practice architecture, practice is a complex form of socially established cooperative human activity composed by people's sayings (their understanding, what they say and think), doings (actions and activities) and relatings (how people relate to one another). Sayings, doings, and relatings are bundled together in distinctive projects, which are the purpose and meaning of the practice (Kemmis and Grootenboer, 2008;Kemmis et al, 2014). These concepts are used to investigate what constitutes the CSTT practice as a site for learning. What the teachers regard as the purpose of participating in the CSTT is the distinctive project that gives meaning to the sayings, doings and relatings. Kemmis et al describes sayings as what the involved say and think in and about practice and doings as behaviour, activity as well as work. How the teachers understand and talk about, as well as within, the CSTT as a site for learning is thereby considered as sayings in this study. The doings of the CSTT are the actions carried out individually or collectively by the group in the medium of work and the relatings are how they relate to each other in the medium of power.
The sayings,doings and relatings are held in place by practice architectures constituted in the cultural-discursive arrangements (e.g. language and ideas), materialeconomic arrangements (e.g. objects, spatial arrangements) and social-political arrangements (relations), already existing or brought to a site, that enable or constrain practice (Kemmis et al, 2014). Transforming practices means changing the way we talk and think, what we do and how we relate to one another, but also changing the practice architectures (Kemmis et al, 2015). Aspects of practice as well as conditions for practice are connected and both shape and are shaped by one another in a complex dialectical interplay (Kemmis & Grootenboer, 2008;Kemmis, 2010;Nicolini, 2013). This means that the relation between agency and structure is at interplay where structure can enable and constrain how practice is carried forward and carried out but that these structures both shape and are shaped by human action and interaction (See . In material ways, practice architectures have a sort of agency in the shaping of practices (Kemmis et al, 2015 p.97).
In this study the CSTT was facilitated by the researchers but the participating teachers worked in different schools. Both internal arrangements provided by the facilitators and external arrangements at respective school are hereby part of the practice architectures and analysed in terms of how they shape, and are shaped by, the CSTT as a site for learning.

The context and conduct of the study
This study was conducted within the frame of the DILS-project where three early primary school classes and their teachers in three public schools in Sweden were followed for three years. The participating teachers were brought together by the researchers and formed the CSTT analysed in this study. In the following section some background information about the context of the study in terms of description of the participants, the organisation of the CSTT and the researchers' roles in the CSTT is given followed by methodological and ethical considerations.

Participants
The selection of participants for the DILS project was based on a wish to engage teachers from different areas with previous experience of, and interest in using technology in their literacy teaching. They were chosen, but participated voluntarily which makes the group different from the local teacher-teams. The differences in characteristics of the schools and the teachers' instructional approach are presented below.
South School, Anita: This school is located in a socio-economically privileged area in a large city. Anita's approach to literacy teaching is inspired by McCormick Calkins (1994;2001) and the Teachers' College Reading and Writing Project. Although Anita is a part of a local teacher team she lacks colleagues to collaborate with on matters related to the subject of literacy and digital tools. Also other conditions such as a change of principal and other circumstances causing temporary disturbances in this team restrained possibilities for local collaboration and support.
North School, Maria: This school is located in a socio-economically underprivileged area in a small town. In contrast to the other schools in this study this is a multicultural school, and Swedish is a second language for all the students in Maria's class. Maria's literacy teaching is inspired by the Writing To Read on Computer-method (Trageton, 2003), a method used by all literacy teachers from grade 1-3 in this school. Maria is both part of a local teacher team that work together closely and other networks in the district. For example, the principal and the ICT-facilitator in this district has initiated cross-school collaboration, professional development initiatives and meetings with teachers from the schools nearby.
West School, Lina and Hanna: This newly initiated school is located in a socioeconomically privileged area in a small town. From the start this school has focused on working with digital development and recruited teachers with an explicit interest for digital tools and school development. The literacy teaching here is not limited to one specific method but flexible. Hanna and Lina share responsibility for this class and work closely together with one another and with their local teacher team. As Maria, they are also both active in other networks in the district.

