Social learning as a catalyst for building resilience among smallholder farmers: Exploring its role in promoting transformations

ABSTRACT

Building resilient agricultural systems in the face of the climate crisis requires transformative action, yet progress has remained slow. The current, dominant food system remains a huge contributor to the climate crisis and is itself vulnerable to its effects, making transformations towards resilience a complex and multifaceted challenge. In response, social learning is emerging as a promising mechanism for inspiring transformations by recognizing the central role of learning in building pathways towards resilience. We conducted a systematic literature review to examine this connection between social learning and transformations among smallholder farmers, addressing three research areas: (1) the factors influencing social learning processes ; (2) social learning outcomes; and (3) the extent to which social learning promote transformations . The review reveals many underlying positive assumptions about the role of social learning in transformations , but there remains little insight as to how or why learning leads to the adoption of transformative practices. We also find that while social learning has a positive impact on participants’ understanding of climate-related issues and resilient agricultural practices, sustained adoption of transformative actions remains a challenge. Participatory on-farm demonstrations and addressing perceived risks associated with making changes can increase the likelihood of adopting transformative action. Drawing on these findings, we propose a conceptual framework for promoting the adoption of transformative actions that take into account differences in socio-economic contexts. This study highlights the crucial role of social learning in building resilience and provides insights to inform future research in this field.

KEYWORD:

• Social learning
• resilience
• transformation
• smallholder farmers
• climate change

1. Introduction

In discussions about the future of equitable and sustainable food production, many hopes are being pinned on small-scale resilient agriculture. Resilient agricultural practices are endorsed because of their potential for achieving social-ecological sustainability in an increasingly vulnerable food system (Orsato et al., 2019).¹ Currently, the global-dominant agri-food system contributes significantly to the climate crisis and, simultaneously, remains vulnerable to its impacts (IPCC, 2022). Smallholder subsistence farmers (SSFs), who work on two or less hectares of land, encounter mounting challenges when faced with the climate crisis as they often lack the necessary tools, skills and resources to adapt to and recover from economic, social, institutional, and environmental shocks and stressors (Gomez et al., 2020).

In spite of these challenges, SSFs play a vital role in local and global food and nutrition security. In Sub-Saharan Africa and Asian countries alone, they contribute to an estimated 80% of food production, and globally, they produce an estimated 35% of the total food supply and account for 84% of all farms (Lowder et al., 2021). However, the scale of negative impacts from the climate crisis and other shocks is detrimental not only to the future of food production for SSFs but also to the equitable access to food, prompting urgent calls for transformations towards resilient agricultural practices. Such transformations are instrumental in addressing environmental, social and economic unsustainability across the agricultural sector. As it stands, these transformations often call for an alternative approach to farming practices that requires many resources including financial resources, labour, time, and diverse knowledge (Boillat & Bottazzi, 2020; Wezel et al., 2020). SSFs find themselves restricted because not only do they lack access to many of these resources that facilitate such transformations, but they are already struggling to cope with existing vulnerabilities (Boillat & Bottazzi, 2020).

As such, social learning is seen as essential in catalysing these transformations towards resilience among SSFs. In this paper, social learning is defined as a change in understanding occurring beyond the individual and within wider social units through interactions (Reed et al., 2010). An underlying assumption of social learning in the literature is that they provide an, otherwise inaccessible, facilitating platform for local farmers and other stakeholders to engage in proenvironmental behaviour, as well as a means of acquiring necessary resources for transformative action at the community level (Orsato et al., 2019; Reed et al., 2010). Recent findings from the International Fund for Agricultural Development (IFAD) found that SSFs only received around 1.7% of the total climate finance tracked by IFAD, covering only a small proportion of their needs, further highlighting the need for social learning processes that provide alternative pathways towards transformations (Chiriac & Naran, 2020).

Literature on social learning has grown significantly in the past couple of decades. This is partly because social learning has previously been identified as essential for transformations in SSF systems through fostering pro-environmental behaviours and decision-making (Armitage et al., 2008; Reed et al., 2010). In fact, many authors argue that when learning is fostered through networks, it can strengthen communication between stakeholders, encourage a shared common understanding of a problem(s), develop solutions to address challenges, adopt adaptive behaviours and actions, and ultimately build resilient networks that can be diffused to the wider society (Nikkels et al., 2021; Ridley, 2005; Thi Hong Phuong et al., 2018; Mekonnen et al., 2018). Drawing on these points, it is often assumed that social learning will inevitably lead to resilience among SSFs, and many studies often conclude that increased access to education and knowledge through social learning will spur transformative action whilst, in fact, providing few insights on the processes that actually lead to successful and effective social learning outcomes.

To this end, the aim of this study is twofold: to provide data on what is known at present on social learning and transformations towards resilient agriculture and to develop a conceptual model of successful social learning processes that promote transformative actions towards resilience among SSFs. While social learning comprises both formal and informal learning processes, this study limits social learning to formal interactions between farmers and stakeholders, whether they are instigated as bottom-up or top-down processes. For the development of such a model, factors that influence social learning as well as the success of social learning outcomes need to be determined. Considering how pressing transformations towards resilient agricultural practices are, and the challenges faced by SSFs, social learning has the potential to provide pathways for these transformations and more information needs to be made available on how to facilitate such processes across various socio-economic contexts.

Therefore, the main research question is as follows: what is currently known about the role of social learning in promoting transformations towards resilience in small-scale farming? This question is further divided into three sub-questions:

1. What are the factors influencing social learning processes in agricultural systems?

2. What are the outcomes of social learning processes?

3. To what extent does social learning promote transformations towards resilience?

A systematic literature review (SLR) method is employed in this article to provide an overview of the existing literature and to address the above research questions. Section 2 outlines the methodology. This study adopts the methods used for a systematic literature review by (Tranfield et al., 2003). Section 3 presents the key findings and discussions and section 4 ends with a conclusion, highlighting key implications for future studies.

2. Systematic literature review method

An SLR was chosen to answer the research questions. The frequently used three-stage approach proposed by Tranfield et al. (2003) was used. This method is briefly summarized in Figure 1. This SLR method has been well established as an effective and useful tool for summarizing and analysing existing literature on a given research topic (Michel-Villareal et al., 2019). Using the approach of Tranfield et al. (2003) ensures a transparent and replicable research process that minimizes bias by making the researcher’s underlying values and assumptions explicit in a predetermined protocol. The techniques employed in an SLR, including the development of the search string, as well as the inclusion and exclusion criteria, provide an overview of the decisions and conclusions made by the researchers throughout the SLR process (Tranfield et al., 2003). With this method, efforts were made to systematically address the objectives from the three research questions: (1) understanding how social learning processes are influenced by a number of factors, (2) outlining various outcomes of social learning processes and (3) determining what role social learning plays in catalysing transformations towards resilience.

