From agency to root causes: addressing structural Barriers to transformative justice in transitional and post-conflict settings

Transformative justice has emerged as a new practice agenda for addressing structural and systemic violence in post-con ﬂ ict and post-authoritarian societies. This article is situated at a critical juncture: while the emerging scholarship has focussed on community agency and action, there is little as yet that has explored the social structures and relations in transition societies that are harm-generating and which constrain action. We argue that a critical social science, grounded in realist social theory, systems thinking and complexity theory, have a vital role to play in rendering transparent the relations and structures that resist change. New knowledge about the ‘ root causes ’ of harm is both conceptually innovative and useful to practice, helping practitioners identify societal arrangements in need of change and informing strategies for action. This article illustrates the approach through its application to a study with poor farmers in post-Revolution Tunisia. The article should be of interest to researchers and practitioners in transitional and transformative justice, con ﬂ ict and post-con ﬂ ict, peacebuilding, and security sector reform, who are engaged with understanding and addressing issues of structural and systemic violence.


Introduction
Transformative justice has emerged as a new practice agenda within the field of transitional justice for addressing structural and systemic violence in societies transitioning away from conflict or repression (Gready andRobins, 2014, 2019). This article builds on a growing number of critical commentaries on transitional justice, which critique its practice as short term, focusing on symptoms not causes, civil-political rights rather than their socio-economic counterparts, and as a result producing disappointing outcomes. In response, agendas for transformative change seek to radically reform the 'politics, locus, and priorities' of existing practice in favour of addressing unequal and intersecting power relations and structures that exclude and marginalise (Gready and Robins, 2014: 2; also Gready, 2019). This article is located at a critical juncture. While the emerging scholarship on transformative justice has focused on supporting and strengthening community agency and action, there is as yet little which has empirically examined concrete examples from practice or the social relations and structures in transition societies that are harmgenerating and which constrain action. As such, it is emblematic of critical engagement with other mainstream approaches, for example, peacekeeping and conflict resolution, which tend to privilege agency over consideration of underlying structures or root causes.
The contribution of this article is to demonstrate how a critical social science can support practice. It can do this by rendering transparent the structural and relational dimensions of harm, developing new knowledge about the 'root causes' of violence and unmet need, and identifying societal arrangements in need of change and strategies for change. By critical social science, we mean harnessing critical theories through a metatheoretical framework supplied by the philosophical position of critical realism (Bhaskar, 2016). A relatively new development in the philosophy of science, critical realism provides a satisfying philosophical base for understanding and dealing with causation in social systems. Our approach works towards producing social scientific explanations of harm that are centred on processes, social relations, structures, context and contingency (Hoddy in press;Gready and Robins, 2020). This perspective is advanced against a background where critical theorising is marginal in the policy-oriented scholarship of transitional justice. The approach applied in the case study highlights issues of class and power, which rarely feature in transitional justice scholarship or practice.
To illustrate the approach in action, we draw on a study into the needs and priorities of poor farmers in post-Revolution Tunisia. Tunisia is a pertinent case for transformative justice. The roots of the 2010/11 uprising lay in long-standing tensions and grievances in the country's agricultural regions where the protests began (Gana, 2012), and its premier sloganemployment, freedom, dignityillustrated the indivisibility of political and socio-economic concerns. However, the country has in many respects witnessed a regrouping of powerful economic and political actors and relationships, and a renewal of social mobilisation around the failure of new economic opportunities to emerge (FTDES, 2018). This article begins by indicating how a critical social science might support transformative practice. Findings from the Tunisia case study are presented and discussed, with some illustrations of how knowledge generated might assist practitioners responding to harm and constraints on action. We step back to assess some of the wider implications of the approach for transitional/transformative justice.

Social research and transformative justice
A law-centred field and approach to practice, the core instruments and interventions of transitional justice have struggled to adapt to their contexts in a meaningful way. Despite concerns about legitimacy and the need for local 'ownership' and adaptation, there remains a sense in which transitional justice interventions continue to be applied on 'blank slates', where particular interventions are expected to produce specific outcomes. For example, the field has continued to neglect the rural social, economic and cultural dimensions of transition and agrarian violence, which in some settings may be critical to securing long-term peace and stability (Hoddy, in press). While the shared ideas, values and assumptions of the field have been unsettled by a critical transitional justice literature, this literature has also tended to address transitional justice in its own terms as opposed to providing a more profound rethinking of problems and possible solutions.
