Introduction

The efforts of all the county’s actors must be combined if we want to support innovation in the county. If there is no shared vision between us, we are losing all the potential for generating a sustainable process for local development.

One of the facilitators of Lankidetza Sarea (LS) made this comment during a regular Monday Meeting of the network in mid-2011. This comment was made more than 2 years after the networking process that supported the local economic development (LED) initiative had begun, in late 2008. The comment reflects this facilitator’s realisation that LED required new modes of social relations among local stakeholders; social innovation was necessary. The articulation of this comment at the meeting raised the question of how social innovation can be facilitated, an idea that had not been addressed adequately by the territorial development literature. In late 2012, the Industrial Forum was created. Various local actors (including representatives of firms, training centres, the county technological centre and the county development agency, and politicians from local governments) became involved in a new collaborative process that would change the governance of LED in that county.

Networking processes can generate new modes of relating among the actors, and social innovation is the process that brings innovation to these social relationships (Moulaert et al. 2013a; Van Dyck and Van den Broeck 2013). Becoming a socially innovative territory is a consequence of a facilitated learning process in which the capabilities of the network members are developed through reflexive action. As Howaldt and Kopp (2012) and Gustavsen (2012) emphasise, social researchers can play a role in the creation and maintenance of spaces for collaboration and learning in social innovation. Moreover, this paper adopts the argument developed by Howaldt and Kopp (2012) that the role of social researchers in facilitating social innovation is to organise the process of change as a learning experience. The researcher’s role concerns to the learning to be gained from action. In this respect, the process should encourage the development of relevant skills in the participants and enhance their ability to together produce the appropriate transformation in social relationships. In particular, social innovation involves learning about the nature of concepts such as collaboration and shared leadership in a collective endeavour. Social innovation is thus not a spontaneous process and requires sensitive facilitation. Social researchers can play a role (albeit not the only role) in the facilitation of social innovation.

This paper reports research that attempts to explore how social innovation can be facilitated. Moreover, an action research approach is adopted, which makes it possible to go deeper into the role of action researchers in the emergence and development of social innovation. The engagement of the author in this action research process permitted her to adopt an “inside-out” position that enabled her to explore the similarities in the assumptions that support social innovation and action research. Through an analysis of facilitated collective action, the nature of the facilitator’s knowing how is made explicit in the process of understanding how social innovation is facilitated.

The case analysis will examine the process of creating the Industrial Forum in Goierri, which constituted a significant step towards social innovation. The participants involved in this network have changed their modes of relating to one another. Before this forum existed, the local development agency (LDA) contacted each of them in order to implement the actions defined in the Strategic Plan for LED. The Industrial Forum offers a common and continuous reflection-action space in which collective decision-making mechanisms have been created and the leadership mode has changed. Instead of the staff of the agency (supported by the President-politician and based on the Strategic Plan) deciding the priorities for the actions that encourage LED in the county, the different actors first reflect on the challenges of LED and later make decisions on the actions that will be taken. New skills to collectively effect this transformation have been developed. Moreover, these reflection–action–reflection cycles result in a shared vision and a new governance mode for LED. A new governance mode is then the outcome of social innovation. Governance is understood here to be the mechanism by which the various stakeholders in the territory interact.

The factors that facilitated the transformation of stakeholder attitudes and skills and generated this change in governance are identified by analysing incidents from the beginning of 2011 to the end of 2012.Footnote 1 Accordingly, the generation of a new mode of governance is the specific outcome that provides an anchor in this analysis, which keeps the researcher on course through the interactive cycles of deduction and induction (Pettigrew 1997). To maintain this course, this paper uses a process perspective in a case analysis.

Moreover, the concept of agora defined by Karlsen and Larrea (2014) is used to conceptualise the case and further understand the role of social researchers. The agora is defined by Karlsen and Larrea (2014) as a space is shaped by the dialogue among different actors and consequently, where different actors meet. It is a space that is generated by interactions and interrelations. The Industrial Forum is an example of an agora where territorial actors including researchers, meet to shape territorial development, which includes both technological and social innovation. Social researchers are critical participants who accelerate learning processes in continuously changing agoras and encourage social innovation (Karlsen and Larrea 2014). The paper analyses how action researchers can facilitate social innovation in agoras for LED. As a result, shared elements between action research and social innovation are identified.

