From Protocol to Practice: Collective Reflections on Navigating Ethical Junctures in (Post-)Conflict Research

Abstract

Research with victims-survivors of severe human rights violations raises complex moral and ethical dilemmas that often extend beyond the principles of do no harm and the risk avoidance and mitigation strategies prescribed by formal ethics reviews and guidelines. Yet, while the latter have been widely critiqued, extant scholarship provides little guidance for developing comprehensive ethics architectures that balance institutional demands with the need for adaptability to complex, culturally diverse research contexts. To address this lacuna, this article foregrounds experience-based reflections and insights on the research ethics of a large, multi-sited transitional justice project. Using an adapted Delphi approach to capture insights from successive team discussions, we examine the benefits, shortcomings and effects of the formal ethics protocols and guidelines established at the project’s outset, alongside the dynamic, context-specific practices that emerged during fieldwork and the self-reflective discussions that accompanied them. Focusing on four critical ethical junctures – (1) communicating ethics and the meaning of consent, (2) encountering and acknowledging violence and trauma, (3) sharing research burdens and benefits, and (4) balancing agency and vulnerability – the article explores concrete challenges, tensions and novel approaches that informed the emergence of a layered ethics architecture. The discussion and conclusion propose further strategies for fostering a closer convergence among these various layers through the adoption of an iterative, reflective and relational ethics of care, that can serve as a basis for responsive and meaningful engagement with participants in vulnerable contexts.

Introduction

Studying how victims-survivors participate and mobilize in the wake of widespread violence raises profound ethical questions and challenges. As a team of transitional justice scholars working in a large multi-disciplinary and multi-sited research project, we have sought to predict, understand, mitigate and address these challenges throughout our research journey. Following a first 5-year cycle of research activities, this article assesses how our ethics architecture shaped and was, in turn, shaped by our research practices. This architecture encompasses the formal and ‘fixed’ ethics protocols, procedures and guidelines developed in response to institutional requirements, but also the dynamic and context-based practices and processes emerging alongside and in dialogue with them, as well as the collective and bilateral conversations we engaged in throughout. By sharing our formal protocols (see online Annex), whilst critically examining their benefits, shortcomings and effects, as well as their relation to the practice-based and reflective layers of ethical engagement that (have come to) surround them, we respond to repeated requests by fellow researchers to share aspects of this framework.

In doing so, we aim to facilitate a better integration of formal ethics requirements and reflective ethical behaviors. This integration is important in light of a persistent tension in human rights research whereby formal ethics requirements, meant to ensure ethical behavior, have often become so complex that they may hamper rather than facilitate a deeper deliberative process that helps prepare researchers for fieldwork. There is a wealth of literature critiquing various dimensions of formal ethics reviews. Yet, limited attention has been paid to translating these critiques into more comprehensive ethics architectures that are both robust enough to meet the demands of ethics committees and funders and ensure the highest ethical standards, while also remaining meaningful and flexible in concrete and culturally different contexts, where researchers are invariably confronted with complex moral questions.

As such, we neither present our ethics protocols as an ideal model to be emulated, nor do we dismiss their practical importance in fieldwork contexts. Rather, we offer an experience-based reflection on how these protocols, along with the associated trainings and documents, supported or complicated our navigation of challenging ethical moments throughout the research process. In this endeavor, we do not position ourselves as ethics experts but rather share insights grounded in our lived experiences as researchers, adhering to the idea of research ethics in themselves as reflective practice. We highlight what worked, what did not, how we adapted or improvised and what lessons we take forward. We aim for this exercise to support other researchers in constructing a layered and responsive approach to ethics – one that acknowledges the inevitable gaps and tensions between formal ethics documents and lived practice, but that seeks to harness their complementarity through a range of guiding principles and generative questions.

This aim is also reflected in our collaborative approach to authorship, which seeks to presence the experiences and insights of researchers with diverse roles and positionalities within our team. Through an adapted version of the Delphi method, which we discuss below, we identified four critical ethical junctures emerging across the various contexts in which we worked: (1) communicating ethics and the meaning of consent, (2) encountering and acknowledging violence and trauma, (3) sharing research burdens and benefits, and (4) balancing agency and vulnerability. The discussion then builds on these themes to foreground how formal procedures and responsive practices can intersect in developing a reflective, iterative and relational ethics of care, and how this requires a commitment of time, resources and an open process of team-based evaluation and support. In the following section, we first provide more context regarding the Justice Visions research and its ethics architecture.

Our Research Context & Ethics Architecture

All authors on this article are affiliated to Justice Visions and conduct research on the role of victims-survivors in the broad realm of transitional justice. This includes textbook transitional justice contexts (like Guatemala, Chile, Tunisia and Cambodia), as well as contexts where the language and tools of transitional justice are mobilized in the absence of a political transition, to address ongoing conflict or violence (like Syria under Assad, Lebanon, Peru, the Philippines and the DRC). Initial research projects analyzed the scope and modalities of victim participation in institutional processes (such as trials, truth commissions or reparation programs), examining how this influences victims’ experiences and perceptions of justice. Later, and in closely linked projects, this focus expanded to include how survivors and communities engage in both organized mobilization – such as protests in the streets or (transnational) advocacy – and everyday acts of resistance, including storytelling, cultural and artistic practices or other informal strategies to challenge impunity and assert their rights. This supported the development of a framework that foregrounds victims-survivors’ agency across formal, informal and everyday justice processes (see e.g., Evrard et al., 2021).

