SCHOOL STAFF-CENTERED SCHOOL DEVELOPMENT BY COMMUNICATIVE ACTION: WORKING METHODS FOR CREATING COLLECTIVE RESPONSIBILITY-FROM THE IDEA TO ACTION

School development is often seen as a concerted (re-)action to educational policies, curriculum development, and change in education laws or regulations, and sometimes, as stakeholders’ reactions to low school performance. Generally, school development incorporates organizational, managerial, and educational activities and measures. This is done to adapt to the new situation to achieve the desired changes and goals initiated by a given curriculum reform. In contrast, this paper focuses on school staff–driven development: It describes how teachers together with paraprofessionals contribute to school-development. Using collective responsibility creating working methods of communication, the enactment of staff’s self-chosen measures will be possible. This school staff–driven school development approach makes use of the Dialogue Café and the Reflection Cycle. The combined use of these working methods is seen as appropriate support for school staff–centered development of the School-In project based on the idea of collective responsibility. The paper presents by the application of a qualitative content and visualization analysis, how the communicative action according Habermas took place. This is the process from the individual participant’s ideas, through individual and group reflections to finally formulated measures that result in the school’s staff joint actions.


INTRODUCTION
School reforms intend to foster school development and improvement, but reforms take often different effect (OECD 2015;Cuban, 1990Cuban, , 2020. Classrooms, schools, and districts do change but they incorporate just some reforms to remain stable (Cuban 2020: 670). Research has shown that participative school development contributes to collective capacity building (Hargreaves 2019;Harris 2011) which is important for professional learning communities (PLCs). Even if participative school development supported by the paradigm of PLCs has become popular, little research (Korthagen andVasalos, 2008, 2005;Korthagen, Greene and Kim, 2013) is conducted on democratic working methods. Besides, starting school development by generating appropriate ideas, activities and measures one final tipping point is its implementation. For instance, the enactment of specific developed and useful measures needs continuous guidance, tracing, and support as a final stage. In Korthagen, Greene and Kim (2013: 115) the 'within-in' approach stresses as one implementation characteristic this kind of 'transfer enhancement'. Therefore, this paper will contribute to fill this research gap and report on the whole working process of staff-driven school development. In specific, the paper analyzes and presents how school staff contributed to school development by using working methods of dialogue and mixed group discussions, and by enacting school staff's self-chosen measures (School-In Consortium, 2016;Hillen, in preparation;Dalehefte and Hillen, in preparation). The study makes use of the constant comparative method and its graphically visualized qualitative analysis conducted by using NVivo (Glaser and Strauss, 1967). The paper applies the concept of school staff instead of teachers because all employees and school-associated members create the school and the school's culture, respectively. This includes paraprofessionals, such as assistant teachers, academic counselors, school psychological and social workers, etc. Generally, paraprofessionals are educated or trained and certified persons who assist and support the professionals' work, in our case the teachers. Focusing on teachers only would create an intra-professional platform for collaboration, whereas including paraprofessionals creates interprofessional groups for school development. Concerning the working methods, the Dialogue Café method ("world café"; according to Brown and Issacs, 1995) and the Reflection Cycle method (Dalehefte and Kobarg, 2014) are used. These methods were further developed by the project team (Dalehefte and Hillen, in preparation) to support the school staff-driven school development. Specifically, the combined use of the working methods is seen as extraordinary support for school staff-driven development. This is based on the idea of creating collective responsibility (see Ingebrigtsvold Saebø and Midtsundstad, 2018), which refers to student learning, students' social and moral development, and teachers' expectations.

THEORETICAL FRAMEWORK
The theoretical platform spans from Habermas' (1984) communicative action, Hargreaves (1995aHargreaves ( , 1995bHargreaves ( , 2014 and Fullan's (2010a) collegiality and collaboration concepts; capacity building, collective responsibility by Kruse, Louis and Bryk (1995) and Whalan (2012) to professional learning communities (Bolam et al, 2005;Kansteiner et al, 2020) for school staff-centered school development. These approaches can be seen from a historical research perspective, but they also mutually build a platform for each other's arguments. These are interwoven and well-known concepts used in the context of school development studies.

Collective responsibility for school development
Collective responsibility (CR) seems to be crucial for school development, but CR must be established first (Whalan, 2012). Collective responsibility is seen as a set of school norms that reflect teachers' willingness to take responsibility for their students' learning (Lee and Smith, 1996: 110, 114). Lee and Smith emphasize that if school staff believe that their contribution is important for student learning, which means they feel highly collectively responsible, then the staff's effort will be increased as well as student achievement. Lee and Smith's (1996) research confirm statistically significantly that collective responsibility is positively related to student achievement. Moreover, the study showed that if the school staff have collective responsibility for all students independent of their students' previous performances or their students' social background, this learning was likely to create more equity among all students. 'These results indicate that schools, where teachers take collective responsibility for their students' learning, are not only more effective but also more equalizing environments for students' learning, …' (Lee and Smith, 1996: 128). A later empirical study conducted by LoGerfo and Goddard (2008) reveals as well that collective responsibility is positively and statistically significantly related to mathematics achievement independent of the students' social background. To summarize the phenomenon of collective responsibility, it appears to be a decisive, valuable, and necessary element for school development in general, and specifically, for the development of an inclusive, equalizing school learning culture. Because of little research on CR, Whalan (2012: 46) uses survey and interview data in multiple case studies to examine the phenomenon of CR. First, she identified with a factor analysis of survey data seven independent variables for collective responsibility (Whalan, 2012: 66): 1. teachers' satisfaction with professional learning on quality teaching 2. impact of professional learning on teaching practice 3. consistency of professional development focus on the school's goals 4. shared commitment to the quality of teaching 5. commitment to the school's shared goals 6. teachers' collective responsibility for student learning 7. teacher-to-teacher trust.
Second, her further analysis of the survey data confirmed research conducted by Louis, Marks and Kruse (1996) on CR, as well as work by Bryk and Schneider (2002) on the importance of trust related to CR. Whalan (2012: 87) study reports in detail that the survey data suggest teacher-to-teacher trust and teachers' shared understanding and commitment to a school's shared goals as elements of a professional learning community, and that they correlate positively with collective responsibility for student learning.