Regarding the paper’s structure, first, a theoretical discussion concerning social innovation is presented to understand how social innovation addresses territorial development and why facilitation is crucial. The contribution of action research to this discussion is presented to close this section. Next, a case analysis from an “inside-out” perspective is described, in which both the process analysis and the theoretical discussion are integrated. This section shows how the action research process facilitated social innovation in the network. These insights are discussed in a later section. The paper concludes with a summary of the main findings, some reflections and new ideas for further research.

Social Innovation and Territorial Development

Social innovation refers to the process of transforming social and power relationships (Moulaert et al. 2013a). As Moulaert and Nussbaumer 2005, 2008 argue, this process involves transforming the social dynamics (and assumptions) that underpin networking, learning, governance or community culture in a territory. Thus, there is a need to understand the nature of the social dynamics that contribute to the building of alternative, more appropriate, social relationships, aspirations and governance practices.

From the territorial development perspective, social innovation has a triple focus. First, social innovation concerns the satisfaction of basic stakeholder needs (Moulaert and Ailenei 2005; MacCallum et al. 2009; González et al. 2010). Second, social innovation empowers local actors and manifests the ‘empowering’ and emancipatory character of transformation processes (Crises 2003; Moulaert and Nussbaumer 2008; González et al. 2010; Jessop et al. 2013). As a result, collective visions and intentions drive collective change, and new capabilities for transformation are developed (González et al. 2010). Third, appropriately transformed social and power relationships among community actors, and the governance modes resulting from them, add new value to the collective endeavour (Moulaert et al. 2005; González et al. 2010).

When linking social innovation and territorial development, the centrality of territory as a field of action in the social innovation literature must be emphasised (Van Dyck and Van den Broeck 2013). Klein and Tremblay (2009) refer to the role of social actors in innovative governance modes for territorial development. Klein and Tremblay also refer to the mobilisation processes of local and extra-local resources to stimulate knowledge-building cycles and to generate change in institutional structures, thus enabling social innovation. These references are related to the argument of Fontan et al. (2013), who argue that social innovation is necessary to assure the vitality and success of territorial economic renewal, or as Van Dyck and Van den Broeck (2013) state, the importance that some authors who work on territorial innovation systems (such as Tremblay et al. 1998) attribute to bottom-up processes where social actors are involved in collective learning.

Social innovation thus does not simply occur in a spatial context but involves changing the specific spatial relationships: that is, it is spatially negotiated and spatially embedded, as Moulaert (2009) notes. This transformation shows that specific contextual elements condition these processes. Accordingly, Van Dyck and Van den Broeck (2013) state that social innovation researchers explain that territorial development is related to the capacity of the territory to transform spatial relationships and to the embeddedness of relational assets and spatial proximity in networks. According to this explanation, territorial development will depend on the creation of networks in which the involved actors can transform their modes of relating to increase their control over these modes and their impact on development trajectories (Van Dyck and Van den Broeck 2013). This means that social innovation is necessary for territorial development. However, in all of these publications in which social innovation is related to territorial development, there is little exploration of how to develop these networking processes and how to transform the social relationships in networks to facilitate territorial development.

The Need for Facilitation of Social Innovation

In addition to arguing the need for facilitation of SI processes, this section describes the types of skill that the facilitation process should encourage. Changing the modes by which local stakeholders relate involves process innovation. This means that innovation occurs in the change process within the actors’ mobilisation–participation process (Moulaert et al. 2013b) and results in new collective skills that facilitate transformation. Social innovation concerns change therefore. Changing the structures of governance, through developing new modes of socially relating and greater collective empowerment, is referred to as a possible socially innovative outcome by Moulaert et al. (2013b).

Describing social innovation as a process innovation (Moulaert et al. 2013b) concerns that social innovation must be facilitated. A changing process within the ways in which actors relate to one another supports a learning process and the generation of new capabilities. It is in the acceleration and organisation of the learning process where the facilitation takes place.

Facilitators are people or groups who provoke continuous reflection–action–reflection processes in actors (ConectaDEL 2013). Facilitators facilitate collective learning processes. Here, the knowledge management literature developed by authors such as Nonaka, Alavi and Leidner, Grant, and Massingham can be referenced (Nonaka and Takeuchi 1995; Grant 1996; Alavi and Leidner 2001; Massingham and Massingham 2014). Facilitators are in charge of knowledge management in participation processes. Knowledge management shares the process perspective of social innovation and action research and includes the process of creating, storing and retrieving, and sharing and applying knowledge (Alavi and Leidner 2001). Facilitators can be considered to be knowledge creation enablers (Nonaka and Takeuchi 1995).

