Can agricultural research and extension be used to challenge the processes of exclusion and marginalisation?

ABSTRACT

Purpose: This paper examines if and how agricultural researchers and extension officers can see, understand and change processes that exclude some people and influence marginalisation.

Design and methodology: We used participatory action research (PAR) in a programme building sustainable farming practices for nutrition and income in Solomon Islands as our case study. Two qualitative PAR data streams were analysed: (i) documentation of community activities over three years including action planning, learning activities, training workshops, focus group discussions, key informant and informal interviews and (ii) documentation of the research teams’ own learning and reflection sessions.

Findings: Agricultural research and learning activities facilitated through PAR can help researchers and extension officers see, understand and challenge processes that cause social exclusion and marginalisation and lead to inequitable access to agricultural opportunities. A combination of (i) starting with a collective vision; (ii) facilitating systematic reflection exercises; and (iii) having locally tuned facilitators creating safe spaces; makes processes of social exclusion tangible, discussable and ultimately actionable, illustrating the potential of the research and extension processes to facilitate social change in real time.

Theoretical Implications: The paper makes a contribution to the growing body of theory and literature on innovation systems and people-centred approaches to agricultural development, by highlighting the facilitation challenges and opportunities that can create more learning focused and power-aware agricultural programming.

Practical Implications: Our approach, examined in this paper, can improve implementation of policies such as the Solomon Islands Agriculture and Livestock Sector Policy (2015–2019), which aims for active participation of women and youth in agricultural development.

Originality: Using a PAR approach to discover how agricultural research and extension activities can help transform the processes that cause social exclusion and create disadvantage and marginalisation.

KEYWORDS:

• Marginalisation
• social exclusion
• participatory action research
• reflection
• collective learning

1. Introduction

Reaching and supporting marginalised people is a goal of many agricultural research, development and extension programmes. A new generation of approaches to agricultural research aims to achieve development outcomes by reaching the marginalised, these include integrated agricultural research for development (IAR4D) (Hawkins et al. 2009; Hall et al. 2014), and research in development (RinD) (Douthwaite et al. 2017). Common among them is their use of participatory methods to engage farmers and policy-makers directly in agricultural research, following the early work of Chambers (1989) that led to a proliferation of methods and policies for putting farmers first (e.g. Scoones, Thompson, and Chambers 2009; Wettasinha and Waters-Bayer 2010). Despite their intent, reaching marginalised people remains a challenge. As cautioned by a number of scholars (e.g. Nelson and Wright 1995; Cook and Kothari 2001; Wollenberg, Anderson, and Edmunds 2001) participation is not a silver bullet and, if used without sufficient attention to power dynamics, it can become a new form of ‘tyranny’ that further reinforces existing power relations that exclude and marginalise some people. Moving away from power blind use of participatory tools requires an understanding of the underlying power dynamics and associated social norms that influence participation. Within the agricultural research and extension sector we see a growing appreciation of the need to engage with power, particularly with a gender lens (Nightingale 2002; Alex 2013; Galiè 2013; Mudege et al. 2016; Lamontagne-Godwin et al. 2018).

Central to the logic of ‘power-aware’ approaches, is the ability to facilitate a process of empowerment that, as Eyben et al. (2008, 6) suggest ‘happens when individuals and organised groups are able to imagine their world differently and to realise that vision by changing the relations of power that have been keeping them in poverty’. Yet, some argue that engaging with multiple forms of power is often forgotten in development programmes (Kabeer 2000; Cornwall and Eade 2011; Cornwall and Aghajanian 2017). They argue that empowerment must be embraced as a process of shifting multiple power relations. Similarly, in practical fields where marginalisation has received more attention (such as health, education and psychology) it is described not just as a condition to be shifted but also as a process that inhibits some people from participating in society, asserting their rights and accessing opportunities and resources (e.g. Hall, Stevens, and Meleis 1994; Dei 2000; Rudmin 2003). Participatory approaches that aim to reach the marginalised, therefore, must be thought of as processes, to engage with the root causes of marginalisation in order to shift underlying power dynamics, and not simply a set of tools for including people in programming (e.g. Kesby 2005).