Context of the CSTT and researchers' roles
As described earlier, the CSTT in this study was brought together and studied within the DILS project. For this reason a brief description of the phases and activities of this project and the researchers' role is provided. The researchers functioned as facilitators for the CSTT by providing support through: • assembling the teachers in workshop meetings • planning the content of each workshop together with the teachers • providing texts for shared reading (research articles, book chapters et cetera) on topics relevant to what the teachers wanted to discuss or work with • setting up shared digital spaces for reflection and communication between physical meetings (a closed Wiki-page and a closed Facebook-group) • funding the teachers through the research project to give them time to work with the project and attend meetings.
This support is to be considered part of the practice architecture of the CSTT, as arrangements brought to the site by the researchers, and as such they will be discussed in the analysis. Throughout the project, researchers recurrently made classroom-observations and interviews with the teachers and examples from these classroom observations were discussed during the workshops. A brief summary of the activities in the DILS-project and the CSTT is here presented to give the reader an overview of the context of the CSTT . The overall activities within the DILS-project are described in three phases: First, an initial phase of "mapping the terrain" during the first year. After researchers' first classroom observations and interviews, the teachers were introduced to one another, met in workshops where they visited each other's schools, discussed teaching and made plans for collaborative projects. Between meetings the teachers wrote logbook reflections and communicated in the Wiki-page. In the second phase of "class and teacher collaboration" during the second year, the teachers developed collaborative projects that they planned and evaluated together during workshops and their shared digital spaces. And finally, there was a third phase of "teacher collaboration" during the last months of the DILS-project, where the teachers worked with collegial assessment of multimodal texts and evaluated their experiences in the CSTT.

Data production and analytic process
The object of study is PD through teacher collaboration in a CSTT and practice is the unit of analysis. CPD in the CSTT is in this study regarded as a certain practice, occurring in the present but oriented to the future in response to the past, which is brought into being by people acting collectively (Cf. Kemmis et al., 2014 p.32-39). Like any social phenomenon, practice is multidimensional. To grasp more than one of these dimensions, an ethnographic approach to data production with multiple data sources has been adopted (e.g., Mason, 2002). Such a methodological approach was necessary in order to capture the semantic, material and relational dimensions of practice related to context. Throughout the two and a half years that the CSTT was active, the researchers recurrently conducted semi-structured interviews with the teachers. From nine of these interviews conducted in phases II-III, the parts where collaboration and learning in the CSTT were discussed, were extracted for analysis. The 7 planned workshops (1-2 days at a time) in which the researchers participated were documented by notes and audio recordings. Furthermore, teachers' written logbook-reflections and communication with one another in the shared digital spaces in Wiki-space and their closed Facebookgroup were collected. The empirical material hereby provides information of both the participants' actions and interactions in the CSTT and their understandings of the CSTT as a site for learning, what they did and their relations. On the one hand data produced in interviews and workshop-conversations provide insights to all aspects including those not possible to extract from observations, as their understandings. On the other hand observations of what they actually did and how they related to one another in their interactions provided information that was not filtered through the teachers' perceptions of these interactions.
Initially, an inductive approach was taken where the research question and the choice to illuminate the collaborative practice as a site for learning through the lens of practice architecture evolved in the process, starting with an initial thematic analysis of the interviews. Themes occurring in the teachers' statements in these interviews were reflections of the collaboration, their own learning, students' learning and prerequisites enabling or constraining collaboration and learning. The final research questions about what constitutes CSTT practice and how this is shaped by as well as shaping arrangements enabling and constraining practice were formulated as a result of this initial preanalysis, the gap in research identified in previous research, and the choice of a theoretical lens. Hereby the approach evolved from an inductive analysis towards a theory-driven approach.
The choice to use practice architectures as the theoretical framework provided the analytical tools for the following steps of the analysis. The first step consisted of using the concepts of sayings, doings and relatings to examine the characteristics of the CSTT. Sayings about as well as within the CSTT were scrutinised both in utterances and content in interviews and workshop-discussions but restricted to sayings about and within the CSTT as a site for learning. Doings and relatings were made visible for analysis both in observations of physical and digital interactions and in the participants' utterances in interviews. As the work and activities of the CSTT included elements like planning and evaluating teaching projects in the classes in order to develop literacy instruction, doings undertaken in the CSTT sometimes overlap doings in other practices as teaching. Doings are therefore sometimes carried out in several steps, placed in various settings and yet part of doings within the CSTT. For example, as they planned, evaluated and reflected upon their teaching together, doings in the classrooms, in the CSTT-meetings and sometimes even in their local teams were all part of the work of the CSTT. The second step, where the practice architectures were analysed, consisted of trailing predispositions and practices and how they relate to each other. The presentation of the findings are organised in two sections corresponding to these two steps.

