Learning effects of interactive decision-making processes for climate change adaptation
Authors and Affiliations
*Julia Baird a , Ryan Plummer a b, Constanze Haug c, Dave Huitema c d
a Environmental Sustainability Research Centre, Brock University, St. Catharines, Ontario, Canada, L2S 3A1. Tel: 905-688-5550 x 4782; E-mail: [email protected] b Stockholm Resilience Centre, University of Stockholm, Sweden c Institute for Environmental Studies (IVM), VU University Amsterdam, The Netherlands d School of Sciences, Netherlands Open University, The Netherlands
Acknowledgements
The authors thank all the participants in the action research project associated with this paper. We are indebted to Kerrie Pickering for her tremendous efforts throughout the research project and are grateful for the insights and contributions of Brad May, Samantha Purdy and Nicole Klenk. Financial support for the research came from Environment Canada through a Grants and Contributions Agreement with Brock University. Contributions by Constanze Haug and Dave Huitema were made possible by support from the Dutch Knowledge for Climate Program and the European Union’s 7th Framework Programme (The Responses project, Grant Agreement number 244092).
1 Abstract
Learning is gaining attention in relation to governance processes for contemporary environmental challenges; however, scholarship at the nexus of learning and environmental governance lacks clarity and understanding about how to define and measure learning, and the linkages between learning, social interactions, and environment. In response, this study aimed to advance and operationalize a typology of learning in an environmental governance context, and examined if a participatory decision-making process (adaptive co-management) for climate change adaptation fostered learning. Three types of learning were identified: cognitive learning, related to the acquisition of new or the structuring of existing knowledge; normative learning, which concerns a shift in viewpoints, values or paradigms, and relational learning, referring to an improved understanding of others’ mindsets, enhanced trust and ability to cooperate. A robust mixed methods approach with a focus on quantitative measures including concept map analysis, social network analysis, and self-reflective questions, was designed to gauge indicators for each learning type. A participatory decision-making process for climate change adaptation was initiated with stakeholders in the Niagara region, Canada. A pseudo-control group was used to minimize external contextual influences on results. Clear empirical evidence of cognitive and relational learning was gained; however, the results from normative learning measures were inconclusive. The learning typology and measurement method operationalized in this research advances previous treatments of learning in relation to participatory decision-making processes, and supports adaptive co-management as a governance strategy that fosters learning and adaptive capacity.
1. Introduction
How contemporary environmental challenges are understood and the corresponding requirements of governance processes are central issues to global environmental change. Crona and Parker argue that “humanity faces increasingly intractable environmental problems characterized by high uncertainty, complexity, and swift change. Natural resource governance must therefore involve continuous production and use of new knowledge to adapt to highly complex, rapidly changing social-ecological systems to ensure long-term sustainable development” (2012, p. 32). Themes of adaptiveness, flexibility, and learning are receiving growing attention in environmental governance scholarship (Folke et al., 2005; Gerger Swartling et al., 2011; Armitage et al., 2012; Crona and Parker, 2012).
Several social processes which purport to enable learning, confer flexibility and encourage experimentation have emerged in relation to environmental governance. A few among the many potential examples include adaptive governance (e.g., Folke et al., 2005; Pahl-Wostl, 2009), adaptive co-management (Armitage et al., 2008; Berkes, 2009) and deliberative approaches, such as focus groups, round tables, social learning groups, and citizen juries (e.g., Rowe and Frewer, 2000; Huitema et al., 2010). Authors have argued that such innovative governance mechanisms which couple the potential for social learning with collaboration build adaptive capacity of individuals and collectives (Keen et al., 2005; Fazey et al., 2007; Plummer and Armitage, 2010).These mechanisms are especially critical for addressing climate change. As Pelling et al. (2008) argue, considering adaptation in terms of learning makes clear adaptation and institutional alteration as genuine adaptation strategies as well as raises questions about the processes by which individuals may learn to be adaptive.
At the same time as interest in environmental governance mechanisms that confer learning and adaptability is growing, several researchers are raising critical questions about the state of
2 scholarship emerging at the nexus of learning and environmental governance (e.g., Armitage et al., 2008; Muro and Jeffrey, 2008; Diduck, 2010; Crona and Parker, 2012). In summarizing these critical questions, Crona and Parker (2012) point out that there is: a dearth of consensus about how to define or measure learning; limited understanding of the relationship between social interactions and learning; amorphous ideas of how environments shape learning; and, little appreciation for how conflict and power dynamics influence learning. Their summary leads them to argue that “new concepts, methods, and metrics for conceptualizing and measuring learning in support of natural resource governance and testing the conditions under which it can be achieved are therefore badly needed” (Crona and Parker, 2012, p. 32). Given the cost, time and effort involved in designing and implementing interactive environmental governance processes one would expect a significant degree of evaluation activity to assess their outcomes. Recent systematic reviews of the burgeoning literature on social learning in natural resource management by Rodela (2011) and Rodela et al. (2012) find that few studies even attempt to empirically assess learning effects of specific interventions on participants.
This paper responds to the imperative for environmental governance mechanisms to bring about learning, the voids in this area of scholarship as noted by Crona and Parker (2012), and the need to assess learning in relation to interactive decision-making processes. It specifically aims to: 1) advance, test and operationalize a typology of learning (cognitive, normative, and relational) in an environmental governance context; and, 2) examine if an adaptive co-management intervention concerned with climate change adaptation fosters learning. The following section positions the research in relation to scholarship on learning and the environment and focuses on the relationship to participatory processes, the call for greater specificity in understanding and measuring learning, and the development of an appropriate typology of learning. These strands are brought together in the methods section where the mixed methods procedures for assessing learning using indicators from the typology in relation to an adaptive co-management intervention concerning climate change adaptation are detailed. Results from the research are presented according to the three-fold typology. The discussion relates the findings from this research to scholarship concerning learning and participatory environmental processes and policies, the value of enhancing specificity about types of learning, and the efficacy of the measurements proposed. Several avenues for future research are set forth in the conclusions.
2. Learning and environmental governance: a selected overview of insights
The breadth and depth of scholarship on learning across disciplines is vast; indeed, even a focus on the conceptual landscape of environmental learning is rich and complex, intersecting with theoretical strands and cutting across education, psychology and social psychology (see Lundholm and Plummer, 2010 for an overview). As a result, a review of all learning related literature is beyond the scope of this study, and accordingly this section positions our specific interest in environmental governance and learning within the broader landscape of scholarship concerning the environment and learning. We concentrate on setting forth the strands of scholarship relating most directly to this research, and in pulling together strands from these related areas we advance a typology of learning.
Many discussions on environmental governance start from the notion of ‘wicked problems’ (Rittel and Webber, 1973), that is, problems that defy easy solutions. It is now almost commonplace that dealing with wicked problems requires an approach that is focused on learning (Stirling, 2006; Voβ et al., 2006), and it should thus not be a surprise that several
3 concepts have emerged in this context. Examples include ‘social learning’ (e.g., Parson and Clark, 1995), ‘collaborative learning’ (e.g., Vernooy, 2010), ‘learning communities’ (Kilpatrick et al., 2003) and ‘communities of practice’ (Wenger, 1998). Social learning is singled out in particular for “….becoming a normative goal in natural resource management and policy” (Reed et al., 2010, p. 15). Muro and Jeffrey observe that, “despite the lack of a coherent theoretical foundation and a clear definition, a common understanding of the process social learning entails, its outcomes and contributions to natural resource management emerges from the literature. At the core of these models is a process of collective and communicative learning, which may lead to a number of social outcomes, new skills and knowledge” (2008, p. 330). Learning, as generally understood in this area of scholarship, complements the shift in focus from management to governance as well as the contemporary emphasis on conditions of complexity, uncertainty and value conflicts. Social learning thus is said to come about through an inclusive, communicative and participatory process; take a systems and integrative orientation; be action oriented and anticipate iterative adjustments; and, develop an understanding of change through multiple means (Leeuwis and Pyburn, 2002; Keen et al., 2005; Diduck et al., 2005; Pahl-Wostl et al., 2007; Muro and Jeffery, 2008).