With an overall concern to address structural and systemic violence, our argument is that transformative justice demands a critical social science that can support its practice. Up to now, the focus of the emerging transformative justice literature has been on the need to support and strengthen community agency and action for securing justice in transition (e.g. Evans, 2019;Gready and Robins, 2017, 2014, with participation and empowerment identified as among the core values underpinning practice. Harms are acknowledged as complex and as potential drivers of conflict and further violence, but there is as yet little that has sought to empirically unpack them, their social origins and their relationship to the state and state processes. What is missing from this literature is the place of critical theorising in transformative practice, that is, where new knowledge is developed about the social systems where change is sought, and how these systems disempower and constrain, and how these features might be challenged. Critical theorising can support practice by helping to expose and explain forms of harm, violence, oppression and constraint that characterise the world of societies and communities, providing a picture of how social systems, structural contexts and cultural assumptions produce particular effects (descriptive theory); and providing some conception of how things might be different from what they are (normative theory). It can shed light on the societal relations and structures that generate harm, how they operate, and their effects concretely manifested in transitional and post-conflict settings. Such knowledge is useful to transformative justice practitioners in several key respects: by providing new conceptual and practical insights that helps them see how things can be different from what they are; by offering understandings of what needs to change; and by informing strategies for action.
Relations, structures, complexity, and 'root causes' What do we mean by 'harm-generating relations and structures and 'root causes'? Our understanding here draws on insights from realist social theory, systems thinking and complexity theory (Archer, 1995;Bhaskar, 2016;Jackson, 2001;Mingers, 2014;Sayer, 2010). Simply put, root causes are the underlying 'generative' or 'causal mechanisms' that make events occur. The mechanisms of interest here are those which link particular societal relations and structures to violent or harmful outcomes and constraints on action (Rylko-Bauer & Farmer, 2016). What makes a structure are various objects and practices that stand in relation to one another, such as the practice of exchange between buyers and sellers that characterise market structures.
Structures steer human activity are continually reproduced and occasionally transformed by the actors implicated in them. The emergent 'powers' of structures to behave in certain ways and their 'liabilities' towards certain kinds of change are what is referred to as mechanisms. Mechanisms lie at the heart of many problems that transformative justice is concerned with, such as indirect violence and poverty, since these phenomena can only occur as a function of societal structures. Both mechanisms and structures are trans-empirical phenomena, that is, they are not readily open to empirical measurement and can only be understood indirectly, through retroductive enquiry that examines their empirical effects. Structures may be social, cultural, political, economic and so on, and are experienced in terms such as poverty, discrimination, and the lack of access to public services.
Crucially, the way mechanisms operate and the effects they have depends on their interaction with various conditions, or other mechanisms, such as material and cultural constraints, political decisions, and how people decide to act (Mingers, 2014). Processes and outcomes are contingent, and change is non-linear and causes difficult to trace. Mechanisms may operate with and against each other to produce similar or different outcomes, and feedback may ensue where outcomes affect change elsewhere in the system. Mechanisms may remain unactivated under particular configurations of conditions. In addition, mechanisms exist at multiple scales, such as from the biophysical level to those found in the structures of the global economy. The association of a given event or phenomenon with mechanisms at different scales reflects the 'laminated' nature of these systems (Bhaskar, Danermark, & Price, 2017).
Human agency is central to the activation of social mechanisms. While social structures are mostly reproduced unconsciously by human actors, such as where the nuclear family is reproduced by people deciding to marry, transformation is possible because structural conditioning is always 'mediated by the actuality or possibility of reflexive deliberation by the agent on the course of action to be followed' (Bhaskar, 2016, p. 55). In social theory, Archer's(1995; esteemed account of the structure-agency dynamic captures the interplay between structures and agency over time as one that is mediated by how people consider themselves in relation to their social contexts and vice versa. People may benefit from social structures like the family, but resistance to, subversion of, or ultimate transformation of structures may follow when people 'recurrently find themselves aggravated, restricted, oppressed, and dehumanised' (Smith, 2011, p. 343).
