The gap between the language and the practice
The word ‘collaboration’ now appears in almost every UK research application, funding guidance document, and institutional strategy. In practice, funders tend to use terms such as ‘partnership’ and ‘collaboration’ when describing relationships with organisations, and co-design, co-production, and engagement when describing relationships with communities. In both cases, these terms can describe either a process in which the subjects of the exercise genuine power in research decisions, or one in which their involvement is secured only after the agenda has already been fixed. The difference matters enormously for what the research produces, for the communities it claims to engage with, and for the institutions that share responsibility for the outcomes.
The most cited framework for thinking about levels of community participation remains Sherry Arnstein’s ‘Ladder of Citizen Participation’, published in 1969 in the context of US federal urban renewal programmes. Arnstein defined degrees of participation across eight rungs, grouped into three registers: non-participation (manipulation and therapy), tokenism (informing, consultation, and placation), and genuine citizen power (partnership, delegated authority, and citizen control). Her central argument was that the lower rungs, the ones most deployed in institutional practice, are forms of legitimation rather than power-sharing. Informing operates as a one-way flow with no channel for feedback. Consultation becomes what Arnstein called a ‘window-dressing ritual’ when no mechanism exists to ensure that what communities say will shape what institutions do.
The International Association for Public Participation (IAP2), founded in 1990, adapted Arnstein’s framework into a Spectrum of Public Participation, converting a hierarchy of power relations into a set of technically equivalent choices and reframing the question of participation as one of technique rather than political stakes. This framework has shaped the professional public participation model now employed by most heritage organisations and institutions that design and run publicly funded research projects.
Oral history has its own tradition of grappling with these tensions, one that predates the formal vocabulary of co-production but engages the same underlying questions. Paul Thompson’s The Voice of the Past (1978) argued that the practice gives history back to the people in their own words, redistributing historical authority from professional historians to working-class communities and those otherwise excluded from the written record. Michael Frisch’s concept of ‘shared authority’ (1990) gave this a more precise methodological formulation, challenging the hierarchies of cultural institutions and moving toward genuine dialogue. The Oral History Association’s Social Justice Task Force guidelines (2019) set out what this means in practice: ethically grounded oral history projects avoid extraction, exploitation, and the entrenched power structures that reproduce inequality within the project itself. These principles are not abstract aspirations. They are measurable features of how projects are designed and resourced. And, as Ross and colleagues have shown, engagement with communities is still too often shaped by ‘shallow frameworks that serve to silence dissent or limit inclusion to those who do not question dominating structures of power’. They are also, as the evidence from UK research funding shows, consistently honoured in the breach.
What the funding architecture does
The current participation funding landscape was not inevitable. It was built through a sequence of policy decisions whose history illuminates why the gap between language and practice has proved so persistent.
By the late 1990s, a series of failures, including BSE, GM crops, the handling of nuclear waste, had badly damaged public confidence in the relationship between science, government, and institutions. The House of Lords Science and Technology Committee’s 2000 report Science and Society diagnosed what it called a ‘crisis of trust’ and called for a shift from the old model of ‘public understanding of science’, which had positioned the public as a deficit to be filled, to a new model of genuine two-way dialogue. For the first time, a major parliamentary report insisted that engagement should become an integral part of how research was conducted and governed, not an optional add-on. Research councils responded within a year with a £4.5 million ESRC programme on science and society. The language of dialogue and participation entered the institutional mainstream.
The second decisive turn came with the introduction of the Research Excellence Framework in 2014. REF2014 was the first national exercise to formally assess the societal impact of research, requiring universities to demonstrate benefits beyond academia. The impact agenda embedded participation requirements in the governance of UK research in a new way, not as an ethical aspiration but as a condition of institutional funding and performance assessment. The effect was to institutionalise community engagement as something universities needed to demonstrate to funders, managers, and evaluators, which is a different thing from giving communities genuine collaborative power over what research does and who it serves. As the National Co-ordinating Centre for Public Engagement has noted, the impact agenda provided a mechanism to recognise the value of engagement, but the incentive structure it created was geared toward satisfying institutional performance metrics rather than redistributing research power. REF2021 increased the weighting given to impact; REF2029 increases it further still. Each iteration has deepened the requirement to engage, without resolving the structural question of who holds power over the agenda.
The dominant frameworks for community participation in UK research exist not only as theory but as requirements built into grant conditions. Understanding the structural relationship between funders and the communities whose participation in the research process they nominally seek to involve is therefore not optional.