Analyses of learning in environmental governance draw on a range of disciplines, from psychology to management, organization theory, and the policy sciences (e.g. Bandura, 1977; Argyris and Schön, 1980; Bennett and Howlett, 1992). Authors use slightly different dimensions to characterize various forms of learning (see Haug et al., 2011). Yet most tend to focus on levels of learning, often distinguishing between a technical level of learning and one or two conceptual levels at which learning can take place (e.g. Fiol and Lyles, 1985; Hall, 1993; cf. Gerger Swartling and Nilsson, 2007). Similarly, the unit of analysis across different studies varies from the level of the individual to groups and networks or even the socio- ecological system as a whole (Diduck, 2010; Rodela, 2011).
Against this background it is perhaps not unexpected that recent efforts to review and synthesize the insights gained from more than a decade of work in this burgeoning area point to a continuous struggle with conceptual imprecision and conflation of positive and normative elements (Reed et al., 2010). Moreover, there are few studies that systematically appraise and evaluate social learning outcomes from the interventions studied (Reed et al., 2010; Rodela, 2011). Indeed, there has been a healthy dose of criticism leveled against the various ways in which social learning is understood (Armitage et al., 2008; Lundholm and Plummer, 2010; Diduck, 2010; Reed et al., 2010); its employment in contradictory ways (Blackmore, 2007; Muro and Jeffrey, 2008); its conflation with other learning concepts (Armitage et al., 2008; Diduck, 2010; Reed et al., 2010); and, abundance of unsubstantiated claims of its existence (Reed et al., 2010; Rodela, 2011).
The scholarship on adaptive management and, more recently, adaptive co-management is of particular relevance to our agenda here because it is identified as a way to make environmental governance operational (Armitage et al., 2009; Plummer et al., 2012) and emphasizes the importance and roles of learning therein (Armitage et al., 2008; Berkes, 2009). Adaptive co-management draws upon the overarching narratives in environment and resource studies of adaptive management (i.e., learning) and collaborative management (i.e., linking) to engender a distinct approach (Armitage et al., 2007; Armitage et al., 2009). Adaptive co-management is defined as “…a process by which institutional arrangements and ecological knowledge are tested and revised in a dynamic, ongoing, self-organized process of learning-by-doing” (Folke et al., 2002, p. 20).
4 In addition to its centrality to the raison d'être of adaptive co-management, learning emerged as the most prevalent variable in the adaptive co-management literature to 2010 in a recent systematic review (Plummer et al., 2012). The following investigations of learning in relation to adaptive co-management inform this study. Berkes (2009) identifies social learning as one of two keys to examining the dynamics of co-management. Plummer and FitzGibbon (2007) explored the relationship between the development of adaptive co-management and the variables of social learning and social capital in three Canadian river corridors. They, as well as Berkes (2009, p. 1699), recognize that “social learning is one of these tasks, essential both for the co-operation of partners and an outcome of the co-operation of partners”. Cundill (2010) investigated collaboration and social learning in three case studies in South Africa employing a novel collaborative monitoring methodology. This work illuminated a small set of variables which influence collaboration and learning, and ultimately led her to conclude that the learning outcomes from multi-level networks are oversimplified. Most recently, Munaretto and Huitema (2012) explored the institutional prescriptions of adaptive co- management, learning, and their relationship in the Venice lagoon. Their qualitative analysis of learning processes and outcomes revealed that outcomes were not being realized because legal constraints inhibit normative changes and network development. While the aforementioned works help to shape this research, the need for robust evaluation of learning in adaptive co-management is clear. For example, Armitage et al. (2008, p. 97) argue that the treatment of learning in adaptive co-management scholarship is paradoxical and “evaluation of learning outcomes linked to a more careful conceptualization of learning types, approaches and challenges will foster more adaptive and collaborative forms of natural resource management”.
2.1. Introducing the typology
The conceptualization of learning used in this paper has arisen out of the desire to build on the literature introduced above while seeking to address some of the perceived shortcomings of existing typologies and their operationalization. As developed in earlier work (Huitema et al., 2010; Haug et al., 2011; Munaretto and Huitema, 2012), we distinguish between cognitive, normative, and relational learning (see definitions in Table 1). These three learning types capture key dimensions of learning in an appraisal context and lend themselves to a systematic and robust assessment across different cases and social units of analysis. The first two, cognitive and normative learning, feature prominently in the literature discussed above. The last one, relational learning, reflects the emphasis on interpersonal, social aspects of learning which transpires from research on learning in natural resource management, with its focus on small groups of actors that become familiar with one another and that (begin to) collaborate. The typology resonates with the definition of adaptive co-management, with testing and revision of ecological knowledge and institutional arrangements strongly linked to cognitive and normative learning, respectively, and the process of adaptive co-management i.e., “learning-by-doing” (Folke et al., 2002, p. 20) closely related to relational learning. It also exhibits connections with attributes of other learning constructs including social learning where interaction and deliberation are highlighted (Plummer and Fitzgibbon, 2007) and linked with relational learning, as well as the double- and triple-loop learning constructs (see for example Argyris and Schön, 1978; Armitage et al., 2008) where single-loop learning may be loosely linked to cognitive learning and double and triple-loop learning are somewhat aligned with normative learning. The notions of normative and relational learning of citizens and policy-makers alike are at the heart of the literature on deliberative democracy (e.g. Dryzek, 2002; Gutmann and Thompson, 2009).
5 The reasons for diverging from dominant conceptualizations of learning in an environmental governance context were twofold. First, from an assessment perspective, it appears preferable to analyze learning effects in terms of their nature (cognitive, normative, or relational) rather than their perceived value. The typology therefore steers clear of the ‘hierarchical’ understanding of learning ubiquitous in the literature. The often visible preference for learning that involves ‘deeper’, normative change in the existing literature may be somewhat questionable, following Owens et al. (2004) who suggest that primarily cognitive change can, at times, generate equally fundamental learning effects. For instance, a different perception of the problem (a cognitive process) may not result in a normative shift, but enable a virtuous cycle of yet more cognitive learning. At the same time, normative learning may not always be beneficial to problem solving. Finally, the implicit preference for normative learning risks inducing a bias in empirical research, leading authors to focus on seeking out evidence for normative shifts, while potentially neglecting cognitive changes which may be just as important and desirable. Therefore, the framework treats the three types of learning side by side on an equal footing. Second, by separating out relational learning, the typology emphasizes a dimension that is key in the context of adaptive co-management, yet that is treated only implicitly in other typologies. Aspects of relational learning are regularly referred to in the literature, but they are often conflated with other forms or levels of learning. For instance, Hisschemöller et al. (2001) and Pahl-Wostl (2009) assume that ‘higher’ levels of learning combine changes in normative frameworks and improved relations or better understandings between stakeholders and participants. Conceptually, these are two separate aspects and there is no apparent reason for connecting them. Relational forms of learning can occur independently from other forms of learning (one can despise a teacher yet learn a lot from him or her in cognitive terms). Moreover, developing a better awareness and understanding of how another stakeholder thinks might enhance the respect for that stakeholder, but does not necessarily lead to changes in one’s own normative framework.
Table 1 includes definitions of the three types of learning as well as the measures used to evaluate the extent to which they were realized over the course of the intervention described below. The different aspects of each definition serve as indicators in the following assessment; rather than being cumulative, they each highlight different dimensions of the respective learning types. While cognitive, normative and relational learning are doubtlessly interrelated, we are convinced that it is possible and sound to distinguish between them analytically using a mixed methods approach with a quantitative analytical focus, and that such an approach should also facilitate cross-case comparison.
3. Background and Methods
3.1. Overview
A participatory process was undertaken with stakeholders in the Niagara region of Ontario, Canada. The Niagara region is well-known for its Niagara Falls, a popular tourist attraction, and for wine production. Water resources play a prominent role in the region (Renzetti et al., 2013) and climate change impacts will affect water and other important resources (Renzetti et al., 2013; Penney, 2012). The objective of the participatory process was to engage stakeholders (i.e., participants) in an adaptive co-management process as a governance approach for climate change adaptation for the Niagara region (Baird et al., submitted; Baird et al., 2014). A key objective of the research, and the specific focus of this paper, was to assess learning as a result of participation in the adaptive co-management process for climate
6 change adaptation. Thirty-two participant stakeholders represented 23 organizations from a range of sectors. Participants were chosen through a social-ecological inventory process that identifies stakeholders with an interest and with the potential for action around an issue, in this case climate change adaptation, by combining aspects of stakeholder analysis and ecological inventories (Schultz et al., 2007). The social-ecological inventory resulted in a sample population appropriate for the adaptive co-management process and not necessarily a representative sample of the population of the Niagara region. All individuals were invited to an initial information session in November 2010, where the results of the social-ecological inventory were shared and the option of engaging in a participatory research process was outlined. For a more detailed description of the process see Baird et al. (2014).