Our starting assumption, therefore, is that processes that determine participation are embedded within and influenced by social exclusion and disempowerment, and that these need to be understood in order for empowerment to be supported, thereby contributing to improved agricultural and social outcomes. We posit that researchers and extension officers can use their own participatory processes to reflect upon and better understand how their actions and decisions influence the inclusion or exclusion of people, and that by doing this in real time, they may improve the reach and depth of their programmes.

In this paper, we explore this proposition, by sharing our learning about how researchers, practitioners and extension officers can identify, engage with and challenge processes that limit opportunities for people to participate in and benefit from development. We examine in depth the case of an agricultural development initiative, using participatory methods that pay explicit attention to social and gender norms, as part of an integrated and systemic approach to agricultural research. Specifically, we focus on a sustainable farming for nutrition and income research initiative with three communities in Solomon Islands.

2. Context

2.1. The programme and its approach

Between 2012 and 2015, as part of the CGIAR Research Program on Aquatic Agricultural Systems (CRP AAS), WorldFish (the lead CGIAR¹ center) worked with three communities in the northern region of Malaita Province, Solomon Islands. The intervention intended to build the capacity of smallholder farmers to innovate and adapt in order to improve productivity and livelihoods. The Sustainable Farming for Nutrition and Income initiative (SFNI) responded to community-defined nutrition and income needs of smallholders. Together with partners, we engaged directly with community members, and a coalition of stakeholders in the province, including government extension officers. The shared goal was to improve fish, vegetable and root crop farming technologies while building capacity for joint inquiry and collaborative learning through Participatory Action Research (PAR).

As noted by Barnaud et al. (2010) participatory approaches in agricultural interventions can focus on creating communicative spaces for ‘dialogue’ between multiple stakeholders as a vehicle for social learning (e.g. Pretty 1995) and/or, take a critical perspective and focus on understanding power relations (e.g. Edmunds and Wollenberg 2001). Our approach aimed to do both. The starting point was to critically engage with the processes of marginalisation and the multiple forms of power that underpin them, and to do so with multiple stakeholders.

In VeneKlasen and Miller’s (2002) framework for conceptualising and expressing power, ‘power over’ is usually associated with repression or coercion, e.g. when a political leader uses their power to dominate others in formal decision-making. Three other forms of power provide more positive expressions, ‘power with’ is associated with collective action, the ability to do more with others than on your own, while ‘power to’ recognises that everyone has the potential to shape their own world, even those who are marginalised. Together, ‘power with’ and ‘power to’, are often referred to as agency. The fourth form of power in the framework is the hardest to understand and the most hidden – ‘power within’ refers to one’s own self-worth.

Across disciplines, scholars have engaged with this hidden form of power, to understand how to open up opportunities for transformative social change for the most disempowered. For example, Bourdieu’s concept of misrecognition argues that the experience of being deprived can become habitus, leading people to adjust to their unconscious situation, creating a barrier to taking action (Gaventa 2003; Navarro 2006). Similarly, the concept of adaptive preference refers to the process through which deprived people, due to the need to survive, adapt to their state of deprivation and adjust their aspirations making them unable to demand change (Nussbaum 2000; Sen 2001). Together, these theories illustrate that marginalised people often unconsciously adjust to their situation and become incapable of change (Sen 2001; Appadurai 2008; Fishkin 2014).

Informed by these theories of empowerment, the design challenge was to articulate a process that would stimulate the collective ‘power with’ and build collective agency in the communities, while also creating opportunities to build ‘power within’ of marginalised people specifically to enable engagement with agricultural development. Two key design features characterised the PAR approach.