Figure 1. The systematic literature review process developed using the three-stage approach proposed by Tranfield et al. (2003).

2.1. Planning the systematic literature review

The first step of planning included developing a review protocol that outlined each step that will be taken during the SLR process. This included carrying out a preliminary search of the literature on concepts of social learning as well as resilient agriculture. This helped establish the key concepts and terms used across the field.

2.2. Conducting the review

The search was carried out in the Web of Science (WoS) Core Collection, a comprehensive database with a multidisciplinary focus (Birkle et al., 2020). To conduct the search, the search string in Table 1 was used and confined to abstracts and titles. The asterisk (*) was used to find variations of select terms (Bezner Kerr, 2021). Under the topic of social learning, the indicated terms in Table 1 were used because, based on preliminary research, it was found that these terms are commonly associated with social learning and its processes. The search was also limited to agricultural research as the topic of social learning is comparatively under-studied compared to other fields such as natural resource management and environmental governance. Under the topic of agriculture, the terms smallholders and small scale were used because the scope of this systematic literature review is limited to the role of social learning among SSFs. Furthermore, although resilience is the key concept used in this study, the term sustain* (i.e. sustainability or sustainable) was used as it is often the case that the terms are used interchangeably in the literature, and therefore to not exclude more articles, it was decided to include the term (Grace, 2022).

Table 1. Keywords used in the search string for the systematic literature review and categorized according to the key topics of the research question.

Social Learning	Agriculture	Resilience
“social learning” OR “communit* of practice” OR “knowledge” OR “co-learn”	“smallholder” OR “small-scale” + “agriculture” OR “farming”	“sustain*” OR “resilien*” OR “adapt*”

Historically, key concepts of social learning have been rather contested across the literature. Therefore, it was decided that Reed et al.’s (2010) frequently cited definition of social learning was to be used for this paper. According to Cundill and Rodela, using Reed et al. (2010) definition allows further research to contribute to “shift[ing] the social learning discourse towards greater empirical analysis” (Cundill & Rodela, 2013, p. 8). As such was chosen as the starting year for the exclusion criteria. Another goal of analysing studies published after Reed’s frequently cited definition was to determine if the definition of social learning is now more consistent across the literature.

The inclusion and exclusion criteria were further determined based on a manual screening of titles and abstracts; the number of articles was reduced to 114. Articles were excluded if the abstract did not include subjects on the transformation, transition or adoption of sustainable or resilient farming practices. Furthermore, articles were not included if they focused on the uptake of information and communication technologies (ICTs) as a tool or platform for knowledge-bridging or the digitalization of agriculture. The reason for this exclusion is that many SSFs in the Global South are not sufficiently or reliably connected to ICTs and adaptation processes would be dependent on the whole community’s access and connectivity (Wyckhuys et al., 2018). The full-text version of the selected articles were read in their entirety. Articles that did not mention knowledge exchange or did not specify information sources were excluded, along with articles that only discussed the relevance of media, television or radio as key sources of information. Most articles were excluded on the basis that they only concluded that networks and social capital were important for resilient transformations, rather than focusing on the processes of transformation. This is the case because such articles offered social learning as a solution but did not provide an analysis on why or how social learning itself could lead to transformative actions. To this end, papers were also included if they focused on concepts of transformations. This final screening and selection of articles yielded a total of 39 articles. The full list of inclusion and exclusion criteria is included in Table 2. An evaluation of the articles extracted key bibliometric information including the author(s), title, year of publication, country of study, method(s), definitions of social learning and resilience, and farming systems and practices (as identified by the researcher). These findings are summarized in the section on the bibliometric data.

Table 2. The inclusion and exclusion criteria applied during the first and second screenings of the literature.

Inclusion and exclusion criteria used to screen the articles
Inclusion	Publication year between 2010 and 2022. Written in English. Peer-reviewed. If the article looks at knowledge exchange, co-production of knowledge within and between organizations. If information sources discussed in the study are peers, community members, or organization(s). If the article discusses adaptation practices or transformative actions that could lead to resilience. Focused on sustainable or resilient farming systems and practices as defined by the author.
Exclusion	If the article does not discuss the transition, transformation to, and/or the adoption of sustainable or resilient farming practices. If the role of organizations, initiatives, platforms or other associated terms are only discussed in the conclusion. If the articles are focused on the uptake of ICTs as a platform for knowledge-sharing or the digitalization of farming systems. If information sources originate from media, television, or radio.

2.3. Data extraction and evaluation of the review findings

The 39 articles were re-read and qualitatively and manually coded using the program atlas.ti. The articles were classified according to, first, the method(s) used by the authors (case study, review, and action research) and, second, the definitions of social learning. The categorization of the methods was based on the authors’ self-reported studies, especially the action research and case study articles. Articles identified as reviews or syntheses were also self-identified as such by the authors and because they provided solely an analysis of existing research. In comparison, the case studies and action research provided new data based on interviews, focus groups, surveys, etc., followed by an analysis of the data. While most of the action research articles were self-identified by its authors, there were a few that were identified as such because they shared the same characteristics based on Waterman et al.’s (2001) definition of action research (problem identification, planning, action and change, evaluation).

Within these classifications, the coded quotes were extracted for interpretation and analysis using an adaptation of Ernst’s (2019) framework on the factors that influence social learning. The framework itself was developed with the goal of better understanding “how and under what conditions participation leads to improve outcomes” (Ernst, 2019, p. 1), without operating under the assumption that social learning can induce positive social change. Other notable frameworks such as those developed by Bullock et al. (2022), Cundill and Rodela (2013), and Suškevičs et al. (2018) also explore social learning and outcomes in the context of climate change adaptation and socio-ecological change, yet we selected Ernst’s framework because of its comprehensive take on influencing factors and their interconnections.

To apply the framework, we first coded for and evaluated the factors influencing social learning processes. To code the literature, the factors and their relevant characteristics were clearly established. Based on the descriptions offered by Ernst, these factors focused the characteristics of learning processes, such as the diversity of participants, the more “theory-driven” characteristics, such as the context or procedural fairness, and finally the factors that evolve and/or demonstrate short-term outcomes (Ernst, 2019). Then, a 1 or a 0 was given to that factor within the article based on whether or not it demonstrated an influence on social learning throughout the article. It was also noted when that factor demonstrated a positive or negative influence, thus highlighting the importance of recognizing the varying socio-economic and environmental contexts of each case.