In a network for territorial development (a local network for LED in the case of this paper), social innovation involves generating collective skills for collaboration and shared leadership. The collective learning processes that occur in the network should follow this objective.

The concept of collaboration must be differentiated from that of cooperation. In addition to exchanging information, cooperation involves aligning or altering activities so that more efficient results are achieved (Camarinha-Matos and Afsarmanesh 2006). The concept of cooperation, however, does not exclude participants from having private goals and using private resources and individual strategies in addressing these goals. Collaboration, however, refers to collective action towards common objectives. As Moulaert et al. (2013b) note, social innovation also concerns collaboratively sharing views and strategies to achieve shared goals.

In the idea of shared leadership, there is no place for absolute or hierarchical power or for a single leader (Karlsen and Larrea 2012). All participants must consider themselves part of the leadership process. For many people who have experienced only hierarchical power relationships, learning how to participate in a shared leadership process will be challenging. This situation is the case for policy makers. As Karlsen and Larrea (2012, 2014) argue, policy makers who have been defining policy in hierarchical structures can be open to the participation of private agents in consultative terms but often resist relinquishing the power to make decisions. Shared leadership approaches will then depend on how policy makers are involved and on the nature and availability of resources with which to formalise these processes. Moreover, the availability of resources in these collaborative processes determines the auto-regulation level of the processes and, consequently, the leadership mode in them. Collaboration and shared leadership processes look for auto-regulation of interaction processes, and dependence on external resources puts the autonomy of these processes at risk. The development of shared leadership processes is significantly affected by the availability of these resources then.

Accordingly, social innovation concerns collaborative processes in which shared leadership modes emerge. The definition of collective action to address shared challenges and the diversity level of the participants will condition the capacity for collaboration. Shared leadership approaches will depend on how politicians are involved and on the nature and availability of resources with which to formalise these processes. This collaboration and the emergence of shared leadership will also alter the institutional capacity and the multilevel relationships among different territorial levels. Furthermore, the role of network facilitators will be critical. This importance raises the issue of the specific forms of knowledge required to facilitate the transformation of traditional, deep-seated assumptions regarding the modes by which actors in a territory interact with one another. The central argument here is that this specific knowledge will be conditioned by the facilitators.

Social Researchers as Facilitators in Social Innovation: Action Research

Howaldt and Kopp (2012) emphasise the social character of innovation processes and the role of social researchers in analysing the social prerequisites for innovation. Together with Gustavsen (2012), Howaldt and Kopp endorse the need for a facilitation role for social researchers with regard to the creation of the spaces for the communication, cooperation and knowledge integration that social innovation requires. Moulaert et al. (2013a, b), who relate social innovation to territorial development, refer to action research as the most suitable methodological approach for understanding and analysing social innovation. If the objective is to critically examine territorial development processes, this approach poses an epistemological challenge for researchers, and action research can meet this challenge (Moulaert et al. 2013b). Given the emancipatory potential of social innovation, the methodology should facilitate the reflexivity of its agents and observers, enabling lesson-drawing and lesson transfer (Jessop et al. 2013). The main argument here is that both social innovation and action research concern change—through cogenerative learning and through the generation of new skills to address shared challenges—and thus a process approach is intrinsically considered by them.

Action research is defined here as a form of praxis in which researchers actively engage in the field with practitioners to solve pertinent and practical problems (Levin and Ravn 2007). The cogenerative process in the agoras for territorial development described by Karlsen and Larrea (2014) is considered to be an analytical framework in the case analysis in this paper. The Industrial Forum is then an agora for LED where social innovation occurs.

The democratic character of the cogenerative learning processes that occur within the groups of participants involved in the agora is emphasised by Karlsen and Larrea (2014) when describing how cogeneration occurs. The democratic agoras discussed by Karlsen and Larrea (2014, p 130) refer to agoras that fulfil Gustavsen’s conditions for democratic dialogue. Democratic dialogues allow the development and acceptance of shared meanings and principles in the dialogue among people and allow the democratisation of communication among these people (Gustavsen 1992, 2003; Gustavsen et al. 2007; Karlsen and Larrea 2014). This cogenerative learning process, however, not only concerns the exchange of rational knowledge but also creates Mode 2 knowledge (Gibbons et al. 1994). Mode 2 knowledge is always produced in a context of continuous negotiation, and it will not be produced unless and until the interests of various actors around a particular application are included (Karlsen and Larrea 2014, p 86).