First, following asset-based community development models (e.g. Mathie and Cunningham 2003), engagement was initiated with a community visioning exercise that anchored the SFNI in locally defined aspirations, and established a collective starting point for co-inquiry and joint action. An appreciative mindset was central to initiating discussions with a broad range of community members and to tap collective ‘power within’ (based on Constellation’s community life competency approach²). Once community action plans were agreed, community members implemented them. An ongoing cycle of planning, acting and reflecting was established and supported by partners, community co-researchers,³ local trainers, government extension officers and lead farmers who had been trained by a Solomon Islands non-government organisation Kastom Gaden Association (KGA).

Second, a critical and reflexive perspective was necessary to respond to the expected exclusion of marginalised people from the participatory process at the community level. Reflection sessions were used to explore who was marginalised, and question how various forms of power were used by different stakeholders. Cycles of review occurred initially every six months and then annually, through after-action reviews that evaluated: what happened, what did not, why, what changes were needed, who was involved, who was not and why not? New action plans were collectively developed. Beyond this, it was also necessary to engage directly and separately with marginalised people, creating appropriate and safe spaces to hear their voices and, as the process evolved and power dynamics shifted, to include them in the collective process.

PAR as a research methodology, embraces ‘reflexive research practice’, whereby the researcher examines how their agenda, assumptions, beliefs and emotions influence the outcome of the research process (Gluck and Patai 1991). The research facilitation team, therefore, explicitly engaged with their own positionality and their power in the process through holding their own reflection sessions.

2.2. Study location

Solomon Islands is a small island state of nearly 1000 islands inhabited by multiple ethnic groups speaking more than 60 distinct languages. About 80% of the country’s estimated 639,147 (SINSO 2017) population are rural smallholder farmers and fishers (SINSO 2015). About 87% of the land is under customary tenure (AusAID 2008). Solomon Islands was ranked one of the lowest countries (156) on the 2015 Human Development Index (HDI) (UNDP 2015). Of nine provinces, Malaita Province has the highest population density and with 20% of the province’s population concentrated in the north, this is the most densely populated rural area (SINSO 2009).

The focal communities were selected because of their high reliance on fishing and farming, the expressed interest in the initiative, the presence of community champions who could help facilitate PAR, and the support of community leaders. Three communities were identified – Fumamato’o,⁴ Suafa/Kwai and Alea – where the majority of people are indigenous Melanesian. Fumamato’o comprises 806 people in four villages, Suafa/Kwai has 1000 people in more than seven villages. Alea has 1200 people in seven villages and hamlets. It had a community member already involved in agricultural development via the Baetolau Farmers’ Association (BFA), a registered community-based organisation (CBO) with more than 300 members (farmers). Alea has been a centre of farming activities and co-ordination of BFA trainings for many years. It has previously collaborated with KGA and Government extension services.

Like most coastal communities in Solomon Islands, the three communities depend on fishing and gardening of root crops, fruits and vegetables for their daily food and income (Molea and Vuki 2008; Boso, Paul, and Hilly 2010; Lawless and Teioli 2015). Common challenges include increasing population, declining coastal marine resources from overharvesting and poor management and declining soil fertility from repeated cropping. Climate change impacts such as sea level rise and changing weather patterns also affect garden yields and community livelihoods (Schwarz et al. 2013). Unsustainable farming practices and high population pressure are exacerbating pressure on food security and livelihoods.

3. Research methods

PAR approaches guided the design and implementation of the SFNI with communities. PAR uses collective inquiry to address real-life challenges (Reason and Bradbury 2008). In this paper, we refer to the WorldFish team as the ‘research facilitators’. Their role was to facilitate the PAR community engagement, document activities, record progress and facilitate reflection sessions. The research facilitation team included three staff from North Malaita who had connections with the communities. In addition, a lead farmer together with AVRDC and BFA supported extension services including training, establishing and monitoring demonstration and test sites, cultivating nursery stocks and implementing seed distributions.

Qualitative data were analysed to answer two research questions: (1) How does the process of social exclusion play out in agricultural research and extension activities? (2) Does PAR help researchers understand and challenge processes of social exclusion, and if so how?