This was followed by an analysis of the resulting outcomes and whether they were considered successful or unsuccessful. Outcomes were considered successful or unsuccessful based on three key indicators as identified in Reed et al.’s definition of social learning: (1) if there was evidence of farmers and stakeholders having undergone a change in understanding and/or having adopted new resilient practices; (2) if these changes have occurred at the individual or at wider social units; and (3) if these processes occurred in a social setting (Reed et al., 2010). This allowed us to critically examine which conditions were present or lacking in successful social learning. Then, we examined to which extent these noted outcomes were consistent with the author’s conceptualizations of resilience in order to find linkages between the two concepts.

3. Results and discussion

3.1. Trends in the literature

Figures 2a - c present a descriptive overview of the SLR results from 39 articles. Figure 2a shows that the number of studies on social learning and resilient agriculture have increased significantly since 2010. This finding is in line with the many claims that participatory governance has been increasingly studied because of its potential role in improving resilience (Ernst, 2019; Reed et al., 2010). The year 2022 was included within the SLR but since this review was conducted between June and September 2022, it is possible that articles published later in the year were excluded.

Figure 2a. Number of Publications in the SLR per Year from 2010 to 2022.

Figure 2b is a representation of the types of studies included in the SLR. In total, 69% of the articles in the review were case studies or comparative case studies, 15% were action research, and 16% were syntheses. The results and analysis are derived from the case studies and action research articles, while the syntheses provide additional information with which to compare the analysis. Since questions on social learning and resilient agricultural practices are often context-specific, it makes sense that most articles, 84%, were case studies that aimed to gain insights on the environments in question.

Figure 2c. Geographical scope of the literature.

Figure 2c shows the geographical scope of the literature. In total, 89% of the articles focused on regions in the Global South, as described by the Global South Studies Centre (Wolvers et al., 2015).² The four articles (in Figure 2c) that focused on the global level were synthesis articles. There are several reasons that could explain why the case studies are concentrated in the Global South. For one, the effects and consequences of climate change are significantly more dire in sub-Saharan Africa and Asian countries.

Figure 2b. Methodologies used in the analysed studies.

The Global South is also where the agricultural sector serves a fundamental role in the respective economies, representing around 90% of all agricultural workers globally (Losch, 2022). As such, it may be that studies are more focused on addressing these pressing issues first. Another reason, however, is that one of the terms used in the search string is “smallholder”, because the scope of this study is limited to SSF systems, and SSFs are largely concentrated in low-income and lower-middle-income countries (Lowder et al., 2021). This is especially the case in sub-Saharan Africa (Giller et al., 2021), where around 57% of case studies were located.

3.2. Definitions of social learning and resilience

3.2.1. Social learning

The two key concepts studied within this SLR are social learning and resilience, both highly contested concepts with varying definitions based on the field of research and the researchers’ goals. As such, the goals were, first, to identify the most commonly applied definitions of either concept to better understand the overarching goals of this field of research, and second, to establish a working definition of social learning in the context of agricultural transformations.

The recent emphasis on the significance and potential of social learning, as a type of learning, results from the increasing recognition that a “self-organizing process, facilitated by knowledge development and learning, has the potential to increase the resilience of resource use systems” (Berkes & Turner, 2006, p. 479). Learning itself can be considered as a process of updating beliefs, where when someone engages in a learning process, they take in knowledge and reflect on it (Dunlop & Radaelli, 2018). There are many ways to conceptualize and understand social learning. For example, several authors (Armitage et al., 2008; Pahl-Wostl, 2009) have drawn on – and expanded upon – Argyris and Schön (1978) and Keen et al.’s (2005) organizational learning frameworks that present social learning in three loops – single, double, and triple-loop. Presented in a stepwise fashion, these levels aim to provide a conceptual framework through which researchers can analyse how individuals and collectives learn about complex issues and to what extent the learning processes and outcomes lead to adaptive behaviour and sustainable transformations. The underlying assumption here is that such an adaptive learning process “may have different levels of intensity and scope” (Pahl-Wostl, 2009, p. 358). Single-loop learning refers to simply adjusting actions and practices without calling into question underlying norms and values, double-loop learning occurs when actors reflect on the norms and values that underpin actions and practices, while triple-loop learning challenges these norms and values (Armitage et al., 2008; Pahl-Wostl, 2009; Reed et al., 2010). According to Pahl-Wostl, the latter “refers to transitions of the whole regime”, without which transformative action in environmental governance would fail to take place (Pahl-Wostl, 2009, p. 359). Other notable definitions of social learning refer to the process as “co-management”, including Schusler et al. (2003) and Plummer et al. (2012). They emphasise the importance of partnerships and links between various actors, involving learning processes that focus on “negotiation and sharing” and “enhancing the capacities of actors” to manage common resources, respectively (Plummer et al., 2012; Schusler et al., 2003). Alternatively, Daniels and Walker (1996) stress the processes involved in social learning and focus less on the stakeholders involved in this participatory process r. Instead, they define social learning as “the process of framing issues, analysing alternatives, and debating choices in the context of inclusive public deliberation” (Daniels & Walker, 1996, p. 73). Also different is the lack of focus on resource management and emphasis on sustainability.

Others, such as Reed et al. (2010), define social learning as a collaborative learning process whereby actors within social networks undergo a change in understanding at the individual level and at the wider societal level (Reed et al., 2010). Reed’s definition draws on Bandura and Wenger’s concepts of social influence and communities of practice, respectively. It also makes a distinction between learning at the individual level and learning that occurs at the wider societal level. As discussed in the methods section, the reason for choosing papers from 2010, was to determine whether social learning definitions are consistent across the literature following the publication of Reed et al. (2010) definition. The SLR demonstrates that this is not the case and that there remain many conceptualizations of social learning within the literature, and authors tend to focus on specific characteristics of social learning. Table 3 summarizes four main defining characteristics of social learning found across the conducted SLR as well examples from those articles and citing references. These defining characteristics are not fixed, and many of them overlap. The extracted characteristics were the most commonly mentioned when authors described social learning.

Table 3. Defining characteristics of social learning extracted from the literature.