The first round of data collection to measure learning occurred in December of 2010 prior to the launch of the year-long participatory adaptive co-management process. The participatory process was then initiated; modeled as a researcher facilitated, participatory space for stakeholders to interact, deliberate, learn, and ultimately act for climate change adaptation for the Niagara region (Plummer, 2012; Baird et al., submitted) using a series of workshops, meetings, and an electronic repository and social space. Workshops and meetings included information sessions about climate change impacts and adaptation, opportunities to learn about each other’s efforts and interests related to climate change adaptation, and the space and time to deliberate about the group’s identity, purpose, and potential actions and outcomes (Baird et al., submitted). The second round of data collection, with an identical format to the first but with the addition of reflection questions, was administered 14 months later to those who completed the pre-study (ex-ante) survey and were willing to respond to the post-study (ex-post) survey. Reflection questions queried the amount of learning that occurred during the participatory process based on the indicators of learning identified in Table 1. Only those participants who completed both the ex-ante and ex-post data collection instruments were included in this analysis.
3.2.1. Participant groups
Participants in the adaptive co-management process were separated into two groups for the purpose of analysis. Those who participated in three meetings or less were considered to exhibit a ‘low activity level’ (n = 12). Those who participated in more than three meetings were considered to exhibit a ‘high activity level’ (n = 11). This method of separation was consistent with self-reported measures of activity in the process. While being cognizant that learning occurs at the individual level, separation of the participants into two groups allowed us to control for external context and identify potential differences in learning as a result of participation in that process. Huitema et al. (2010) found that patterns of learning were different for those active in a participatory process than those in a control group.
3.3. Measures
Measures of learning were assessed using the typology described in Table 1. Indicators were used as measureable proxies for cognitive, normative and relational learning, and describe the dimensions of each. As such, the indicators need not all be measured to the same extent, nor must they all be present for learning to have occurred. Rather, indicators offer a way to capture changes in one or more aspects of learning types. Baseline cognitive, normative and relational states were assessed at the outset of the study and learning that occurred during the course of the adaptive co-management process study was measured at the study’s conclusion. Methods to measure the three learning types are explained below.
7 3.3.1. Cognitive learning
Two indicators of cognitive learning were evaluated. First, the acquisition of new knowledge was tested using questions adapted from a survey of America’s Knowledge of Climate Change (Leiserowitz et al., 2010). Questions with correct or incorrect answers clearly established in the literature were scored and participants accumulated a percentage score pre- and post-study. A statistically significant difference for the change in score between groups was tested using SPSS Statistics 19 (IBM). A self-reported indication of learning using a Likert scale was included in the post-study survey. Differences for all self-reported learning measures were tested for significance between groups using the confidence interval of the median to assess the magnitude of difference.
Restructuring of existing knowledge, the second cognitive learning indicator, was tested by evaluating ex-ante and ex-post concept maps created by each participant. Concept maps are hierarchical, graphical representations that reflect knowledge of an individual or group about an issue or topic (Novak, 2010). They allow participants to structure knowledge about an issue or topic, in this case what participants think about when prompted with the phrase ‘climate change and the Niagara region’, as a series of nodes branching from, and connected to, the central topic that become more detailed and refined with increasing number of nodes away from the centre. Analysis of group patterns in individual concept maps followed the method of Haug et al. (2011). Briefly, knowledge items, or nodes, on the concept maps were transcribed into a spreadsheet (MS Excel) and coded as one of 11 categories (developed by content analysis) by two researcher coders. Intercoder reliability was tested using percent coding similarity (90%) and percent category distribution similarity (±1.3%); both indicated a high degree of reliability. A change in the structuring of existing knowledge was identified where there was a change in the categories in the ex-ante and ex-post concept maps, where categories became more or less central (centrality), and how often categories were mentioned (specificity). Categories were plotted according to their centrality score (x-axis) and their specificity score (y-axis) and positions were compared between the two maps. Concept map methodology and analysis is described in greater detail in Morine-Dershimer (1993) and in Haug et al. (2011). Shifts in structuring of existing knowledge demonstrated by the concept maps were compared with meeting minutes for evidence of discussion of the themes.
3.3.2. Normative learning
Four indicators of normative learning by participants (Table 1) were probed in the surveys. A shift in viewpoint was evaluated ex-ante and ex-post using the New Ecological Paradigm (NEP) tool developed by Dunlap et al. (2000) and modified by Stedman (2004). The NEP tool presents a series of statements about the environment that focus on interactions between the social and ecological systems. Statements were used to probe six aspects of climate change core beliefs and values which together measure an ecological worldview and are described in Table 2 (Dunlap et al., 2000; Stedman, 2004; Dunlap, 2008). Participants rated the series of 18 statements in terms of the degree to which they agreed or disagreed with them. Mean standard deviations from NEP responses were used to measure the second indicator of normative learning, consensus building. A reduction in the standard deviation across participants indicated convergence, or consensus building (Huitema et al., 2010). The final two indicators of normative learning, a shift in values and a shift in paradigm, were queried ex-post using reflective questions. Normative learning was further probed using the meeting minutes for evidence of consensus building around actions and outcomes.
8 3.3.3. Relational learning
Three indicators of relational learning were evaluated using ex-post reflective questions. Participants were asked to rate the degree to which participation in the study contributed to enhanced trust and building of relationships, an increased network of people and organizations they communicated with, and enhanced interactions among organizations.
Relational learning was also evaluated using social network data. Network size (number of actors in the network) and number of ties (identified instances of communication between two actors), the proportion of ties that were reciprocal, the proportion of multiplex ties reported (communications between actors for more than one reason) and the number of reported collaborations were used to further investigate indicators of relational learning that included building relationships, enhanced trust, and enhanced ability to cooperate (Lin, 1999; Pretty and Ward, 2001; Prell et al., 2009; Newig et al., 2010). It is important to acknowledge that respondents provided these social network data in terms of communication with organizations rather than individuals, and so the analysis may underestimate the total number of communications occurring if more than one person was contacted at a single organization.
3.3.4 Validity, reliability and triangulation
Multiple methods were used to measure indicators of cognitive, normative and relational learning. Several of the instruments chosen have been used elsewhere and the validity and reliability of them has been established and is described in the above sections. The methodology described was employed in order to triangulate data and gain empirical evidence of learning from multiple sources, thereby strengthening the validity of the results and providing a richer dataset from which to measure and discuss learning effects (Mathison, 1988; Miles and Huberman, 1994). However, we were cognizant of the potential for participants to over-report learning as a symptom of common method biases such as a form of ‘social desirability bias’ where participants may have felt a desire to report learning as a result of participation in the study, or ‘transient mood bias’ where the context within which participants responded affected their responses (Ganster et al., 1983; Podsakoff et al., 2003). Triangulation provided several methodologically different sources of evidence to mitigate the potential for systematic errors in learning measurement and erroneous conclusions, including the use of multiple sources (participants and meeting minutes) (Creswell and Miller, 2000; Podsakoff et al., 2003). In addition, the inclusion of a pseudo-control group increased the potential to identify threats to validity via the Hawthorne effect where researcher contact with individuals may influence their responses (Huitema et al., 2010).
4.0. Results
4.1. Cognitive learning
4.1.1. Acquisition of new knowledge
Results of the survey questions to test knowledge gains showed that the ex-ante score for percentage of ‘correct’ answers was nearly identical for the two groups (Table 3); however, after participation in the study, respondents in the high activity level group generally scored higher. Testing showed that the difference in the mean change of the group scores was statistically significant (95% confidence level) (Table 3; Figure 1).
9 Self-reported measures of the acquisition of knowledge supported the climate change knowledge scores: the median degree to which participants reported knowledge was fostered was ‘a little’ by the low activity level group, while the high activity level group median response was ‘moderately’ on a four point scale from ‘none’ to ‘a great deal’. The difference was not statistically significant and there was no significant correlation between the acquisition of new knowledge and self-reported measures.