UKRI spent £12 million on public engagement activities in 2024-25. This is a significant figure, but modest relative to its total research investment of £9.9 billion in the same year and engagement requirements vary substantially between individual research councils, concentrated with a few significant exceptions at the inform and consult end of Arnstein’s ladder. The AHRC’s Mobilising Community Assets programme uses an assessment process with equal numbers of academic and lived-experience reviewers on the funding panel. This is one of the few instances in UK public research funding where community representatives have formal decision-making parity rather than advisory status.
Since REF2014, the language of ‘impact’ has been embedded across the governance structures of UK research. As Cohen, Williams, and Grant (2025) argue, participation has been institutionalised largely through mechanisms designed to satisfy funders, performance evaluation, and university management systems, not to transfer agenda-setting power. Partnership becomes a way of demonstrating institutional value rather than meeting the needs of partners. Collaborative projects also require time and resources that continue to be unevenly distributed within research teams. The structural issue of project staff at different career stages, from those in their first job as research associates, to professors in the later part of their careers, having different needs, is beginning to be acknowledged by initiatives such as the AHRC’s Thrive, but has not yet been resolved.
The National Lottery Heritage Fund (NHLF) is a key funder specifically supporting oral history research. Its ten-year strategy Heritage 2033 includes ‘inclusion, access, and participation’ as one of four investment principles. Yet applications are prepared and submitted by institutions, not communities. The project framework, timeline, and intended outputs are established before community involvement is formally required. Communities appear in applications as beneficiaries and participants of projects whose design is controlled by the applying organisation.
UKRI’s own published evaluation of its Community Knowledge Fund (2026) acknowledged candidly that ‘the field of community-led research is less well developed than traditional research approaches and its importance and value less understood by the wider R&I system.’ That a major public funder finds it necessary to recommend making community-led approaches ‘business as usual’ is itself evidence that the dominant business of research remains something else.
The structural pattern across funders is consistent. Funding flows via institutions. Communities participate within frameworks they did not design, on timelines they did not set, towards outputs whose ownership was determined before they entered the process. This matters because people take part imagining that outputs will be used to change something, that their knowledge will carry weight. When projects end and communities realize they had no real say over how findings would be used, or what the research had been designed to show, that expectation is disappointed. The system produces disillusionment, not because individual funders are careless but because the rules of the funding structure generates it.
Some funders have begun to acknowledge this. Lankelly Chase, a major social justice funder, announced in 2024 its intention to relinquish control of its assets so that money could flow freely to those doing social justice work, and subsequently committed to ‘finding ways that resources reach communities in ways designed by and for them.’ The New Philanthropy Capital’s 2024 report identifies similar pressures across institutional philanthropy. These are significant departures from the prevailing model. They also confirm, by their rarity, that the dominant model remains otherwise.
What oral history projects reveal
The Newcastle University Oral History Collective has spent nearly a decade testing what genuine partnership looks like in practice. The five projects below illustrate what working within, and sometimes against, the structural constraints of research funding look like.
Foodbank Histories (2018-2024) was one of the Collective’s earliest collaborative projects, funded by the ESRC, the Newcastle University Social Justice Fund, and Newcastle’s own QR-SPF fund. Working with West End Foodbank in Newcastle alongside Northern Cultural Projects, the project recorded the experiences of food poverty and mutual aid in the North East of England, experiences that were happening in real time and not safely in the past. The research produced a wide range of outputs including exhibitions, youth theatre performances, sound art commissions, and workshops with foodbank clients, culminating in public installations, outreach activities, and a co-produced cookbook that amplified lived experiences of food poverty, as well as academic publications, a History and Policy briefing and a theatre production at Live Theatre Newcastle. The foodbank’s own staff, volunteers and service users shaped what questions were asked. What made this project distinctive was not the involvement of community members (that is common enough), but that the three organisations worked together on an agreement before the project started, guaranteeing a fair and equitable collaboration that extended beyond the project’s formal end date. The agreement included provision that any profits generated from creative outputs would be for the benefit of Newcastle West End Foodbank. The outputs were designed to be used by the communities whose histories they recorded, not only by academic audiences.