The qualitative data from the PAR process included documentation of programme activities (14) during 2013–2015 with an average of 20 participants per activity, such as reflection sessions, action planning, training workshops, sex-disaggregated attendance records, and facilitator observations. A second PAR data stream came from documentation of the research facilitation teams’ reflection sessions as they engaged in reflexivity for their own first and second person inquiry as is common for action researchers (Reason and Torbert 2001; Coghlan and Brannick 2005). We analysed the data through manual open coding (descriptive) to identify what helps people participate and what leads to exclusion (Boychuk Duchscher and Morgan 2004).

We deepened the emerging findings through axial coding and use of ethnographic methods for supplementary data collection including:

1. Key informant interviews (KII) with six lead farmers (two male and two female) from each of the three communities; four participating farmers (male & female); a government extension officer, and, a representative from AVRDC.

2. Two focus group discussions (FGDs), one with 17 men and one with 16 women farmers.

We triangulated results on power dynamics and marginalisation with findings from a qualitative social and gender analysis study conducted in Suafa/Kwai and Fumamato’o and published elsewhere (Lawless and Teioli 2015).

4. Results

4.1. How does the process of social exclusion play out in agricultural research and extension activities?

Our analysis suggests two main factors influence how social exclusion plays out: (i) underpinning social and gender norms, and (ii) the role of lead farmers⁵ in the engagement strategies used by external agents.

4.1.1. Social and gender norms

Community members described different ways that social norms and customs resulted in some people being excluded from agricultural research and extension activities. Women’s roles are often defined by domestic and family responsibilities and men’s roles defined by income-generating activities, as shown by Lawless and Teioli (2015). Participants further explained that it is culturally more acceptable for men to join activities, meetings and trainings than women and youth. Men said women were excluded because of family responsibilities. In Kwai, despite a collectively developed vision, some activities were thought to not be suitable for women. This was evident in meetings and activities where women often stayed in the background observing activities and deferring questions to men.

Women’s access to information was also affected by attitudes about men’s and women’s roles. Participants said it was customary for men to be the primary source of information and to communicate with people outside the community – both important for engagement in agricultural development activities. If women had questions about crop problems and their husbands were not around they tried sourcing information from other women or waited until their husbands returned but would not seek support from extension agents. Similarly, men led inter-community training sessions and were the primary point of contact with AVRDC for seed distribution activities. ‘Men should attend workshops as they are more “savvi” (knowledgeable) and they can pass the information to their wives’ (Male participant, Informal discussion in Suafa/Kwai, 2015).

We also found that young people had limited access to information and knowledge. Again, adult men were the primary contact points especially for information from outside the community or from people in authority such as extension officers. These findings are confirmed by Lawless and Teioli (2015) who found that women and young people are prevented from sourcing information outside the community.

4.1.2. Role of lead farmers in engagement with external agents

Community members referred to both formal and informal leaders. In the sustainable farming activities, they mostly talked about lead farmers. Perceptions about lead farmers greatly influenced processes of exclusion. For example, several lead farmers, who had been trained in sustainable farming by KGA and were involved in extension activities such as training others, had a strong influence on who attended SFNI activities. Among the lead farmers we also found different levels of power and influence. ‘People know we have skills and knowledge as lead farmers but only believe what they are told if it said by one particular lead farmer’ (Male lead farmer, KII, Suafa/Kwai, 2015).

People shared how the engagement strategies of external agencies could positively and negatively influence exclusion. Some people felt they could not join sustainable farming activities unless they were members of farming associations. Some thought the research ‘belonged’ to these groups and was not for everyone. During a farmer field day, several people from surrounding communities mentioned they only came because they had a personal invitation from an association member. Further, during informal discussions in Suafa/Kwai, farmers explained that membership fees had been imposed by a local farming group which excluded some people from SFNI activities. The same perceptions affected participation in activities run by WorldFish.