It is in these complex social systems where root causes can be identified through research, in terms of the causal mechanisms of interest, the structures they are embedded in, the conditions affecting how they operate and the harmful effects they have when activated. What do these notions bring to fieldwork? This mechanism-based perspective here represents a novel approach to causality that can bring to light the underlying or 'real' structures and mechanisms in social systems alongside an account of how they behave. These remain elusive in the literature, as mentioned above. What makes the perspective novel is its understanding of causation transcends conventional ideas about causation that are ill-suited to identifying root causes in complex systems (Sayer, 2010): the positivist empiricist approach, which claims a cause where statistical regularities among sequences of events are found; and that of the interpretive social sciences, which rejects the idea that causes can be identified in the social world and which focuses instead on understanding meanings, symbols, identities, and so on.
As such, applying a mechanism-based perspective can render a more nuanced picture of these systems and how they behave. It can assist transformative practice in several respects. In the first instance, building a picture of the mechanisms that produce harms and constraints on agency can help practitioners and communities identify and reason over which might need to be removed, blocked or modified, or where new ones can be established for offsetting the effects of another. This will involve shedding light on the components making up the systems in question, in particular the various social relations and the powers or capacities of different actors that these relations produce and sustain; and how those capacities are distributed. Mapping these out can help practitioners and communities identify potential courses of action in favour of reworking particular social relations, where that action might be directed and what consequences may follow. Gready and Robins (2020) have argued a similar case for greater awareness of system complexity in transitional justice. Drawing on insights from the development field, where systems thinking and complexity science have been applied to work on change processes, they argue for context-dependent interventions that challenge 'the very idea of a single practice relevant in all contexts and spaces' (Gready and Robins, 2020, p.294). Such an approach can help overcome the way inequalities are unintentionally reproduced through transitional justice. A lack of consideration for power and social relations in transition societies has led for example to the field's privileging of formal NGOs and groups that already have a voice as opposed social movements or grassroots groups. This understanding has particular practical traction for the design and delivery of transformative reparations (de Waardt & Weber, 2019;Manjoo, 2017). It has been argued for instance that reparations, particularly for women, should seek transformative redress rather than restitution so that women are not returned to the conditions of structural discrimination that characterised the past. Scholars have analysed whether a corrective/ restorative or transformative ambition is preferable (Urban Walker, 2015); and the potential of transitional justice mechanisms, such as courts, to deliver transformative reparations (Durbach, Chappell, & Williams, 2017). Weber's( 2017) useful work on reparations with women in Colombia suggests there is a need to restructure patterns of social relations around education, organisational skills and income-generating projects to enhance women's social and economic agency and autonomy. While Weber does not draw on the language of mechanisms, the broader point is that contesting these structures and their effects involves practical projects that seek to reposition groups more favourably within their relational contexts.
Other important components include the cognitive categories and normative and moral beliefs sustaining particular structures and practices. In a number of other fields, such as rights-based development, community reflection permits a questioning of these beliefs and categories, and, if the community wishes, subverting them (Gready and Ensor, 2005). For instance, denaturalising and overturning shared beliefs about gender that limit the agency and participation of women and normalise sexual violence. Boesten and Wilding's( 2015, p. 1) suggestion for more transformative modes of gender justice captures this for example, where interventions should seek to respond to 'understandings of gendered roles in society and the perceived links between reproduction and community. ' Lambourne and Carreon (2016, p. 83) also suggest that securing gender justice requires subverting and modifying existing beliefs in societies where 'patriarchal structures and attitudes' persist.

Methods in search of root causes
The following illustrative study demonstrates how social scientists can use the lens of transformative justice to identify and unpack root causes. The study took needs as a starting point and sought insight into the root causes of unmet need among farmers in a rural community in Tunisia. It also sought to identify farmers' priorities for change. Participants were interviewed about their experiences, needs and priorities in the context of political transition rather than in relation to transitional justice. The purpose of this was to develop an account of the societal roots of unmet need and to allow perspectives to emerge that are not necessarily framed around transitional justice and its mechanisms.
