Using Participatory Scenarios to Stimulate Social Learning for Collaborative Sustainable Development

Interdependent human and biophysical systems are highly complex and behave in unpredictable and uncontrollable ways. Social and ecological challenges that emerge from this complexity often defy straightforward solutions, and efforts to address these problems will require not only scientific and technological capabilities but also learning and adaptation. Scenarios are a useful tool for grappling with the uncertainty and complexity of social-ecological challenges because they enable participants to build adaptive capacity through the contemplation of multiple future possibilities. Furthermore, scenarios provide a platform for social learning, which is critical to acting in the face of uncertain, complex, and conflict-laden problems. We studied the Minnesota 2050 project, a collaborative project through which citizens collectively imagined future scenarios and contemplated the implications of these possibilities for the adaptability of their social and environmental communities. Survey and interview data indicate that these participatory scenario workshops built and strengthened relationships, enhanced participants’ understanding of other perspectives, and triggered systemic thinking, all of which is relevant to collective efforts to respond to social-ecological challenges through sustainable development activities. Our analysis shows that participatory scenarios can stimulate social learning by enabling participants to engage and to discuss options for coping with uncertainty through collaborative actions. Such learning can be of value to participants and to the organizations and decisions in which they are engaged, and scenario processes can be effective tools for supporting collaborative sustainable development efforts.


INTRODUCTION
Many critical environmental and social challenges such as climate change, poverty, ecosystem degradation, and widespread hunger are "wicked" problems (Rittel and Webber 1973) that defy simple solutions.These complex issues emerge from interdependent ecological and human systems and unfold across multiple interacting political, social, and biophysical scales (e.g., Berkes andFolke 1998, Gunderson andHolling 2002).Such challenges that arise from complex social-ecological systems cannot be addressed effectively with narrow "command and control" management (Holling and Meffe 1996), but instead require innovative approaches that incorporate scientific and other forms of knowledge, recognize competing values, and foster communication and negotiation to build collective capacity (Norgaard 1989, Bawden 1991).Engagement of stakeholders and institutions is a critical component of approaches to build adaptive capacity and is essential to conducting effective science for sustainability (e.g., Clark 2007).
Participatory social-ecological research can lead to improved decision-making by developing actions that are more practical and more acceptable to those potentially affected by policies and management (Reed 2008).Involvement of stakeholders in collaborative decision-making also can lead to more equitable governance and support a deliberative democratic approach to governance (Elster 1998), whereby a diversity of values and perspectives are integrated through public deliberation (Yankelovich 1991).Furthermore, application of expert-driven, science-based solutions to sustainability challenges, developed without the involvement of citizens and practitioners, may be untenable in a "shared-power" world (Bryson and Crosby 1992) where actions and decision-making happen at multiple levels of governance and across overlapping public and private institutions.Complex, dynamic, and multiscale social-ecological challenges are illsuited to unilateral regulatory or policy mechanisms; instead, these challenges call for an ongoing process of "socialenvironmental learning" (Finger and Verlaan 1995).learning through the imitation of role models (Bandura 1977) and of experiential learning by adults as they form and reform ideas by testing them against prior experiences (Kolb 1984).Scholars of organizational management expanded discussion of the concept, i.e., beyond analysis of individual cognition, to consider learning within and by interacting groups and organizations (e.g., Argyris andSchon 1978, Senge 1990).The concept of social learning holds promise for sustainable management of complex social-ecological systems (Steyaert and Jiggins 2007) as researchers and managers seek to understand the mechanisms behind effective participatory environmental management processes.Recently, social learning has been studied in community forest management (Wollenberg et al. 2000), water resources (Ison et al. 2007, Steyaert and Jiggins 2007, Pahl-Wostl et al. 2008), the use of natural resources (Rist et al. 2006), wildlife management (Schusler et al. 2003), and environmental risk assessment (Dana and Nelson in press) among other contexts.In particular, social learning is central to the concepts of adaptive management (Holling 1978) and adaptive comanagement (Olsson et al. 2004, Berkes 2009) because learning among groups fosters adaptive capacity to cope with socialecological complexity and to respond to an uncertain future (Tompkins and Adger 2004).