WorldFish should hold separate meetings and activities to reach different people otherwise some will miss out. Many people will not feel welcome if they do not get direct invitations from WorldFish. (Male and female farmers, FGD, Alea Farmer Field Day, 2015)

Another dimension of exclusion was location. ‘Some people cannot come to the activities and meetings because they cannot get to the place as they live too far away or they have to work’ (Young man, informal discussion, Alea, August 2015).

We also found some evidence of processes that help build inclusion. For example, in one community, the church had established agricultural groups involving women and youth who worked on Fridays to help families that were usually excluded from farming activities. These groups helped everyone and did not differentiate between people’s socio-economic status or circumstances (Lawless and Teioli 2015).

4.2. Does PAR help researchers understand and challenge processes of social exclusion, and if so then how?

Our findings suggest that research processes can enable real-time understanding of the processes that contribute to marginalisation. We share how PAR helped the research facilitation team pay attention to social exclusion and see ways to challenge this. We highlight three aspects: having a collective vision, facilitating reflection sessions and building capacity for facilitation and communication.

4.2.1. Having a collective vision

Reflecting on progress toward goals in the collective vision was an entry point for discussing who was being left out and created a safe way to talk about unequal power relations. For example, people reflected on what it would take to increase productivity collectively for the whole village, and spoke about needing to involve more farmers who were currently not engaged. This revealed who was not getting nursery stocks, seeds and information about new crops and participants then discussed how this affected their collective vision. In this simple way, linking our questions about exclusion to the collective vision made it relevant for people to talk about exclusion, and it led to problem-solving discussions. Having a shared vision, as we experienced, helped build relationships between people based on similarity rather than difference, and in this way highlighted the ‘power with’ of the collective.

4.2.2. Reflections

Taking time to share what people noticed, learned and experienced was the basis of reflection sessions, and central to our PAR approach. Facilitating discussions also helped the research facilitation team see who was influencing and making decisions and who was being left out. They could intervene and facilitate inclusion and begin to see how power was working. During action planning in Suafa/Kwai, farmers discussed ways to increase production and attract bigger buyers. Facilitators asked if there were enough farmers involved, which led to discussion about membership fees for a local farmers’ association and how this was preventing some from joining. Reflections helped stimulate such thinking about participation and created opportunity for greater inclusion.

One of the most helpful findings from reflections was learning how WorldFish and other partners could develop more inclusive practices. For example, research facilitators initially worked through contact points and community leaders. While this relationship was crucial for starting community engagement it also contributed to creating information and communication bottlenecks for women and youth, because it relied on working with those who already had more power. Such bottlenecks can be hard to see unless researchers anticipate and look for them. Reflections were effective opportunities for finding out about such issues and exploring ways to overcome them. During a reflection session in Alea, for example, farmers came up with ways to involve people who were otherwise being left out such as holding separate meetings for young people, meetings on Sundays when women had more time and expanding the number of contact points in the community. The participatory reflection process was a way to see how to adapt the approach in real time as it was unfolding.

Reflections were also important for stimulating ‘out of the box’ thinking. For example, in Kwai, people saw how action planning motivated people to work together, which led to a discussion about how to become better at it and how to scale it to Faufanea, the largest community in Suafa Bay, thus creating the opportunity to expand the ‘power with’ to a broader geographic area.

4.2.3. Capacity of research facilitators and co-researchers

Using PAR as an approach for research in aquatic agricultural systems was a shift for most WorldFish researchers trained in the natural sciences. Although participatory approaches were well embedded in the Solomon Islands programme (Boso, Paul, and Hilly 2010) additional skills and capacity were required. This involved moving from the action research approach which focused on a particular research question (Andrews et al. 2007) to using PAR on collectively agreed community development goals. Importantly, the research facilitation team was no longer just researching fish or agriculture in isolation of social processes. It was a multi-disciplinary research team learning how to practice PAR and through it, to embed research in community development processes with partners and community members. Two capabilities the team was building were important: locally tuned facilitation skills and creating communication spaces for open dialogue.