The study's methodological detail has been presented and discussed in depth elsewhere (Hoddy, 2019). In brief, enquiry proceeded from an examination of participants lived experiences towards the underlying relations, structures, and mechanisms in this particular rural context. A retroductive research design was adopted for 'imagining a model of a mechanism that, if it were real, would account for the phenomenon in question' (Bhaskar, 2016, p. 55). Since investigating 'laminated systems' supposes 'no a priori account of what levels or the number of levels that may be involved in any particular explanation' (Bhaskar et al., 2017, p. 5), our emphasis is on the operation of mechanisms at the mesoor community-level, as this is the site where social change projects are initiated. The study acknowledges however that higher-level structures and processes that have a bearing on this rural setting as well as rural settings more generally (Hoddy, 2019, in press).
For synthesising data and making inferences about causal mechanisms, a realist application of grounded theory techniques and procedures was chosen (Hoddy, 2019). 42 semistructured interviews were conducted between September 2015 and January 2016 and analysed with these techniques and procedures. 1 The rural focus of the study meant drawing on critical social theories, namely from rural poverty studies and critical agrarian studies for developing a causal explanation for unmet need.

Case study background
The research took place in a rural community on the Cap Bon peninsula, Tunisia, with small landowning farmers, renters and sharecroppers who pay rent to landlords as a portion of the crop yield or its market value equivalent. These structural positions emerged in the context of the region's 'market gardening revolution' in the 1940s and 1950s that saw peasant farmers shift from extensive farming to intensive irrigated agriculture that was becoming integrated into domestic markets (Sethom, 1977a(Sethom, , 1977b(Sethom, , 1992. In the community, fast-maturing and labour-intensive crops are grown in the main and production is dependent on chemical fertilisers, pesticides and modern seed varieties. Agricultural inputs and equipment are acquired at several private sector suppliers and nurseries in the community, some of which are affiliated with agri-food processors that occasionally sign production contracts with farmers and which absorb and transform large volumes of local produce. Farming supports household consumption needs indirectly, through exchanges of crops and livestock as commodities with other actors for money. The cash acquired through these exchanges is used for buying other commodities, such as food or possibly farming inputs such as seeds and machine services. Production is integrated into commodity chains that provision agri-food processors, wholesale markets, supermarkets, and some foreign buyers.
The rural focus and choice of case for the study has to do with the rural nature of poverty in Tunisia and in this community (Hoddy, 2018) and because rural economic issues were a source of significant grievances leading up to the Tunisian Revolution in 2010/11. The uprising in 2010/11, as studies have indicated (Ayeb, 2012;Elloumi, 2013;Gana, 2011Gana, , 2012, was rooted in long-standing tensions and grievances in the agricultural regions where the protests began. Liberalisation and restructuring of agriculture from the 1970s and 1980s onwards gradually moved the sector away from state-led development towards more capital-friendly regime, spawning new rural tensions and patterns of rural inequality (Sethom, 1977a(Sethom, , 1977b(Sethom, , 1992. The phenomenon of rising farmer debt and difficulties making repayments was a key source of grievances in the lead up to the uprising, particularly among small farmers facing lawsuits from the banks and who were at risk of losing land.
The 'explosion of demands for social justice in rural areas' (Gana, 2012, p. 210) in 2010-2011 saw farmers of different sizes mobilise around various social, economic and political agendas. Protests contested agricultural organisations and the Tunisian Union of Agriculture and Fisheries (UTAP), effectively a power relay of the government; inequities in several agricultural sectors (tomato, milk etc.); and there were demands for access to land; and, in the case of farmworkers, improved salaries and working conditions (Gana, 2011(Gana, , 2012. In Cap Bon, the most significant protests were organised around the tomato sector where sitins and factory blockages led to the closure of several processors. Much of the transition period has seen the continuation of popular protest of varying intensity around local and national issues. The renewal of social mobilisation, especially since 2015 has come as a response to the failure of new economic opportunities to materialise over the period, with actors organising around the same grievances and demanding that the state address persistent poverty and unemployment, and improve access to resources and services (FTDES, 2018).