In this paper, we understand social learning to occur when group interactions change individual knowledge and understanding, and this individual learning subsequently influences and informs the group's knowledge and actions (Reed et al. 2010).In particular, social learning emerges among groups of individuals who share differing knowledge and experiences, and it involves the revelation and integration of different and often contrasting participant viewpoints (Mostert et al. 2007).Social learning is both an outcome of, and an essential input to, effective cooperation within a group (Berkes 2009).It arises from a process through which individuals become aware of how others understand reality and reflect upon the alternative ideas and experiences they encounter (Schusler et al. 2003, Keen andMahanty 2006).Such learning occurs when people interact and share diverse perspectives and experiences, and thus build relationships and develop networks through the process of engaging with others (Schusler et al. 2003).Social learning has the capacity to transform a group of individuals into a community that shares a common interest or goal (Webler et al. 1995), or into a group that can undertake collective action (Wenger 1998, Röling 2002).As such, social learning is an iterative and continuous process that is thought to enhance the flexibility of a socialecological system and increase its ability to respond to change (Pahl-Wostl and Hare 2004).
The growing body of scholarship about the value of social learning has generated important insights for natural resource management and sustainable development, and it has stimulated interest in "new platforms and processes for facilitating social learning" for coping with the complexity of interdependent human and ecological systems (Woodhill and Röling 1998, p. 47).We examined participatory scenarios as one such potential platform, and evaluated the social learning that emerged from a collaborative scenario process in the state of Minnesota.

Scenarios
Scenarios, as we employ the term, are narratives deliberately crafted to describe multiple plausible futures.They are alternative "possible views of the world" created to help improve understanding and decision-making (Ringland 2002, p. 3).Scenarios allow for expressions of ambiguity through "qualitative causal thinking" (van der Heijden 1996, p. 15) and thus have proven to be well-suited to tackling complex social-ecological problems that are fraught with uncertainty (e.g., Peterson et al. 2003, Bohensky et al. 2006, Carpenter et al. 2005).
Scenarios also offer a vehicle for learning; preparing for an uncertain future shaped by the complex behaviors of socialecological systems requires such learning on an ongoing basis (van der Heijden 1996).Scenarios help integrate varied and diverse ideas by enabling users to coalesce a range of insights into coherent narrative frameworks.They also can be an effective means for individuals to encounter others' viewpoints and expand their own mental models in response (Schwartz 1996, Garb et al. 2008).Scenarios about humanenvironment interactions help participants learn not only about their social-ecological systems, but also about the values and worldviews that shape their own and others' approaches to addressing sustainable development challenges (Bawden 2007).Furthermore, the orientation towards multiple possible futures provides a nonthreatening environment in which to engage other opinions (van der Heijden 1996, Ringland 2002).Scenarios are often created by "futurists" and "professional builders of scenarios" (Schwartz 1996 p. 10).Yet scenarios need not be the domain of experts alone, rather they can be used to enhance collective action and serve as a platform for collaboration between scientists, citizens, and other stakeholders.Unlike forecasts, which are a "statistical summary of expert opinion" (Berkhout et al. 2002, p. 87), narrative scenarios often do not quantify uncertainty and can be powerful tools for stimulating learning that is independent of technical expertise (Carpenter et al. 2006).Indeed, participation of groups with diverse knowledge, values, and expertise is fundamental to effective scenario development because the strength of scenarios derives from incorporating a breadth of ideas about the future.In recent years, participatory scenarios have been developed in a range of management and decision-making contexts (e.g., Berkhout et al. 2002, Hulse et al. 2004, Kok et al. 2007, Patel et al. 2007, Enfors et al. 2008, Bohensky et al. 2011, Kok et al. 2011).http://www.ecologyandsociety.org/vol17/iss2/art9/Such participatory scenario processes can provide a platform for dialog among citizens, technical experts, and decision makers, and help develop shared capacity to address sustainable development challenges.
We launched a participatory scenario development project in the state of Minnesota in 2007 to evaluate the potential of scenarios to foster participants' capacity to pursue sustainable development in a context of social-ecological uncertainty.We carefully designed participatory scenario processes to stimulate social learning among participants.The key social learning objectives were: q provide a forum for active, equitable and creative discussion of complex social-ecological systems and inherent future uncertainty; q build participants' capacity in systemic and strategic thinking related to regional sustainable development projects; q link different forms of knowledge, including scientific and experiential, and enable participants to share and negotiate different worldviews and perspectives; q enable participants to create new relationships, and to enhance existing ones, to build their capacities for collective action; and q inform individual behavior and guide decision-making by individuals and participating organizations and institutions In this paper we first discuss the unique participatory scenario approach we employed, and then we present results related to the achievement of the above social learning objectives.