Within and across individual case studies, we employed a wide range of – primarily qualitative – research methods, including in-depth interviews, focus groups, participant observation, learning circles, collaborative making, co-creative mapping, etc. ¹ Our normative orientation has been (implicitly) rooted in the local and transformative turns in critical transitional justice scholarship, and its emphasis on victim-survivor participation and protagonism (Firchow & Selim, 2022; Gready & Robins, 2019; Tsai et al., 2018). This emphasis on protagonism translated in the adoption of a victim- and actor-centered approach, that sought to yield space to research participants, by acknowledging their interests and agency and considering how their input might drive our actions and future research agendas, even within the constraints of projects largely predefined in terms of research questions and methods (e.g., Krystalli et al., 2021; Pittaway et al., 2010). While facing certain limitations, for example related to the adoption of more far-reaching participatory-action methodologies in early case studies, we argue that also within more conventional qualitative methods there are underexplored opportunities for action-oriented or emancipatory practices.

The concrete design of our ethics architecture drew on literature on the implementation of human rights principles in scientific research, guidelines on conducting research in complex and remote sites, our faculty’s ethical protocol, as well as requirements of the European Research Council (ERC), which funded the initial research. These requirements included full ethics clearance prior to project start, the instatement of an ethics advisor and yearly ethics reports to the ERC. Furthermore, our ethics protocols were inspired by other ERC-funded research projects with potentially vulnerable people at our university, and similarly they fed forward into the ethics protocols of colleagues with similar projects at and beyond our university. This repurposing and borrowing from existing ethics protocols illustrates the influence of these documents, which are often emulated and adapted in subsequent projects to ensure compliance with formal requirements, but without necessarily being accompanied by reflective assessments of their efficacy. Their longevity underscores the importance of revisiting, interrogating and dialoguing with these documents and architecture for constant and future-oriented learning.

Reflecting formal requirements, our ethics architecture has been grounded in a do-no-harm-principle, and focused primarily on anticipating and managing risks for research participants, ensuring full and transparent information sharing and engaging with potentially vulnerable research participants in mutually beneficial ways. These objectives were concretized and formalized in a Safety and Harm Reduction Policy and Data Collection and Processing Protocol, accompanied by informed consent forms, a secure database for managing research data and intermittent monitoring. Additionally, we organized team meetings and trainings on various topics (e.g., data management, working with potentially vulnerable groups, vicarious trauma, etc), and developed concise factsheets and guides to translate the protocols into more actionable tools. This formal ethics framework, established before data collection and subject to official review, was mostly shaped by the project’s Principal Investigator, the researcher acting as the ‘ethics and data management’ coordinator, the project’s ethics advisor, university experts (the data protection officer and data steward) and ethics reviewers at both the ERC and our faculty. However, as we will discuss below, dynamic on-the-ground approaches and processes developed by the researchers in response to, and interaction with, this formal framework have been crucial in shaping the emergence of ethical research practices in specific contexts, which fed into a more layered and encompassing ethics architecture.

Critiques of Formal Ethics Requirements

The insights emerging from our collective self-reflection both respond to and draw on several key trends and gaps in the literature related to opportunities and shortcomings of formal ethics requirements characterizing predominantly Global North ethics procedures. Three lines of argumentation that run through much of the scholarly literature problematizing institutionalized and standardized ethics requirements merit attention in the context of this article. A first critiques the extent to which formal ethics procedures may in fact stifle reflective thinking about ethical behaviors (e.g., Haggerty, 2004; Hannes & Parylo, 2014; Murphy & Dingwall, 2007); a second problematizes the extent to which assumptions about vulnerability lock people up in static identities that disregard their agency (Fassin, 2014); and a third focuses on the ill-alignment of these frameworks and their principles with the epistemologies and value systems of the people we work with (Mavhandu-Mudzusi, 2023; Mungwini, 2022; Owor, 2022). The latter includes literature problematizing the highly individualized understanding of the core principles of ethics proceedings, such as consent, agency, harm, rights or benefits (Caeymaex et al., 2023; Martín de Almagro et al., 2024).

In response to these critiques, a vibrant scholarly debate on sound ethics practices has emerged, notably from within the field of anthropology, which has foregrounded relational philosophies, feminist approaches to ethics (Eidenskog et al., 2024; Martín de Almagro et al., 2024; Vivas Romero, 2017), decolonial and Global South perspectives (Cumes, 2015; Guzman et al., 2021), experience- and community-based approaches (Sultana, 2007) and a wider range of ontologies and epistemologies that can inspire ethics approaches (like Ubuntu or Pacha Mama) (Meadows et al., 2003). These proposals foreground notions of agency, care and self-reflexivity, describing in deeply contextual ways how these orientations may shape up in practice (Ansoms et al., 2021; Dunn et al., 2012; Iphofen, 2009), including when working in conflict contexts (Douedari et al., 2021; Dunia et al., 2023; Wood, 2006)

In this article, we seek to bridge these two strands of writing (one being more critical of formal ethics protocols, the other mostly looking at contextual practices) by teasing out meaningful opportunities and moments of convergence between these situated and relational practices and the formal ethics protocols that govern them. By resituating these practices within the ethical considerations of human rights research projects, we seek to move beyond the idea that drafting ethics protocols is a stale exercise, and explore how to turn these protocols into more pertinent tools for anchoring ethical reflexes during fieldwork. In doing so, we steer clear from more managerial approaches, and seek to enrich existing literature, notably by focusing specifically on vulnerabilities that emerge in the context of massive human rights violations, which complements a dominant focus in current literature on vulnerability as an attribute of minors, elderly or patients in health settings.

We also propose a meta-reflection that surpasses the positionally of any individual researcher or specific context. As such, the article contributes to current debates by using practice-based insights and critiques as stepping stones for a more responsive and relational ethical engagement, by teasing out logics and dynamics that are relevant across a wide range of contexts.