Debt, uncertainty, instrumentalisation, and future prospects
Research findings indicate that, for participants in the community, the rural economy is an accumulation of present and longstanding issues and grievances having to do with their everyday insecurity and enduring poverty. Participants identified themselves and the community in general as worse off in the transition period, reasons for which they attribute to the Revolution. The structural positions they occupy were unchanged since the Revolution, but they faced worsened conditions for their farming livelihoods, difficulties repaying rising debts, and lower household incomes.
Four specific issues were prioritised in relation to needs and livelihoods that shed light on the relational context in which participants are embedded: debt; ongoing uncertainty in various forms; a sense that they were instrumentalised as a means of profit-making by and for others; and a sense that farmers are 'going nowhere' and have no future prospects. The general phenomenon of low incomes among farmers in the community, and growing limits on what can be done with those incomes in terms of their purchase power and rising prices leads them to describe farming as a livelihood that is 'just for living'. Especially for sharecroppers, they obtain either a seasonal net income which is enough to get by on but little more than that, or they obtain a net loss and debt which threatens to push them under. Low incomes, net losses and indebtedness can or has led them to depress household consumption. In more serious cases, depression of household consumption may also affect food and nutrition of the family. Farmers may turn to landlords and friends to borrow money for essentials or buy food from grocers through credit.
Farmers worry about the material and social implications of debt, in particular the debt they owe to suppliers of agricultural inputs. Knowing whether they can repay their seasonal debts on time or not depends on the season's output and prices. Debts can extend over multiple seasons, with the present season's harvest used for servicing the debts of the previous. If participants are unable to pay these debts or repayment is delayed for too long, creditors can impose higher rates of interest that further erode incomes. There are other forms of sanction that concern farmers too: suppliers may impose more stringent lending conditions in future or may refuse to lend at all. There is also the prospect of being brought before the court for failing to return what they owe.
Participants have to contend with ongoing uncertainty in their livelihoods, which come in a number of forms. The risk of crop diseases and pests, and associated economic costs, are one source of uncertainty occurring in the absence of accessible insurance. Another important source of uncertainty concern end-of-harvest selling and markets, such as where more powerful economic actors have the capacity to renege on production contracts or to buy back the value amount that covers the debt amount but leave everything else behind. Livestock thefts and fears for personal security have also been a particular source of concern since the Revolution, with many having abandoned livestock or taking new measures to protect it, such as bringing it into their homes at night.
Participants perceive that they are being instrumentalised or objectified by other actors for the purpose of profit-making. 'We are working for others' was a commonly recurring phrase among tenants and small farmers alike. Other actors were setting the terms and conditions of their work and these actors were its main beneficiaries. Grievances are especially directed towards agri-input suppliers and agri-food processors in the community, perceived as exploitative. One research participant described these actors as 'colonising' the community while others referred to them as 'exploiters.' This instrumentalisation or objectification also has non-material dimensions. Farmers sense their control over both their work and the way resources and services are exchanged has been transferred to these other actors. Landlords, suppliers and processors, shape production and exchange through various means that involve their ownership of productive resources and their ability to grant access to those resources and to markets. They define for example when farmers must repay; which seed varieties and treatments they must use; when they must be harvested; and in some cases how farmers should dispose of the crop. Dependent on these actors, farmers view them as distant from the community, and the individuals running them as uncaring and unwilling to listen to their concerns.
Finally, there is a shared sense that farmers are 'going nowhere'. Farming livelihoods have come to mean little for participants beyond satisfying basic needs for food, shelter and so on. A general dislike for agriculture and a longing for regular, salaried (predictable) work outside the sector was common. For tenants, the distance between their incomes and price of land has placed land beyond their reach. Even the utility of land as a mechanism for advancement in the wider region looks ambiguous. Small and medium sized farmers, who are better positioned than tenants and farmworkers, are themselves grappling with the same issues of rising production costs, low incomes and debt.
This sense that farming is 'just for living' and 'going nowhere' is accompanied by descriptions of it as a 'struggle' or a 'fight'. Farming livelihoods require contending with overwork and fatigue to attain the best possible yields and incomes. Some farmers would prefer to stop working in agriculture altogether but lack alternative sources of income.