4.1.2. Restructuring of existing knowledge
Concept maps showed evidence of shifts in the way existing knowledge was structured for the two groups. The mean number of levels was calculated by counting the degrees of separation from the central node ‘climate change in the Niagara region’ to the node connected to it by the greatest number of others. (i.e., length of chains of nodes extending from the central theme). The mean number of levels for the low activity level participants’ ex-ante mind maps was 3.4, and decreased to 3.0 ex-post. The high activity level group, however, maintained the mean number of levels over the course of the study, with a very slight increase from 3.9 to 4.0, indicating that the cognitive complexity around the central topic of climate change adaptation stayed the same over the course of the participatory process for this group.
Knowledge items provided by all participants were coded into one of 11 categories, and these are presented and briefly described in Table 4. For the low activity level group, the most substantial mean decrease in centrality occurred for the ‘mitigation’ category (t=-2.4; df = 6; p = 0.05) (Figure 2). In terms of specificity, the only statistically significant shift occurred for the ‘socio-econ’ category (t = 2.5; df = 6; p = 0.05). Shifts toward increasing specificity and centrality also occurred for ‘adaptation’ and ‘education’ (Figure 2). For the high activity level group, there were no significant changes over time; however, several trends were evident (Figure 3): the ‘impacts’ category remained highly central with a high specificity score; the ‘socio-econ’ category shifted to a lower specificity; ‘education’ shifted higher; and, categories of ‘adaptation’, ‘mitigation’, and ‘opportunity’ all became more central. The categories of ‘mitigation’ and ‘education’ correspond to the focus of several discussions that occurred during meetings held throughout the study. Meeting minutes indicated that much discussion focused on categories that became (or remained) prominent over the course of the study including: understanding climate change impacts (first four meetings); issues of funding (three meetings near the end of the study); actions the group should take (including a climate change action plan) (11 meetings); and, communication of the group’s objectives and activities to the broader community (eight meetings).
4.2. Normative learning
4.2.1. Shift in viewpoint
A shift in viewpoint was measured from level of agreement with the NEP statements grouped into six facets for analysis. The mean score for each facet was calculated for the two groups (Figure 4). A change in the score given to a category of statements indicates a shift in viewpoint. The only statistically significant difference between the two groups in terms of their normative views occurred at ex-ante for the ‘rejection of exemptionalism’ facet, where the high activity level group agreed more strongly with the statements within this facet. Within each group there was no significant shift in viewpoint over time (Figure 4). However,
10 self-reported measures of the extent to which individual views shifted illuminated a perceived change in views by both groups with a greater change reported by high activity level participants (‘moderately’) than the low activity level group (‘a little’).
4.2.2. Building of consensus
Standard deviations to the NEP questions were calculated to assess the potential for building of consensus (Figure 5). Differences occurred in standard deviations of individual facets of the NEP; of particular note was the ‘limits to growth’ facet where consensus was built within the high activity level group. There were no statistically significant changes in standard deviation for the low activity level group. When the total mean standard deviations for the two groups over time were tested, the differences between them were insignificant within and between groups.
The adaptive co-management process brought together participants who, at the baseline level, were undertaking disparate climate change-related activities in relative isolation of one another. The outcomes of the process offer additional evidence of the ability of the high activity level group to build consensus. Meeting minutes indicated that a charter outlining the group’s common understanding of climate change and its impacts, goals, and commitments to action was accepted by members at the ninth meeting, and the group developed working principles, a strategic plan, and a group name and logo.
4.2.3. Shift in values and shift in paradigm
Self-reported measures were used to identify a shift in values and a shift in paradigm. A perceived shift in values was reported by high activity level participants (median response ‘a little’), but less so by low activity level participants, as was expected (median response ‘none’ to ‘a little’). For questions related to a shift in paradigm, both groups indicated that they experienced a shift to some degree (Table 5). The degree to which the two groups questioned governance norms and protocols was significantly different, with high activity level participants reporting a high degree of questioning and potentially a shift in paradigm.
4.3. Relational learning
Relational learning was measured using three indicators: building of relationships; enhanced trust; and, enhanced ability to cooperate. In addition, the question of how useful discussions with other stakeholders were for learning was posed to participants in the ex-post survey, and the low activity level group median response was ‘useful’, while the high activity level group median response was ‘very useful’.
4.3.1. Building of relationships
Building of relationships was measured using self-reported degree of enhanced interactions among organizations. The median response for the low activity level group was ‘a little’, while the high activity level groups was ‘a great deal’. The difference between the two groups was statistically significant based on the confidence interval of the median (Table 6).
Further insight was gained into the nature of relationship building using social network data. Network size and number of ties for low activity level and high activity level participants (i.e., the number of organizations connected to participant groups through communication
11 pathways or ‘ties’) was investigated to understand how the number of relationships changed from the study outset until its conclusion. Participants in the two groups reported a relatively similar total number of ties to organizations in 2010 (218 ties to 58 organizations and 252 ties to 79 organizations, respectively). In 2012, however, low activity level participants collectively reported a much lower number of total ties (85 ties to 49 organizations) than the high activity level group (177 ties to 56 organizations). Further, the high activity level group reported more ties with one another than the low activity level participants reported with the entire participant group (46 and 31, respectively) in 2012.
4.3.2. Enhanced trust
Enhanced trust was self-reported as a result of participation in the study. The low activity level group reported a median response of ‘a little’ to ‘moderately’, while the high activity level group’s median response was ‘a great deal’ (Table 6).
Trust may be identified by the presence of reciprocal relationships in the participant networks. The proportion of ties reported by groups that were reciprocal (with the exclusion of unhelpful reciprocal ties) is presented in Table 7. Briefly, the high activity level group increased their proportion of reciprocal ties over the course of the study, while the low activity level participants’ were reduced over time. Differences were not statistically significant due to high variation in responses.
4.3.3. Enhanced ability to cooperate
The final relational learning indicator measured by self-reporting was an enhanced ability to cooperate. Participants were asked to rate the degree to which their participation in the study increased the network of people and organizations with which they communicate. The low activity level group reported a median rating of ‘moderately’, while the high activity level group’s median response was ‘a great deal’ (Table 6).
From the social network data it is evident that the high activity level group held and maintained more multiplex relations (two or more reasons for communicating with other actors) over the course of the study (mean of 16 per person ex-ante and 12.5 ex-post), while the low activity level group’s mean number of multiplex relations per person dropped over time (from 12 to 5). In addition, the number of reported collaborative ties by high activity level participants stayed constant over time at a mean of 21 per person, while the low activity level group reported a mean of 18.5 collaborative ties ex-ante and 11 ex-post. Given that the total number of reported ties (communication for any reason) dropped for both groups in 2012, the high activity level group demonstrated an increasing trend in the proportion of ties reported that were considered collaborative and the proportion of multiplex relations. However, the change in the number of multiplex relations and collaborative ties was not statistically significantly different between the two groups due to a high degree of variation in reported ties.
5. Discussion
5.1. Study design and learning typology
The learning typology described and operationalized in this study provided a valuable multidimensional picture of learning. The concepts of cognitive, normative, and relational
12 learning have been used in other studies (see for example Huitema et al., 2010; Haug et al., 2011). However, the identification of specific, measureable indicators for each type of learning, and measurement of indicators at the outset and at the conclusion of the study provides a robust approach to trace different types of learning effects over time, which is relatively rare in the learning literature (Rodela et al., 2012).
As a result of the typology and study design, we identified several indications of learning (Table 8). It is clear that cognitive and relational learning occurred for the high activity level group, and less so for the low activity level group. Measures of normative learning were much less certain and thus not conclusive. Interestingly, similar results were reported by Haug et al. (2011) in a policy games exercise over a short period of time, and by Munaretto and Huitema (2012) in an analysis of longer-term management of the Venice lagoon. Ascertaining normative learning effects is certainly challenging methodologically, given the difficulty of identifying appropriate measures that capture such learning as suggested by Brossard et al., 2005 in an attitudinal change study of environmentally aware participants of a citizen science program. Yet it is also conceivable that normative learning does not occur with similar frequency and / or under similar conditions as cognitive and relational learning, particularly when participants are knowledgeable about the subject matter prior to participation (e.g., Haug et al., 2011). A substantial body of evidence suggests that social norms are institutions that change very slowly (e.g., Williamson, 2000; Roland, 2004). Environmental psychology has taken up this question of attitudes, values and norms in relationship to the environment. Heberlein (2012, p. 34) argues that attitude change “often takes time, direct experience, and social influence.” Thus it is not surprising that a one-year intervention like ours would not show a major shift in values and norms.