Communicative action and collegiality
Before diving deeper into the search for appropriate methods for school development, there is a need to emphasize the nature of collective responsibility which is related to teachers' shared understanding and trust. Bryk and Schneider (2002: 8) show that relational trust is dependent on shared commitment. For this, the paper introduces Habermas' (1984) paradigm of Electronic ISSN 1803-1617 "communicative action" which is an understanding-oriented approach. His approach implies social action as a prerequisite for consensual collaboration based on free will. The concept assumes that people coordinate their interaction through linguistic communication. This communication provides people with the opportunity to resolve conflicts, but also to live together peacefully on reason and justice (Aakvaag, 2011: 189). Habermas' theory (1999b: 138) includes both non-social or social actions, and non-linguistic or linguistic actions of human beings.

Reaching success/results
Reaching understanding Non-social action Instrumental action Social action Strategic action Communicative action

Non-social action and social action
Non-social actions are described as instrumental actions. This implies that people do not need to interact with others to realize their action plans. In contrast, social actions are characterized by people's dependence on others to realize their action plans (Aakvaag, 2011: 190). Habermas (1999b: 153) also distinguishes between two main types of social actions: strategic action and communicative action. Table 1 illustrates these types of actions.

Strategic action
A strategic action is a result-oriented action (Habermas, 1999b: 139). Through this type of social interaction, the actors try to influence each other through positive or negative sanctions, such as a reward or punishment (Aakvaag, 2011: 189). In other words, this is a goal-oriented communication in which the actors try to reach an objective by influencing each other in a certain direction; there is an intention behind the action. Through a strategic action, understanding and consensus are only a means of achieving a goal (Habermas, 1999a: 16). The actors adopt a result-oriented and rational attitude (Habermas, 1999b: 154). In this case, the participants can approach each other only as objects. Thus, language is used only as a medium for transmitting information (Habermas, 1999b: 140-142).

Communicative action
Language is used as a source of social integration (Habermas, 1999b: 142-143). This means that the listener of free will is motivated to discuss and recognizes the matter being discussed (Habermas, 1999b: 140). This creates a prerequisite for productive, dialogue-based group work. School development has been criticized by researchers to pressure teachers (Hargreaves, 2014). This is not in line with Habermas claimed free will participation and is counterproductive for mutual understanding. Habermas points out that agreement cannot be imposed on one party by the other, but it must come from within. This understanding builds the platform for the actors' actions. This is a prerequisite for collaboration on agreed objectives. Schaefer et al (2013: 1) summarize it as 'communicative action is oriented towards mutual conflict resolution through compromise. Actors here do not primarily aim at accomplishing their own success but want to harmonize their plans of actions with the other participants.' Hargreaves (1995a) distinguishes collaboration from collegi-ality. He argues that collegiality is far from being a synonym for collaboration. An institutional structure is needed for collegiality: the collegium, or an organized society of persons performing certain common functions (Hargreaves, 1995a). The school staff's collaboration understood as collegial collaboration (Hargreaves, 1995a(Hargreaves, , 1995b can become a communicative action where the actors jointly try to create a mutual understanding and by this a reflected, unsolicited, negotiated agreement. Hargreaves (1995a) points out the need for a structure for collaboration. 'Collaboration does not necessarily involve an institutional base to its structure but refers to a disposition towards, or the enactment of, a style of relationship which may take place in a very wide range of structural conditions' (1995a: 31). Fullan (2010a) also states that collaboration if purposefully organized is an opportunity to ensure that there is coherence inside the targeted reform process. 'The right drivers -capacity building, group work, instruction, and systemic solutions -are effective because they work directly on changing the culture of school systems (values, norms, skills, practices, relationships)' (Fullan, 2011: 5). Interestingly, Fullan uses collaboration concepts of "within school" or "intraschool collaboration." He mentions its importance for facilitating professional learning to be based on teacher needs and to ensure consistency of practice within and across grades (Fullan, 2010b: 19). Fullan mentions as one key inhibiting factor in the grammar of schooling is the egg-crate classroom led by individual teachers (2020: 654). Practice within and across grades prevents the "privatization practice" of teaching and instruction. In contrast, it creates coherence for school development activities . 'Within-school (or intraschool) collaboration, when it is focused, produces powerful results on an on-going basis' (Fullan, 2010b: 36). It is important to annotate how different school staff act and react to top-down and self-made contribution to school change. An unsurprising finding by Hargreaves (2019: 608) was that teachers were overwhelmingly positive about changes they had initiated themselves, but they were equally critical of changes that had been imposed from the district or the government above.