4.2.4. Locally tuned facilitation

Critical for embracing PAR was the research facilitation team understanding their role as facilitators of joint learning rather than leaders of a research process. To facilitate quality and critical conversations about power it is important to use appropriate language and find the right questions to ask, at the right moment (Hagmann et al. 2002). Having three research facilitators from the region who were conversant in local language helped simplify concepts of social exclusion and marginalisation into a subject that could be discussed with community members.

Timing is one of the more challenging aspects of facilitation in reflexive PAR practice and involves being able to think on your feet. The locally tuned facilitators were open and responsive to following the conversation rather than leading it or reframing answers in their own words – they were learning about managing their own power in the process. Being versed in the local language meant facilitators could pick up the vernacular used to help engage others in the conversation. ‘To me marginalisation means being left out. Doesn’t matter why. It should not happen. Instead we should focus on involving people more’ (Male lead farmer, KII, Alea, May 2015). The value of local knowledge in supporting community observations and deeper insights is something others have noted too (e.g. Harrington, Taolo, and MacLaren 2014), as well as the importance of tuning into local meanings and resisting temptation to paraphrase and use words that researchers think might be more eloquent or accurate (Hooks 1994).

4.2.5. Creating safe communication spaces

Much has been written about communication spaces that help people feel comfortable about sharing (for example, Cornwall 2004; Classen et al. 2008; Wicks and Reason 2009). Without these spaces, it is almost impossible to get to deeper understanding of how hidden forms of power are working. Testing and comparing different communication spaces such as informal discussions and reflections were ways to see who came, who was controlling the conversation and who was left out. Informal discussions elicited greater involvement of weaker voices. For example, people often sat around in the evenings telling stories and were more comfortable to speak up and share opinions than in structured meetings. Women spoke more candidly when they were either not with men or when they were in smaller groups. Sensitive topics were easier to discuss individually such as how leaders influence social exclusion and what they could do to better involve people. Such discussions motivated actions – they led to finding more ‘power to’. For example, as a result of such reflections, one farmer began working with women he thought were isolated and left out, teaching them farming techniques – he was expressing ‘power with’. In his own words:

I got a vision, thought about people who hadn’t had any connection and communication with the sustainable farming work before, they were isolated and not involved in anything in the community so I began to reach them by sharing information at first and then learning about their situation. (Male lead farmer, KII, Alea, May 2015)

Informal discussions were a discrete way of hearing women’s views. ‘This is the first time we had tried sustainable farming methods and established small test plots in our own gardens. We try new vegetables (new seed varieties) of cabbage and we test organic mulch’ (Female farmer, informal discussion, Alea Farmer Field Day, 2015).

They wanted to learn more and preferred to work in a small group together, away from others, until they had more knowledge and confidence. The research process led the team to appreciate what a more fruitful empowerment strategy looked like in situ, real time.

5. Discussion

Our starting proposition is based on evidence that PAR has the potential to facilitate direct engagement with power and reach those who are marginalised, and support transformative social change (Hickey and Mohan 2004; Apgar, Allen et al. 2015). We argued this was possible, as Barnaud et al. (2010) suggest, through both supporting social learning for collective change, and, taking a critical perspective to engage with underlying power dynamics. In testing this approach through implementation, we found that the more the research facilitation team understood their dual role of open dialogue and critical reflection, the deeper their insights and actions to address social exclusion became. What we evidenced was building the ability of a team to facilitate a process of empowerment (Kabeer 2000) embedded within their implementation of agricultural programming. This finding, we suggest, is relevant for agricultural research, development and extension programmes that aim to reach the marginalised.