Ethical considerations
All participants including the pupils' parents have been informed and have given their consent to participate in the study. During the process the researchers have both recurrently asked for consent to record observations and presented data and preliminary results to the teachers. The informed consent was hereby handled as an on-going process in line with the guidelines for participating observations from University of Toronto (SSH REB, 2005). Furthermore, all names of informants and schools used here are fictitious. Research ethical principles and guidelines concerning information, participation, consent and confidentiality have been considered (Hermerén, 2011).

The practice architectures of a CSTT as a site for learning
The results of the study are presented in two parts and discussed in the following sections. The first section concerns what constitutes the CSTT practice as a site for learning and the second section deals with how this was shaping as well as shaped by enabling and constraining arrangements.

What constitutes practice and the distinctive project of the CSTT
The sayings, doings and relatings, analysed in observations, interviews, logbook reflections and written communication, are here presented separately. Yet, as sayings, doings and relatings are bundled together in the distinctive project they sometimes overlap. For example, the concept of sharing is referred to in both the section on sayings (as their understanding of the CSTT as a site for learning) and doings (as a kind of action within or connected to the CSTT).
The distinctive project in the CSTT, the purpose as articulated by the participants, was PD. The purpose could also be described as two-fold; breaking isolation or extending access to knowledge and diverse perspectives for professional development depending on the individual teacher's needs. The CSTT was either described as a substitute for a lack of colleagues with a shared interest to collaborate with locally (South School) or as a complement to local groups (North and West School). For Anita the CSTT provided the colleagues to collaborate with and discuss technology-mediated literacy teaching, which she lacked at South School. The complementary function concerned aspects of the CSTT that were different from their local teams. These aspects will be further elaborated in the following sections as they unfold in sayings, doings and relatings.
Talk about and within the CSTT as a site for learning -Sayings The teachers described the CSTT they participate in as PD where they said they learned through articulating their teaching and their thoughts, sharing knowledge and experiences, and connecting theory to their teaching experiences. All these components have been highlighted as essential for learning in previous research (e.g., Meirink et al, 2007;Lohman, 2005;Little, 2002;Stedt, 2013). Learning was articulated as a result of two categories of activities in the semantic space: As sharing knowledge and perspectives, and as distancing in terms of making one's teaching the object of reflection. The sharing helped distancing by articulating and through comparisons and perspective-taking. This was something the teachers described as desirable or something they felt a need to continue once they had this experience.
In terms of the content of the discussions, the CSTT was focused purely on pedagogical, rather than practical, issues. This was not just visible in the observations but highlighted in interviews and conversations within the CSTT as a difference between this team and their local teams. Likewise, they described the content of their discussions as providing more diverse knowledge and perspectives than in their local teams where the members are part of the same school-culture (e.g., Kemmis et al, 2014;Meirink et al, 2010).