Minnesota 2050 project
The project-Minnesota 2050: Pathways to a Sustainable Future (hereafter referred to as MN 2050)-emerged in 2007 as a collaborative effort between researchers at the University of Minnesota and citizens working with the Regional Sustainable Development Partnerships (RSDPs).The University of Minnesota is a major research university in the United States and it is a public land-grant institution dedicated to substantial engagement of the citizenry of Minnesota.The RSDPs are a citizen-driven network affiliated with the University of Minnesota that works to foster sustainable development around the state.Citizen and faculty boards guide five semi-autonomous regional partnerships and identify community priorities in renewable energy, sustainable agriculture and local foods, natural resource management, and tourism, and leverage the University's resources and expertise to implement local projects.The RSDPs essentially function as a "bridging organization" (Berkes 2009, Folke et al. 2005)

Workshop design
Social learning does not simply emerge from every group interaction; rather, it depends on the political, institutional, and social contexts, and on the nature and structure of the participatory process employed (Tippett et al. 2005).With this in mind we designed our scenario workshops to include the participatory process elements that are necessary to stimulate social learning (Fig. 1) (Schusler et al. 2003, Mostert et al. 2007, Muro and Jeffrey 2008).The key process components we chose were: facilitation, democratic structure, and diversity of participants (Daniels andWalker 2001, Mostert et al. 2007).Skilled facilitation is critical to participatory processes because facilitators set the tone for interaction and ensure that the physical and social space is tailored to the needs of the conversation and the participants (Daniels and Walker 2001).A democratically structured process is one which ensures that all participants are able to contribute equally and feel that their input will be heard and respected (Dana and Nelson, in press).Diverse participation promotes inclusion of a greater breadth of worldviews, mental models, and lived experiences in the process (Jones et al. 2011).
A participatory process that incorporates these essential components can generate process attributes that support social learning among participants, including: diversity of knowledge, open and effective communication, unrestrained thinking, constructive conflict, and extended engagement (for example, Daniels and Walker 1996, Schusler et al. 2003, Mostert et al. 2007, Muro and Jeffrey 2008).Diversity of participants enhances the likelihood that the process will incorporate a range of information and varying forms of http://www.ecologyandsociety.org/vol17/iss2/art9/experiential and scientific knowledge (Berkes 2009).Open and effective communication is critical for honest discussions and frank engagement to occur among various knowledge types and those with differing values and opinions (Kahane 2004).Unrestrained thinking fosters creativity, encourages openness to new ideas, and fosters learning across knowledge types (Mostert et al. 2007).Conflict can easily emerge from open communication about contested values and worldviews, yet if managed appropriately, constructive conflict can foster new understanding among diverse participants (Daniels and Walker 2001).Finally, extended engagement enables ongoing and iterative interactions to occur among participants and is also important for fostering social learning (Dana and Nelson, in press).In all, nearly 200 individuals participated in the workshops.Among the participants were: farmers; small business owners; health care providers; natural resource professionals; county, municipal, and state government leaders; city planners; religious leaders; teachers; high school and college students; and corporate executives.
During Phase I, we convened 1.5-day workshops to develop exploratory scenarios.Participants, working in 4 to 6 small groups of 5 to 6 persons each, used the focal question "How are we interacting with the landscape and natural resources in 2050, and how is the environment affecting our quality of life?" to imagine multiple regional scenarios.The year 2050 was chosen because the researchers and the RSDP leadership wanted to challenge participants to think beyond conventional planning time scales, and in some cases beyond their lifetimes, while still imagining futures that would be very real for their children or grandchildren.We employed the In Phase I workshops, small groups completed the [I]NSPECT process, with 20 to 30 minutes allotted for each of the six categories.First, individuals independently envisioned and noted possible future conditions and important drivers of change.Then the small group shared and collected individual insights at the end of working through each category.The [I] NSPECT process generated a series of flip charts, one from each small group, filled with sticky notes containing numerous distinct future possibilities for each category.Following the [I]NSPECT process, the small groups rotated around the room to a different group's set of flip charts.Participants then worked in pairs or small groups to quickly create "scennarratives" by weaving together meaningful future possibilities from the other groups' notes.Then these scennarratives (usually from 1 to 3 per small group, totaling approximately 8 to 15 per region) were shared and discussed among all participants in order to collect additional reflections.