The territory-oriented approach developed by Karlsen and Larrea (2014), however, proposes two main differences from the workplace-oriented cogenerative action research model defined by Greenwood and Levin (2007). First, Karlsen and Larrea state that researchers contribute with both disciplinary field knowledge of territorial development and process knowledge of these cogenerative processes. Process knowledge refers to research techniques, work forms, management or facilitation of the collaborative research process or documentation and synthesis of results and action implications (Greenwood and Levin 2007). Action researchers also have expertise that contributes to and is necessary for the cogenerative learning process. This expertise is why the contribution of disciplinary field knowledge is emphasised by these authors. The case analysis shows how the disciplinary field knowledge of action researchers advances social innovation in the agora.

Second, Karlsen and Larrea (2014) refer to the nativeness of researchers in these cogenerative processes. This means that researchers, being territorial actors, are also owners of the challenges and problems proposed by practitioners and, moreover, researchers also change during the research process. In this way, as natives, the researchers are also part of the change process (Karlsen and Larrea 2014, p 101). The concept of insiderness defined by Massingham (2014) and related to Community Action Research (CAR) theory (Senge and Scharmer 2001) reinforces this argument. Social researchers participate in the learning process that occurs in agoras, and researchers may influence the change process (Karlsen and Larrea 2014). This concept changes the idea of participation as it relates to influence in this framework. The choice of an action research approach to research allows deeper exploration of the ‘researcher-as-facilitator’ role and shows how action research can facilitate social innovation through an in-depth understanding of the change process experienced by the researcher.

Social researchers can accelerate the learning processes and contribute to social innovation in the agora helping to change the rules of the game used by actors to relate to each other, the way they learn together and make decisions and the way they turn decisions into action (Karlsen and Larrea 2014, p 42). The following case analysis allows a deeper understanding of this argument and shows how action researchers facilitate this change process – social innovation in an agora. According to the basic concepts of social innovation, this paper permits an understanding of how action researchers, through participation in cogenerative processes, can facilitate the generation of capabilities for collaboration and shared leadership in participant actors.

In this context, this paper responds to the following two research questions:

  • What facilitates collaboration and shared leadership in networks for LED?

  • How can action researchers contribute?

Industrial Forum: The Case Analysis from the Inside Out

Presentation of the Case

The Industrial Forum is currently a local network derived from LS. LS is a network of local firms created in 2008 to promote LED by Goieki, the county development agency in Goierri County.Footnote 2 LS began with a market-oriented approach: its objective was to develop local industrial firms’ capabilities to increase their global competitiveness and, consequently, facilitate a LED process.

Goierri is an area located in the southern part of Gipuzkoa province in the Basque CountryFootnote 3 and it has approximately 42,000 inhabitants and is 271.3 square kilometres in size. Economic activity in Goierri occurs mainly in the industrial sector, in which 45 % of the working population is employed. Industrial activity is predominantly shaped by subcontracting relationships between large firms competing in the global market and their dependent suppliers, most of which are small-to-medium enterprises (SMEs). This unique situation facilitates specific inter-firm relationships based on confidence in, and long-term experience with, one another. However, this situation also limits the innovation capabilities of the SMEs, who have become dependent on the custom of the large companies.

LS was attempting to transform the traditional role of the agency from being a service provider to becoming a booster of LED by creating a new interaction dynamic among all local firms (Estensoro and Larrea 2012). However, at this stage, LS was far from constructing a new governance mode for LED in Goierri County. This gap led to the resultant action being limited to firm-based interests without strategies shared among the rest of the county’s actors and to the focusing of county development on technological innovation.

In 2012, the Industrial Forum (IF) was created by key territory stakeholders, including the author as a researcher. This forum intended to create a ‘multi-actor’ space within which shared strategies for industrial development could be created. At the same time, the creation of the IF was supposed to generate an agora for LED.

Data Analysis

The process approach argued in social innovation and action research is reflected in the methodological approach for the case analysis. On the one hand, the ‘inside-out’ perspective is applied. On the other hand, the processual analysis approach defined by Pettigrew (1997) is used to structure the data. The definition of social innovation and action research both as change and the goal of connecting them necessitate an approach that includes the process approach in data analysis. As Pettigrew argues, it is not possible to consider the change process as though it has a clear beginning and a clear end. The processual analysis (Pettigrew 1997) applied in this paper argues that the research is capable of generating knowledge not only of processes and outcomes but also of why and how outcomes are differentially shaped by processes.