Rural structures and mechanisms
Retroductive analysis involved theoretically redescribing the data emerging about these phenomena through existing frames of reference provided by the critical literature mentioned earlier. As trans-empirical phenomena, the presence and role of relations, structures and mechanisms is not self-evident and has to be inferred by combining observation with theory. Findings suggest participants are inserted into particular patterns of social relations, or structures, which play a causal role in the four issues described above. Three main structures were distilled out through this process which are of causal interest and which comprise sets of relations and causal powers and liabilities. These structures connect and interpenetrate, with each serving as the basis for a particular mode of surplus extraction.
The first structure is the landlord-tenant relation which underpins sharecropping as a system of agricultural production and is the basis rent payments to landlords as a mode of surplus extraction. The second structure comprises internal relations around farmers and private sector 'suppliers' in the fieldwork location which control access to other productive resources, such as seeds and fertilisers, and is the basis of a different mode of exploitation based on debt and interest. The third set of relations refer to those between farmers and industry, where production has been incorporated into commodity chains that link farmers to consumers through agri-food processors and private storage bodies. These relations, which are the basis of a third mode of exploitation based on resource extraction for profit-making, see the delivery of farmers' produce to industrial and commercial actors who store or transform it into new products that acquire higher value over time or as they are moved along the commodity chain. These structures guide production and exchange activities in the fieldwork location and form the context or conditions in which livelihoods are pursued. What sustains and links these structures together are material resource dependencies and flows, and shared cognitive categories such as knowledge of who does what in production and exchange and an understanding of the 'rules of the game'. These structures are constitutive of particular structural positions and they produce a variety of potentials, or powers and liabilities, that are unevenly distributed across these positions.
Retroductive inference about causality indicates the presence of two main causal mechanisms that are activated to deprive people of access to material and non-material objects for satisfying needs: wealth extraction and social subjugation or subordination. The first refers to a mechanism through which more powerful economic actors have been routinely able to exploit poor farmers. This mechanism emerges through the intersection of the three main structures of interest. For example, through these structures agri-industrial actors acquire indirect access to land and labour but are able to insulate themselves from various (structural) risks and costs associated with production, which farming households are required to shoulder instead. These structures were found to set limits to participants' autonomy over their work and lives and to diminish the prospects for alternatives. Social subjugation or subordination was a second identified general mechanism that accompanied wealth extraction, and which captured its non-material dimensions. It is experienced by participants in various subtle ways, such as by having to forfeit decision-making over planting; by having to cultivate clientelist relationships with other economic actors; and of having little or no means of mounting an effective challenge or seeking redress.
These structures and the mechanisms they sustain do not stand alone but are nested within structures and processes at higher scales. The movement of crops, land, and other productive resources between actors in the region occurs within a broader context of generalised commodity production as an outcome of a historical process of agrarian change (Hoddy, 2018). Farmers are integrated into capitalist social relations where securing the conditions for life and for future farming takes place to a large extent inside of 'commodity relations and the disciplines they impose' (Bernstein, 2010, p.102). These features also distinguish sharecropping in the region from its more 'feudal' or 'pre-capitalist' variants, though farming there shares with other instances of sharecropping elements of rural patronage and exploitation through transactions tied across the input and output sides of production.
The landlord-tenant relation which governs the exchange of land, labour and other resources predates the turn toward producing for the market, but it continues to endure alongside other structures of agrarian capitalism that arose in the region the second half of the twentieth century (Sethom, 1977a;1977b). This began with early stirrings towards producing for the market in the late 1940s which deepened in the context of national development in the post-Independence period. This intensified in the period of liberalisation after the 1970s and restructuring from the 1980s onwards. The abandonment of 'traditional' farming practices accompanied the emergence of agro-industry in the region geared toward predictable and fast-maturing cultures, an expansion of trade in inputs such as high yielding seeds, chemical fertilisers, pesticides and treatments, and new equipment, varieties and products. Much like has been observed elsewhere in the region (e.g. Bush, 2007), the subsequent process of liberalisation and restructuring has seen the state oversee its own withdrawal from direct involvement the sector in favour of fostering a more business friendly environment, though it has retained an indirect role in elaborating frameworks and mechanisms to support and regulate tenure, enterprises, and to coordinate sectoral actors in line with a global process of agricultural neoliberalisation.