Our Methodological Outlook

The authorship of this article reflects our ambition of engaging in a multi-voice, multi-perspective exercise that reflects how different (professional and other) identities affect our relation to, and navigation of, ethical research engagement. In doing so, we seek to identify dynamics and observations that speak to a wider range of cases. To engage in this collective self-reflective exercise, we adapted the Delphi method to our context, to inform the insights underpinning this article.

Identities and Positionalities

The authors of this piece include researchers with diverse roles, ranging from local research assistants involved in fieldwork, to the Principal Investigator, to Ph.D. researchers from within, outside and in-between local contexts (e.g., being from the country but not the same ethnic group, or being raised in the country but enrolled in a program in the Global North) with some working on the project for which the original ethics protocols were developed and others drawing on these protocols in a more ad hoc way for their individual projects. ² We are all funded through a European university and European grant makers, which affects our formal ethics protocols, but also our positionalities – something which we reflect upon throughout the body of the article. We have had various kinds and levels of cooperation throughout the project, with the Ghent-based team having regular meetings, including on ethics, and local research assistants mostly interacting with their direct counterpart in Ghent. However, we also facilitated peer-to-peer exchanges and training between local research assistants and other researchers, which reflected our ambition of multi-directional learning. ³

An Adapted Delphi Method Focused on Collective Learning

A Delphi method is a pragmatic research approach that facilitates structured, iterative communication to arrive at action-oriented insights about real-world issues by fostering expression and convergence of opinion among stakeholders within a specific area of inquiry (Brady, 2015; Hsu & Sandford, 2007; Linstone & Turoff, 1975). It has been widely used and adapted in qualitative research on public policies, social justice and human rights, and can accelerate an actionable dissemination of findings. In this study, we adhered to several key criteria typical of qualitative Delphi approaches, including its emergent design, three waves of data collection and the use of thematic analysis. However, we did not maintain anonymity, as all participants were known to one another, and we integrated both individual (written) and collective (oral) contributions during the data collection waves. This adaptation emphasized trust and inclusivity, ensuring the process was as responsive as possible to the diverse realities and insights of the team members.

The process began with an initial exploratory round aimed at identifying the most pressing ethical challenges encountered during the research. To this purpose, a limited set of open-ended questions was formulated by the first and last author. These were first answered in writing, which allowed researchers with practical constraints (e.g., research assistants with other jobs, different time zones, different languages) to provide input in a convenient way. The inputs were collected by the first and last author, who formulated further guiding questions that structured an in-depth focus group, organized in English and Spanish. The two lead authors then engaged in a thematic analysis to identify the overarching themes, which were summarily described in a first draft report and shared with the team.

The second round of collective discussion focused on further developing the identified themes, linking them to specific components of the ethics framework and various stages of the research journey. This allowed for assessing past practices and identifying concrete areas for improvement in the research ethics architecture, as well as pinpointing which components of that architecture had been useful during fieldwork. This resulted in the creation of a detailed outline for the article. The third and final round involved refining the structure and content of the envisioned article, in addition to addressing any remaining ambiguities, to ensure the document accurately reflected the team’s collective insights. This phase also marked the transition from dialogue to collaborative writing, with individual team members adding, commenting on and revising the consecutive drafts proposed by the first and last author.

The Delphi method thereby facilitated a consensus on ways forward and the formulation of actionable recommendations for future research (Dunn et al., 2012). We acknowledge that our adaptation of this method is only useful in a context where people feel free to share openly, including about failures, and that more anonymized approaches may be more useful in other settings. Moreover, the Delphi method also faced certain limitations. Multiple languages, physical distance and the passage of time since fieldwork created challenges for participation, particularly for local research assistants. These barriers were partially addressed by combining opportunities for written inputs with sufficient time for explanation and follow-up discussions. In the next section, we present key insights emerging from our iterative thematic analysis.

Ethical Junctures and How We Navigated Them

In this section, we identify four critical ethical junctures in the research journey, where tensions or gaps emerged between the codified protocols and meaningful ethical engagement on the ground, spurring reflection and adaptation through myriad responses. This section cannot do justice to the complexity of concrete situations or coping mechanisms developed in response to them. Yet, these four junctures provide an opportunity to examine how different layers of the ethics architecture interact and intersect – sometimes constructively, sometimes tensely – and offer insights for designing more responsive and dynamic architectures going forward.

Communicating Ethics and the Meaning of Consent

In the run-up to fieldwork, ethical questions emerged about how we enter a research context, communicate and make our purposes intelligible and relevant for the people we hope to work with. In this regard, the introduction of an ethics protocol required sidelining pre-existing or emerging relationships with (potential) research participants, as the procedures for ensuring full and voluntary informed consent consider the beginning of the research activity as the formal start of the relationship. The existence of a detailed step-by-step process, elaborate consent forms and information sheets and several team trainings on the topic ensured that we all paid great attention to transparently outlining the scope, potential risks and envisioned outputs of participation, and to ensuring participant autonomy and protection during data collection and in publication of results. Yet, the timing, formality and terse language of the consent procedure also introduced tensions between procedural adherence and relational dynamics of trust-building. ⁴

It is important to note that several research participants explicitly stated that they appreciated the demarcated consent procedure (for example in contexts where their photographs or data had been used without such consent in the past). Other experiences, however, illustrated how the requirement of obtaining full consent before any data collection begins and through standardized forms and signatures, may misalign with how respondents and researchers seek to foster clear and common understandings about the type of relationship they are establishing, for how long and what the (mutual) benefits may entail. The consent procedure often disrupted the natural flow of early conversations and the building of rapport, or inadvertently (re)enforced an unequal power relation that can cast participants as vulnerable respondents. Some participants problematized signing their name on a form requesting anonymity, or distrusted signing forms more generally, due to negative experiences in other contexts in the past. Moreover, the individuated practice of consent sometimes ill-aligned with preferences for seeking and establishing collective consent, for example when engaging with Indigenous communities or victim collectives.