Collaboration and collegiality
In the paper of Hargreaves in 2019, he summarizes 30 years of research on teachers' collaboration. The research results reported there showed that collaboration can increase student achievement and reduce teacher conservatism towards change as well (2019: 618). The collaborative working methods of the project (School-In, 2016) presented here, also focus on within-school or intraschool collaboration. Even if this kind of collaboration is a genuine collective or joint approach, it does not neglect individual needs or ideas. The individuals' voices must be heard to create the collective (see working methods chapter; Hillen, 2019). Interestingly, in a school where professional interdependencies are strong, which means strong collegial relationships, the school staff focus not only on the overall collective (performance) of the school but also on their own efficacy (Louis, Marks and Kruse, 1996: 764).

Capacity building and professional learning communities
Collective responsibility is seen as a prerequisite for school staff's capacity building. Capacity building by school staff is a major aim, a process as well as a result of school development. It is necessary to increase the ability to cope with the school's needs and the school's desired (future) goals. The crucial concept is "mutuality" and commitment, that is, where professionals work together to improve practice through mutual support, mutual accountability, and mutual challenge (Harris, 2011: 627). Fullan (2010a: 57) highlights that "capacity building concerns competencies, resources and motivation. Individuals and groups are high in capacity if they possess and continue to develop the knowledge and skills (…) if they are committed to putting the energy to get important things done collectively and continuously." Capacity building can focus on the individual teacher, but organizational capacity building "of a school to improve students' learning and equity outcomes is influenced by the extent to which collective responsibility is a feature of the school's culture" (Whalan, 2012: 4). Thus, there is a need for communicative action to enable capacity building at the individual and organizational levels for school development by creating collective responsibility. Lee and Smith (1996) state that collective responsibility is essential for school reforms and is a characteristic of professional (learning) communities (Bolam et al, 2005).

Professional learning communities, interprofessionality, and collective responsibility
Professional learning communities as a concept has undergone multifaceted discussions and developments (Kansteiner et al, 2020). Despite this variety, one overarching aspect is unambiguously the collective in these diverse approaches (Vescio, 2020). Research has shown that PLC's have a positive impact on both teaching practice and student achievement (Vescio, Ross and Adams, 2008: 86). As Resnick notes already in (2010: 183) in her model of pedagogy and content routine (PCR), collaborative routines among teachers are important for student learning. These collaborative routines have been described in various ways but are best described as "professional learning communities" (Harris, 2011: 628, Kansteiner et al, 2020. One of the main characteristics is not the improvement of individual teachers' competence and teaching quality presumably induced by specific teacher further education courses but collective improvement. 'The most powerful strategy for improving both teaching and learning, however, is not by micromanaging instruction but by creating the collaborative culture and collective responsibility of a professional learning community (PLC)' (DuFour and Mattos, 2013: 37). Regarding PLCs, one needs to ask, who are the professionals to be addressed? Stoll and Louis (2007: 225) distinguish among PLCs only for teachers, all employees of a school, and those who do work at a school plus all those who are related to all kinds of school's activities.
Reducing the focus of school development only to teachers as the professional target group would hinder addressing the school's potential coherent development. Johnson et al (2020: 45) review-study on interprofessional partnerships reported for instance on school and community joint work trying to best serve students with special needs. They mentioned that collaboration is the key. The study stressed that the students' needs can be addressed differently through the capacity created by interprofessional partnerships. In contrast, the School-In project created interprofessional workgroups. Wackerhausen (2009) describes that sometimes profession(al)s are trapped in their own knowledge-building community. Wackerhausen terms it as not overcoming the profession's immune system. The professionals maintain their attitudes, traditions, and habits which lead to only first-order reflection when looking at issues. To climb over this "self-created wall," all professionals participating in school tasks and routines must be questioned and heard. This type of interprofessional group structure contributes as well to second-order reflection. Second-order thinking overcomes reflective activities that are predetermined to just stabilize the already stabilized (tradition). Barr (2013: 5) discusses in general interprofessional education (IPE) and refers to Wackerhausen (2009) and Dahlgren (2009). He points out that working just in your own professional community will support first-order reflections that are based on own personal and professional views. It is self-affirmative; it is within your professional surroundings. Whereas second-order reflections are transformative reflections were the professionals step back to become aware of their own frames of reference. You "de-centre" the learning by considering points of views other than your own. Korthagen, Greene and Kim (2013: 128) refer as well to reflections by teams and on school level including pre-service teachers, remedial teachers, and school principals to discuss their educational identity and mission of the school. Hence, Wackerhausen's (2009) second-order reflection is decisive for restructuring established professionals' attitudes, behaviors, and values. Therefore, this paper includes Wackerhausen even if it is used here from an organizational perspective, as a structural prerequisite for interprofessional learning communities. This is close to Fullan's (2010b) idea of within-school collaboration. These approaches influence the construction of processes and groups of the working methods developed (see working methods chapter).