Defining characteristics	Description from References in the SLR	Frequency in the SLR
Co-production and Exchange of Knowledge	“Co-production of knowledge has been used to enhance the adaptive capacity of communities facing uncertain environmental change” (Boillat & Bottazzi, 2020, p. 666).	24 articles
Participatory Process	“These participatory approaches mobilized capacity and expertise […] to provide mitigation, adaptation, and food production services” (Chandra et al., 2017, p. 226). “Innovative platforms and cooperatives are reported to be improving capacities for farmers in […] sustainable solutions to identify problems in farming communities” (Makate, 2019, p. 277).	15 articles
Influence of Social Norms	“The opinion leaders within communities also significantly influence their fellow farmers […] they hold different positions and frame different communication channels” (Tamako et al., 2022, p. 2).	9 articles
Transformative Learning	“Transformative social learning denotes the process of gradual change in views and ways of acting when social actors are confronted with challenges in their environment” (Kansanga et al., 2021, p. 2).	3 articles

Reed et al. (2010) argue that conceptualizations of social learning are often conflated with processes of participation (Reed et al., 2010). As they put it, participation is a process that facilitates social learning, but it does not inherently mean that learning will occur in such a process. In fact, a couple case studies in the SLR reported that in the interviews following participatory social learning, some farmers indicated they did not acquire new knowledge (Salvini et al., 2016). Nonetheless, it remains commonly used and around 15 articles mentioned participatory processes as the defining characteristic of social learning. In other words, without participatory processes, social learning cannot occur.

In contrast to the emphasis on participation, nine articles in the SLR focused specifically on the influence of social norms and peers on individual behaviour. In these articles, social learning is understood as external influence as the motivation for change in an individual’s behaviour. There are several ways in which farmers are influenced by their peers. Sarkar et al. (2022) highlight the importance of demonstrating positive attitudes towards new or sustainable practices for them to be socially accepted and adopted by other farmers. Buyinza (2020), Saint Ville et al. (2016) and Mutyasira et al. (2018) further stress the role that social norms play in building social pressure among farmers in a community to adopt certain practices.

Based on the findings represented in Table 3 and the analysis above, an adaptation of Reed et al. (2010) definition is proposed for future studies on social learning in the context of agricultural transformations. Reed et al. (2010) highlights three key elements in their definition, including a change in understanding; learning that goes beyond the individual to reach wider social units; and that this process occurs through social interaction (Reed et al., 2010). To begin with, the co-production of knowledge was the most commonly, and explicitly, cited characteristic of social learning, where 24 articles mentioned it as the defining characteristic. In the context of this SLR, the co-production of knowledge refers to the act of bringing together multiple sources of knowledges and methodologies with the “goal of co-creating knowledge” (Djenontin & Meadow, 2018). In many of the articles, knowledge exchange, building on participants’ knowledge and problem solving are necessary for social learning to occur (Boillat & Bottazzi, 2020; Henderson et al., 2021; Bezner Kerr et al., 2018).

Following the analysis of the articles in the SLR, it can thus be argued that knowledge co-production is an important element that is not implicit in the existing definition. Furthermore, an interesting observation, considering how the literature in the SLR focuses on agricultural transformations, is how infrequently transformative learning was discussed. However, it is possible that because the literature on transformations remains relatively recent, and other terms, such as adaptive co-management, have previously been used to referred to similar processes (Plummer, 2009). As succinctly explained by Plummer (2009), the concept of adaptive co-management has evolved to increasingly consider adaptation for transformation (Plummer, 2009). To this end, the use of the concept of transformations encourages a contribution to further this field.

In the recent literature, transformative learning has been used to identify and emphasize outcomes of social learning including a “change occurring in individuals’ attitudes, values, and frames of reference, but also the change of practices or individual actions” (Suškevičs et al., 2017). Despite social learning being seen as essential for transformative learning, there are only a handful of articles, three in total from the SLR, that focus on this connection. The example quoted in Table 3 illustrates how Kansanga et al. (2021) conceptualize social learning with the specific aim of changing understandings and behaviours among participating stakeholders. However, this definition may not reflect the deep-seated and radical change that other authors believe transformation learning should bring about (i.e. Diduck et al., 2012; Markard et al., 2012), indicating that there remain continued discussions on how transformations are conceptualized across the literature.

Ultimately, this is an assumed outcome of social learning; however, making the distinction between social learning and transformative social learning recognizes that not all participatory processes will lead to changes (Reed et al., 2010; Kansanga et al., 2021).

3.2.2. Resilience

The second key concept in this research is resilience. Historically, the term has been largely contested as early thinkers (Holling, 1973) argued that resilience as a concept concerns the durability of a system and its ability to maintain its original functions under duress (Folke et al., 2010). However, in the context of complex systems, any disturbance can have unprecedented and unpredictable ripple effects, making it difficult for such a system to return to its original state. Furthermore, there is a considerable push to move away from currently intensified farming practices because of the myriad of negative impacts on the environment and society (Beyer et al., 2022; Langsdorf et al., 2022). To this end, in recent literature, more focus has been placed on the adaptive capacity and the transformability of a system in the face of unexpected and unprecedented internal and external shocks and stressors (De Steenhuijsen Piters et al., 2021; Folke et al., 2010; Meuwissen et al., 2019; Masson-Delmotte et al., 2021; Tittonell, 2020).

Czekaj et al. (2020) suggest that in farming, concepts of resilience can be approached in two distinct ways: a holistic approach that views the potential of farms as resilient systems or the capacity of a farmer to adapt to and cope with challenges on their farm (Czekaj et al., 2020). Within this SLR, it was found that the latter definition is more commonly used. Based on the literature, resilience is defined as the capacity (64%) to adapt (90%) in the face of external pressures (79%), where the percentages represent how often that concept was included in the articles across the SLR. This definition presents several interesting finds. First of all, the aspect of transformability is clearly lacking in the literature, despite it being akey component of resilience according to the 2021 IPCC report (Masson-Delmotte et al., 2021). However, this can potentially be explained by the nature of case studies, in that it is often amore tangible goal to study adaptive capacities and the adoption of practices over long-term transformations of systems. Among the 39 selected articles, only five of them specifically referred to farms as systems that would benefit from transformations. Three of those articles focused on the need for new pathways to reshape the food system, a process which can be achieved primarily through learning (Andreotti et al., 2020; Kansanga et al., 2020; Henderson et al., 2021). Two of those articles are self-categorized as action research articles, which accurately reflects the goals of such a methodology that “seeks to understand and improve the world by changing it” (Baum et al., 2006, p. 854). Also, one article emphasized the importance of inclusive knowledge production and encourages “‘proactive’ adaptations alongside ‘reactive’ strategies to prompt more transformative action” (Silici et al., 2021).