Importantly, starting with a collective vision was an explicit design choice to put action before inquiry. This is often a tension experienced by PAR practitioners (Arieli, Friedman, and Agbaria 2009) and can be perceived as a barrier to engaging with power. We argue, it is a tension worth navigating, because, as we found, the strengths-based approach of starting with collective visioning made visible, tangible, discussible and ultimately actionable the process of social exclusion with communities themselves. Together, we built a better understanding of how to build collective action and motivate the community’s ‘power with’ in a way that could also reach the marginalised. The more community members, lead farmers, researchers and extension officers saw the drawbacks of exclusion in relation to community farming goals, the more they appreciated the need to find strategies for inclusion. The example of farmers realizing that increasing production would involve making a conscious effort to include more people, is a clear case of needing to build relationships of solidarity. This required those who were excluding, or had ‘power over’ the marginalised, to create space for them. This echoes the finding that hierarchical processes of power can be transformed by strengthening the horizontal power and focusing on and drawing attention to the commonalities between people (e.g. Vasas 2005). In other words, social change, as a process of empowerment requires engagement with multiple forms of power (VeneKlasen and Miller 2002).

A central feature of our PAR practice was the use of reflection, aligning with what others have called the ‘serendipity’ in action research, which moves beyond the application of a structured design (e.g. Burns 2007). Systematic reflection exercises using evaluative questions with those working on the SFNI were opportunities to collectively reflect on behaviours that continued to exclude, and to explore how these affected the achievement of goals (akin to double-loop learning proposed by Argyris and Schon 1978). This deeper reflection was able to then surface the power that was held, or that was lacking. In one case this led to a lead farmer changing his practice and beginning to involve and train women and youth who were being left out. There is evidence here that changes in behaviour can, in fact, occur relatively rapidly when the actors themselves are reflecting in ways that are ‘safe’ and meaningful to them. A second level of reflection occurred when the research facilitation team practiced researcher ‘reflexivity’ (Gluck and Patai 1991). As Fazey et al. (2010) have noted, understanding multiple layers of reflection when implementing participatory vulnerability assessments in the Solomon Islands can lead to broader community outcomes. Our findings suggest this is also true for agricultural initiatives.

Those who have championed the use of participatory approaches within agricultural research argue that capacity of local facilitation teams can be a major challenge (e.g. Kristjanson et al. 2009). Moreover, the particular worldviews, attitudes and values of researchers can influence the success of participatory methods (Chambers 1997, 2017). Some argue that researchers need to bring a critical stance to move beyond superficial and sometimes inadequate group processes to understand power differentials and their effects (Maguire 1987). As Neef and Neubert (2011) indicate, this is an area that requires further exploration. We have shown that opening up ‘safe’ or ‘communicative’ spaces (Cornwall 2004; Wicks and Reason 2009) was possible with teams that had local knowledge and the right skills. Further, the use of unstructured conversations that emerged opportunistically in more comfortable settings for community members complemented structured facilitations (see also Apgar, Cohen et al. 2017; Apgar, Bastakoti et al. 2017). Building capacity to implement a contextualised approach required researchers and extension agents to build personal relationships with community members. This can be a challenge for researchers in general (Burns 2007) and particularly agricultural scientists who tend to be more comfortable as external observers.

Kaplan (1999) claims that shifting the paradigm of research and development from delivery of resources to facilitating resourcefulness requires understanding and ‘reading’ the development process as it unfolds. The skills required include some of those we saw the research facilitation team develop over time, being open and non-judgmental, quickly making sense of, and drawing insight out of, discussions and using this to guide conversations.

6. Concluding remarks

Reaching the poor and marginalised is an imperative for many agricultural research and development programmes. The Food and Agriculture Organization of the United Nations, for example, advocate policy and programmes to support poor and marginalised farmers to improve productivity and increase employment opportunities. They have explicit policies targeting gender, acknowledging more needs to be done to include women in agricultural development (FAO 2013). Locally, the Solomon Islands Agriculture and Livestock Sector Policy (2015–2019) is explicit about engaging in ‘active participatory approaches’ to effect delivery of extension services and recognises the invaluable role that women and youth, as well as men, play in agriculture. The policy aims to support the implementation of the National Development Strategy through a focus on promoting and empowering women, youth and people with special needs to participate in all agricultural developments and activities through various channels, including extension and research. Achieving these policy goals requires an in-context understanding of underlying social processes and the power dynamics that influence who is marginalised and why in order to build an appropriate empowerment strategy.