How to address school staff for school development? Working methods for school staff-driven development by communicative action and collective responsibility -working interprofessional from within
Consequently, the question emerges of how to address school staff to participate in school development activities. Moreover, how to foster them to take initiative, and how to make them collectively responsible. These challenges introduced and prepared the creation of the main research question for this paper: "How to initiate, support, and develop collaborative responsibility through communicative action?" Sub-research questions which specify the analysis are as follows: What kind of working methods are needed for school staffdriven school development, and what are their characteristics? How can the developed working methods and its combination contribute to support these desired different processes that span from the individual idea to collective action? One approach is democratic participation which means to actively include the school staff in decision making for their own school development objectives (Dalehefte, Kristiansen and Midtsundstad, 2018). It seems to be a trigger for sustainable collective responsibility to use a participation and meaningmaking school development approach. Geijsel and Meijers (2005: 426) point out that educational change is supported by collective meaning-giving and personal sense-making. One could conclude that the less meaning one sees in school change-related activities, the lower one's engagement will be. Korthagen, Greene and Kim (2013: 5) mentioned that a 'within'-development is decisive for transforming education by taken the inner life of teachers seriously. 'Evoking and nourishing the inner life of teachers can provide them to revisit their commitment to and the passion for teaching because it re-connects them with their core qualities.' This will affect the quality of teaching. Zwart, Korthagen and Attema-Noordewier (2015: 580) asked for instance, how one can create working environments for teachers and students where they can thrive and flourish, and where teachers believe they can make a difference to the academic performance of their students. Korthagen and Vasalos (2008) were inspired by these kind of questions and develop a professional development approach the 'Quality from Within' [QfW] approach, focusing on growth, starting from, and building on the inner potential of teachers and students. In contrast to Korthagen, Greene and Kim (2013) where the focus was more on the instructional side, our approach included the individual (I), the group level (G) and the school level learning reflection process (P) in one but stepwise working approach by using the so-called IGP stages of the 'Mental Mapping Response' (MMR) working method. This democratic, reflective approach of the Mental Mapping Response-method will enable the staff to participate in decisions about the overall school development area (main objective) which should initiate the responsibility-creating process. The MMR method is not discussed further here (Hillen, in preparation). It is one of the project's working methods applied in the innovation (Table 2). Concerning Korthagen, Greene and Kim (2013: 128) the importance of a learning process on the school level including teachers, school principals remedial teachers and preservice-teachers as well is close to the interprofessional group learning approach mentioned as decisive for secondorder reflection (Wackerhausen, 2009). This idea of a "within" school development approach by Korthagen, Greene and Kim (2013) can be found in all our democratic teacher-driven school development working methods, MMR, DC, and RC. The combined working methods use the democratic communicative action paradigm which serves and continues the process for creating collective responsibility that begun previously with the MMR method.

The intervention processes and the working methods
To initiate, support, and develop collaborative responsibility through communicative action, appropriate participative working methods for school staff are needed. The whole intervention approach of the School-In project is divided, into six different working days and activities (Dalehefte and Midtsundstad, 2019). We call these working days together with the school staff "innovation days". On these days, the school staff worked partly in groups guided by the project team. These groups consisted of school staff members from different divisions, grades, and subjects they teach. This was done in line with the intra-and interprofessional group composition paradigm for coherent school development.  On day 2, the combined use of the Dialogue Café (DC) and the Reflection Cycle (RC) started. The DC should support communication, ideas, reflection, and understanding connected with the Reflection Cycle to make a collective decision about the objectives and measures to be enacted across the school. In addition, the working method of the Reflection Cycle included staff-suggested indicators to prove that their measures had taken effect. The rationale behind the combination of the working methods and its cyclic use is to create room and time through an easy-to-apply procedure. School staff communication and reflections led to the measures to be jointly taken. To enable intra-professional discussion, the group included teachers from different grades, subjects, and divisions. This is similar to what Fullan (2010a) calls "'within' school collaboration." Moreover, paraprofessionals participated across all groups which created opportunities for inter-professional discussion and reflection. To use Wackerhausen's (2009: 468) term, this should contribute to jeopardize and to overcome the profession's "immune system" to enable reflective growth from the first to the second order. The combination of the working methods and the mixed group approach intended to contribute to de-stabilizing the stabilized professionals' identity (Wackerhausen, 2009: 466f). This effect (the second-order reflection) was analyzed in the research project using focus group interviews (Ingebrigstvold Saebø and Midtsundstad, in review) which is not further described.