These different approaches to understanding transformations towards resilience in farms are also reflected in the varying SSF systems and practices identified and discussed by the researchers within the literature, illustrated in Figure 3, Figure 4. Case studies (in blue) often looked at how learning affected the adoption of certain practices that would enhance a farmer’s adaptive capacity such as improved water irrigation, crop diversification, or sustainable soil management. Other articles either did not mention a specific farming system, but rather the focus of the paper remained on social learning.

Figure 3. Farming systems and practices identified by researchers in the literature.

Figure 4. Conceptual model presenting social learning factors, processes and outcomes in the context of transformations towards resilient agriculture.

Second, concepts of resilience are, arguably, frequently associated with external shocks and stressors such as climate change and only recently have other dimensions of social and economic resilience been discussed (Czekaj et al., 2020). This observation is also reflected in this SLR, as 79% of the articles examined farm system’s resilience in the face of climate change. This finding highlights a focus in the literature on the immediate threats posed by unpredictable climatic events, and the need to address those events through the adaptive capacity of farmers. While these threats are undoubtedly a pressing matter, risking the livelihoods and food security of many SSFs, there is the risk that such a focus can overlook the need for social and economic dimensions of resilience in research. Nonetheless, a handful of papers did broaden their scope to discuss wider societal issues, such as the COVID-19 pandemic (Tamako et al., 2022), or social shocks and stressors, such as gender inequality (Mathez-Stiefel et al., 2016), trauma and tragedy (Maltitz & Bahta, 2021), and issues of food justice (Millner, 2017). Within these articles, the inclusion of other dimensions allows for broader resilience strategies as the authors focus specifically on how the influence of societal norms on learning, as well as the role of equitable access to learning opportunities, directly affect the adaptive capacity of all members of the community. Interestingly, three papers look at resilience in terms of the long-term availability of resources, based on the definition proposed by the World Commission on Environment and Development in 1987. In these case studies, these authors also highlighted the importance of long-term resilience by acknowledging that while the present populations’ needs must be met, it is critical to ensure that future generations also have equitable access to resources.

3.3. Social learning as promising pathways towards transformative action

3.3.1. Research question 1: What are the factors that influence social learning?

Social learning, as a collaborative learning process that involves numerous stakeholders and is subject to the goals of participants, is inherently dependent on a number of factors that can impact these processes, whether positively or negatively. Above we discussed the varying ways in which authors across the SLR articles defined, and prioritized aspects of, social learning, highlighting the lack of a common definition. Also apparent in the literature is that conceptualizations of learning are highly dependent on underlying goals and assumptions. In this section, we now look at the factors that influence social learning. Similarly, the presence and discussion of influencing factors varies across the SLR. To better understand and categorize these factors, we adapted Ernst’s framework from her review on factors that influence social learning. There exists a wide range of factors, moving from more descriptive ones such as participation format and diversity of participants to more conceptual ones such as procedural fairness and legitimacy. The purpose of using Ernst’s review is twofold: first, it serves to frame the SLR coding and data analysis, and second, it aims to confirm Ernst’s findings in the context of social learning in agricultural transformations, as hers is centred around participatory environmental governance. Table 4, below, summarizes these factors, along with the respective characteristics (based on Ernst’s review and coded for in the articles) that distinguish them. For the most part, the literature is strongly in-line with Ernst’s review, with evidence supporting the presence and importance of these factors in social learning. Nonetheless, two additional factors have been included in this framework: access to resources, in the participation process characteristic category, and innovation, in the intermediate process outcome category. These two factors were included because they were also frequently mentioned throughout the literature. Also included in the table is a representation of the presence of these factors across all articles in the SLR, indicated by percentage. To analyse the coded data, a score was given to each factor based on whether that factor was absent (0) or present (1) within a given article. The presence of these factors was noted when, or if, the authors clearly indicated that they influenced social learning, both positively and negatively. By using this method, we were able to determine the extent to which factors were considered influential by the authors. For example, while almost all authors found that the participation format was deemed a key influencing factor, mentioned in 98.6% of articles, only 43.7% of articles specifically spoke about the role of conflict resolution.

Table 4. Factors influencing social learning adapted from Ernst (2019).

Factor	Ernst’s Definition of the Factor	Characteristics	Presence of factors across all articles in the SLR
			L.R.	C.S.	A.R.
Participation Format	The approach to organizing and structuring processes of learning.	feedback, active participation, spatial distribution, visualizations, simulations, experimentation, and games	67%	63%	3%
Access to Resources	The access to financial and material resources.	Internal credit loans, external funding, and availability of resources	83%	67%	83 %
Access to information	The access to external scientific and technical resources.	Education, raising awareness, and building technical capacities	17%	81%	100 %
Diversity of participants	Representation of interests, values, and knowledge.	Network size, multi-stakeholder partnership	83%	70%	100 %
Facilitation	The mediation and structuring of discussions.	use of facilitators or mediators	33%	78%	100 %
Participant Characteristics	This includes gender, age, educational experience, political attitude and other behavioural patterns.	Time, workload, family member involvement, lived experiences, individual circumstances, motivation	67%	74%	83 %
Context	Pre-existing social and cultural contexts, types of issues, pre- existing relationships, and institutional settings.	safeguarding local cultures, faith-based organizations, dominant narratives, local political and economic environment	67%	74%	100 %
			L.R.	C.S.	A.R.
Procedural Fairness	Aspects of fairness and justice such as equal participation and political equality.	flexibility, blending scientific and local indigenous knowledge, addressing gender (in)equalities, inclusive actions to mitigate power imbalances	83%	74%	83%
Legitimacy	Exercising sovereignty over decision-making and validating processes.	Governance, democratic, agency, capable of self-organizing, relevant	67%	44%	83%
Effectiveness, efficiency, satisfaction	Relating to the perceived output and outcome of participation.	Influence of social norms, perception of risks, recognition of challenges, perception of implications for livelihoods	67%	78%	100 %
Trust	Trust between the stakeholders during learning processes	safe environment for collaborative action	50%	44%	100 %
Network Building	Processes that establish new or existing relationships between participants	sharing and collaboration, membership	83%	93%	83%
Conflict Resolution	Conflict as an attempt for participants to articulate and advocate for their concerns and interests. Consensus is understood as having overcome these conflicts or that all participants have reached an agreeable decision	problem solving and overcoming conflicts	33%	15%	83%
Innovation	The act of co-producing knowledge, encouraging creativity, innovating, and reflecting on participants own perspectives and knowledge	creativity, reflection and common vision	67%	52%	83%