The technical and legalistic language of consent forms, required to meet institutional review board standards, often posed further barriers. Clauses about personal data processing and preservation (reflecting GDPR requirements) or data publication in different channels were not always intelligible or relevant to respondents and required extensive contextualization. At a more fundamental level, it was often difficult to explain the concepts of academic research and publication in remote communities where people could difficultly imagine what this implied, despite best efforts at making this concrete. Jointly, these challenges point to the inherent limitations in the formal component of our ethics architecture, and notably of the pre-set and pre-approved documents and procedures that had to be frontloaded and strictly adhered to. This impeded a more dynamic and responsive approach to what full and free consent meant in specific collaborations with research participants.

In response to this, several researchers adopted an approach to iterative consent, whereby consent was framed as a continuous and evolving dialogue rather than a one-time agreement. This iterative engagement involved revisiting consent at multiple stages, creating opportunities to re-explain the research’s purposes and outputs, to address emerging concerns and to adapt to participants’ evolving understanding of their involvement. This strategy materialized in various contexts, each time with different modalities, from the light interpretation of going back to research participants every time before publishing anything with their name or insights in it, to a more extensive approach of revisiting consent before every interaction in case of multi-stage fieldwork activities.

Going forward, such an iterative approach may also structure the formal procedures, for example by applying for ethics clearance at the start of every work package, rather than upfront, so that progressive insights from previous fieldwork can feed into the formal protocols, and empirical insight can support proposals for less conventional approaches. During the collective reflection process for this article, other concrete proposals also emerged for revisiting the formal consent process. These emphasized the importance of embedding responsiveness, relationality and co-creation as constitutive components to consent in the development of formal protocols – to enhance their capacity for serving as a stable anchor point in navigating complex research contexts.

For longer collective research processes (e.g., with a community or group of respondents) and to be able to engage in a more open-ended process, a strategy of delayed consent was proposed, whereby signed consent forms were only collected after multiple meetings aimed at familiarizing participants with the project, its focus and the envisioned method and outcomes. This approach shifts the focus from an early opt-in to a more informed (or even collaborative) decision-making process, thereby making the initial consent more meaningful and vastly reducing the hurdles which could be associated with retracting consent later. However, this also means that any activities conducted before the formal consent procedure cannot be explicitly used in the research, which can be a significant limitation.

Reflecting on the power dynamics inherent in formal consent processes, we also considered the potential for reciprocal consent, whereby researchers themselves sign a more extensive agreement, outlining the researcher’s commitments to participants. This process could be imagined as part of a broader strategy meeting, as this acknowledges the important role played by collectivities and the extent to which these can enhance agency of research participants, for example in jointly discussing and deciding on ethical guidelines and mutual expectations, which can complement the formal ethics architecture established by the project. Both delayed and reciprocal consent processes acknowledge the need for formal individual consent, but create a space for discussing the meaning and modalities of consent. When implemented at a collective level, these approaches can moreover acknowledge collective decision-making preferences in some communities, and bridge the formal requirement of individual consent with reality on the ground, including traditional leadership structures.

Lastly, researchers also emphasized the importance of recognizing and incorporating cultural and implicit ways of expressing (non-)consent. This requires remaining attuned to different “ways of saying no”. Researchers also stress the need to translate or operationalize the consent clauses into locally intelligible and relevant scenarios, with concrete examples of what participation, risks and outputs could look like. These relational and context-sensitive approaches stress an interactive communication regarding ethics and consent, which may often clash with resource constraints and timeframes imposed by ethics committees, institutions and funders. For instance, iterative consent processes demanded significant time, effort and skills, both from researchers, assistants and participants. It requires rethinking and committing to consent as a fundamental process that comes to constitute one of the main research activities and that requires flexibility to accommodate dynamic realities on the ground, as well as sufficient resources and support from local counterparts to ensure sound implementation.

Encountering and Acknowledging Violence and Trauma

Research with survivors of severe human rights violations often triggers strong concerns over the risk of re-traumatization. To mitigate this risk, our ethics protocol, in line with mainstream perspectives, prescribed the development of topic guides focusing on justice-seeking agency and aspirations, the selection of victims-survivors who had shared their stories in public before and the establishment of referral networks for research participants showing signs of distress. In practice, however, across nearly all research settings, participants wanted to use the space and time provided by the researcher to share personal stories of harm, loss and injustice. Interviews and focus groups were often perceived as spaces for reflection and acknowledgement, and sharing these stories was considered crucial for understanding the justice activism and priorities that we sought to foreground. In most contexts, detailed recounting of human rights violations was part of almost every interview, and reflected participants’ search for acknowledgement of injustices that often remain un(der)recognized.

Designated trainings on trauma-sensitive research, a hands-on guide for identifying and reacting to respondent distress and the use of referral partners proved to be valuable starting points, both in terms of raising initial awareness and in offering a framework for action. Yet, the dominant focus of our ethics protocols on risk management and risk mitigation – which assumes that ethical concerns are predictable, avoidable and resolvable through a broad set of procedural safeguards – prescribed us to bracket off, avoid or defer engagement with stories of violence. This approach left us feeling ill-prepared and, more fundamentally, discouraged deeper engagement with complex moral questions, the development of (soft) skills or attention to small-scale but immediate benefits of participation, such as creating a space for validating experiences.