We convened subsequent 1.5-day workshops, 8 to 12 months later during Phase II.Prior to the Phase II workshop in each region, we integrated the scen-narratives generated during Phase I into 4 to 5 more coherent and internally consistent scenarios (Appendix 1).During the Phase II workshops, participants were asked first to individually "inhabit" the scenarios by imagining themselves living in the future time and under the natural, social, political, economic, cultural, or technological conditions described.Participants first acclimated to the potential futures by imagining routine and daily aspects of life, such as how they might travel or prepare meals in the potential future scenarios.Then participants imagined pursuing tasks such as ensuring clean water, securing ample healthy food, or providing sufficient energy.Each small group inhabited two of the 4 to 5 alternative scenarios, and then worked to "backcast", or identify policies or actions that, if taken soon, would make the accomplishment of these important tasks easier.Significantly, the small groups were not asked what could be done in the present to achieve a desirable future or avoid an undesirable one, because in these exercises the core challenge was adapting to, rather than creating the future.Phase II workshops also incorporated a modeled scenario segment in which researchers presented quantitative analysis of critical trends as they might unfold under the alternative scenarios (Schmitt Olabisi et al. 2010).
Two core assumptions of the MN 2050 scenario approach were somewhat distinct from other scenario development processes.First, we characterized the future as being not only uncertain but also to a great extent uncontrollable.We asked participants to reflect on the limited geographic and decisionmaking domains over which they could exercise influence and to understand that national and global social, economic, and biophysical forces would significantly determine future conditions.Rather than seeking to design and create a preferred future, their primary task would be to respond and adapt appropriately.The second distinguishing assumption central to this scenario exercise was the recognition that the process, not the scenarios themselves, was the primary goal.
Participants were fairly unconstrained in the scenario creation process; all scenarios were presumed to be plausible in the exercise, and for the purposes of deriving insights about resilient actions and policies (Appendix 1).Following from this assumption, we do not present the scenarios as results in this paper; rather we discuss in the Results section our analysis of the social learning outcomes.The scenarios appear in Appendix 1.
Both Phase I and Phase II workshops included dedicated time for individual and group reflection, to enable participants to derive meaning from the multiple scenarios that were developed and ponder the divergent values and perspectives surfaced by the group.The workshop elements and the repeated opportunities for engaged discussions within and across groups of participants were specifically designed to foster social learning among participants.

Evaluation of social learning outcomes
We evaluated social learning outcomes using researcher observation, surveys, and interviews, which are methods Table 1.Relationships and appreciation of others' perspectives.Percentage of participants who 'strongly agree' (5 on 1 to 5 scale) and 'mildly agree' (4 on 1 to 5 scale) with the following survey statements (n = 39 surveys), and the total number who agree (either 4 or 5 on 1 to 5 scale).comparable to those employed in other assessments of social learning (Schusler et al. 2003, Tippet et al. 2005, Mostert et al. 2007).We analyzed data from only the final three workshops because ongoing adaptation of the scenario process in the earlier workshops rendered the process insufficiently uniform to evaluate social learning outcomes at that point.We collected 39 end-of-workshop surveys and completed 14 semistructured phone interviews with select survey respondents about a month after their participation in Phase II workshops.The surveys elicited responses to seven statements about the four social learning outcomes (Tables 1,2, and 3).The interviews included general questions about the scenario workshops and process, and the interviews targeted questions about relationships, workshop interactions, and systems thinking, and actions and outcomes.Multiple researchers observed participant behavior and communication during each of the three workshops, noting evidence of key social learning outcomes as well.Additionally, we conducted follow-up interviews with the five RSDP directors approximately one year after the conclusion of the MN 2050 project to investigate longer-term impacts attributable to the scenario workshops.
We evaluated four commonly studied, and critical, social learning outcomes: ability to think systemically, appreciation of others' perspectives, new or deeper social relations, and anticipated behavioral impacts and outcomes (Muro and Jeffrey 2008).Systems thinking is a crucial component of social learning because individuals are better able to adjust their own mental models when they understand multiple facets of an issue (Daniels and Walker 2001).An appreciation of others' ideas and perspectives is an important outcome because individuals who understand and value different and potentially contrasting worldviews are more likely to learn together and from one another (Mezirow 1996).New and enhanced relationships within the group can indicate that in addition to gaining an understanding of others' perspectives, individuals are also building a foundation on which to deepen collaboration and collective action (Schusler et al. 2003).