Overall, it was found that almost all factors were discussed and demonstrated at least some influence on social learning in the six action research articles. In particular, participation format, access to information, diversity of participants, facilitation, context, effectiveness, and trust were present in all articles. Many of these factors are inherent to action research precisely because such a research process demands diverse participation and recognition of socio-economic contexts in order to promote transformative change (Mathez-Stiefel et al., 2016). The more descriptive factors, such as participation format, access to resources and information, and processes of procedural fairness, were regularly noted as influential in the case studies and action research articles, ranging from 67% to 100% of articles. On the other hand, while the more normative factors including legitimacy to innovation are found to be varyingly influential in case studies (with ranges from 15% to 78%), remain highly influential in the action research articles (between 83% and 100%). It is possible that, given the nature of outcome-oriented action research, more attention was paid to issues of trust between participants and the role of conflict in discussions, while innovation, through creativity and reflection, was highly encouraged in order to achieve successful outcomes for both participants and researchers.

In general, the case studies offered significantly more varied results compared to the action research articles. This could be because in these case studies, researchers were often seeking to better understand what these characteristics are and to what extent they influence social learning processes and the outcomes. The most commonly discussed factors included participation format, access to information, participant characteristics, diversity of participants, facilitation, context, procedural fairness, and effectiveness. Access to resources and innovation were mentioned in over half the articles, hence why they were introduced to Ernst’s original framework. Access to resources was particularly relevant as it served as a big indicator of whether or not an individual will commit to engaging in social learning. It was often the case that organizations, either self-organized or not, could offer access to increased credit, tools and machinery as well as agricultural inputs which would significantly decrease an individual farmer’s or community’s risk aversion and encourage them to participate. It also serves as an indicator of how well a community involved in social learning would be able to increase their adaptive capacity and transform their practices. In one case, a researcher noted that local social learning initiatives struggled to adequately diffuse adaptive technologies to the community, resulting in low intentions to adopt adaptive practices among community members (Li et al., 2020). Another researcher noted that community members deliberately did not join social learning initiatives precisely because they did not have the resources to participate in activities, resulting in low levels of self-reported innovation in this community (Saint Ville et al., 2016). Another interesting finding is the apparent lack of focus on the role of conflict resolution throughout social learning processes. Cuppen (2018) argues for the value of conflict in social learning, also in line with research by Daniels and Walker (1996) who highlight the importance of constructive approaches to conflict that foster open dialogue and learning. Cuppen suggests that conflict in itself is a form of “self-organized participation” that encourages otherwise excluded dialogue between participants by shedding light on varying perspectives and, ideally, allowing participants the space to jointly seek a common vision (Cuppen, 2018). Nonetheless, one article did highlight that social learning initiatives within the community actually decreased conflict, in terms of disagreements between farmers, because participants were able to adopt improved irrigation schemes that helped mitigate previous disagreements (Parry et al., 2020). Thi Hong Phuong et al. (2018) in fact, also point to the importance of constructive conflict as critical to social learning.

Several studies specifically highlighted the need to address gender inequalities, noting that in some cases, women would not actively participate in social learning processes unless efforts were made by facilitators to address gender inequalities (Gotor et al., 2021; Kansanga et al., 2020, 2021, Maltitz & Bahta, 2021; Nyasimi et al., 2017; Settle & Garba, 2011; Shaw & Kristjanson, 2014; Thi Hong Phuong et al., 2018; Tume et al., 2019). There were several limiting factors for women noted among the studies. For example, it was difficult to encourage women to actively participate, some cropping systems were traditionally associated with one gender, and, in one paper, mixed-gendered groups reinforced gender inequalities by ignoring power relations that dictate social status (Tume et al., 2019). Importantly, Zeweld et al. (2019) found that farmers who did not value relationships with community members were less likely to be influenced by them and thus less likely to learn from them.

However, studies in both the case studies and action research categories found that sometimes, farmers lack intrinsic motivation and, despite social learning interventions, did not learn or that the social learning may simply not lead to change (Salvini et al., 2016). This lack of intrinsic motivation is also reflected in some case studies where farmer field schools or other similar initiatives were abandoned by the participants shortly following the departure of trainers and experts, promptly ending these educational and communication initiatives between participants and the provision of guidance (Nyasimi et al., 2017; Osumba & Recha, 2021). However, it is also possible that in some cases, SSFs did not have the time and resources themselves to maintain such initiatives without external support (Saint Ville et al., 2016). One article also found that social learning was, in fact, not the farmers’ preferred way of learning, but rather they preferred learning through media such as television and radio (Popoola et al., 2020). Also, important to acknowledge, is the fact that all of these factors listed in the adapted framework can also negatively influence learning processes and that social learning itself might lead to negative outcomes with regard to resilience. Similarly, to how conflict and trust can provide a pathway for a joint vision, they can also lead to tensions and unhealthy competition among participants (Kansanga et al., 2021). Additionally, without explicit procedural fairness and facilitation, it can often be the case that social learning reinforces social inequalities and addressing these requires a deep understanding of socio-economic contexts. For example, while some case studies highlighted the importance of grouping men and women together to promote equitable access to information and resources, one demonstrated that women-only groups fared better in terms of adopting new practices because women felt safer in such environments, thus improving their self-esteem (Tume et al., 2019). However, the latter was not representative of the overall case studies and there were far more examples of mixed groups benefiting minorities as long as a facilitator was present to mediate unequal power dynamics. Bezner Kerr et al. (2018), for example, illustrated how including women and AIDS-affected community members in experimental social learning encouraged inclusivity in other aspects of community life (Bezner Kerr et al., 2018).

Other factors that had a negative influence were participant characteristics and participation format, where several case studies found that large differences in age between participants, network size and lack of familial ties in membership were found to reduce trust, and thus learning (Saint Ville et al., 2016; Wossen et al., 2013). In terms of impact on resilience, one article found that membership with formal associations was negatively correlated with the adoption of resilient agricultural and natural resources management practices because membership provided farmers with access to inputs and resources for intensifying agricultural production (Wossen et al., 2013).