The benefits and challenges of our approach in relation to measuring learning and a discussion of the implications of the findings for adaptive co-management follow.
5.2. Benefits of the approach
The learning typology presented and employed in this study builds on previous learning constructs while avoiding the ‘leveled’ or ‘stepwise’ understanding of learning presented in double- and triple-loop learning models (e.g., Argyris and Schön, 1978; Armitage et al., 2008; Pahl-Wostl, 2009). Importantly, the learning typology we use identifies that normative learning (loosely aligned with double- and triple-loop learning) did not occur to the same extent as other learning types. Defining learning with our three-fold typology also allows for identification and measurement of relational learning, which is not explicitly included in the above-mentioned models. Inclusion of multiple indicators and methods for each learning type improved the robustness of the measures and highlighted the multi-dimensionality of learning types.
The use of a pseudo-control group that exhibited a low activity level provided a relative ‘baseline’ measure of change that could not be attributed to participation in the participatory adaptive co-management process and a benchmark for the identification of true trends and learning change. Calls for a control group highlight the potential benefits, including controlling for external context and the learning effect of instruments used in evaluation (Haug et al., 2011; Huitema et al., 2010) and thus increased the validity of the results in this study.
13 Many of the evaluative tools used were a good fit with the three types of learning. Concept maps as a tool to measure cognitive learning proved to be valuable. The general approach has been utilized in various related ways, for example in mapping ecosystems and creating management plans (Özesmi and Özesmi, 2003) and future scenarios (van Vliet et al., 2010), mapping individual and group conceptualizations of social-ecological systems (Gray et al., 2012) and to uncover individual perspectives for an understanding of stakeholder values, goals and aspirations in a stakeholder analysis (Hjortsø et al., 2005) among others. However, concept, or cognitive, mapping is used in these examples as a tool for learning and is used less often as a tool to measure learning over time as we have used it here. While differences of theme centrality and specificity were minimal in the context of this study, the analytical potential of this tool in terms of the ability to directly link indicators of cognitive learning to the measures from the maps make it highly suitable for this task.
Social network analysis provides measures of the structure and processes of interactions among participants and with the broader community. Increases in the incidences of reciprocity and multiplexity (i.e., meaningful interactions) as well as the density of the network of participants in the high activity group as opposed to the low activity group indicated that relationship and trust building had occurred, and that cooperation was enhanced; all indicators of relational learning (Newig et al., 2010; Crona and Parker, 2012). Social network analysis has rarely been used in the context of measuring learning (Newig et al., 2010). We see social network analysis as an integral tool to measure relational learning using the typology employed in this study.
5.3. Challenges of the approach
The study design also resulted in some challenges in measuring learning. First, the selection of participants was targeted to those within the community with an interest, and undertaking activities, in climate change adaptation. Participants in both the high and low activity level groups originated from this sample of the population, and thus brought a baseline knowledge of climate change adaptation and activity level that was potentially greater than the average community member. However, our results are consistent with those of Huitema et al. (2010) who used a random control group in citizen jury learning study and found that learning effects were greater for participants than for the control group, but that the difference between the two groups’ responses was not vastly different. A potential influence on this result may have been the process of providing responses, which was lengthy and participants may have experienced some fatigue (Baird et al., submitted). The social network analysis results in particular presented some challenges in terms of interpretation as a result of difficulty in obtaining the ex-post data and the tendency of most participants who responded to report a smaller network at that time. This challenge is described in more detail in Baird et al. (submitted).
Question design proved to be a challenge in the measures of normative and relational learning. The NEP, used to measure normative learning, evaluates environmental attitudes and has been used, in various forms, extensively in the literature for this purpose (Dunlap 2008), and specifically within the context of climate change (e.g., Stedman, 2004; Zahran et al., 2006; Dietz et al., 2007; Takahashi and Meisner, 2011), but less so for measures of learning by changes in environmental attitudes (Harraway et al., 2012 is one exception). In the context of normative learning, the NEP was used with reasonable success in our study to identify consensus building and the measures from it were consistent with self-reported normative learning. However, the NEP failed to conclusively measure a shift in viewpoint for
14 participants in both groups and alternative methods for measuring normative learning may be required. The relative similarity of environmental attitudes exhibited by both groups of respondents may be indicative of a true similarity and an artifact of the use of a pseudo- control group who were active in climate change adaptation activities rather than a random sample of the community. Deyle and Slotterback (2009), in a study of group learning through participatory planning processes, identified the design of questions as a key challenge of measuring learning, and also noted that external contextual factors are unavoidable and difficult to control for in learning measurement.
5.4. Adaptive co-management, learning and climate change adaptation
The findings from this research empirically show learning occurred as a result of implementing an adaptive co-management initiative for climate change adaptation. The results indicate that the operationalization of the multidimensional learning typology set out, using measurable indicators for each type, provided measures of learning with a robustness not achieved in previous studies. According to these empirical measurements, the adaptive co-management intervention fostered learning, with the cognitive and relational types of learning being most evident.
Although relational ambiguity between environmental governance and adaptive co- management exists (Huitema et al., 2009), it is generally considered a mechanism to make governance operational (Plummer et al., 2012). Learning is central to this interactive process for decision-making. A recent systematic review by Plummer et al. (2012) revealed that learning is frequently highlighted in definitions of adaptive co-management and is the component garnering the most attention in the adaptive co-management literature. Despite this strong association, the treatment of learning in relation to adaptive co-management has been variously defined, differently understood, and rarely assessed (Armitage et al., 2008; Plummer et al., 2012). The typology of learning, associated measures, and, methodological procedures employed in this research advance considerably the previous paradoxical treatment of learning in relation to adaptive co-management.
Routine calls have been made to robustly assess the outcomes of adaptive co-management (e.g., Plummer and Armitage, 2007; Cundill and Fabricius, 2010) and learning frequently emerges as a potentially valuable product from the process in the literature (Plummer et al., 2012). This phenomenon is widespread. Rodela (2011) and Rodela et al. (2012) recently conducted two systematic reviews of the literature on social learning in natural resource management and found that only a few studies attempt to empirically assess learning effects of specific interventions on participants. Efforts by Cundill (2010) and Munaretto and Huitema (2012) are exceptions to the dearth of assessing outcomes in regards to learning from adaptive co-management. Cundill’s (2010) experimental investigation of three case studies in South Africa led her to conclude that rhetoric regarding learning outcomes from multi-level networks is oversimplified. Munaretto and Huitema’s (2012) study revealed that learning (cognitive, normative and relational) outcomes do not accrue due to the Special Law in the Venice system that inhibits evolution of values, beliefs and networks. The three-fold typology and methodology pioneered in this research offers a multi-dimensional way to quantitatively assess learning outcomes from an adaptive co-management initiative. In so doing, it overcomes the shortcomings of considering learning in hindsight, solely through researcher experiences, or extrapolated from other data (Rodela et al., 2012).
15 While it is important to re-assert the caution expressed by Berkes (2009) that learning does not always lead to adaptation, the results from this study support the possibility of adaptive co-management as a process for climate change adaptation (Locatelli et al., 2008; May and Plummer, 2011; Plummer, 2013). Most interestingly, this research lends empirical support to conceptual work by Plummer (2013) that adaptive co-management can be implemented as a governance strategy which supports climate change adaptation as it builds generalized adaptive capacity and offers a novel institutional arrangement for generating adaptive responses.
Learning is an important part of the adaptive co-management process. Research about learning in relation to the process have to date largely come from qualitative analysis of case studies (e.g., Plummer and FitzGibbon, 2007; Armitage et al., 2008; Berkes, 2009; Cundill, 2010). In concentrating on the interplay of variables concerning collaborative interactions and learning in adaptive co-management, Plummer and FitzGibbon (2007) and Cundill (2010) have provided insights into the dynamism of this relationship and the relatively short time frame (three years and 18 months respectively) in which change may be apparent. The results of this research confirm the interactions between deliberation among actors and learning. The findings also suggest that learning in association with adaptive co-management may take place more quickly than previously thought and may be evident even in the early stages of adaptive co-management. Finally, the finding that normative change was the least evident is consistent with other studies of learning types in participatory processes (e.g., Munaretto and Huitema, 2012; Haug et al., 2011) and with the literature regarding change in environmental norms and values (e.g., Heberlein, 2012).