The change process is defined by Pettigrew (1997:338) as a sequence of individual and collective events, actions, and activities unfolding over time in context. In this analysis, several important incidents that refer to these events are identified. These are incidents for which the progress is transparently observable (Pettigrew 1990), recognising that these incidents are not the only conditioning elements of change. Each of these incidents provides a transparent look, such as a clear point of data collection, and an important practical consideration in an extended stream of time, events, people and processes. This analysis is an exercise that reveals the driving mechanisms of the process, going beyond a mere surface description (Langley 1999). Consequently, all of the analyses of the data and their bracketing facilitates a deeper examination of the evolution of the network and allows the detection of the main elements that condition the network’s development.

These incidents describe specific events that change the development path. According to this ordering criterion, a shapeless mass of process data is transformed into a series of more discrete but connected blocks (Langley 1999) while considering that the driving assumption behind process thinking is that social reality is not a steady state but a dynamic process (Pettigrew 1997). Subsequent incidents do not always involve sequential phases. It is necessary to question simple process models that assume neat linear progressions and causation models of well-defined phases leading to well-defined outcomes (Langley 1999; Pettigrew 1997). Accordingly, one incident can provoke or facilitate a subsequent one, with some incidents creating transversal conditions that fit the emergence of a new governance mode.

The case analysis below is based on critical incidents and is structured in subsections that address the following information. First, how each incident emerged is clarified. The mechanisms that drive each incident and, accordingly, the features of the immediate context are explained. Second, the event in which the incident took place is described. After this description, the event is analysed. This analysis allows the identification of the factors that facilitate social innovation and the role of the action researcher in the event. Shared elements between social innovation and action research relevant to each incident are then identified. The content of this analysis is embedded in a table at the end of each section where each column refers to a particular subsection. An extra column that shows sample data is included in these tables. The most significant sentences have been selected from the available data for display in this column.

A ‘first person’ perspective is used to facilitate an understanding of the role, contribution and engagement of the researcher (Herr and Anderson 2005; Foss and Moldenaes 2007). This perspective corresponds to the epistemologies of the action research paradigm and the self-reflexivity that this research requires. Here, reflexivity means establishing relationships between the rigor and the relevance of the resultant research. As emphasised by Herr and Anderson (2005), action research activates an on-going reflection process in a researcher community. This reflection contributes new theoretical insight and practical knowledge concerning social transformational processes. The ability to explain the knowledge generated through action research processes is then a requirement for effective action researchers (Greenwood and Levin 2007; Levin and Ravn 2007). In this way, following York and Nicolaides (2007), engaged researchers can make sense of one another’s experiences and of the new co-generated knowledge produced through the conversations within the group.

Recognition of the Need for Social Innovation

Emergence

In the first half of 2011, the lack of a shared vision among the various county actors and their low engagement in the networking process created an increasingly evident limitation on the network’s ability to support LED. The awareness of this limitation activated a deep reflection process in the facilitators’ group of LS. My role in this group was focused on facilitating the action research process and particularly continuous collective reflection in the group on the development of the networking process and possibilities to improve it. This role involved raising questions for reflection, stimulating thought, and challenging assumed views concerning the governance of the group. Beyond these activities, specific analytical frameworks and concepts were shared to facilitate a better understanding of the challenges faced by the group and to develop effective decision-making processes to address these challenges. Furthermore, my role included building learning processes by questioning our ways of thinking. These reflection exercises led to an awareness of the lack of a shared vision among local actors, which activated a new action research cycle.

Event

The comment by the representative of Zerbitzualdea (the Department of Services for Companies inside the agency) and referred in the top of this paper mobilised the facilitators’ group to begin work on a shared vision for the entire county (as opposed to a vision that served only the interests of the firms).

Analysis

The realisation that a shared vision was required led to the recognition of the need to build a shared leadership process among local actors. This process required more democratic modes of relating within the county to encourage LED; that is, new modes of relationships were necessary. Furthermore, the reflection process showed that the technical capabilities of the SME representative (a member of the facilitators’ group) were insufficient to generate sustainable LED processes. The group became aware that the transformation of social and power relationships among local actors (i.e., social innovation) and the generation of a new governance mode for LED would require new facilitation capabilities.

Relating Social Innovation (SI) and Action Research (AR)

The reflection–action–reflection cycles facilitated by the action research process that occurred in the facilitators’ group was a key factor in the learning process. This continuous learning process was based on democratic dialogue (Gustavsen 1992; 2003; Gustavsen et al. 2007), which facilitated inclusion and knowledge integration (in which a variety of forms of knowledge were socialised) and avoided monopolisation of the process by some individuals (Moulaert and Nussbaumer 2005). In this way, the facilitated action research process established the early foundations for social innovation to occur in the territory. The first shared element of the social innovation and action research processes was established: the democratic principle of reflection–action–reflection cycles (Table 1).