Finally, changes in individual behavior or the actions of the group thought to arise from a participatory process can provide strong evidence that social learning has occurred (Muro and Jeffrey 2008).

Process components and attributes
Drawing on researcher observation of the scenario workshops, we reflected on the extent to which the workshops provided the critical process components and attributes (Fig. 1).The MN 2050 workshops were intentionally designed to provide a democratic structure for the purpose of enabling participants of varying ages, professions, and expertise to engage actively in small and whole-group discussions.Individuals were given equal opportunities to ask questions, provide feedback, and respond to other participants' perspectives.The MN 2050 workshops benefited from excellent facilitation, provided by a professional facilitator with more than 25 years of experience in natural resource planning and environmental management.
The facilitator created a comfortable environment and encouraged participants to embrace the improvisation and playfulness of the scenario exercises while acknowledging the seriousness of core values expressed through the process.The democratic structure and excellent facilitation stimulated open communication throughout the workshops.The process was structured to provide repeated opportunities for feedback and questions, and participants were also able to communicate in writing with workshop organizers through packets, evaluations, feedback forms, and group email.No doubt issues of power and status among individuals and between participants and researchers influenced interactions (e.g., Lyon et al. 2010), but the structure and tone of the MN 2050 workshops were designed to be inclusive and open for all participants.Consequently, individuals actively communicated with each other and workshop organizers expressing differing understandings about the topic of sustainability, responding to the opinions of others, and commenting about the scenario process itself.The nature of the MN 2050 scenario process also allowed for unrestrained thinking by prompting individuals to imagine multiple future possibilities for both social and ecological systems, and participants testified to the creative and free thinking encouraged by the workshops.
Table 2. Systems thinking.Percentage of participants who 'strongly agree' (5 on 1 to 5 scale) and 'mildly agree' (4 on 1 to 5 scale) with the following survey statements (n = 39 surveys), and the total number who agree (either 4 or 5 on 1 to 5 scale).Although some individuals initially found the task of imagining a distant future very challenging, the process quickly stimulated creative brainstorming among participants.
The MN 2050 scenario workshops were less effective at stimulating constructive conflict, perhaps because discussions that were focused on future possibilities were removed from current politically charged issues.For example, conflicting viewpoints about current energy policy were uncovered but conflict was muted because of general agreement that in future scenarios energy would be used more efficiently and would come increasingly from nonfossil fuel sources.The workshops also were less successful at ensuring diverse participation, as nearly all the meetings were convened in rural regions and involved participants who shared some interest in sustainable development and a connection with the RSDPs.Despite this common affiliation, the workshops still included individuals from a variety of professions, with a range of personal interests and experiences, and from various ideological and political perspectives.Nonetheless, more diverse participation, particularly by younger people and underrepresented groups such as new rural immigrants and members of neighboring First Nations, likely would have enhanced the diversity of knowledge types included in the discussions.Finally, the MN 2050 process was also less successful at providing for extended engagement because participants interacted through the MN 2050 workshops for only three days over the course of a year and a half.With significant additional project resources, more extended engagement could have been supported and potentially more opportunities to foster social learning could have been offered.

Social learning outcomes Enhanced relationships
The majority of survey respondents (32 of 39) agreed strongly that participating in the scenario workshop enhanced existing relationships and helped create new ones as well (Table 1).
Although workshop participants were affiliated with the RSDPs, interviewees indicated that they had not known many of the other individuals prior to the workshop but that became acquainted with new people very quickly through the process.
One respondent stated that he had known about a quarter of the participants before the scenario workshop and continued his relationships with new acquaintances afterwards.Another interviewee made connections with people that he felt he would not have met otherwise, and believed that the scenario process fostered rapid networking within the group.Several interviewees further stated that the process built or enhanced relationships such that they can contact other workshop participants to exchange ideas or seek collaborative opportunities.Additionally, there was agreement among survey respondents (36 of 39) that the scenario workshop increased their belief that sustainable development challenges are most effectively addressed collectively (Table 1), and as one interviewee stated, nurturing connections among people is "extremely important for sustainability".Four of the five RSDP directors also indicated that the scenario workshops had provided potent networking and relationship-building opportunities that will prove valuable over time.One director referred to a community "green fair" that had organized by newly connected workshop participants in the months following one workshop.Another director, however, indicated that the workshops had minimal relationship-building value because many participants already knew each other, and busy schedules and project overload make it difficult to cultivate new partnering opportunities.