We have shown in this paper that using reflexivity and adaptive practice focused on achieving a community-determined goal, provides an opportunity to engage in-depth, to appreciate, and to begin to challenge patterns of power. This is not easy to do and indeed it was not always the case that there were positive outcomes or certainty about our impact. We are not arguing that this is the only way to research marginalisation, rather, what we have shown is evidence that PAR is a useful approach to help move from an analysis of power relations from an external perspective to challenging and shifting behaviour real-time within programming.

It has been recognised for some time now that such learning and reflexive approaches to agricultural research require a ‘new professionalism’ (Pretty and Chambers 1993) and researchers and extension officers need to engage stakeholders meaningfully within context. Importantly also, there is a need for their institutional environments to shift and nurture more flexible and poor responsive, and ‘power-aware’ practices (Thiele, van de Fliert, and Campilan 2001). Allowing reflexivity to expose internal power dynamics and shift them is not an easy process to support, and indeed the lead organisation in this case at times found it difficult (see Douthwaite et al. 2017; Apgar, Bastakoti et al. 2017). Our experience in the Solomon Islands, however, gives us confidence that researchers can influence social processes and improve the way agricultural research and extension engage with them.

Implications of these findings for agricultural policy in general, such as the FAO and its goal of enabling marginalised farmers and women to better access opportunities in agricultural development, are twofold: (i) more attention in programme design needs to go in to building the capacity of people who are on the ground and working with farmers directly, to ensure they have the skills to be reflexive, see and expose power dynamics and open up spaces to deepen programme implementation towards the marginalised; and (ii) using participation as not only an implementation approach, but also as a research methodology, as we have shown to be possible, can help build across-context understanding of how to turn poor responsive policy in to power-aware practice on the ground.

Ethical approval

This study was conducted in accordance with the ethical standards of WorldFish and Community Research Agreements established with the three communities in North Malaita.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Michelle Rice was a knowledge sharing and learning researcher working with WorldFish at the time of this research. She is currently with UNRWA in Gaza working with Palestine refugees. She has extensive experience working on international humanitarian and development programmes and with marginalised people. Her research experience and interests include participatory action research and collaborative learning.

J. Marina Apgar was knowledge sharing and learning scientist at WorldFish and responsible for the design and implementation of PAR in the AAS program. She is currently Research Fellow in Participation at the Institute of Development Studies and her research interests are in learning from use of participatory approaches as vehicles for engaging with complex social-ecological systems and supporting transformative change. She has extensive experience working in multiple contexts with marginalised communities.

Anne-Maree Schwarz is a scientist working in small-scale fisheries and was Program leader at WorldFish Solomon Islands at the time of this research. She is currently Team Leader Mekem Strong Solomon Islands Fisheries Programme and is a Senior Fellow at the Australian National Centre for Ocean Resources and Security, University of Wollongong. She has extensive experience working in research teams with rural communities in Solomon Islands.

Enly Saeni was a social science researcher at WorldFish, Solomon Islands during this research. He is currently completing a Masters in Sociology at the University of Hawaii. He has experience of research and development with organisations in Solomon Islands.

Helen Teioli is a gender researcher at WorldFish, Solomon Islands. She has been part of an international team developing the WorldFish gender research program over recent years.

Notes

1. The CGIAR is an international consortium of 16 independent agricultural research organisations focusing on poverty reduction.

2. See http://www.communitylifecompetence.org/en/ for more information.

3. We saw community members – farmers, fishers and community groups – as our partners in the research and called them co-researchers. The intention was to build their research capacity and our relationship as a partner.

4. Fumamato’o can be spelled in several ways, including Fumato’o which is considered an abbreviated version.

5. Lead farmers is a term used within Solomon Islands to refer to farmers trained by local NGO Kastom Garden Association to teach others.