Table 1 Critical incident 1: recognition of the need for social innovation
Full size table

Generation of Disciplinary Field Knowledge

Emergence, Event and Analysis

The research for my Ph.D. Thesis yielded disciplinary field knowledge on social innovation for territorial development. My research focused on understanding how local networks to encourage LED can facilitate social innovation in a territory. LS was the main case in the thesis, and part of the writing period coincided with my role as a facilitator in LS. Accordingly, the thesis contributed an analytical framework with a process perspective in which action research and social innovation were related. This knowledge introduced new concepts, analytical frameworks and work methodologies that were shared with local actors, which was knowledge that contributed to the development of the social innovation process.

Factors for SI and AR Elements

This Ph.D. research increased knowledge concerning social innovation in the local network. The action research process and, specifically, the reflection loops that occurred in the writing period of the thesis made this increased knowledge possible and involved learning how to facilitate the transformation of relationship modes in the network (Table 2).

Table 2 Critical incident 2: generation of expert knowledge
Full size table

The Political Credibility of Social Innovation Expertise

Emergence and Event

The change of government in 2012 was fortuitous for the action research process because it provided political (institutional) momentum for a change in the governance method for LED. The starting point of this connection was a meeting with the mayor of one municipality in the county. The mayor is also the President of the county development agency. The mayor had heard about my research, and he was interested to learn more. Bildu, the nationalist left-wing party that he represented, had won the elections in almost all of the municipalities in the county, and significant reflection regarding how Bildu would address its political challenges was occurring. An invitation to meet with me served this purpose. The mayor showed strong interest in the analytical framework that I had developed in my Ph.D. research and presented at the meeting. I used my knowledge concerning social innovation to provoke reflection on the need for new approaches to relationships among local actors to encourage LED, how this need could be facilitated, and the role of policy makers in creating these new relationships. The mayor committed to the application of the social innovation approach to implementing Bildu’s policies for county economic development.

Analysis

As a researcher, the adoption of my analytical framework for social innovation gave me the opportunity to participate in a politically supported transformation process. The mayor’s credibility, in supporting my knowledge, was key to connect the formal political process to the action research process. Shared leadership and a shared language based on trust followed and led to a shared vision for the requisite transformation of social/power relations in the territory. The need for SI was fully recognised. Furthermore, the mayor recognised and legitimised my role as a researcher. This meeting was a critical step to increase my nativeness (Karlsen and Larrea 2014) and insiderness (Massingham 2014) in the change process. Asymmetries in our relationship decreased.

Factors for SI

A policy making process to share the design and implementation of local development policies among local actors was prioritised. In LS, policy makers are involved in shared leadership processes. LS, therefore, shows that political openness to the participation of other local actors in the LED process is a condition for SI.

Relating SI and AR

The LS case shows a facilitated learning process, in which the researcher guides reflection and contributes the requisite theoretical and conceptual frameworks within which action-initiated insights are considered and socialised. The socialisation of knowledge is another element that is shared by social innovation (Moulaert and Nussbaumer 2005) and cogenerative learning in agoras (Greenwood and Levin 2007; Karlsen and Larrea 2014) (Table 3).

Table 3 Critical incident 3: the political credibility of the expertise in social innovation
Full size table

Motivation of the Actors Involved

Emergence

The motivation of local stakeholders boosted the creation of and the participation in the Industrial Forum. The recognition of a shared problem (the future of the university campus in the county) led to the realisation of common interests among the participants.

Event

Although managed by Mondragon Unibertsitatea (a university inside Mondragon Cooperative Corporation and related to the cooperative tradition in the Basque Country), several members of the territory participated in the process that created a campus of this university in Goierri County. However, the decreasing number of students in the 2011–2012 academic year motivated a meeting of interested people to consider the future of this campus and its role in the county development process.

This collective reflection led to the creation of the Industrial Forum, a new collaboration that committed local stakeholders to act regarding this specific challenge. With this initiative, a valuable opportunity for SI was created. Furthermore, the participants in these initial meetings of the Forum recognised that this new collaboration could solve broader LED issues.

Analysis

The participants recognised the interconnectedness and interdependence among them, and accordingly, the need for a new governance mode was asserted. Instead of the agency inviting the participants to take part in a one-day reflection to define the Strategic Plan for LED, the first step in changing the decision-making mechanisms and channels, to move from reflection to action, was made with the creation of this collaboration space.