Appreciation for different perspectives
Nearly two-thirds of survey respondents (24 of 39) strongly agreed that the scenario workshops improved their understanding of others' perspectives; while about one-third indicated mild agreement, and only one respondent thought the workshop had not improved this understanding (Table 1).Many interviewees indicated that they enjoyed and appreciated the "good mix of people" participating, and one person said that the workshop participants were not the people with whom she "normally hangs out".One interviewee stated that because he had not encountered many other participants "like him" at the workshop, he left with a greater understanding Table 3.New or changed actions.Percentage of participants who 'strongly agree' (5 on 1 to 5 scale) and 'mildly agree' (4 on 1 to 5 scale) with the following survey statements (n = 39 surveys), and the total number who agree (either 4 or 5 on 1 to 5 scale). of others' values.Another participant admitted that he began reflecting on his notions about the role of economics in supporting sustainable development after being challenged during the scenario process.A third interviewee observed that talking with others allowed him to understand "their vision of reality" and that the scenario process gave contrasting opinions "more legitimacy".Three interviewees also stated that the workshops were much more inclusive of participants' opinions than other meetings about sustainability in which they had participated.They responded that "no one was the boss" during conversations, that all individuals' experiences were highly valued in the process, and that the open nature of discussions led to a "rich tapestry of ideas".Two interviewees also indicated that working in fiction and the future helped them to dissociate from their own ideas and that the scenario process allowed people with disparate views to sit at the same table and talk.Our analysis suggests that participants in the scenario workshops not only became better acquainted with one another, but they also, in a short time, gained a greater understanding of the opinions and perspectives of participants with different worldviews.

Systemic thinking
The MN 2050 scenario development process was quite systemic in nature because individuals focused on conceiving the wholeness of future possibilities.The [I]NSPECT process emphasized the interdependence of various components of social-ecological systems and obligated participants to reperceive their worlds as manifestations of interacting components of human and environmental systems.The majority of survey respondents (32 out of 39) either agreed or strongly agreed that the scenario workshops had improved their ability to think systemically about connections between humans and the environment (Table 2).Four interviewees stated that the scenario process and its longer time horizons stretched their ability to think systemically.Three interviewees indicated that the workshops and visioning were unlike any other conversations in which they had participated and that they were a great way to break free from "normal thinking".One participant was "jazzed up" by the imaginative systems thinking of the process and said that the scenarios helped him see beyond his own time and place.Another respondent indicated that the process allowed him to "feel" rather than simply think in typical ways, and to flex a "part of his brain not normally used".Several interviewees and survey respondents claimed that they were already thinking systemically prior to the workshops but that the scenario exercise reinforced their perception of interdependence between social and ecological systems.The directors observed that systemic thinking skills carried forward into other discussions after the workshop, with three directors noting that board members seemed more willing to have broad discussions of the RSDPs' long-term mission and goals following participation in the scenario workshops.

New or changed actions
Another important outcome of social learning is the use of newly acquired individual and collective knowledge to inform and motivate action (e.g., Muro and Jeffrey 2008).To evaluate the scenario workshops' potential influence on subsequent actions, we asked participants whether they thought their participation might influence their personal and professional lives.The majority of survey respondents agreed or strongly agreed that the scenario workshops would influence their professional work (35 of 39) and personal lives (34 of 39).Several interview respondents indicated that the scenario workshops had already influenced them.One related that he had approached meetings with farmers differently following his participation, and that he had introduced the idea of being adaptable to future uncertainty into these usually conventional and conservative conversations.Another stated that he now thinks differently about priorities and regards resilient local food systems to be equally as important for sustainability as renewable energy.Two other interviewees claimed that they had made some modest lifestyle changes such as reducing water use or increasing recycling, but that the impact of the scenario workshops was receding as they returned to their busy day-to-day lives.Three interviewees indicated that they are now reconnected to the RSDPs and more engaged with these groups' sustainable development activities.The directors echoed this, indicating that some individuals became actively engaged or newly connected with the RSDPs following the workshops.