In practice, such deeper engagement was crucial, as the encounter with and retelling of traumatic life experiences (and aspects of vicarious trauma emerging in response to this encounter) became an unavoidable reality that we had to navigate. This reality required revisiting our assumptions regarding participants’ agency in deciding when and how they would claim time and space to tell their stories, without assuming that doing so would retraumatize them or reduce them to those experiences. It also illustrated how Western conceptualizations of trauma and trauma responses were, at times, ill-suited and inappropriate from cultural and religious perspectives in different fieldwork settings.

In response, we began building on the most relevant elements of the formal ethics protocols to understand how we could complement and go beyond them, often with guidance from the research assistants and local partners. To engage with traumatic experiences, we started from a personal place of commitment and care to develop a practice of active and empathetic listening. Acknowledging (our own and others’) emotions was key in engaging with potentially vulnerable participants, requiring both the strength and sensitivity from the researcher to hold space for those who chose to share their feelings and experiences. However, this meant the affective impact on researchers was profound, with compassion fatigue or vicarious trauma occurring in several cases. In some settings, the inclusion of a support person (like a psychologist, social worker, or trusted individual) during and after research activities proved to be beneficial, whereas in others, this would have been considered paternalistic, or risked displacing other ways and sources of coping. ⁵

Across these different scenarios it was crucial to expand our understanding of what forms support and coping practices can take, beyond a Western-centric focus on trauma and how to address it. This includes more attention for the role of spiritual leaders, elders or other trusted figures in a community, as well as religious or cultural practices related to coping. We also had to revisit the false dichotomy regarding who we consider vulnerable (for example when identifying some participants as experts or gatekeepers), and remained attentive to less considered forms of trauma, such as the denial of experiences of harm and the impact of epistemic violence, i.e. where victims’ knowledge systems, experiences or voices are dismissed, ignored or invalidated (Fricker, 2007). Moreover, questions of local resources and burdens should be carefully considered. We should question the expectation of simply identifying a local referral network, since these (typically gendered) actors are often under-valued, under-resourced and sometimes themselves traumatized. Providing psycho-social support for research participants also raises questions regarding scarcity, prioritization and the consequences of (voluntary or involuntary) non-participation in areas or communities where trauma and historical marginalization are widespread.

Looking ahead, we find it crucial to integrate meaningful forms of psychosocial awareness, care and support into each research activity, and to make space for this within the formal ethics architecture – taking into account culturally sensitive understandings of care, who is best positioned to provide care and how this role is shaped by context, existing relationships, language and translation skills. In addition, our experience hints at the relevance of strengthening or leveraging the social fabric of groups themselves to provide ongoing support. Notably when working with collectives, this can be more relevant than ‘bringing someone in’ or establishing a formal referral network.

Lastly, when it came to researcher well-being, interpersonal relations and communications among researchers, with local research assistants and with fieldwork partners became vital. Pre-departure training and subsequent peer-support, however, did not always prove sufficient to cope with the emotional and psychological burdens of fieldwork. This was particularly true when confronting extreme injustices, whether during conversations, while transcribing or while revisiting interviews and focus group discussions through repeated readings. The standard psychological support services available at the university were not sufficiently attuned to these needs, nor could the Principal Investigator take on the role of a mental health professional. Researchers had to invest significant efforts (and sometimes resources) to establish psychological support structures and coping mechanisms. Looking forward, we recommend proactively acknowledging the likely toll of this kind of work on researchers, rather than framing it as a preventable risk, so that systematic exchanges and tailored psycho-social support can be foreseen with particular attention to the specific challenges faced by researchers from, and/or working in, the Global South.

Sharing Research Burdens and Benefits

Working in contexts where people have endured extreme harm, and where their demands for justice and redress are often inadequately addressed or entirely overlooked, the commitment to minimizing the burden of research participation and ensuring meaningful benefits was a constant concern throughout the research process. This dual commitment frequently served as both a motivator and a source of distress for researchers. Our discussions on this topic centered around several key questions: What is a burden and what is a benefit and who decides this? At what stage of the research process should benefits materialize? And how do we reconcile this commitment with high workloads and feasibility concerns?

In a minimalist sense, and in alignment with our formal ethics protocols, we aimed to avoid extractivism and minimize the burdens on participants, recognizing that the time they dedicated to our research could itself be a burden. Yet, this view was challenged by one of the research assistants who emphasized that creating a space where people are heard, where their stories acknowledged on their own terms and their suffering made visible, can itself be a positive and beneficial experience. In other contexts, research participants from Indigenous communities considered such spaces as a starting point for building a more long-term relationship that could provide benefits, future opportunities and contribute to their own networks and agendas.

Both positions underline the importance of acknowledging that benefits may materialize across different levels and timelines throughout the process, rather than being something to only work towards after the conclusion of research activities – as we initially envisioned in our societal value creation strategy. Several research participants notably expressed a preference for immediate rather than postponed benefits that would only materialize long after the initial interactions, just as most researchers acknowledged the limited bandwidth for developing comprehensive benefit-sharing strategies in the last stages of their Ph.D. journey (when some participants may, moreover, no longer be reachable). This also speaks to the broader concern regarding the time and resources needed for such activities, and the limited (institutional) acknowledgement of or support for them.

In our work, both immediate and larger initiatives emerged organically, often through attentiveness to participants’ questions, suggestions, needs and justice activism. This included attention to immediate and concrete benefits within the confines of specific research activities, for example, by reserving time to articulate what participants gained or sought to gain from their engagement, by making time for discussions or knowledge sharing on topics that participants themselves brought to the table, by creating a caring environment for active listening or by incorporating collaborative approaches and joint meaning-making. In addition, across different contexts, researchers also worked on more extensive non-academic research outputs, often in co-creation with research participants. This included support in drafting legal documents or in civil society advocacy, developing an illustrated memory booklet, facilitating intergenerational dialogues with younger community members, broadcasting a radio series, co-creating artwork and more. In this way, process-related benefits and societal value creation became closely connected, and sometimes also embodied in specific activities, such as strategy meetings held with participants midway through the research process.