Working Methods -Dialogue-Café and Reflection Cycle
During the innovation days, the school staff worked in mixed groups after the main issues for the school and its development had been discussed and chosen with the MMR method on innovation day 2 (see Table 2). Dialogue and reflection are core elements of the working methods. In addition to Korthagen's (2017: 392) reflection model 'ALACT' where the model helps teachers in developing their own personal theories about teaching and learning, the reflections during the Dialogue-Café and Reflection Cycle are focusing as well on the entire school activities, the school's context, expectations, and its culture asked in the end for the enactment of new, alternative or changed staff behavior. The Dialogue Café and the following Reflection Cycle method are applied as a communication platform for reflective talks based on free will (Habermas, 1996). The whole school staff had the opportunity to communicate in interprofessional mixed and changing groups. That is, the school staff included paraprofessionals, for example, assistant teachers. Even if a huge group was meeting (Lagrosen, 2017), the dialogue café method allowed to let everyone work together with someone else by dividing all participants into smaller groups and by changing the group composition at the round tables. First, the school staff was introduced to the Dialogue Café method (Ingebrigtsvold Saebø, in preparation). Each round table was provided a thematic question for discussion. One scribe (host) stayed at one table to support the discussion process and to summarize the discussion verbally. The other table-mates changed each time when they joined another table. In general, the Dialogue Café is an approach to create mixed intra-professional groups. By including paraprofessionals, we enabled them to work as inter-professional groups, too, to enhance the opportunity for second-order reflection (Wackerhausen, 2009) over time. After the school staff had discussed and reflected on the critical issues in the Dialogue Café, with the questions given, they summarized it on a poster or paper. Then, the Reflection Cycle followed (Figure 1) as a group working method. The Reflection Cycle is an approach that follows the general problem-solving process. It has a cyclic structure (see Figure 1; Trepke, 2014: 35). In step 1 of the Reflection Cycle, the objectives were selected, based, and derived from ideas developed from the Dialogue Café results. The last step of the work with the Reflection Cycle (Dalehefte and Hillen, in preparation) was that the school staff groups decided and defined the measures to be enacted in their own school. The objectives were operationalized by the formulation of the measures. The group planned how to put these measures into action until the next innovation day. The groups received a Reflection Cycle form, as paperand-pencil work, to fill out the different steps 1 through 6. In step 3, they needed to write down beside the measures, appropriate indicators. The intention was to be able to see how the measures took place and how they showed an effect. A spokesperson, chosen by the group, was responsible for summarizing the results, sending a feedback protocol, and being the contact person for the group and the research team.

Dataset and data analysis process
Fourteen schools were included in the project. Seven were reference schools, and seven innovation schools. Innovation schools were those schools actively participating in the school development process.
To become an innovation school, schools had to apply for taking part in the project. The intension was to gain acceptance and motivation of these schools. The reference schools were those schools, not actively participating in the innovation process that is, not having used the developed working methods of School-In, but having contributed with questionnaires of the staff to serve with reference data about the school in the same district. This is done because one knows that school development is contextdependent. This context perspective helps to consider the given contextual situation for a reference school as well as it helps to validate the quantitative data collected but not discussed here in detail. For the analysis, one innovation school called 'Beeland' for anonymization reasons was chosen with a two-time analysis of the combined working methods Dialogue Café (DC) and Reflection Cycle (RC).
This school was selected because a full data set was available. During the research project, the project team not only developed but also improved the working method.
In this case, we added feedback protocols to offer better bidirectional communication and support. The school staff comprised (n = 35) participants, 40% of whom were paraprofessionals. Both working methods were conducted together two times, that is, following two innovation days of the innovation phase. Table 3 presents the teacher and paraprofessional groups (G2/G5) and the time points (T1 and T2) of the innovation day's activities when the DC and the RC were applied. For instance, DC21 represents the recorded discussions during the Dialogue Café of Group 2 on innovation day T1.  To conduct the Dialogue Café, questions for each table had to be developed. The questions were created by the research team to enable and foster discussion and reflection. Although the questions for each Dialogue Café table were created by the project team, they were derived from the school staff's self-chosen area of development (Table 4).
Beeland had decided to work on "Local community and parents used as resources for the school" as a development area. Additionally, the school staff had received short presentations from the project team, as knowledge input (Table 2) related to the development area and as followup to the measures conducted to deepen the school staff knowledge and understanding of the related pedagogical content knowledge. The conversations in response to the questions (Table 4) were recorded. Later, the discussions were transcribed digitally and analyzed with a constant comparative method with open coding (Glaser and Strauss, 1967). In addition, the feedback protocols written by the school staff were collected as additional data to show how the planned measures were actually enacted (Table 5 and Table 6). Q2 How can one make use of people who are familiar with the local environment to engage the students in instruction? The different school staff groups developed measures that they reported on the following innovation day during a plenary session. The intention of the plenary session was twofold: to make visible what the different groups had initiated and conducted thus far and to inform and distribute the activities and measures further across the whole school. This was also structurally laid down and implicitly implemented by the fact that the group members belonged to different divisions and taught different grades and subjects. This distribution approach made the measures easier to share. The school, the school district, and the local environment are named "Beeland".