6. Conclusion
This research provides a way to assess learning in adaptive co-management specifically and in natural resources management and environmental governance more broadly. Prospects for more robust assessments of learning in these contexts are advanced by its sound conceptualization and specificity, clear operational measurements and analytical techniques, transparent methodological decisions, and replicable design. The research demonstrates the feasibility of empirically capturing outcomes from adaptive co-management and in particular illuminate nuances associated with learning brought about through the process.
Learning is essential to building adaptive capacity (Fazey et al., 2007; Plummer and Armitage, 2010) and is a genuine adaptation strategy (Pelling et al., 2008). Policy makers and practitioners concerned with generating adaptive responses to climate change may want to consider adaptive co-management as a way to engender learning or as a governance strategy which supports climate change adaptation (as suggested by Plummer, 2013). At the same time, the findings from this research clearly show differences among types of learning. Appropriate expectations about learning, in particular different types of learning during relatively short temporal periods, are thus important considerations for the design and practice of adaptive co-management for climate change adaptation, adaptive co-management in other contexts, and more broadly for participatory processes for environmental governance.
The conceptual framework, methodology, and findings from this study culminate in several new directions for research about learning in relation to adaptive co-management as well as natural resources management and environmental governance. The typology of learning opens questions for empirical investigation concerning adaptive co-management and the
16 speed at which learning effects are evident as well as the type(s) of learning brought about. This study and others (e.g., Haug et al., 2011; Munaretto and Huitema, 2012) have found that normative learning is less evident than other learning types. This leads to research questions that include: do normative learning indicators shift more gradually than cognitive and relational learning indicators? Are measures of normative learning sufficiently sensitive to identify subtle changes in norms and values, convergence of group opinion, and shifts in paradigms? Are norms and values of individuals too homogenous prior to participatory processes to show change? In pursuing these questions, researchers of adaptive co- management and environmental governance interface with work on slow changing institutions (e.g., Williamson, 2000; Roland, 2004), contribute to the rich literature on changing norms in environmental psychology (e.g. Heberlein, 2012), and shape practitioners’ expectations about the speed of normative change from participatory processes. Understanding and further untangling types of learning and relationships among them, probing the relationship(s) between other variables and learning within the process of adaptive co-management, as well as making comparisons among the efficacy of learning brought about by various environmental governance mechanisms, are fruitful avenues of research.
References Argyris, C., Schön, D.A., 1980. Organizational Learning. Addison-Wesley, Reading, MA. Armitage, D., Berkes, F., Doubleday, N., 2007. Introduction: moving beyond co- management. In: Armitage, A., Berkes, F., Doubleday, N. (Eds.), Adaptive Co- management: Collaboration, Learning and Multi-level Governance. UBC Press, Vancouver, BC, Canada. Armitage, D., de Loë, R., Plummer, R., 2012. Environmental governance and its implications for conservation practice. Conservation Letters 0, 1-11. Armitage, D., Marschke, M., Plummer, R., 2008. Adaptive co-management and the paradox of learning. Global Environmental Change 18, 86-98. Armitage, D., Plummer, R., Berkes, F., Arthur, R., Davidson-Hunt, I., Diduck, A., Doubleday, N., Johnson, D., Marschke, M., McConney, P., Pinkerton, E., Wollenberg, L., 2009. Adaptive co-management for social-ecological complexity. Frontiers in Ecology and the Environment 2 (7), 95-102. Baird, J.M., Plummer, R., Bodin, Ö., submitted. Experimenting with adaptive co- management for adaptation to climate change: a social network perspective. Baird, J.M., Plummer, R. and Pickering, K., 2014. Priming the governance system for climate change adaptation: the application of a Social Ecological Inventory to engage actors in Niagara, Canada. Ecology & Society. 19(1), 3 Available from: http://www.ecologyandsociety.org/vol19/iss1/art3/ (accessed 01.20.14). Bandura, A., 1977. Social Learning Theory. General Learning Press, New York, NY. Bennett, C.J., Howlett, M., 1992. The lessons of learning - reconciling theories of policy learning and policy change. Policy Sciences 25, 275-294. Berkes, F., 2009. Evolution of co-management: role of knowledge generation, bridging organizations and social learning. Journal of Environmental Management 90, 1692- 1702. Available from: http://www.sciencedirect.com/science/article/pii/S0301479708003587 (accessed 10.08.12). Blackmore, C., 2007. What kinds of knowledge, knowing and learning are required for addressing resource dilemmas? A theoretical overview. Environmental Science and Policy 10, 512–25.
17 Brossard, D., Lewenstein, B., Bonney,R., 2005. Scientific knowledge and attitude change: the impact of a citizen science project. International Journal of Science Education 27(9), 1099-1121. Creswell, J.W., Miller, D.L., 2000. Determining validity in qualitative inquiry. Theory Into Practice 39 (3), 124-130. Crona, B., Parker, J.N., 2012. Leaning in support of governance: theories, methods, and a framework to assess how bridging organizations contribute to adaptive resource governance. Ecology and Society 17(1), 32. Available from: http://dx.doi.org/10.5751/ES-04534-170132 (accessed 08.08.12). Cundill, G. 2010. Monitoring social learning processes in adaptive comanagement: three case studies from South Africa. Ecology and Society 15 (3), 28. Available from: http://www.ecologyandsociety.org/vol15/iss3/art28/ (accessed 15.09.12). Cundill, G., Fabricius, C., 2010. Monitoring the governance dimension of natural resource co-management. Ecology and Society 15 (1), 15. Available from: http://www.ecologyandsociety.org/vol15/iss1/art15/ (accessed 17.09.12). Deyle, R., Slotterback, C.S., 2009. Group learning in participatory planning processes: an exploratory quasiexperimental analysis of local mitigation planning in Florida. Journal of Planning Education and Research 29, 23-38. Diduck, A., 2010. The learning dimension of adaptive capacity: untangling the multi-level connections. In: Armitage, D., Plummer, R. (Eds.), Adaptive capacity and environmental governance. Springer, Berlin, Germany. Diduck, A., Bankes, N., Clark, D., Armitage, D., 2005. Unpacking social learning in social- ecological systems: case studies of polar bear and narwhal management in northern Canada. In: Berkes, F., Huebert, R., Fast, H., Manseau, M., Diduck , A. (Eds.), Breaking Ice: Renewable Resource and Ocean Management in the Canadian North. Arctic Institute of North America and University of Calgary Press, Calgary, AB. Dietz, T., Dan, A., Shwom, R., 2007. Support for climate change policy: social psychology and social structural influences. Rural Sociology 72 (2), 185-214. Dryzek, J., 2002. Deliberative democracy and beyond. Oxford University Press, Oxford. Dunlap, R.E., 2008. The new environmental paradigm scale: from marginality to worldwide use. The Journal of Environmental Education 40 (1), 3-18. Dunlap, R.E., Van Liere, K.D., Mertig, A.G., Jones, R.E., 2000. Measuring endorsement of the New Ecological Paradigm: a revised NEP scale. Journal of Social Issues 56 (3), 425-442. Fazey, I., Fazey, J., Fischer, J., Sherren, K., Warren, J.M., Noss, R., Dovers, S., 2007. Adaptive capacity and learning to learn as leverage for social-ecological resilience. Frontiers in Ecology and the Environment 5, 375-380. Fiol, C.M., Lyles, M.A., 1985. Organizational learning. The Academy of Management Review 10, 803-813. Folke, C., Carpenter, S., Emqvist, T., Gunderson, L., Holling, C.S., Walker, B., Bengtsson, J., Berkes, F., Colding, J., Danell, K., Falkenmark, M., Moberg, M., Gordon, L., Kaspersson, R., Kautsky, N., Kinzig, A., Levin, S.A., Mäler, K-G., Ohlsson, L., Olsson, P., Ostrom, E., Reid, E., Rockstöm, J., Savenije, S., Svedin, U., 2002. Resilience and sustainable development: building adaptive capacity in a world of transformations. Report for the Swedish Environmental Advisory Council 2002. Ministry of the Environment, Stockholm, Sweden. Folke, C., Hahn, T., Olsson, P., Norberg, J., 2005. Adaptive governance of social-ecological systems. Annual Review of Environment and Resources 30, 441-473. Available from: URL: http://www.annualreviews.org/toc/energy/30/1 (accessed 14.09.2012).