There was not necessarily a fundamental tension within the formal ethics requirements at this juncture, largely because the specific obligations or contours of benefit-sharing were vaguely defined – reflecting the extent to which this is hardly emphasized by academic institutions or funders. Our experience highlights the need to emphasize benefits-sharing more systematically in formal ethics requirements, ensuring it is recognized as an integral part of the research process, rooted in commitment and solidarity, with a framework for context-sensitive initiatives. This context-sensitivity is crucial since it would be difficult and counterproductive to formalize specific activities and strategies in a rigid way.

While benefits-sharing activities were critically important to our work and normative orientation as human rights researchers, we struggled with significant constraints of finite time and resources. This challenge is particularly pronounced in the context of Ph.D. research, for which deliverables are formulated in academic terms, and often most intense in the final stage, which is typically also when social value creation is planned. The reality of time and resource constraints underlines the importance of foregrounding process-based benefits, and, in planning fieldwork, to foresee the co-creation of safe spaces. While specific training may be helpful in this regard, in our context we benefitted most from open discussions between the Principal Investigator and researchers, and the willingness for collectively thinking and supporting each other, as well as having the required resources. This supportive team environment, and the allocation of project resources, allowed and encouraged researchers to be attuned to identifying and seizing opportunities for benefits-sharing as they arose.

Balancing Agency and Vulnerability

The modalities of our formal ethics framework reflected the involvement of participants who, because of their victimization, were considered to be potentially vulnerable and in need of protection. Yet, critical literature on ethics in human rights research underscores the importance of balancing ethical safeguards (such as informed consent, trauma-sensitive methodologies and data protection measures) with preserving participants’ agency and autonomy, to avoid locking participants into positions of vulnerability or victimhood. Despite efforts to balance these in our ethics approach, it was difficult to adequately account for the nuanced ways in which agency, vulnerability and resilience are shaped by complex subjectivities and power hierarchies, especially when developing protocols before actual fieldwork took place. This presented significant dilemmas for researchers, who were aware that, on one hand, failure to protect participants could result in significant harm, but on the other, indirect and long-term harm might result from the way participants were (not) included in the research and represented in its outputs, as a result of this protective reflex.

In fact, concerns of balancing agency and vulnerability permeated many ethical dilemmas encountered at the different junctures described above. The formal consent procedure, with its focus on risks and standardized forms, could inadvertently reposition participants as vulnerable individuals. The initial strategy of avoiding stories of violence and trauma precluded participants from exercising their agency in choosing to (not) recount these experiences as part of their healing and justice-seeking journeys. In benefits-sharing, concrete tensions included situations when we decided not to cite (by name) despite their explicit consent, or when participants wished to share images of research activities on online platforms in ways that challenged anonymity. In addition, at the crucial stage of analyzing and writing up findings, recognition of (potential) imbalances in agency and vulnerability between researchers and research participants also necessitated a critical re-examination of our own positionality in the research context. ⁶

Complex questions of agency and vulnerability were tied to varying and intersectional identities, for example, regarding gendered dynamics in research participation, which had not been addressed in the ethics protocol. Gendered power dynamics were found to be both obstructive and enabling throughout the research process. For instance, local leaders who acted as gatekeepers in Indigenous communities were predominantly men, and accessing women from those community required leaders’ consent. Related to this, when some women research participants became more vocal through the process of iterative collective research activities, researchers wondered what would happen after the end of the research activities, because such research-induced dynamics could generate opportunities as much as tensions in the social contexts of these participants. In some cases, gendered patterns also facilitated access, for example, when women who stayed home were more easily accessible during field visits, or more at ease with talking to women researchers.

The political subjectivities of respondents added another layer of complexity. In some cases, research activities involved participants who brought diverging political affiliations and opposing priorities to the research environment, which sometimes challenged an open and safe discussion environment, or even prevented others from participating. At the same time, expecting respondents to come into the research as apolitical participants, as we implicitly did in our formal ethics protocols, was neither feasible nor desirable. Occasionally, respondents would steer the research activity toward their own priorities, reorienting the conversation to topics or narratives that diverged from, or were not immediately related to, the questions we sought to foreground. Such moments underscored how engagement with research activities – and the researchers – could be leveraged for different agendas, emphasizing the need for awareness and flexibility in navigating participants’ inherently political engagement.

In these and other instances, not just the ethics framework, but we ourselves too, erred on the side of caution and protection, either by not citing research participants by name even when they asked for this, or by using cautious language when referencing sensitive political contexts, such as when participants criticized authorities. Equally challenging where instances where we had to decide whether to draw attention to divisive or harmful dynamics among research participants (e.g., tensions within or between victim collectives and NGOs) and our concerns that doing so may discredit them in public discourse or undermine their broader justice struggle. At the same time, this caution raised questions about the duty to engage with (potential) conflicts or harms, for example when prominent men in CSO networks displayed problematic behaviors that went unaddressed. Such dilemmas were not anticipated in incidental findings policies, and complicated what incidental findings may entail.

This highlights the limitations of ‘anticipatory regulatory regimes’ (Murphy & Dingwall, 2007), which are constrained by what we can plausibly imagine agency and protection to look like before entering the research context, and which tend to prioritize protection in ways that preclude more nuanced, intersectional understandings of vulnerability and agency. Our protocols used a blanket categorization of vulnerable people, distinguishing only between expert and victim profiles, but without considering the diverse sources, modalities and hierarchies of agency and vulnerability, including gendered and political subjectivities, or different sources of power and knowledge. Unsurprisingly, participants often rejected the identities projected upon them through these protocols, for example by claiming space to tell their stories of harm, urging to be quoted in outputs, bringing politicized interests into research activities, questioning the informed consent process and so on. Moreover, the protocols were not responsive to participants’ evolving needs or desires throughout the research process, to shifting risk assessments caused by significant social and political developments or to participants’ consent considerations varying in light of the language or context in which their names or insights would be presented.