Mutual Aid (2020-21), two related projects funded by the Newcastle University Social Justice Fund and QR-SPF extended this model into an even more immediate context delivered with Northern Cultural Projects. The projects recorded the oral histories of mutual aid groups that had formed or expanded rapidly during the pandemic. They documented practices of community solidarity that were unfolding as the interviews happened. The distinctive feature was that the projects had activists as community researchers, people who were themselves engaged in mutual aid rather than researchers brought in from outside. This level of genuine embeddedness would not have been possible without the Collective’s relationship with communities already doing that work. Working at speed, with communities whose capacity was already stretched, the projects made remuneration of participants and shared ownership of recordings explicit conditions of the work from the outset. The experience directly informed the Collective’s later Paying Our Way project, which addressed remuneration for research participation as both an ethical and a practical question. If community partners are contributing knowledge and labour, not paying them is not a neutral act.
Living Deltas Hub (2019-2024) UKRI’s Global Challenges Research Fund supported this international project, which brought oral history and environmental justice research together across delta communities in multiple countries including Bangladesh, Vietnam, India and East Africa. The Voices sub-project, led by Siobhan Warrington with Graham Smith, recorded the memories of older people as active managers of natural resources and carriers of long historical knowledge of environmental change. It demonstrated that the gap between institutional research timelines and community timelines is not simply a practical inconvenience. In communities living with immediate environmental pressures, a researcher who arrives, collects, and leaves has a very different relationship to ‘participation’ than one who sustains a long-term connection. The Living Deltas research established methods for intergenerational oral history that took seriously both what older narrators knew and what younger participants were already experiencing. The project’s impact phase continues into 2026.
Positively Spoken (2021-2023), funded by the NLHF, was a youth-led participative oral history project gathering life stories from young people who had grown up living with HIV. It involved them in co-producing, collecting, archiving, and creatively sharing their own oral histories. It drew on the 30 years of experience accumulated by Chiva (a national UK voluntary agency supporting the health and wellbeing of children and young adults living with HIV, formerly titled the Children’s HIV Association) of working with children, young people, and their families around HIV and AIDS. Wendy Rickard and colleagues have argued that oral history has a structural bias toward older narrators. Most of what gets archived is the voice of people reflecting on long lives. Young people’s voices are systematically under-represented in oral history collections, a ‘sonic silence’ that mirrors wider exclusions in whose histories get preserved. Positively Spoken challenged this by making young people agents of the research process, not its subjects.
Byker Community Archive (2025-2028) funded by the NLHF and Karbon Homes is led by Silvie Fisch of Northern Cultural Projects with Sally Watson and Hannah James Louwerse as archivists. It is the Collective’s most sustained current experiment in community-led oral history. The archive is physically located on the Byker Estate in Newcastle. Rather than documenting a settled past community, it sets out to assemble memories that challenge official narratives, including histories of environmental activism, community arts, and resistance to redevelopment that remain unrecorded elsewhere. The archive is opening in 2026. The long-term plan is for it to be 100% community-led and community-run, with the project progressively upskilling residents to take full control. It is funded through a heritage funder institution,, but its design operates against the grain of what heritage funding typically produces, which is safe, consensual narratives that serve institutional purposes. The Byker project achieves its level of shared authority despite the funding architecture, not because of it.
What the evidence shows
Across these projects, several patterns are consistent.
Communities are most genuinely involved when research questions belong to them from the beginning. In Foodbank Histories, the question of how people in the North East understood and experienced food poverty was shaped by people who worked with and used foodbanks, not only by academics. In Byker, the question of which memories of the estate deserve preservation is answered, in the first instance, by residents.
Remuneration matters. The asymmetry between academic institutions that hold grants and community partners who contribute knowledge and labour is not resolved by goodwill. It requires structural mechanisms. That is, budgeted remuneration, shared ownership of outputs, and a continuing relationship rather than a single contractual transaction. Remuneration raises practical questions that funders currently leave unresolved. Individuals’ involvement must be acknowledged appropriately, and participation needs to lead to tangible benefits. But the implications for people receiving means-tested benefits are serious. Cash payments can affect benefit entitlements, and responsibility for navigating this is typically shifted by funders onto community partners. The result is that people often will not engage at all. Tailored advice, flexible and non-cash forms of reimbursement, and the administrative capacity to support this, need to be built into project budgets from the outset. These questions become particularly complex for projects involving children and young people, where payment for participation raises additional ethical and legal considerations that the sector is resistant to, but projects like Positively Spoken show they can be worked through.