3.3.2. Research question 2: What are the outcomes of social learning processes?

Determining the outcomes of social learning, and whether or not they are successful, is arguably a subjective goal that depends on the objectives of the researcher(s) and study as well as the goals and objectives of any initiative (Muro & Jeffrey, 2008; Nikkels et al., 2021; van Epp & Garside, 2019). In this review, the success of social learning outcomes is determined based on whether a change in understanding occurred, if there was a noted change in practice, and whether this change occurred at the individual or at the wider societal level (Herrero et al., 2019; Reed et al., 2010, Siebenhüner et al., 2016; Thi Hong Phuong et al., 2018). In the context of achieving transformations towards resilience, outcomes were successful if it led farmers and stakeholders to acquire the knowledge to improve their practices or if there was evidence that they did adopt such practices. For example, Osumba and Recha (2021) found that in farmer field schools, while farmers did show evidence of learning and acquiring new knowledge, there was little evidence of adopting actions as they preferred to be guided by facilitators. On the other hand, Chandra et al. (2017) found that following social learning, farmers adopted organic farming practices. Both of these examples illustrate varying levels of success in social learning processes, compared to Saint Ville et al. (2016) who found that farmers did not have the resources nor desire to participate in learning and were even less responsive to change (considered unsuccessful learning processes).

Although a change in understanding is only one aspect to Reed et al.’.s (2010) definition, arguably it does represent a positive outcome for social learning because of the emphasis on learning in resilience literature. According to Oteros-Rozas et al. (2019), processes of social learning, which involve both learning as well as knowledge co-production and exchange, have the potential to be transformative through its efforts to challenge dominant paradigms and worldviews (Oteros-Rozas, et al., 2019, p 5). This further supports the proposition that knowledge co-production should be emphasized more within definitions of social learning, especially when there are underlying assumptions that social learning catalyses transformations. In the SLR, almost all case studies and action research articles reported at least a change in understanding that occurred following social learning (32 articles out of 39). However, in 17 of the articles, farmers moved beyond just a change in understanding and reported a change in practices following social learning, compared to 11 that only reported the change in understanding. Of the articles that demonstrated increased adoption rates of resilient farming practices, a critical component of social learning was the ability to have on-farm trials and demonstrations that prompted farmers to learn-by-doing. Another critical factor that played a significant role in changing practices among farmers was efforts to address perceived risks. Often, case studies found that farmers were less likely to engage in social learning initiatives if they were not fully aware of any added benefits. Many of these farmers are SSFs in low-income countries who depend almost entirely on farming for their source of income and food security (Lowder et al., 2021). Therefore, efforts to change farming practices should be accompanied with demonstrations of livelihood benefits otherwise farmers are less inclined to invest resources into practices that they are less familiar with. Among the action research articles, almost all articles reported that the studies resulted in a change in practices.

There are several reasons why some of the case studies reported only a change in understanding. Two articles, for example, only interviewed farmers on their intention to adopt practices following participation in social learning initiatives (Li et al., 2020; Occet et al.e, 2021). In these cases, no follow-ups were conducted to determine whether or not those who indicated such intentions actually followed through with them. Another article cited lack of collaboration among stakeholders following the end of the project, while other articles did not include the adoption rate of new practices within the scope of their studies (Habtu, 2018; Tume et al., 2019). Interestingly, one article examined how farmers changed their practices, but only ranked social learning from social interaction as 6^th when asked about their preferred modes of learning, favouring media sources such as television and radio instead (Popoola, 2020).

Looking at the second element in Reed’s et al. (2010) definition, we analysed if the change in understanding seemed to have moved beyond the individual or not. A number of articles, 18 in total among the case studies, demonstrated a change in understanding beyond the individual level and moving into the community. In some articles, this was very clearly the case. For example, one article discussed how a social learning project on organic practices by a Farmer Field School not only increased adoption of organic practices but also shaped new approaches to policy development at the national scale (Settle & Garba, 2011). Two articles in particular illustrated how social learning among women-only groups or in projects that were specifically inclusive, led to changes in practices but only among women and minority groups (Maltitz & Bahta, 2021; Tume et al., 2019). Many of the articles where understanding was widespread involved projects like Farmer Field Schools. Much like those articles that demonstrated increased adoption of practices, learning-by-doing and demonstrating benefits was critical in gaining community members’ trust.

Considering the proposed addition to Reed et al.’s definition, we also measured to what extent processes of co-production of knowledge occurred. Twenty-one of the 39 articles analysed in the SLR included processes of co-production of knowledge. Many specifically defined social learning as the co-production of knowledge such as Boillat and Bottazzi (2020) who write that their study was “based on processes of co-production”. Others highlighted it as an aspect of social learning that is key in influencing collective action, such as Bezner Kerr, et al. (2018) who acknowledge that knowledge co-production serves as a political act in-itself.

3.3.3. Research question 3: To what extent is social learning contributing towards building resilience?

Following the review of the literature, this section concludes with a conceptual model outlining key factors and processes that lead to positive and successful social learning outcomes. Nonetheless, it does remain difficult to provide a definitive answer to this question for several reasons. First of all, not all authors measured outcomes in the same way, thus, the results above are interpreted based on the developed definition of social learning in this study – if a change in understanding occurred, if there was an uptake in adoption of new practices, and, or, if processes of co-production of knowledge occurred. Furthermore, it is also important to recognize that only one study in the SLR took place over a period of more than one season. Almost all case studies only looked at the immediate outcomes following social learning processes and, therefore, it is difficult to draw conclusions on the long-term impacts. This is in-line with Measham’s (2013) study that sought to address a gap in the literature on social learning time frames. One study highlighted that adaptation projects developed by social learning processes may not always be effective long term as they often focus on and adapt to immediate threats instead of projected changes (Silici et al., 2021). As such, while many studies had fairly positive outcomes that corroborate with many of the assumptions on social learning in the short term, there is still a significant gap in the literature on the long-term outcomes of social learning and its role in contributing towards building long-term resilience. Considering the results discussed in the two previous sections, it could be argued that assumptions should not be made that learning inherently leads to the adoption of resilient practices as there are many situations which can negatively or positively influence outcomes.

Nonetheless, as discussed in the section above, almost all case studies found evidence of changes in understanding of the issues at hand among participants, and around half of the case studies noted changes in practices. Some researchers specifically highlight improvements in resilience in their study areas. For example, Chandra et al. (2017) found an increase in adoption of organic farming practices after attending farmer field schools, Gotor et al. (2021) noted benefits for participants' livelihoods and ability to recover from shocks after diversifying their crops. Others illustrated how sharing resources and information among community members in social learning networks also helped improve resilience (Maltitz & Bahta, 2021) as well as bargaining power among agricultural traders, thus further benefiting livelihoods in the long term, in addition to ecological resilience. On the other hand, Chandra et al. (2017) also noted that despite visibly improving yields, no immediate climate change mitigation benefits were found following the end of the study, further highlighting the need for more studies on the long-term effects of adoption and transformation from social learning.