Verbal and graphical analysis of the combined working method process and accompanying questionnaire results
This subchapter, show and explain the innovation, the working method's process, and its results. The working method approach is a procedural and combined one. Therefore, the qualitative process presented and analyzed here is to be seen as a formative research result itself. As well as this kind of qualitative, hermeneutic study describes and highlights the processes and stages necessary to be conducted with the interprofessional staff. For reasons of triangulation, the dataset was accompanied by quantitative data (see Table 7) which were collected at the end of the innovation phase. The scale spans from 0-5. (0 = I do not agree at all; 5 = I totally agree). The quantitative analysis is not discussed here further. These descriptive findings do support by its triangulation the qualitative study results and its interpretation. That is, that innovation and school development by creating and enacting the measures has taken place using these working methods. For instance (see Table 7), there was an increased reflection on the school activities (5.1), more initiatives and processes towards school development were perceived (5.2), and more collaboration and participation of the staff was mentioned after the innovation phase of half a schooling year. Desimone (2009: 184) highlights that professional school development or change are less likely by short or one-time interventions. Duration is needed, frequency as well as the timespan is decisive that professional development can take place and show effect.    Using NVivo, specifically, the graphical approach contributed to visualize the qualitative analysis of what had been discussed applying the different working methods. At a glance (Figure 3), one can compare graphically, in addition to the textual overlapping references (ratio %, Figure 2), what topics were discussed during the different working methods (DC, RC). Open coding was applied (Glaser and Strauss, 1967) to understand the socially constructed meaning by avoiding reduction with pre-structured concepts or labels of a codebook (Creswell and Poth, 2017). The open coding process resulted in the definition of the nodes. The coded compared nodes showed what kind of topics were taken up in the discussion during the DC and in the RC or what was discussed only in the RC. The

1803-1617
screenshot of the NVivo application ( Figure 2) reveals that, for example, "cooperation with parents" (in Norwegian, Foreldresamarbeid) is named 5 times, and 26 references can be found to the "natural environment" (in Norwegian, Naturområde) in both group work methods. The same nodes are found in both files (Figure 2). One file (file 1) relates to the transcription and coding of the Dialogue Café talks and one to the file of the Reflection Cycle (file 2) talks.

Dialogue Café Example 1
Discussion in the DC, Group 2, Innovation day T1: Coded Node: "Natural environment (Naturområde)" Group 2 reads question Q2: What in the local community can engage students, and how to take advantage of this? P 1: Generally, the nature in the local environment could engage the students, I think,… could use nature in different ways… P2: Are there "outdoor mailboxes" in our district? (guestbooks to write your name in)… P3: …our school could have our own tour-book to write one's name inside to engage the students in their leisure time to get them out into nature…

Summary Example 1
The core content in the Dialogue Café's discussion and the objective of the activity described were to motivate the students to experience and become engaged with nature.

Reflection Cycle Example 2
Talks during the follow-up of the RC, Group 2 on the same Innovation day (T1): Coded node "measure (tiltak)" (see Figure 3)

Summery RC Example 2
The core content of the discussion and the objective (Example 2) for the activity described, formulated as a measure to implement in the future, was to use excursions not only for walks and to be outside but specifically to connect them to the intended teaching purposes. Coming back to Habermas communicative action, one can realize that in these dialogues often the participants, either repeated what was said before or asked for further explanation, seeking to understands the other's utterances. This is a prerequisite for communicative action by trying to recognize and to understand the matter being discussed.

Dialogue Café Example 3
Talks during the DC, Group 5, Innovation day T1 was as follows: Coded node "talking positively about the school (framsnakke)''

Summary DC Example 3
The core content of this Dialogue Café talk (example 3) was to possibly let parents indirectly contribute to the school's teaching. One participant stated it could be risky to let parents participate. The challenge was to find the key for how to let them in the school's teaching. During the discussion, the group found out that parents could motivate their children for schooling and make them more interested.

Reflection Cycle Example 4
Talks in the follow-up RC of group 5 of the same Innovation day (T1): Coded node "talking positively about the school' Person 1: Parents talk positively about the school, students, and staff and the parent association. The objective of the Reflection Cycle discussion (Example 4) was how to get parents actually involved. There was a shift in the discussion. Firstly, when they were reflecting to take the school staff own's attitude into account that would make a difference in parents' motivation to participate, they became aware they should not continue talking negatively about the parents themselves. Secondly, the staff decided as a future measure to conduct a dialogue café with the parents to get them involved. Figure 3 shows the discussion process in the different group working methods using the DC and the RC (examples 1 and 2). On the left side of Figure 3 are coded nodes referring to the Dialogue-Café. The arrows point to the symbol loudspeaker of the DC. In the middle of Figure 3, the bubbles represent topics that were discussed in both group working methods (the DC and the RC), and on the right side are topics discussed only in the RC. The graphical comparison shows that not all topics were taken up again in the Reflection Cycle, and that new or additional topics (nodes) appeared in the discussion. If one opens the node "Measure (tiltak)," one finds the topics "nature and teaching purpose". The content of the talks transcribed in the node "Measure" is shown above in Example 2. Summarizing the different processes and results, one can see the working methods combined steps for school development as follows:

Graphical analysis
Mental Mapping Response (1) See Table 2: The staff chose the core development area for the whole school: "Use the local environment and the parents as a resource for the school." Dialogue Café: (2) See Example 1: Discussion on how to use the environment to motivate students to make them interested in and become acquainted with nature.
Reflection Cycle: (3) See Example 2: Discussion about whether outdoor activities can be used as outdoor school or for teaching purposes and developed a measure to conduct outdoor teaching.
Feedback protocol: (4) See Table 2: The group reported what they had enacted as a measure for outdoor teaching.
The working methods process produced results in such a way that measures were formulated by the school staff in line with the overall formulated goal for their own school development. Out of this goal, the development area "to use the parents and the local environment as a resource," they discussed and reflected on the use of nature (26 references). This topic was taken up again in the Reflection Cycle where outdoor activities were planned to use now as a relevant topic for teaching. Communicative action took place among the school staff. Interestingly, by analyzing the raw data similar kinds of outdoor activities had been done before but were not used for teaching purposes. The reflection led to a new perspective on the same school activity. Specifically, the need to formulate and specify what the group would conduct, by using the RC form, which they needed to fill out, and the later feedback, functioning as a reminder, supported the enactment of the measures.
The results of the NVivo analysis and discussions related to the school staff's reflection and decision process qualitatively analyzed and visualized showed that the method could help support the formulation and implementation of measures for the whole school and its development. The analysis shows that the combined methods lead to support communicative action, which can also be procedurally proved by the measures, objectives, and indicators formulated and defined by the staff themselves during the discussions. The group work was open for each participant's contribution based on free will. One of the challenges was that the school staff had difficulties distinguishing the concepts measure, objective, and indicator. This difficulty became obvious in some of the transcriptions but was solved by the support of the project team facilitating the working group methods. Besides the main findings, one obvious result was that those measures were preferably enacted which needed less preparation but could, therefore, become included in an almost daily routine. Unexpectedly, some of the school staff felt less satisfied with measures they enacted themselves if the measures seemed smaller than colleagues' measures. Anyhow, the project team tried to emphasize the importance that small changes can make a difference in school development. One explanation lies in the fact that these smaller measures can be applied with less effort and easily every day. Presumably, the preparation time was less, and the measure, therefore, quickly became a routine in the new school culture.

DISCUSSION
The paper's focus is on school development by communicative action, which is implicitly based on a democratic understanding and its support for collective responsibility. The school staff-driven development approach is used based on the paradigm of PLC's (Viscio, 2020; Viscio 2008) but extended with the characteristics of interprofessional group work (Barr, 2013;Wackerhausen, 2009;Johnson et al, 2020). The analysis and discussion were divided into two sub-research questions: "What kind of working methods are needed for school staff-driven school development, and what are their characteristics?" The first sub-research question addressed the characteristics of the working methods. The intention with the DC and RC was to enable support for participative democratic school development by working methods fostering collegial collaboration and collective responsibility. Using the presented theoretical framework above, one can contend that: (1a) school-wide participation with within-school collaboration (Fullan, 2010b), (1b) or more specifically, interprofessional participation (Wackerhausen, 2009, Barr 2013), (2) by using a democratic, free will-based, participative, and collective approach as communicative action (Habermas, 1984, Hargreaves 2019: 617), (3) through an explicit or implicit organized society for performing certain common functions (Hargreaves, 1995a) is needed.
These participative characteristics (1a,b) and (2) as well as (3) certain functions of an organized society, which presumably enabling collegial collaboration (Hargreaves, 1995b) and creating collective responsibility, are recognizable features of the working methods used in the School-In project for teachercentered school development. The open (1b) interprofessional dialogue (Barr, 2013) and mutual understanding process (2), as a genuine communicative action (Habermas, 1995) was served especially by the DC opportunity, of continuously mixing the composition of the groups of teachers with paraprofessionals. Specifically, the opportunity to create mutual understanding expanded by the interprofessional views (Johnson et al, 2020) through discussion and reflection on questions related to school issues is provided. The discussed questions were also rooted in the school staff's chosen developmental area. This can create responsibility for future activities to intentionally involve school staff by free will (Habermas, 1999a). Repeating Hargreaves (2019), teachers' self-made contributions to school change showed that teachers were overwhelmingly positive about changes they had initiated themselves. As well, the specific group approach's function (3) as a kind of structural condition (Hargreaves, 1995a: 31) was provided. For instance, every single group has had a host person (DC) as well as a contact person (RC) to report, to survey, and to distribute the results. This supports purposeful and directed collaborative work as Hargreaves emphasizes. Collaboration understood as collegial collaboration needs an institutional structure or an organized society. This organized society of persons performing certain functions is needed, and one can find these implicitly not only by the group compositions but also by the group process structured DC and RC activities. Hargreaves (1995a: 40f) states that collaboration is wasteful and pointless if it lacks purpose and direction.
How can the developed working methods and its combination contribute to support these desired different processes which span from the idea to action?
The second sub-research question addresses the potential support of the whole process from the idea to final action, the enactment of the measures for school development by combining the DC and the RC. The direction of the school staff's work, based on the former jointly defined development area, is encompassed by applying the RC. The RC follows a general problem-solving (Betsch, Funke and Plessner, 2011;Borko, 2004), development (see Trepke, 2014: 37), or innovation process (Midtsundstad, in preparation). This process ends in the implementation of the innovation or more specifically, in the enactment of the measure. One important part Midtsundstad (in preparation) states is the formulation of objectives or goals. Researchers have shown that goals are less explicitly addressed in innovation processes outside the private sector (DeVries, Bekkers and Tummers, 2016). This neglect makes innovation projects in the public sector, which school development belongs to, less effective (DeVries, Bekkers and Tummers, 2016). In this case, the RC addresses explicitly the formulation of a goal, of course, on a micro-level. Moreover, it includes the description of its enactment, the measure itself. Thus, the RC takes up the formulation of how-to, that is, the measure itself, and the goal (what) to be accomplished. Even more, the paper-and-pencil form applied for the RC activities requires filling in indicators (see Trepke, 2014: 37) as well. Later, the school staff themselves can trace if their measure has taken effect. The Reflection Cycle in step 6 has closed now and can restart. Innovations are seldom linear and require several efforts to take effect. The paper-and-pencil documentation by the RC form and the follow-up by the project team and the contact person and his or her later plenum's talk serve as safeguards as well. One needs to keep in mind that the measures, and the objectives were negotiated and decided within the groups themselves, that is, to follow the idea of 'within' or democratic staff-driven school development.
Combining the working methods, specific structural conditions are created which are needed for collaboration as Hargreaves (1995a: 31) points out. Referring to the combined processes, one activity (e.g., talk on a school issue) is to be initialized for discussion and reflection processes across both methods. Example 1 and Example 2 show discussions and reflection in each working method, but both are needed to reach finally the insight for the enactment of a measure. The DC (Example 1) showed that outdoor activity can be used to engage students with nature, but during the RC talks (Example 2), the school staff concluded that it could be used for educational purposes as well. The DC and the RC (examples 3 and 4) showed that the focus of the persons' talks about the goal to better involve parents changed by the reflection on the second-order (Wackerhausen, 2009). The school staff detected on their own the need to change their own attitude toward parents to get them actively involved (Example 4). These communicative actions (examples 1, 2, 3, and 4) were fostered to become more coherent by the mixed and interprofessional group process approach ended in a measure documented in the RC and in the feedback protocols. The measures were reported as enacted by the contact person in the plenum of the school staff meeting during the following innovation days.