Memory is not a collective possession. This is a theoretical point with practical consequences. Our oral history methodology insists that memory is individually formed, shaped by personal experience, by social relationships and material conditions, and by the contested historical narratives that circulate in any given time and place. What people share are the conditions of their remembering, not a single account of the past. In Foodbank Histories, participants who had all experienced food poverty in the same city produced accounts that diverged in analytically significant ways, shaped by their different relationships to the dominant narratives about poverty and dependence. In Living Deltas, older people in rural Vietnam and Bangladesh produced divergent accounts of environmental change shaped by gender, age, health, and relationship to land. In Positively Spoken, young people born across several countries produced accounts of a pandemic’s impact shaped by public health understanding, governmental resistance, activist zeal, and pharmaceutical inequalities. The differences between individual historical accounts are as important as the convergences.
The implication is that participation processes which treat communities as unified, internally consistent entities (which can be engaged, consulted, or empowered as a whole) are likely to miss what matters. Communities are traversed by contradictions of class, age, race, and gender. Any participation process raises questions about who represents the community, whose voice is amplified, and whose account of the past shapes the research agenda. These questions cannot be answered by selecting the appropriate point on a participation spectrum. They require ongoing analytical attention throughout the project and across projects.
Young people’s voices are notably scarce. As Positively Spoken demonstrated, meaningful youth participation requires challenging the archival conventions, ethical frameworks, and project timelines built to fit the norm of an older narrator. Planned future work addressing the low engagement of social housing tenants from ethnically diverse backgrounds points to a further such gap. The communities most affected by the decisions that research is supposed to inform are frequently least represented in its design and execution.
What needs to change
The following changes to funding architecture would make a material difference. None requires abandoning research quality standards. Several would strengthen them.
Budgeted remuneration for community partners should be a condition of funding, not an optional inclusion. The current norm, in which community knowledge is extracted and institutional researchers are paid, is not ethically neutral. Funders should require that grant budgets include explicit, adequate provision for community partner participation, including payment for time. They should also lobby for payment-for-involvement not to be classified as income for benefit purposes, a step that would remove one of the most significant structural barriers to genuine participation.
Agenda-setting should be separated from application design. The current model requires communities to participate within a framework established before they are formally involved. AHRC’s Mobilising Community Assets programme is one model with lived-experience voices brought into assessment at the funding decision stage, not only into project delivery. Ideally, communities should have a meaningful say in deciding which research is relevant in the first place. At the very least, they should be part of the framework at the initial fundraising stage. This approach should be extended across research councils, not treated as exceptional.
Ownership of outputs should default to narrators and communities. This means recordings, transcripts, and archival materials. Oral history methodology already treats this as an ethical standard. It should become a funding condition, with explicit long-term access provisions designed into projects from the outset rather than managed retrospectively against GDPR requirements and open licensing obligations. This applies equally to impact funding, which is currently also researcher-led. Even where community partners are formally involved, communities should have a meaningful say in how research findings are used, not only in how data is gathered.
The reduced capacity of community organisations under austerity must be factored into project design. Continuous cuts to local authority finances and services have significantly reduced the ability of community organisations to engage as equal partners in research. Funders should account for this explicitly, building resourcing for community partner participation, including time, travel, and administrative support, into standard grant timelines and budgets.
The UKRI Community Knowledge Fund’s recommendation that community-led research become ‘business as usual’ should be treated as a structural commitment, not a series of pilots. That means dedicated, sustained funding and an explicit recognition that the research and innovation system is impoverished when community knowledge is treated as an input to institutional goals rather than as a resource with its own claims on how research is designed and owned.
Conclusion
The participation problem in UK research funding has never been primarily a problem of practitioner technique or of selecting the right point on a spectrum. It is a historical problem of who holds the agenda. The funding flows via institutions. The intellectual property remains with institutions. The timeline is set by institutions. Communities participate within a framework they did not design, toward outputs they do not control. In the most meaningful sense, the research agenda itself (what questions get asked, what problems get defined as researchable) remains a decision made by institutions rather than by the communities most affected by the issues at stake.
The projects described above, from a foodbank in Newcastle’s West End to delta communities in Vietnam, from a housing estate in Byker to young people from around the world living with HIV in the UK, demonstrate that genuine shared authority, meaning real partnership in shaping what research does and who it serves, is achievable and that it requires working against the grain of how research is currently funded and structured, not simply along it.
The question facing funders, research institutions, and the communities who engage with them is not whether to use the language of collaboration. That language is already everywhere. The question is whether the structures that govern how research is funded, designed, and owned will be changed to make the language mean something.