In navigating these dilemmas, we adopted a number of (often implicit) strategies. During the actual interviews, for example, we sought to center participant voices and perspectives, recognizing that participants often had both a deep understanding of their own vulnerabilities and a desire to shape the research process and be acknowledged for their contributions to it. It required rethinking the notion of expert and victim to recognize all research participants as knowers, and to be more aware that institutional or civil society experts were often also direct or indirect victims of violence and subject to risks in their justice struggles. Moreover, decisions about what to share, were typically not only shaped by security considerations, but also by a concern over not violating the intimacy and trust of the safe spaces for sharing that were co-constructed with participants. For instance, one researcher chose not to use artwork and poems created by young students during collaborative research activities, to preserve the intimacy and closeness among those who participated in the original experience.

At a deeper level and looking forward, we experienced a need for opportunities to co-create ethics protocols with participants, for example during initial collective meetings. Involving participants in this exercise recasts who has the authority to assess who is considered vulnerable and in need of protection, on what grounds, and what options are available when different parties’ assessments do not align. ⁷ Rather than thinking about this in terms of who has the last word (and eventually who is (legally) accountable), co-creation of ethics architectures (ideally with the inclusion of the funder) offers a more nuanced and complete set of scenarios and options for navigating them in context-sensitive, responsive and iterative ways. This would also align with the care-based approaches mentioned in the previous sections, including through active listening and collaborative decision-making that looks beyond the (false) binary of agency versus protection.

Towards a Reflective, Iterative and Relational Ethics of Care

In spite of the rounded critiques of formal ethics protocols and requirements, our experience is that having an encompassing formal framework in place offered important benefits. In a very concrete sense, the availability of a detailed data management plan and architecture, along with trainings, forms and guidelines, established a reliable framework for handling sensitive data, which reinforced participants’ sense of being taken seriously and protected throughout the research process – something they themselves sometimes explicitly commented positively on. Beyond this practical utility, the protocols also fostered an awareness and heightened sense of responsibility within the research team that influenced broader decisions and actions during fieldwork, and integrated ethical consideration at every stage of the research, as well as offering a concrete framework to guide our actions.

Researchers were thus inspired to build upon ‘what worked’ in the formal ethics framework, while complementing and adapting these policies where they were ill-aligned with their research context, thus effectively building a more encompassing and layered ethics architecture. These processes of practice-based adaptation were normatively guided by a reflective, iterative and relational ethics of care, ⁸ and significantly influenced by researchers’ personal commitment, positionalities, situated understanding of the fieldwork context, inspiration from peers, relations of trust with research participants and the expertise brought in by local research assistants and grassroots partners. In this discussion section, we seek out ways in which stronger convergence can be achieved between situated proposals and practices rooted in an ethics of care and the mainstream ‘do no harm’- principles typically underpinning formal ethics procedures. By identifying dynamics that cut across the four themes discussed above, we explore how care-driven ethical principles and reflexes can permeate the different layers of an ethics architecture.

As many of our examples illustrate, a strict compliance-oriented approach to predefined ethics could have stifled or marginalized participants whose desires and agency transcended predefined risks, while deprioritizing relational and context-sensitive ethical considerations that fall outside the logic of calculable risks. Based on our discussions within and beyond the context of this article, we therefore encourage ethical engagement that starts from core principles rather than standard procedures, that emphasizes relationality rather than solely relying on externalized rules, that adopts an iterative and continuous process rather than one rigidly defined in advance and that is attuned to both collective and individuated (bilateral) processes. In order to embed these dynamics across the formal, practice-based and reflective layers of an ethics architecture, the following overarching suggestions emerged alongside the concrete recommendations outlined above.

Developing a Shared Language of Core Ethical Principles

During our team meetings and bilateral conversations, shared concepts and commitments often surfaced that inspired our ethical considerations and choices in concrete ways, but were not explicitly codified in the formal ethics protocol. To bridge the gap between formal protocols and their practical implementation, the articulation of this shared language of ethical principles can be integrated explicitly at the start of a research project. This exercise can draw on institutional expectations and guidelines, and on theory and best practices available in literature, but also build on and reflect team conversations and culture. Frontloading the explication of shared ethics principles as the basis for a formal framework offers a sound basis for developing this framework itself, but also offers more concrete pointers when that framework later on has to be adapted to specific situations. For example, developing a Lexicon of Ethical Principles that can formalize this shared understanding. ⁹ Also individual researchers may benefit from explicating these principles as part of their required ethics procedures, and in the process of establishing and consolidating partnerships with research participants and collectives.

Including Guiding Questions Regarding Core Principles

An approach rooted in relationality rather than externalized rules requires a willingness to engage in an honest self-reflective assessment of one’s own actions and attitudes. Most of us do this in implicit ways, which have limited affinity with the formal nature of an ethics protocol. One way of exploring possible convergences between these prefixed protocols and our dynamic practices and responses in the field, is the explication of a comprehensive set of questions regarding the alignment of our core ethical principles and the practical decisions we take as part of our operationalization thereof. Foregrounding these self-reflective questions in the development of the formal layer of a comprehensive ethics architecture can be a means of overcoming the risk of externalizing or standardizing ethics in ways that may create a sense of detachment. By (re)visiting these questions throughout the research cycle, ethical practices are consistently considered and reflected upon throughout the project in ways that are adaptable in order to become locally and culturally relevant. In addition to explicitly foregrounding self-reflectiveness, these kinds of questions could also speak to the notion of reciprocal consent, as they could also be thought of as tools for our research partners to hold us accountable, demand a discussion about ethical principles and ensure the transparency and collective reflection that can be a basis for further engagement and relationality.