The actions and activities of the CSTT -Doings
The teachers interacted in planned workshops and shared digital spaces. During the workshops the teachers visited each other's schools, shared and discussed their experiences, ideas and reflections on teaching and digital approaches to literacy and discussed literature. Furthermore, they planned, conducted and evaluated collaboration projects between the classes and the teachers. Between the face-to-face meetings they used the digital spaces of Wiki and Facebook for discussion, the sharing of material, and individual logbook-reflections.
Through face-to-face discussions, reading each other's logbook-reflections and conducting collaboration projects, the group shared and learned from good examples. They also provided different perspectives on matters discussed which expanded the accessible recourses compared to local teams. The Wiki-page was used as a documentation of what had happened and their thoughts throughout the process, which they both used as a way to lighten their workload and returned to for inspiration. While local teams interact in physical meetings, some of the actions and interactions of the CSTT took place in this digital writing and as such were accessible and visible regardless of time and space. Documentation and reflective writing might be rare in local teams but were a recurrent and ongoing activity in the CSTT.
Furthermore, sharing reached outside the CSTT as the teachers, either on their own initiative or by request from school leaders, shared their experiences and insights from the CSTT with colleagues. The participants' experiences of finding others to be inspired by the knowledge and experiences shared contributed to processes where they saw their own practice in a new light. For example, the teachers at West School repeatedly noted in interviews and observations that they, through the conversations and discussions, had realised that their pedagogical approach in fact was not as self-evident or insignificant as they had thought. Hereby the CSTT practice connected to and interrelated with the local practices.
The articulation of teaching and the transparency providing the CSTT access to one another's teaching, knowledge and experiences led to the teachers making comparisons where they noticed things, named what they saw and reflected upon it. A process Smith (2008: 77) describes as "noticing, naming and reframing", where things noticed are brought forward for closer examination. Participants of the CSTT, contrary to their local teams, represented a spectrum of diverse experiences, student groups and school cultures. As a result of these diversities, comparisons and contrasts were highlighted, enabling the process of noticing and reframing. For example, through comparing teaching at North School with mainly second language learners with the other classes, the teachers re-evaluated their teaching concerning the level of demands on the students. While the teacher at North School asked herself whether she had lowered her expectations on the students in areas other than language development, the teachers at West School discovered they needed to pay more attention to students' preunderstanding of words. In this way, the interaction between the teachers from various backgrounds and contexts influencing each other brought out the expertise of the teachers (e.g., Mak & Pun, 2015).
The Facebook-group and Wiki-page were used for different purposes. While the Wikipage was used for discussion, sharing material, and individual logbook-reflections, the Facebook-group was used to contact each other on practical matters that required a quick answer. When writing logbook-reflections the teachers narrated events from their teaching, reflected upon them, shared thoughts, and posed questions and feedback to each other. They also uploaded examples of students' work, examples of their own teaching-material, and gave each other suggestions of further reading of pedagogical literature. Logbooks, observations as well as interviews showed examples of how the teachers made use of the shared material and ideas in their teaching. While they all made use of information shared in this arena and expressed in interviews that they valued the opportunity to reflect, there were differences in how often they used the Wiki-page for writing reflections. Maria turned out to be an avid writer, posting more text and more often than the others, and even continued writing after the research project was finished. The other teachers' writing was more sporadic but intensified during collaboration projects.
Creating a "safe space" within the CSTT -Relatings Relational aspects were expressed in interviews as crucial for two reasons. First, building a relation by getting to know one another and visit each other's schools was described as necessary for them to discover that they needed the knowledge they could gain from collaborating. Secondly, the collaboration climate within the CSTT was referred to as supportive and without prestige, characterised by interdependency and mutual responsibility. These relational aspects were emphasised by the participants as prerequisites for, and benefits of the CSTT. Maria stated that starting up with new collaboration partners requires support to become a group and Anita made a similar comment saying: "We have come to know one another. To just get a name and an e-mail address wouldn´t really be the same thing".
To share one's experiences and thoughts and make them the object of conversation and reflection in the group is in a way exposing oneself to criticism. However, the teachers' statements in interviews as well as their digital interaction and interaction in observed workshops, shows a supportive way of relating lacking prestige. The permissive climate is manifested in Maria's statement from the last workshop-discussions in the excerpt below as a view of practice as context dependent where there is not one "right way" that is generally valid or adequate: A: It was a very relaxed atmosphere. There is no prestige in this group. That is very nice.
H,L,M: (Yes… Mmmm…yeah…) @< talking simultaneously> M: What is right for you doesn´t have to be right for me and some things I can take on because it would suit my group The importance of creating a safe space for learning, reflection and risk-taking has previously been highlighted as an important issue in various contexts such as organisational learning and e-learning (Falconer, 2006) and mentoring practices (Langelotz, 2013;Smith, 2007). However, these relatings, especially as manifested in the written communication in Wiki and Facebook, were based on noticing and acknowledging the positive rather than to assume a critical perspective as in what Handal (1999; refers to as "critical friends". The conversations in the CSTT were characterised by the members acting curious and interested, asking questions for clarification. When challenging someone's interpretation or view they did this in an argumentative way without the disciplining components noticed in Langelotz's (2013) study of peer-group mentoring. Langelotz problematises the view of collective PD as unequivocally "good" by illuminating the risk for stigmatisation of individuals by collective disciplining through making judgments in the mentoring process. Critical reflection and discussion in the CSTT were instead devoted to comparisons and concerned neither judgments nor suggestions of how the others "should" teach. The critical reflection, emphasised by Handal (2007) as crucial to avoiding stagnation and reproduction in learning communities, concerned one's own, rather than one another's, teaching. This self-reflective re-evaluation was however evoked by the comparisons of different perspectives and interpretations that were displayed.
The relations within the CSTT were furthermore characterised by mutual responsibility andinterdependency where all participants were involved as illustrated in this excerpt from a workshop conversation where the relational aspects of the CSTT were discussed: …that everybody contributes. No one sits and just listens, but you contribute. Much depends upon that. There´s no point if one half [of the group] sits there and just listens. (Hanna) This way of relating to one another displays the teachers' shared interest and interdependency. To be part of a group with shared projects to complete, where they all needed the contributions of the others, was also highlighted by Anita as an encouragement to bring the teaching projects to completion.
While the teachers in their local teams shared employers, workplaces and pupils with their colleagues, the CSTT shared nothing but the interest and the purpose of their collaboration with their fellow members. That shaped the relations into something other than what characterises the local team. First, their interest focused on their shared purpose in this group. Secondly, they were not exposed to the potential positioning in the group that might occur when you need to negotiate your wage and working conditions at the workplace. Pedagogical issues and development of literacy teaching in a digital classroom were focused upon in the discussions, not as a result of researchers' guidance but because the shared interest for these questions was what united them.