The directors also reported that the scenario exercise helped focus and prioritize the RSDPs' efforts.In particular, four of the five regions became more focused on water issues because http://www.ecologyandsociety.org/vol17/iss2/art9/ of the prevalent discussion of water quality and availability across multiple scenario exercises.One region reformulated a committee to focus more on water and others developed plans to fund more water-related research and monitoring or to support education and outreach activities with local partner organizations.Another region began developing a new program to address ecological health that was, according to the director, partly inspired by the scenario workshop discussions.Although the directors all indicated that it was difficult to pinpoint direct impacts from the workshops on the RSDP programs and activities, each of them stated emphatically that the scenario exercise was extremely worthwhile and should be a regular and recurring tool for their organization.

CONCLUSIONS
The participatory scenario process developed and implemented in the MN 2050 project provided a valuable experience for the individuals involved and stimulated important elements of social learning.Previous social learning evaluations related to environmental decisionmaking or natural resource management have looked for a mix of changes in actions (e.g., new policies, management practices, or pilot projects) and collective capacities (e.g., changes in relationships, knowledge, perspectives, or trust) (Mostert el al. 2007).The social learning outcomes that we observed are similar to those found in studies that investigated changes in the collective capacities of stakeholders after engagement in a participatory process.Scenario workshop participants reported increased knowledge, improved systemic thinking, enhanced relationships, and awareness of new perspectives, all of which are valuable for developing adaptive capacity.These outcomes are similar to those identified in other studies, for example in Measham (2009) where stakeholders worked to manage dry land salinity, and in Schusler et al. (2003) where stakeholders developed Lake Ontario management options, among others.Some social learning evaluations have focused on changes in environmental policies and decision-making, but incorporation of new knowledge and perspectives from other individuals has been identified as a prerequisite for these subsequent higher level institutional changes (Pahl-Wostl et al. 2007).
Although we found some evidence of changes in the organizational practices and decision-making of the RSDPs after the scenario workshops, we did not evaluate such potential longer term impacts in detail, but we focused instead on the emergence of critical precursors that support ongoing social learning.However, paper makes a new contribution to social learning scholarship specifically by suggesting that participatory scenarios can support collaborative sustainable development by nourishing changes in a group's fundamental collective capacities.
Limitations in the design and implementation of the MN 2050 process likely diminished the extent to which the workshops stimulated social learning and limited the researchers' opportunities to evaluate the process.We encountered project challenges that are likely to confront other participatory scenario efforts, and we offer some recommendations based upon our analysis and reflections of the MN 2050 project.Attention to these issues may help other researchers maximize social learning among participants and enhance the impacts of future scenario processes.
The first issue is the importance of ensuring the group of participants is diverse.Balanced and appropriate social diversity is a core component of resilience and foundational for enhancing collective action (Nelson et al. 2011).Groups with too little diversity converge quickly and without exchanging contrary ideas, and will necessarily develop fewer options for action.Yet too much diversity can hobble a group and overwhelm a process with too many conflicting values and perspectives.In our experience, the MN 2050 project had too little diversity given the demographics of the regions and the issues of concern; younger participants, immigrants, tribal members, and other rural citizens likely would have brought different perspectives to discussions of local foods, and energy and natural resource use.Our project, and potentially other future participatory scenario efforts, would benefit by ensuring diverse participation that is carefully attuned to the social context and the issues of interest.
The second challenge is the effective promotion and management of constructive conflict.Scenarios allow participants to engage with each other through hypothetical futures and, as a result, may offer a potentially unique vehicle to balance tension among participants (Garb et al. 2008).If channeled properly, conflict can lead to a sharpening of participants' insights and an articulation of mutually acceptable pathways for action (Daniels and Walker 2001).Some scenario development projects have engendered quite a bit of conflict, for example when resulting scenarios were to be used to plot the course of a new government (e.g., Kahane 2004).However, minimal tensions arose in the MN 2050 workshops, likely because participants were discussing distant futures and were not required to choose a particular scenario or identify a preferred policy or decision-making pathway.Future applications of scenario processes could realize potentially greater social learning benefits by stimulating constructive conflict.
A third, and critical, consideration in developing and evaluating processes to promote social learning is the length of engagement among participants (Muro and Jeffrey 2008).