The action research process that was initiated to address the participants’ shared problems caused a change in their language; the use of the first-person plural became the common mode of speaking. This change in language reflected the emergence of an intuitive sense of ‘all being in it together’ (Dovey and Muller 2011). Moreover, a new agora for LED had been created and was contributing to the building of mutual trust and to creating an empowerment process. This changing process was already active when I entered the Forum.

Factors for SI

As a collective action space, the creation of the Industrial Forum established another foundation for SI. Furthermore, the recognition of shared interests by a diverse range of committed actors was an important contributor to the process of social innovation. Collective learning, as an alternative to the conventional dyadic relationship between the subordinate individual and redefined power structures, became another important task requiring facilitation in the action research process. Steps for shared leadership were then taken.

Relating IS and AR

The aim of meeting the needs of the community emphasised by the territorial development approach for social innovation (Moulaert 2000; Moulaert and Ailenei 2005; MacCallum et al. 2009; González et al. 2010) coincides with the argument that underlies the primary reason for activating an action research process: solving participants’ problems or needs (Greenwood and Levin 2007; Karlsen and Larrea 2014). Similar to action research, the starting point in social innovation processes is the recognition of a common need or problem among a community of people. This recognition demonstrates another shared aspect of action research and social innovation (Table 4).

Table 4 Critical incident 4: motivation of involved actors
Full size table

Formalisation of the AR Project

Emergence and Event

The linking of the research process to the political process facilitated my reincorporation into the social innovation process. This reconnection occurred when the President and the General Manager of the agency asked me to commence the action research process in the Industrial Forum with the support of another staff member from the agency. The action research process was expected to lead to the collective redefinition of a new governance mode for county development and a shared vision for LED. This objective constituted a key step: the consolidation of the action research process for social innovation had occurred. The proposal to begin this action research project was accepted on the 24th of October 2012 (meeting minute, 24-10-2012).

Analysis and Factors for SI

The formalisation of the action research process led to specific resources being dedicated to this project. This acceptance of resources was a major boost for social innovation because it increased autonomy for action research and it supposed another step towards my nativeness (Karlsen and Larrea 2014).

Relating SI and AR

Participants in action research processes must recognise and legitimise the role of the researcher and must believe that the researcher’s involvement helps overcome the challenges that they face. In this respect, my participation in the action research process was initiated by the President and General Manager of the agency, and I was well known to other participants through my involvement in LED activities several years earlier. These factors led to my acceptance by all of the participants as a capable and legitimate facilitator of the process.

In addition, my mode of facilitation directly concerned the realisation of the emancipatory potential of the Forum as it is found in social innovation processes. The action research process in the Industrial Forum was based on the empowerment of all actors to advance the collaboration and on a shared leadership process to encourage LED. My intention was to facilitate a self-managed transformation process with the development of mutual trust, a common identity and cogenerative learning that resulted in new skills. The shared empowering effect of both social innovation and action research is identified here (Table 5).

Table 5 Critical incident 5: formalisation of the AR project
Full size table

Discussion

In this section, a discussion of the key factors that facilitate social innovation in agoras for territorial development is presented. In the case of Goierri, these factors facilitated a new governance mode for LED because the actors involved in this agora developed collaboration and shared leadership skills through the learning process facilitated by the action researcher. The role that action researchers can play in each factor is also discussed based on the case analysis. The change in the action researcher’s role is shown. As Massingham (2014) discusses, this research approach leads to the generalisability of the research findings through a reductionist approach while being conscious of the findings’ limitations. According to Pettigrew (1997), trajectories of change are probabilistic and uncertain because of changing contexts, and processes not only constrain contexts but shape contexts, either by preserving them or altering them.

The content of each of the following factors that facilitate social innovation is related, and these factors have no linear or fixed order. The factors are not independent; instead, convergent interaction and interconnected loops exist among the five factors.

Territorial Actors must Feel the Need for Building a New Governance Mode Based on Shared Leadership and Collaboration

Network facilitators (critical incident 1), the main policy maker (critical incident 3) and the rest of the local actors (critical incident 4) were conscious of the need to change governance modes to foster LED in the county. The awareness of the need for change can be provoked by various means. The AR process can be helpful because action research can have a catalyst role. This role requires action researchers to focus on opening a dialogue and raising questions to obtain information on actors’ perceptions about an issue, resulting in a shared understanding and a commitment to act on this understanding. Karlsen and Larrea (2014) conceptualise that step as the dialogue of awareness when referring to the sequences of coding and decoding that, through posing problems, make participants realise that there is a need for change. Without this awareness, the change process is paralysed.