Enabling and constraining arrangementsthe practice architectures of the CSTT
The practice architectures holding the sayings, doings and relatings of the CSTT in place, were constituted in the set-ups facilitated by the researchers and arrangements outside the team itself, such as the members' workplaces. The facilitations brought to the site by the researchers were the same for all the group members. Yet different practice architectures were present within the practice of the CSTT as a result of differences in the workplace conditions for the teachers.
The spectrum of diverse experiences, student groups and school cultures represented in the CSTT, provided more diverse knowledge and perspectives than the local teams. Even if the members in local teams are not all alike, the participants in this study all described a certain uniformity in local teams due to cultural-discursive and material-economic arrangements as shared in-service training and local policy with specific goals.
Collaboration between these geographically widespread schools was enabled through material-economic arrangements, such as access to shared digital spaces and work-shops facilitated by the researchers. These interactional spaces enabled contact, documentation as well as sharing between meetings and relation-building work necessary to create a safe space.
The cultural-discursive arrangements (i.e. language and ideas) brought to the site first and foremost concerned the knowledge, language and ways of discussing pedagogical issues that were made available. In evaluating the CSTT, the teachers expressed that they not only had been given opportunities for reflection and discussion but learnt how to do so. As Anita expressed it, "I have learned to talk more about what I do and put it into words. That is something you´ve guided us into further somehow". The shared reading of research articles was put forward as helpful to theorise their teaching experiences as well as to further develop the ability to articulate by providing vocabulary and theoretical concepts. Hereby the teachers acquired tools for the process Smith (2008:76-79) describes as "noticing, naming and reframing" constructing "narratives" essential to reflective and inquiry-based learning. This is a process that requires putting one's concerns, thoughts or experiences into language, recontextualising one's understanding, and making conceptual connections to bring them forth for examination and theorisation.
The importance of language in meaning-making was recurrently present in the teachers' statements about the CSTT as a site for learning. Articulating their thoughts and experiences was not something they were used to doing. As Anita explained, "When you get to put what you do and why into words. We´re not used to doing that. We just teach". Hanna and Lina from West School compared the conversations within the CSTT with the conversations they are used to in relation to the pupils' parents: H: We´re used to becoming defensive, in relation to parents and such, but it´s not the same thing.
L: Because it´s not the same questions you´re asked then H: No, when they question it´s not because they´re interested in what you do, but it feels more like they're criticising then The climate in the CSTT created a safe space for sharing. Questions were formulated in a way that signalled interest and a wish for understanding. They were put forward as both shaping how to relate to one another and modelling how one could put one's thoughts and experiences into words. What shaped the sayings and enabled the doings and relatings were the language and ways of discussing made available to them in the semantic as well as in the social space: group-conversations, shared reading and the ways the researchers asked their questions.
Other cultural-discursive arrangements brought to the site concerned what understandings of the CSTT that were communicated to the members by both the researchers and the principals. The teachers highlighted the fact that this CSTT was part of a research project where researchers took interest in their educational practice and development, making them feel that what they did was important. In other words, the interest and attention from researchers signified an enhancing of the importance of their work. Similarly, an understanding of the CSTT communicated by the principals, as a question of school development in the North and West Schools rather than just of individual PD as in South School, had the same effect. In addition, the idea of the CSTT as a question of school development led to transparency and further sharing outside the CSTT interrelating teaching and PD at different schools and the CSTT. Maria described how the fact that she was asked by the principal to share and inform her colleagues of her experiences made her "…feel that it´s more for real. It´s like what we talk about concerning the children… that you get to show it to someone". She emphasised the importance of support where one is given a mandate as well as legitimacy by the principal and colleagues. While such a mandate and legitimacy was present at both North and West Schools, Anita at South School lacked both collaboration partners and a mandate to contribute to the local school development. In her case absence of a general interest and focus on digital tools in literacy instruction, in combination with a change of school leaders and a turbulent situation in connection with this were constraining factors.
The understanding of the CSTT communicated by the principals of the schools as a cultural-discursive arrangement, interrelated with material-economic and socialpolitical arrangements. These varied, both between the schools and through the project. At North School, Maria experienced a shift when the researchers intervened by contacting the principals and invited them to visit some meetings to see the work that was carried out there. As a result, Maria had reduced teaching hours and was commissioned to share her experiences from the project with her colleagues, making her participation in the CSTT of interest for the school's development. The researchers here acted as what Aspfors, el al (2015) refer to as a negotiator, contributing to the processes where arrangements are negotiated but with the important addition of making the work of the CSTT transparent. Maria described the difficulties for teachers themselves to argue for time to reflect and write when they knew that it meant that their colleagues would have to do more teaching hours. Support from the facilitators as negotiators as well as transparency of the CSTT helped her legitimise making reflection a priority. In other words, arrangements brought to the site by facilitators affected the local practice architectures both in the semantic, material and relational space.