CONCLUSION
Returning to the main research question, "How to initiate, support, and develop collaborative responsibility through communicative action (sensu Habermas)?," one needs to take another critical look at the examples and the school's context again. Coming back to Habermas' communicative action, one can see in the dialogue examples (1-4) that the participants, either repeated often what was said before or asked for further explanations, seeking to understand the other's utterances to later harmonize their plans of actions, which are the measures. This is one prerequisite for communicative action by trying to recognize and to understand the matter being discussed.
Even if the analysis of the data, specifically, the examples, show a clear picture, one can critically ask if this has been communicative action by free will and if collaborative responsibility has had developed. How to guarantee that the discussion and final measures made were performed without an strategic intention? One cannot control that a school development situation is not perceived and experienced as forced. There is no guarantee that individuals will not have been working strategically using these methods or even just act instrumental to get the work done. The setting a school offers for applying these methods, for instance, time and room (Hargreaves and O'Connor 2017), will influence the productive effect on the working methods, as well as every single participant and the school leadership's attitude toward school development and the methods used, respectively. This is in line with Park, Lee and Cooc (2018: 769) quantitative study which concludes that principals should give more attention to exerting supportive and egalitarian leadership. The study from Park showed as well an indirect effect of principal support on student achievement through shaping professional learning community, collective responsibility, and collective teacher expectations rather than directive or restrictive leadership. That contrasts with a democratic participative understanding. Generally, Park and Byun (2020) mentioned that transforming a school into a PLC generates many positive effects for both teachers and students. Nonetheless, the project team created working methods that enable and foster communication and reflection processes. These collaboration approaches can support the development of collaborative responsibility by opening up for communicative action. To open up is for instance if teachers are asked to bring in their own ideas. This is a clear signal for being important for the school and to e.g. value their teaching, respectively. This is in line with Lee and Smith when they mentioned that if school staff believe that their contribution is important for student learning -it creates responsibility. Revisiting Hargreaves' (2019: 608) 30 years of research experience on teachers' selfmade contributions to school change it showed that teachers were overwhelmingly positive about changes they had initiated themselves compared to top-down experienced changes. Coming back to leadership and direction, the direction is driven by the school staff themselves. This supports free will participation. Also, the DC and the RC are tools for starting and following these directions and bring it to an end, that is, joint action. The framework, the described processes, for instance, the intra-and interprofessional professional group working process across the school, can serve as one safeguard for a democratic, social situation for school development. The same is due to collective responsibility. It is a social process that cannot be forced but encompassed with these working methods. The analysis of the use of the combined working methods approach showed that school staff was able not only to work collaborate collegially but also to enact measures based on their initiative and engagement to participate in school development. This study is also important as it shows that and how collective responsibility as one driver for coherent school development can be created by school staff-centered working methods. The approach, the combined working methods themselves, tried to create a framework that offers the opportunity for social actions, framed by a democratic, communicative process such that school staff-driven school development can take place.

OUTLOOK
Language is a decisive factor for the 'collective approach' because it is a source of social integration (Habermas, 1999b). One might keep in mind that different professional groups make use of language in different ways. For instance, there is a variety in the professional understanding of concepts. Hence, language can separate and invite as well. Lagrosen (2017Lagrosen ( : 1518 stresses that Team learning implies that every member needs to see the full picture and speak the same language. One might propose this challenge as a future research desideratum that is to analyze the meaning and its effect on the use of language in interprofessional workgroups. Korthagen, Greene and Kim (2013: 128) mentioned that "The development of a common language was essential in this process, a language that supported the teams' discussions on the relationship among theory, vision, and practice at the school level, and it also deepened individual reflections". Hence, the conscious use of language is decisive for interprofessional working methods and the development of collaborative responsibility, respectively.