While these structures are need-frustrating and constraining, they are, in other respects, enabling however. Participants' activating the powers of these structures and performing accordingly is a means of pursuing money incomes for satisfying other basic needs, such as for food and clothingeven if that level is considered by them as unsatisfactory. Needs are pitted against one another and other mechanisms are lacking that might block or cancel out the effects of exploitation and subjugation, such as insurance, forms of social protection and rural representation. In more abstract terms, research participants are active in reproducing these structures and (re)acquiring their structural positions as labourers over time. A combination of powers and liabilities is enabled and activated through these structures, while others are produced or exist but remain in potentia.
For participants, these trade-offs may be expressed only at the level of practical consciousness but they are consistent with findings from other studies which show how the rural poor enter into relationships of dependency, experiencing 'truncated ambitions of self-improvement and advancement' (Wood, 2003: 456) in order to secure the conditions for their livelihoods. Grievances and priorities for change suggest some degree of reflexively over social circumstances, yet the organisation of farming practices around these structures and the absence of organised opposition suggests passive acquiescence. Cooperation might yield certain benefits but alternatives are few, and failing to perform or even resisting can undermine 'trust' and invites costs and punishments. Participants indicated these costs and punishments could come in the form of compromised dependencies, such where lenders refuse to continue lending or where more stringent conditions are applied. So, while these structures exert causal influence on the people implicated in them, the decisions and actions of those inserted into them contribute to their reproduction.
In terms of farmers' priorities for change, participants emphasised improved incomes and consumption, conditions for developing their agriculture and/or employment opportunities; having fairer and more equitable relationships between actors in the community; and improving personal safety and security. Various practical suggestions were put forward that implied three types of intervention: the state playing a more direct role in agricultural development (such as by providing and making seasonal loans accessible); setting limits on what more powerful economic actors could do (such as through regulation and state monitoring); and developing new forms of independent representation or channels for communication (such as creating a genuinely independent farmers' union). Tenants indicated there was little the state would or should change on land ownership matters, though the idea of owning land was viewed positively and land reform was raised several times. Like other farmers however, their priorities lay on addressing the fault lines between farmers and industrial and commercial capital rather than on land ownership.
Stepping back, the knowledge generated lends itself to social and political action for addressing and strengthening the rights of farmers. For community practice mapping out the social relations comprising the system, its mechanisms and the capacities or powers these sustain, provides an indication of where practitioners and communities might consider directing action that interferes with these mechanisms and relations and which might seek to generate new ones. The indicative priorities for change help here as well, although these are indicative only and would need to be examined further through action-oriented research. In particular, new associational powers appear useful in the form of an independent organisation to represent farmers or, at the very minimum, a forum in which farmers can express their views. By building on enhanced freedom of expression since the Revolution, this action has the potential to rework existing vertical relations. For instance, relations between farmers and agri-industrial actors might be reorganised in favour of more horizontal ones that provide farmers with greater autonomy. Doing so would involve retaining existing structures intact but circumventing or creating blockages on more powerful actors' existing capacities as well as generating new capacities for farmers or activating existing ones.
More ambitious action would involve systemic change that replaces these structures entirely over time. This may include equitable and sustainable farming systems which overturn existing dependencies, such as agroecology and diversified farming systems that are oriented towards sustainable technologies and practices and food sovereignty, biodiversity and more direct relations with consumers (Kremen, Iles, & Bacon, 2012;McAllister & Wright, 2019;Rosset & Altieri, 2017). We mention them here for several reasons: because knowledge about them was low; because participants' priorities tended to involve working within the same structures and with the same material resources with which they were already familiar; and because there are environmental pressures bearing on the region, which in the long-term looks set render the existing model of industrial agriculture unworkable (e.g. Bargaoui, Tramblay, Lawin, & Servat, 2014).