Discussion
The results show that what constitutes CSTT practice as a site for professional development are ways of talking, acting and relating that are different from those in local teams. The different characteristics of the CSTT offer an alternative or complementary form of CPD by bringing forth diverse perspectives and a shared focus on specific and professional, rather than practical, issues as well as a safe space for sharing and reflection. Hereby the results illustrate how a CSTT can provide a complement to local teams for work integrated learning and educational development.
Furthermore, the results show how these sayings, doings and relatings were shaped by as well as shaping the enabling and constraining arrangements. Arrangements in the semantic, material and relational space as support, time et cetera found here are well documented in previous research on local teams. Through the lens of practice architecture the analysis however also uncovered the specific characteristics of the CSTT in terms of diversity, interdependency, visibility and transparency. All these characteristics illuminate how different practices as local education and teaching, local PD and organisation all became interconnected and interrelated through the work of the CSTT. The findings highlight that practices external to the CSTT might be prefigured but not fixed (e.g., Kemmis et al, 2014) as actions in the CSTT proved to affect changes at the managerial level in cultural-discursive, material-economic as well as socio-political arrangements. A key finding here is how internal and external transparency worked as a catalyst both for processes of reflection and transformation of local arrangements and practices.
Access to shared digital spaces not only enabled participants to continue to interact between the face-to-face-meetings but worked also as a tool for transparency, documentation and reflection. Technology development and large-scale investments in digital tools in the schools provide access to digital spaces that enable contact and interaction that do not require geographical proximity. Yet, the findings of this study show that the asynchronous interactions needed to be combined with face-to-face meetings in order to achieve the indepth discussions and building of relations that enabled sharing and distancing. Hence, digital tools and spaces could be helpful in creating opportunities for teachers to collaborate and learn, but the use of such spaces would benefit from blending digital interaction and physical meetings as has previously been highlighted by Matzat (2013). To conclude, I argue that a meta-practice as a CSTT, connecting several local schools, teams, teachers and classes due to its characteristics has potentials to enable further critical reflection on both teaching and organisational matters. Knowledge of the practice architectures of the CSTT and how it works as a site for PD offer tools to elaborate upon alternative or complementary forms of, and spaces for, collaboration to promote work integrated learning. To develop CSTT for CPD, a raised awareness of the interrelation of practices is hereby suggested to elaborate on internal as well as external transparency to prompt reflection through comparisons and make the CPD a matter of importance to the organisation as a whole.

Funding
This work was supported by the Marcus and Amalia Wallenberg Foundation: [Grant Number MAW 2011.058].