In the MN 2050 project each regional community participated in three workshop days over the course of a year and a half, and groups often lacked time to fully discuss important issues and reflect on strategic actions identified through backcasting.We suggest that providing more frequent and ongoing interaction enables participants to get to know and understand other individuals' perspectives and to build trust in each other and the participatory process.Allowing time specifically for informal engagement is particularly important, in order for participants to seek out information or expertise from others and to further build relationships (Dana and Nelson, in press) 2011).However, it could be problematic to attribute social learning outcomes to a particular process rather than to other incidental and independent factors.This challenge notwithstanding, additional longer-term social learning outcomes could be evaluated, including increased trust between participants, the empowerment of communities or enhanced institutional effectiveness and changes in formal environmental decision-making or policies (Daniels and Walker 2001, Pahl-Wostl and Hare 2004, Tippett et al. 2005, Mostert et. al. 2007, Steyaert and Jiggins 2007).Although such outcomes often take significantly more time to emerge than those we assessed during the MN 2050 project (Measham 2009), extended engagement of participants could enhance opportunities, for nurturing social learning as well as for evaluating potential outcomes.
A final and related element that is important to the implementation of scenario processes is the direct and explicit connection between qualitative scenario exercises and concrete decision making.Social learning likely can best support sustainable development when it is embedded in robust institutions and is directly linked to specific decisionmaking (Rist et al. 2006).Although the MN 2050 scenarios generated insights useful to the ongoing work of the RSDPs, the project was directly linked with the formal decisionmaking process of the Statewide Conservation and Preservation Plan through only a single scenario workshop.This workshop was quite effective for the participants who attended but it was just one exercise and occurred fairly late in that natural resource planning process.More iterative connections between the discussion of quantitative scientific information and the qualitative scenario conversations could have stimulated a robust cross-pollination of knowledge types that might have generated useful insights for the Statewide Conservation and Preservation Plan.
However, the incomplete link between the MN 2050 scenarios and direct actions is also partly a function of the fundamental difference between scenario thinking and typical decision processes.Scenarios employ open-ended, creative systems thinking while policy discussions and institutional decisionmaking tends to focus on individual issues or specific nearterm problems.Recent research has begun exploring ways of using visualization tools (e.g., Shaw et al. 2009), mixed qualitative-quantitative methods (e.g., Kok 2009), and other approaches to link insights from scenarios with decisionmaking, but the fundamental differences between scenarios and decision processes make the linking of the two types of processes challenging.
Overall, our analysis adds to the literature by suggesting that scenarios can facilitate social learning about social-ecological systems, and therefore that scenarios constitute a potentially important tool for enhancing collaborative sustainable development efforts.However, to further expand the potential power of this tool, scenario thinking must be more integrally designed into scientific research and traditional decisionmaking processes.Future research that evaluates a variety of scenario processes-employed as part of a strategic participatory approach and explicitly linked to decisionmaking processes-would be useful for enhancing our understanding of the value of scenarios in supporting collaborative sustainable development.

Fig. 1 .
Fig. 1.Analysis framework for social learning process components, resulting attributes, and potential outcomes.Drawn from Dana and Nelson in press,Mostert et al. 2007, Schusler et al. 2003, and Daniels and Walker 2000.
[I]NSPECT process (developed by R. Bawden (personal communication) following from Schwartz 1996), which asks participants to address the focal question by imagining and describing the separate categories of the Natural, Social, Political, Economic, Cultural, and Technological aspects of the possible futures.The "I" in the [I]NSPECT acronym denotes the Interpretation of the individual undertaking the scenario-building exercise, thus explicitly recognizing the importance of individuals' varying worldviews.
Fig. 2. The MN 2050 scenario process.Gray shading highlights the elements of the process that are discussed in this paper.Other papers focus on the interplay between qualitative scenarios and quantitative modeling (Schmitt Olabisi et al. 2010) and on the design and effectiveness of the overall learning system (Kapuscinski et al. unpublished manuscript).
Steyaert and Jiggins 2007)used analysis of completed project documents, information about changes in environmental policies and practices, and the emergence of new organizations as evidence of social learning.Longer engagement with participants also could allow iterative collection of survey or other data, enabling analysis of potential changes in participants' perspectives or identification of emerging consensus (e.g., Stone-Jovicich et al.