Throughout the case studies in the SLR, there is demonstrable evidence that social learning can play a role in building resilience, but there are factors (as outlined in Table 4) that influence these processes. While all the noted factors play a significant role in determining whether or not stakeholders will participate in social learning initiatives, let alone learn from them, the degree of impact depends significantly on the socio-ecological contexts of each study. Thus, a critical component of studying social learning should include developing an overview of the local contexts in order to better understand which factors are more likely to influence learning process, and how. In the case that such an overview is overlooked or not addressed, it is less likely that there are positive outcomes from social learning. Future studies should take into account the adapted framework presented in Table 4, focusing instead on relevant factors and socio-economic contexts of each case in order for social learning to be fostered appropriately.

It is thus worth developing a comprehensive conceptual model (Figure 4), presenting the most commonly cited influencing factors and processes throughout social learning that lead to successful outcomes, based on the analysed literature. Since achieving resilience is not a static state, but rather a dynamic one, the most important outcome that could be argued is the continuous ability and desirability for farmers to continue participating in and learning from social learning processes to better develop their capacity to adapt and transform.

The outer part of the model represents the cyclical process of a farm system’s transformations towards resilience. Resilience is a dynamic one that reacts to the demands and consequences of shocks and stressors, and systems are in a constant state of learning and adapting. Social learning is at the centre of the model, highlighting its role as a vital process connecting farm systems to transformations. The connections between social learning and farm systems and transformations are represented by a double-sided arrow because of their ability to influence one another. Below social learning has highlighted some of the key take aways from the SLR. The factors listed are those considered among the most influential catalysers of social learning and its processes. Among the factors, gender played a significant role depending on the context as did the availability of resources and information and the presence of facilitators. It was also noted that effectiveness, such as motivation to participate and addressing farmers' concerns regarding risks, also played significant roles in encouraging social learning and guiding participants in positive directions. Throughout the social learning process, trust between members and conflict resolution, knowledge co-production and exchange and experimentation also all influenced the outcomes. Procedural fairness also plays a critical role, its current position in the model illustrates that this factor plays a role in both driving social learning and as an ongoing characteristic throughout social learning processes.

The most common outcomes are also included in the model. The most recognized outcome was a change in understanding among farmers, for 32 of the articles. Almost all action research articles moved beyond these outcomes between farmers to also change the researchers’ future approaches to studies. A significantly positive outcome was found in Chandra et al. (2017) study, where the authors reported that due to local social learning initiatives between farmers and relevant stakeholders, “the national and municipal governments were able to incorporate community vulnerability into relevant regulatory plans” (Chandra et al., 2017, p. 226). These last two outcomes are highlighted because they primarily represent outcomes that went beyond a change in understanding and the adoption of practices to demonstrate a change at the societal level and not just in the local community.

A noted challenge often faced by policymakers is the lack of scientific evidence on the benefits of resilient approaches to farming, such as climate smart agriculture (CSA). Instead, much of this evidence remains anecdotal and no single pathway can be replicated across all farming systems as the success of CSA, agroecology and regenerative agriculture, among others, depends on the socio-environmental contexts in which they are based (Chandra et al., n.d.; Lundgren et al., 2021). With the support of such studies, like Chandra et al.’s, local governments can find the necessary scientific evidence to put forth policies that integrate more specific and targeted support for their communities. With this increased support from governments, many of the perceived risks faced by farmers can be lowered, thus hopefully increasing the rates of adoption of transformative action towards resilient agricultural practices.

While this conceptual model does provide support for the argument that social learning can serve as a promising catalyst towards transformative action and the adoption of resilient agricultural practices, there remain several things to consider. As explored earlier, it is essential to recognize specific socio-economic contexts, and yet, in some cases that may not be enough as there is an inherent lack in intrinsic motivation or external support which can influence not only the outcomes of social learning, but whether or not it will occur.

4. Conclusion

The goals of this SLR were twofold: to provide a brief overview of the role that social learning plays in the transformation towards resilient agriculture and to develop a conceptual model of the drivers and social learning leading to successful outcomes. As such, several conclusions can be drawn from this study: 1) conceptualizations of social learning across the literature remain contested, and researchers emphasize the importance of certain components depending on their normative goals; 2) although concepts of resilience have grown increasingly broader to include social and economic concerns, the majority of case studies continue to focus solely on environmental resilience thus sometimes overlooking more holistic and interdisciplinary approaches to addressing the complex relationships generated from interactions between humans and the environment; 3) social learning has demonstrated the tendency to improve and change participants’ understanding about issues at hand, and several factors such as experimentation and addressing farmers’ concerns were pivotal in encouraging the adoption of resilient practices. These conclusions provide several insights for future studies, especially with regard to providing more comprehensive information on the socio-economic contexts of study sites. The climate crisis is impacting smallholder farmers’ ability to adapt to increasingly unpredictable shocks and stressors and greater efforts need to be made to explore the potential of social learning in catalysing necessary transformations towards resilience among the most vulnerable. The conceptual model developed in this SLR serves to assist researchers on how to navigate challenging contexts in order to help facilitate and promote future successful social learning. Further research could serve to empirically analyse social learning in the context of smallholder farming systems to further explore the application of the developed model and Ernst’s framework, furthering discussions on factors that influence social learning. The current choice of Ernst’s framework also guided the data analysis and outcomes. It could also be of interest to explore the use of other notable frameworks such as those presented by Bullock et al. (2022), Cundill and Rodela (2013), and Suškevičs et al. (2018). Future research could benefit from a comparison of such frameworks to understand how they may overlap or explore their differences. Furthermore, few studies have currently looked into the long-term outcomes of social learning, indicating a significant gap in the literature where more research could be done on following up with social learning processes to determine their success in the long-run and elaborate on any new challenges that may have arisen. In spite of this, the SLR continues to demonstrate significant positive outcomes of social learning in catalysing transformations towards resilience and more efforts should be geared towards continuing to facilitate these processes.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.

Notes

¹ Examples of resilient agriculture include organic farming, agroecology and regenerative farming and, according to several IPCC reports, definitions of resilience focus on both the adaptive capacity and transformability of a system.

² The concept of the “Global South” entails many political connotations and implications including a country’s economic standing, different relation between regions and countries, as well as a country’s geographical location (.