Motivation, Common Identity, and Trust Among Territorial Actors

The actors who are involved in collaborative processes that intend to change the modes of interaction, and the institutions that they represent, must believe in the process and feel that they are part of it. This feeling of belonging appears to builds honest communication channels and allows the participants to share a principal motivation (shared challenge/problem) that it makes sense to invest time in the problem. This situation requires participants to be able to confront one another over the value of ideas, without the process degenerating into ego battles and sectarian conflicts. Dialogue among the participants is required in order to construct this base. Action research processes can build cogenerative, democratic and emancipatory learning mechanisms that allow knowledge to be managed in order to activate continuous reflection–action cycles. Action research processes can also facilitate the generation of new skills that permit the shared challenges to be tackled.

Political Compromise and Common Language in the Policy-Making Process

This lesson is illustrated by critical incident 3. The involvement of policy-makers in collaboration processes for territorial development can be counter-productive because hierarchical policy-making processes can be replicated. As the case analysis shows, beyond knowing the need for a change in governance modes to encourage sustainable LED, policy-makers must feel that their political objectives would be achieved and that the action research process would provide ways to advance these goals. The direct dialogue-knowledge cogeneration between the action researcher and policy makers is crucial. The specific approach that would be followed in the action research process must be clarified, and the dialogue between action researchers and policy makers can facilitate this clarification. Likewise, the role of each actor in the action research process must be defined, especially the role of policy makers in the territorial development process. This definition is facilitated by the use of common language and trust, and it also contributes to the nativeness of action researchers.

Formalisation of a Project with Resources Dedicated to SI and Confidence that these Resources will be Used Effectively

Critical incident 5 shows the value of formalising social innovation processes and how this formalization increases the importance of the process both internally (empowerment) and externally (institutionalisation of collaboration and shared leadership). Inversion to change through learning is formalised, and it encourages reflection–action processes, which often require time to avoid what Massingham (2014) considers a fragmented shot gun approach. The action research process can increase the confidence that the resources dedicated to the process will be used properly. Action research organises change as a learning process, and consequently, the cogeneration process progresses though several loops but with a clear objective.

Facilitators Who have Both the Disciplinary Field and Process Knowledge Required by the Social Innovation Process

Facilitation accelerates social innovation. Furthermore, the advantages of both disciplinary field and process knowledge for facilitating social innovation have been shown in the case analysis. In agreement with Karlsen and Larrea (2014) and as this case has shown, action researchers can contribute with both disciplinary field and process knowledge. Disciplinary field knowledge fits the reflection phase of the collective learning processes that occur with agoras, whereas process knowledge makes it possible to activate reflection and take action.

Conclusions

This paper has shown the importance of facilitation in the development of social innovation capabilities, in terms of both collaboration and shared leadership. The facilitation role can be enabled through the allocation of sufficient resources and the provision of appropriate political support and trust. The recognition of the need for social innovation and motivation and trust among involved actors are critical conditioning elements too. Furthermore, facilitation becomes more effective when the facilitator has developed both suitable disciplinary field and process knowledge. Facilitation for social innovation requires then constantly changing contributions that involve knowing how to create a social environment in which democratic and inclusive dialogue supports a self-managed transformation process.

Furthermore, the relation between action research and social innovation has been explored in depth. Beyond the argument of authors who relate both concepts, such as Fontan et al. (2013) who argue that action research is a practical social innovation that contributes to more cooperative, sustainable and convivial research environments, it can be concluded that action research can facilitate social innovation. The importance of knowledge cogeneration and socialisation, shared problem-setting, mutual empowerment, and the development of specific capabilities that facilitate self-managed transformation processes are the elements common to social innovation and action research that have been identified through the analysis. These elements reinforce the argument that social innovation does not occur in abstract spaces by chance but must be facilitated and that both social innovation and action research concern change. In addition, this opens further lines of research related to these shared elements of social innovation and action research.

These conclusions support the argument that an understanding of how social innovation can be facilitated requires a process perspective. Action research is not just a constructionist research paradigm with markedly different ontological and epistemological assumptions that fits better with social innovation research. As social innovation concerns change, action research must include a process perspective in order to understand how that change takes place. This means that beyond a mere methodological approach, the ‘inside-out’ and processual analysis methods contribute by including that process approach in action research analysis. This makes a difference to understanding how social innovation occurs by clarifying why and how the relationships among the actors involved in this action research process.