18 Ganster, D.C., Hennessey, H.W., Luthans, F., 1983. Social desirability response effects: three alternative models. Academy of Management Journal 26 (2), 321-331. Gerger Swartling, Å., Lundholm, C., Plummer, R., Armitage, D., 2011. Social learning and sustainability: exploring critical issues in relation to environmental change and governance. SEI Project Report, Mistra-SWECIA Working Paper Nol 5. SEI, January 2011.Hall, P.A., 1993 Policy paradigms, social learning, and the state: the case of economic policymaking in Britain. Comparative Politics 25, 275-296. Gerger Swartling, Å., Nilsson, M., 2007. Social learning and EPI: communicative governance in Swedish climate policy formation. EPIGOV Paper No. 12. Institute for International and European Environmental Policy, Berlin, Germany. Gray, S., Chan, A., Clark, D., Jordan, R., 2012. Modeling the integration of stakeholder knowledge in social-ecological decision-making: benefits and limitations to knowledge diversity. Ecological Modelling 229, 88-96. Gutmann, A., Thompson, D., 2009. Why deliberative democracy? Princeton University Press, Princeton. Harraway, J., Broughton-Ansin, F., Deaker, L., Jowett, T., Shephard, K., 2012. Exploring the use of the revised New Ecological Paradigm Scale (NEP) to monitor the development of students’ ecological worldviews. Ecology and Society 43 (3), 177-191. Haug, C., Huitema, D., Wenzler, I., 2011. Learning through games? Evaluating the learning effect of a policy exercise on European climate policy. Technological Forecasting and Social Change 78, 968-981. Heberlein, T.A., 2012. Navigating Environmental Attitudes. Oxford University Press, Oxford. Hisschemöller, M., Tol, R.S.J., Vellinga, P., 2001. The relevance of participatory approaches in integrated environmental assessment. Integrated Assessment 2, 57-72. Hjortsø, C.N., Christensen, S.M., Tarp, P., 2005. Rapid stakeholder and conflict assessment for natural resource management using cognitive mapping: the case of Damdoi Forest Enterprise, Vietnam. Agriculture and Human Values 22, 149-167. Huitema, D., Cornelisse, C., Ottow, B., 2010. Is the jury still out? Toward greater insight in policy learning in participatory decision processes – the case of Dutch citizens' juries on water management in the Rhine Basin. Ecology and Society 15, 16. Available from: http://www.ecologyandsociety.org/vol15/iss1/art16/ (accessed 19.09.2012). Huitema, D., Mostert, E., Egas, W., Moellenkamp, S., Pahl-Wostl, C., Yalcin, R., 2009. Adaptive water governance: assessing the institutional prescriptions of adaptive (co-) management from a governance perspective and defining a research agenda. Ecology and Society 14 (1), 26. Available from: http://www.ecologyandsociety.org/vol14/iss1/art26/ (accessed 16.09.12). Keen, M., Brown, V., Dybal, R., 2005. Social Learning in Environmental Management. Earthscan, London. Kilpatrick, S., Barrett, M., Jones, T., 2003. Defining learning communities. CRLRA Discussion Paper Series. Center for Research and Learning in Regional Australia, Launceston, TAS. Leeuwis, C., Pyburn, R. (Eds.), 2002. Wheelbarrows Full of Frogs, Koninklijke Van Gorcum, Assen, The Netherlands Lesierowitz, A., Smith, N., Marlon, J.R., 2010. Americans’ knowledge of climate change. Yale Project on Climate Change Communication. Yale University, New Haven, CT. Available from: http://environment.yale.edu/climate/files/ClimateChangeKnowledge2010.pdf (accessed 01.10.12). Lin, N., 1999. Building a network theory of social capital. Connections 22 (1): 28-51.
19 Locatelli, B., Kanninen, M., Brockhaus, M., Pierce Colfer, C.J. Murdiyarso, D., Santoso, H., 2008. Facing an uncertain future: how forests and people can adapt to climate change. Forest Perspectives 5. Available from: http://www.cifor.org/publications/pdf_files/media/CIFOR_adaptation.pdf (accessed 08.08.12). Lundholm, C., Plummer, R., 2010. Resilience and learning: a conspectus for environmental education. Environmental Education Research 16 (5-6), 475-492. Mathison, S., 1988. Why triangulate? Educational Researcher 17 (2), 13-17. May, B., Plummer, R., 2011. Accommodating the challenges of climate change adaptation and governance in conventional risk management: adaptive collaborative risk management (ACRM). Ecology and Society 16 (1), 47. Available from: http://www.ecologyandsociety.org/vol16/iss1/art47/ (accessed 15.09.12). Miles, M.B., Huberman, A.M., 1994. Qualitative Data Analysis: An Expanded Sourcebook (2nd Ed.) Newbury Park, CA: Sage. Morine-Dershimer, G., 1993. Tracing conceptual change in preservice teachers. Teaching and Teacher Education 9 (1), 15-26. Munaretto, S., Huitema, D., 2012. Adaptive comangement in the Venice Lagoon? An analysis of current water and environmental management practices and prospects for change. Ecology and Society 17 (2), 19. Available from: http://dx.doi.org/10.5751/ES-04772-170219 (accessed 15.09.12). Muro, M., Jeffrey, P., 2008. A critical review of the theory and application of social learning in participatory natural resource management processes. Journal of Environmental Planning and Management 51, 325–344. Newig, J., Günther, D., Pahl-Wostl, C., 2010. Synapses in the network: learning in governance networks in the context of environmental management. Ecology and Society 15(4): 24. Available from: http://www.ecologyandsociety.org/vol15/iss4/art24/ (accessed 17.09.12). Novak, J.D., 2010. Learning, creating, and using knowledge: concept maps as facilitative tools in schools and corporations. Journal of e-Learning and Knowledge Society 6 (3), 21–30. Owens, S., Rayner, T., Bina, O., 2004. New agendas for appraisal: reflections on theory, practice, and research. Environment and Planning A 36, 1943-1959. Özesmi, U., Özesmi, S., 2003. A participatory approach to ecosystem conservation: fuzzy cognitive maps and stakeholder group analysis in Uluabat Lake, Turkey. Environmental Management 31(4), 518-531. Pahl-Wostl, C., 2009. A conceptual framework for analysing adaptive capacity and multi- level learning processes in resource governance regimes. Global Environmental Change 19, 354–365. Pahl-Wostl, C., Craps, M., Dewulf, A., Mostert, E., Tabara, D., Taillieu., T., 2007. Social learning and water resources management. Ecology and Society 12. Available from: http://www.ecologyandsociety.org/vol12/iss2/art5/ (accessed 05.09.12). Parson, E.A., Clark, W.C., 1995. Sustainable development as social learning: theoretical perspectives and practical challenges for the design of a research program. In: Gunderson, L.H., Holling, C.S., Light S.S. (Eds.), Barriers and Bridges to the Renewal of Ecosystems and Institutions. Columbia University Press, New York, NY. Pelling, M., High, C., Dearing, J., Smith, D., 2008. Shadow spaces for social learning: a relational understanding of adaptive capacity to climate change within organizations. Environment and Planning 40, 867–84. Penney, J. 2012. Adapting to Climate Change: Challenges for Niagara. Niagara Region. Available from:
20 http://www.niagararegion.ca/government/planning/pdf/climatechangerport.pdf (accessed 10.05.13) Plummer, R., 2012. Fostering governance and building capacity for climate change adaptation: can adaptive co-management help?. Governance of Adaptation Symposium, Amsterdam, The Netherlands. March 22-23, 2012. Available from: http://www.adaptgov.com/wp-content/uploads/2012/03/Plummer-Keynote- adaptgov.pdf (accessed 16.09.12). Plummer, R., 2013. Can adaptive co-management help to address the challenges of climate change adaptation? Ecology and Society 18 (4), 2. Available from: http://dx.doi.org/10.5751/ES-05699-180402 (accessed 01.01.14). Plummer, R., Armitage, D., 2007. A resilience-based framework for evaluating adaptive co- management: linking ecology, economics and society in a complex world. Ecological Economics 61, 63-74. Plummer, R., Armitage, D., 2010. Integrating perspectives on adaptive capacity and environmental governance. In: Armitage, D., Plummer, R. (Eds.), Adaptive Capacity and Environmental Governance. Springer-Verlag, Berlin, Germany. Plummer, R., Crona, B., Armitage, D., Olsson, P., Tengö, M., Yudina, O., 2012. Adaptive co- management: a systematic review and analysis. Ecology and Society 17 (3),11. Available from: http://dx.doi.org/10.5751/ES-04952-170311 (accessed 12.04.13). Plummer, R., FitzGibbon, J.E., 2007. Connecting adaptive co-management, social learning and social capital through theory and practice. In: Armitage, D., Berkes, F., Doubleday, N. (Eds.), Adaptive Co-Management: Collaboration, Learning and Multi- Level Governance. UBC Press, Vancouver, BC. Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y., Podsakoff, N.P., 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. Journal of Applied Psychology 88 (5), 879-903. Prell, C., Hubacek, K., Reed, M., 2009. Stakeholder analysis and social network analysis in natural resource management. Society and Natural Resources 22 (6), 501-518. Pretty, J., Ward, H., 2001. Social capital and the environment. World Development 29 (2), 209-227. Reed, M., Evely, A.C., Cundill, G., Fazey, I.R.A., Glass, J., Laing, A., Newig, J., Parrish, B., Prell, C., Raymond, C., Stringer, L., 2010. What is social learning? Ecology and Society 15. Available from: http://www.ecologyandsociety.org/vol15/iss4/resp1/ (accessed 02.10.12). Renzetti, S., Dupont, D., Plummer, R., Heinmiller, T., & Baird, J. 2013. Liquid Assets: Assessing Water’s Contribution to Niagara. Niagara Region. Available from: http://www.niagararegion.ca/living/water/pdf/NR-Liquid-Assets-Paper-2013- FINAL.pdf (accessed 10.05.13). Rittel, H.W.J., Webber, M.M., 1973. Dilemmas in a general theory of planning. Policy Sciences 4, 155-169. Rodela, R., 2011 Social learning and natural resource management: the emergence of three research perspectives. Ecology and Society 16. Available from: http://www.ecologyandsociety.org/vol16/iss4/art30/ (accessed 13.01.13). Rodela, R., Cundill, G., Wals, A.E.J., 2012. An analysis of the methodological underpinnings of social learning research in natural resource management. Ecological Economics 77, 16-26. Roland, G. 2004. Understanding institutional change: Fast-moving and slow-moving institutions. Studies in Comparative International Development, 38(4), 109-131
21 Rowe,G., Frewer, L.J., 2000. Public participation methods: a framework for evaluation. Science, Technology and Human Values 25 (1), 3–29. Available from: http://www.ecologyandsociety.org/vol15/iss4/resp1/ (accessed 09.09.12). Schultz, L., Folke, C., Olsson, P., 2007. Enhancing ecosystem management through social- ecological inventories: lessons from Kristianstads Vattenrike, Sweden. Environmental Conservation 34 (2), 140-152. Stedman, R.C., 2004. Risk and climate change: perceptions of key policy actors in Canada. Risk Analysis 24 (5), 1395-1406. Stirling, A., 2006. Precaution, foresight and sustainability: reflection and reflexivity in the governance of science and technology. In: Voss, J.P., Bauknecht, D., Kemp, R. (Eds.), Reflexive Governance for Sustainable Development. Edward Elgar, Cheltenham, England. Takahashi, B., Meisner, M., 2011. Comparing influences on Peruvian climate change policy: information, knowledge, and concern among political elites. Journal of International Communication Research 40 (3), 181-202. van Vliet, M., Kok, K., Veldkamp, T., 2010. Linking stakeholders and modellers in scenario studies: the use of Fuzzy Cognitive Maps as a communication and learning tool. Futures 42, 1-14. Vernooy, R., 2010. Collaborative learning in practice: Examples from natural resources management in Asia. International Development Research Centre. Cambridge University Press, New Dehli, India. Voß, J.P., Kemp, R., Bauknecht, D., 2006. Reflexive governance: a view on the emerging path. In: Voss, J.P., Bauknecht, D., Kemp, R. (Eds.), Reflexive Governance for Sustainable Development. Edward Elgar, Cheltenham, England. Wenger, E., 1998. Communities of Practice: Learning, Meaning, and Identity. Cambridge University Press, Cambridge, UK. Zahran, S., Brody, S.D., Grover, H., Vedlitz, A., 2006. Climate change vulnerability and policy support. Society and Natural Resources 19, 771-789.
22 Table 1
Type Definition/ indicators of learning effects Measures of indicators Cognitive Test scores; change in centrality learning Acquisition of new knowledge; restructuring of and specificity of knowledge existing knowledge presented in concept maps
Change in, and convergence Normative Changes in norms; change in values; change in around, environmental beliefs; learning paradigms; convergence of group opinion participant reflections; meeting proceedings Improved understanding of mindsets of others; Change in social network structure; Relational building of relationships; enhanced trust and participant reflections learning cooperation
23 Table 2 NEP facets Statement example for rating agreement or disagreement We are approaching the limit of the number of people the Limits to growth earth can support Humans have the right to modify the natural environment to Anti-anthropocentrism suit their needs When humans interfere with nature it often produces Fragility of nature’s balance disastrous consequences Human ingenuity will insure that we do NOT make the earth Rejection of exemptionalism unlivable Possibility of an ecocrisis Humans are severely abusing the environment Decisions about development are best left to the economic Economic domain market
24 Table 3 2012 Mann me Whit an Mean Range of ney 2010 mean % chan chan U % co ge in ge in test cor rr score score statis Participant group1 rect ect s s tic Low activity level 65.25 62.92 -2.33% -15% to 6% 0.037 (n = 12) High activity level 64.45 70.36 5.91% -5% to 29% (n = 11) 1 Only paired 2010 and 2012 scores were used, participants who completed only one survey were excluded from the analysis.
25 Table 4 Category Description Impacts Climate change impacts on a range of sectors, broad mentions of sectors, climate change impacts on the environment generally Adaptation Climate change adaptation actions (underway or suggested, need for) Discontent Statements of discontent with institutional norms and values (formal and informal), no mention of improved approaches Inst. Broad responses that institutions at any level, government and/or NGO, should Res take pon se Governance Identification of possible collaboration and statements of need for inter- organizational efforts (where at least one actor is non-governmental), stakeholder involvement and consensus building Education Better/different methods of communicating, outreach, needs identification for education, citizen's attitudes Opportunit Perceived benefits of climate change to society, or specific sectors y Socio-econ Items related to economic impacts of climate change; economic adaptations General Broad mentions of climate change and global warming, and aspects that cause or exacerbate climate change Mitigation Climate change mitigation activities (underway or suggested, need for) Unrelated Respondents' mentions of specific sectors/issues outside of the climate change context
26 Table 5 95% Con fide nce To what extent did inte you question… Group Median response rval low activity level 2 (‘a little’) ±1.37 Procedure and policies high activity level 3 (‘moderately’) ±0.95 Governance norms and low activity level 1.5 (‘none’ to ‘a little’) ±1.25 protocols1 high activity level 3.5 (‘moderately’ to ‘a great deal’) ±0.60 1 Significant difference between groups, based on confidence interval of the median
27 Table 6 Indicator Group Median 95% CI1 Building of relationships2 low activity level 2 1.25 high activity level 4 0.48 Enhanced trust low activity level 2.5 1.37 high activity level 4 0.95 Enhanced ability to cooperate low activity level 3 0.8 high activity level 4 0.95 1 Confidence interval of the median 2Significant difference between groups, based on confidence interval of the mean
28 Table 7 Group Change in reciprocal Proportion of reciprocal ties (reciprocal/total (%)) ties 2010 2012 Low activity 6/24 (25%) 1/10 (10%) -15% level High activity 15/32 (47%) 11/21 (52%) +5% level
29 Table 8 Was there evidence of learning? Learning indicator Low activity level High activity level Cognitive Acquisition of new knowledge No Yes Restructuring of existing knowledge Somewhat Somewhat
Normative Shift in viewpoint No Somewhat Shift in values No A little Shift in paradigms A little Somewhat Convergence of group opinion (consensus) No Somewhat
Relational Building of relationships No Yes Enhanced trust No Yes Enhanced cooperation Somewhat Yes
30