Developing such responses requires community-based processes that can ensure strategizing and programming are appropriate to the practical contexts for action. However, there also emerges a need to link transformative practice to empowerment, as a response to some obvious obstacles in the path change and for reducing the risk of accommodation and incrementalism. 2 For instance, shifting to equitable and sustainable systems requires awareness-or even consciousness-raising and radically reorganising the way action is presently directed. It involves incorporating different material objects (such as new crops, technologies and equipment), as well as new knowledge, skills and bodily dispositions among farmers in the community. This is no small feat and would likely require concerted advocacy work in favour of supportive public policies and programming and the forming of independent farmers' organisations. Change is also likely to be resisted or constrained, such as through formal institutional and governance arrangements and associated actors (for example, those concerned with the management of natural resources and of food and commodity chains). Empowerment work offers the potential to navigate and address these concerns, for instance by fostering processes that emphasise collective power and organising, 'critical consciousness', and alliance-building among civil society actors (Cornwall, 2016;Gready and Robins, 2020).

Concluding discussion
In terms of research findings, the case shares with other studies and surveys in transition societies (Pham, Vinck, Balthazard, Hean, & Stover, 2009;Vinck, Pham, & Kreutzer, 2011, e.g. Robins, 2011a, 2011bVinck & Pham, 2014), as well as elsewhere in Tunisia (Andrieu, Ferchichi, Robins, Aloui, & Ben Hamza, 2015), the finding that individuals, communities and groups raise poverty and insecurity as issues they would like to see addressed. The study has done something different however by connecting these issues to the structural context of the community. It shows that the root cause lies in the way farmers are inserted into enduring rural social relations that deprive them of access to material and nonmaterial objects for satisfying needs. The structures and mechanisms identified are nested within structures and processes at higher scales, namely the system of generalised commodity production that arose in the region post-Independence and a policy context beginning in the 1970s that is characterised by liberalisation and agricultural restructuring. The findings also point to the role of farmer agency in the reproduction of these structures as a matter of trading off some material and non-material resources for others.
One key contribution that a critical social science can make is that it provides a means for engaging with wider systems and structures not just as 'inert' background and context for interventions, but as clear root causes of human rights violations. Doing so can yield various entry points for challenging root causes through community-based practice in manageable ways, helping resolve in concrete settings and through practice the tensions between structure and agency and between the multiple scales (from meso to macro) that facilitate violations. The above example of more equitable farming systems that reduce dependencies is a case in point. Processes of critical consciousness-raising can promote reflexivity over existing practices and examinations of individual and collective powers and capacities (for example, resources and networks) for navigating these tensions and challenging violations. While the focus of this study has been on a rural community, the mechanism-based approach can be applied with other populations, settings and issues. Knowledge and theories from other disciplines can be harnessed through a for shedding light on manifestations of structural and systemic violence and their intersection through relations of class, gender, race and ethnicity, and for identifying the various mechanisms at play.
Lars Waldorf (2019, p. 161), in a recent critique of transformative justice, suggests that 'when populist authoritarianism threatens the modest but real achievements of the liberallegalist human rights framework … what's needed is less hubristic overreaching and more principled pragmatism.' We agree that ambitions and expectations need to be carefully negotiated and managed. Yet the growth in support for 'strongman' politics in Tunisia in recent years, as well as the rural roots of authoritarian populism in other agrarian settings (e.g. Scoones et al., 2018), is further reason to view the present conjuncture as one in urgent need of practical projects, supported by research, that can respond to structural and systemic issues. In terms of broader implications for transitional and transformative justice, we suggest that applying a critical social science may yield new knowledge that assists truth commissions doing a better job of addressing basic needs and everyday insecurities by providing a socio-economic 'diagnostic lens' on the past, and links between the past and present. Documenting forms of rural resistance can also be used to identify ongoing channels of resistance, and damaging forms of continuity accompanying change. Participation and empowerment can be saved from technical and superficial formats to instead require a radical relocation of agenda-setting and decision-making to victim and survivor groups, and local communities in concrete settings. Notes 1. Most research participants (35) were male, which reflected difficulties recruiting female participants. Beliefs about the status of male knowledge meant that potential female research participants tended to defer to their husbands. A further round of fieldwork in the region is being undertaken in 2020 that will examine the research findings through a gender-sensitive lens and all-male/all-female focus groups. 2. In rights-based development (Gready and Ensor, 2005;Ensor et al., 2015), empowerment has tended to go hand in hand with participation as a means of surmounting such obstacles.