Collaboratively Constructing a Layered Ethics Architecture

The iterative development of an ethics architecture can be further supported through a more co-creative approach to ethics protocols – one that acknowledges that decisions around how we engage with participants, share findings and understand consent should be made in conversation with those participating in the research. While technical matters, such as GDPR compliance, may be difficult to address co-creatively, the joint identification and reflection on core ethical principles, risks and dilemmas can meaningfully involve participants to achieve agreement and common understanding. This approach is not only aligned with the ‘risk anticipation’ logic of formal ethics protocols but would also produce more relevant guidance and support for researchers working in complex contexts. It allows for a more nuanced reflection of the intersecting vulnerabilities and forms of agency of participants and fosters more equitable practices for sharing benefits and burdens. Such a co-constructive process should not be treated as a one-time exercise, but warrants an understanding of ethics protocols as living and responsive frameworks – as is today already the case for data management architectures, which are accepted to benefit from periodic revisions that reflect progressive insights and complications emerging along the way. Co-creation also changes the power relationship between the researcher and the research participants (who often face a binary choice in signing pre-set consent forms), opening space for more equal collaboration and for acknowledging the importance of the collective.

Creating Value-Based Spaces for Bilateral and Collective Engagement

The three proposals above encourage, and rely on, the creation of collaborative and reciprocal spaces for engaging in research on the basis of research participants’ (collective) agency, resilience and negotiated consent. Foregrounding and integrating these shared and dialogical spaces with research participants, local researchers and partners throughout the research process requires both an awareness of the role of these spaces in specific contexts and the ability to navigate them with respect and empathy, and constitutes another concrete way in which a relational ethics of care can inform formal protocols. Such space- and process-based forms of engagement heavily rely on values such as recognition, respect, empathy and solidarity. These so-called soft issues are indeterminate and intimately connected to personal dispositions, and therefore typically omitted from formal protocols and requirements (Ulrich, 2017). Yet, they often matter greatly to research participants and are particularly crucial when navigating traumatic experiences and cross-cultural boundaries. In addition to codifying these values within the shared language of ethical principles, more tailored training and capacity-building should be organized for researchers to become attuned to the emotional dimensions of engaging through value-based ethics of care.

Concluding Remarks

The four proposals above start from a reality of having to bridge institutional requirements with realities on the ground within a context of finite resources. They also reflect progressive insights gleaned from real ethical dilemmas and complex contexts, where we were not always able to respond in an adequate way, for example when many of us were confronted with vicarious trauma. The four proposals seek to identify ways to better align formal procedures with the dynamic processes of ‘doing ethics’ in practice, in ways that do not necessarily require more resources but rather a rethinking of our relationship with formal ethics protocols, to move them closer to context-sensitive and relational ways of engaging, rather than perceiving them as an impediment to such engagement. This approach also stems from the realization that there is often no strict enforceability or accountability regarding ethical practices. Protocols represent a baseline or minimalist thinking – and even this is not always easy to enforce. Institutions may be very involved ahead of the stages of ethical clearance, but where there is typically little monitoring or evaluation of ethical practices later on.

In light of this, this article serves as an exercise in collective knowledge generation on research ethics, not just within our team, but also in response to numerous requests from other researchers facing similar challenges when working with people in situations of vulnerability. We have sought to approach research ethics as a layered architecture, as visualized in Figure 1. Ethical decision-making occurs continuously – before, during and after fieldwork – requiring a robust (formal) framework of guidance, context-sensitive and empathic approaches to implementation and a commitment to (collective) self-evaluation and forward-looking learning. This layered ethics architecture should enable the individual researcher to make decisions at difficult ethical junctures, since “the responsibility to act ethically rests ultimately on the individual researcher” (Fujii, 2012, p. 718). In proposing this layered architecture, and through the approaches and recommendations outlined in this article, we seek to add to the thriving literatures on ethics of care and relationality, by seeking a rapprochement between practice-based proposals and formal ethics protocols, and embedding both in a process of (collective) self-reflection, which is often left implicit.

Figure 1.

A Layered Ethics Architecture

We acknowledge there may be vast differences in how open ethics committees and universities are to engaging in this conversation and adapting their approaches to ethical review. In our team’s upcoming projects, for example, our faculty’s ethics committee has expressed interest in collaborating on a more iterative approach, and participating to a collaborative dialogue with participants and stakeholders to discuss initial documents and protocols. Furthermore, our findings are contributing to informing new training initiatives and support trajectories set up by the faculty and university, particularly aimed at assisting early-career researchers in conducting fieldwork in difficult settings and with potentially vulnerable populations. For researchers or teams facing institutional resistance, demonstrating the benefits of a more iterative and context-sensitive approach (e.g., case studies, successful examples from other institutions or pilot projects) and fostering alliances with like-minded colleagues, research teams and external ethics bodies can provide important leverage.

Supplemental Material

Supplemental Material - From Protocol to Practice: Collective Reflections on Navigating Ethical Junctures in (Post-)Conflict Research

Supplemental Material for From Protocol to Practice: Collective Reflections on Navigating Ethical Junctures in (Post-)Conflict Research by Elke Evrard, Kim Baudewijns, Marit de Haan, Miriam de Paz Brito, Brigitte Herremans, Farah Hindawi, Sarah Kerremans, Gretel Mejía Bonifazi, Sofie Verclyte, Daniela Zubicueta, and Tine Destrooper in International Journal of Qualitative Methods