Developing a Women's Thought Collective methodology for health research: The roles and responsibilities of researchers in the reflexive co‐production of knowledge

Abstract Background Co‐produced research holds enormous value within the health sciences. Yet, there can be a heavy focus on what research participants think, do and know; while the researcher's responsibility to explore and re/work their own knowledge or praxis tends to escape from view. This is reflected in the limited use of co‐production to explore broad structural distributions of health and risk(s). We argue this missed opportunity has the potential to unfold as what Berlant calls a ‘cruel optimism’, where something desirable becomes an obstacle to flourishing and/or produces harm. We explore challenges to involving lay populations meaningfully in health research amidst a neoliberal cultural landscape that tends to responsibilise people with problems they cannot solve. Methods and Findings Drawing together principles from hermeneutic and feminist philosophy, we develop a novel methodology for co‐producing research about determinants of health and health risk (using a case study of alcohol consumption as an example) that centres on what researchers do, know and think during research: Women's Thought Collectives. Discussion Keeping the constructed nature of social systems—because they shape ideas of value, expertise and knowledge—in view during co‐produced research illuminates the potential for cruel optimisms within it. Such reflexive awareness carves out starting points for researchers to engage with how social hierarchies might (tacitly) operate during the co‐production of knowledge. Our work has broad utility for diverse population groups and provides important considerations around the roles and responsibilities for reflexive co‐production of knowledge at all levels of health systems. Patient or Public Contribution The development of these ideas was sparked by working with lay participants during the Women's Thought Collectives for Kristen Foley's doctoral research 2021–2023, but undertaken without their direct involvement—in accordance with the responsibilities of researchers in the reflexive co‐production of knowledge. Forthcoming publications will address the outcomes and processes of this work.


| INTRODUCTION
It is clear that co-produced research should happen on topics about health 1-3 and it has become an expected feature of many healthrelated research grant applications and funding awards. The benefits of co-produced research are thought to have value across three axes 4 : (1) it builds relationships with people that live with the issues being researched; (2) develops more robust knowledge to inform interventions that will in turn be more effective or acceptable and (3) upholds an ethical mandate to broaden research participation and knowledge development. Much is written about how variations of co-production can achieve these values in practice. 2,5,6 Yet, there remains more focus in this research on what lay populations think and do, rather than consequences for health researchers such as how their values or relationship to the social order must be problematised and changed through these encounters. 7 Co-produced research is desirable because it aims to explicate how power and health are unevenly spread across society (according to social hierarchies). Yet, researchers and the systems they work in (including various knowledge 'economies') are directly embedded in this unevenness. When the focus of co-production then is exclusively on what lay populations do (or do not do), it unfolds the potential for cruel optimsm-where something that is desirable because it will promote flourishing, instead, paradoxically prohibits or dissolves it. 8 When lay populations are only invited to participate in research in ways that fit research/er expectations and environments, it can risk situating them as problematic and dismiss their knowledge (ironically, the very potential for transformation)-while responsibilising them with the burden of solving complex multilevel research issues that require their insight in the first place. Reflexive awareness of the social systems in which health researchers are embedded, including the knowledges which shape perceptions of value, 9,10 is critical to guard against the unfolding of cruel optimism 8 during the coproduction of knowledge in health research.
This manuscript discusses a novel methodology that aimed to enhance opportunities for lay women from different walks of life to co-produce research about health risks, without burdening them with problems they cannot solve: Women's Thought Collectives. This methodology was inspired by journalist Ariel Gore's 11 project to critique happiness through assembling a 'council of experts' who informed and reviewed her thinking about western cultural obsessions with happiness. She invited women to think alongside her about how happiness works (or not), when, how, and for whom-akin to Ahmed's 12 exploration of how happiness gets stuck to different objects and economies to various ends. The council helped collect data and shared their ideas (via keeping a journal), then were invited to contest, extend and corroborate her ideas about happiness over a period of 2 years. This type of co-production methodology could expand investigator knowledge over time while also giving participants space and time to adjust and inflect interpretations. Her book included limited methodological details, so rather than following her approach, we attempted to develop our own, although loosely based on her council of experts.
The flipped positioning of this approach places the researcher's learning, thinking and reflections in view and open to exploration by research participants. We explored how these ideas sat with key principles from hermeneutic and feminist philosophy, then developed a novel methodology for Kristen's doctoral research, which focused on illuminating the commercial determinants of alcohol consumption for Australian women in midlife. The purpose of our paper is fourfold.
1. Reflect on the flipped positioning of a council of experts methodology, which brings into view what researchers think, do and know (reflexivity) when co-producing research in health settings 2. Explore how contemporary feminist and hermeneutic philosophical thought might add to, or refine, methodologies for coproducing health research 3. Outline the Women's Thought Collective methodology which will be of use and interest to researchers in diverse health settings 4. Discuss how to pursue co-production research that does elevate the role of lay populations in research but does not reconstitute a cruel optimism by eliding roles and responsibilities for researchers during the co-production of knowledge.
We, therefore, distill key ideas from hermeneutic and feminist philosophy, then reflect on critical ethical considerations for health researchers to keep in mind when co-producing research in neoliberal settings-a cultural landscape that tends to reconstitute the problem and responsibility for solution 8 within groups and individuals already experiencing oppression. We hope that this structure will showcase the flow of ideas from the macro context of philosophy (presented in the methods) into the specific methodology developed (outlined in the findings) and residual questions (raised in the discussion) for co-produced research about upstream distributions of health resources and health risks.
We weigh co-produced research against its potential to enable different groups to exert control over knowledge as a resource and determinant of their lives and living conditions. 13 This focus on power and social structure illustrates the democratic and politicallymotivated nature of co-produced research and demarcates it from 'cobiquity' where broad and diffuse research partnerships might constitute co-production. 14 Aligned with the 'values and principles' typology of co-produced research, differences between researchers and lay experts are democratised, and the co-production process is treated as evolving rather than a rigid or fixed concept. 2 We recognise there are diverse histories and disciplines which inform the words used to describe involved participants in co-produced research, and that rather than these terms (i.e. lay, expert, research participant) having objective meanings we argue they are made meaningful through their contextual relationship to the researcher/s and research process/es. In the co-production sphere, a key feature of this contextuality can include whether participants are paid and how (much), as well as what is possible in the cramped spaces of academia. 15 While these terms matter, we use several within this manuscript but do not see this to contradict our central argument: that keeping the contextuality in view enables exploration of how health researchers, lay knowledges and co-production initiatives can be situated around socio-political structures including historical trajectories and the distribution of power and resources 16 -the key undercurrent of co-production research. Our research context is a useful area for this, in exploring alcohol consumption (a health risk in a healthiest society that carries sociomoral dangers 17  Early arguments for involving lay populations in health research responded to the recognition that they hold expertise that will enhance understanding of health and its social genesis. 1,24 The distinct, plural and free-form ideas from lived experience are now seen as critical to realising epistemic justice within democratic methods, 25 yet this sits in tension with ingrained logics of natural science which pursues the exclusion of participants and their subjectivities. 26 It may also irritate tightly wound socialisation(s) around the authority and expertise of health professionals and systems, to the extent that both lay experts and health researchers tend to misinterpret and downplay what and how to lay populations can contribute to health research. 9 These knowledge contests are scaled up during power-focused co-production methods, 5 because they explicitly seek to elevate the power participants have in research processes and even help to reimagine 27 and reorganise 28 power dynamics. This radical politic likely underpins why turbulence will unfold if co-production is done 'properly'. 29 Researchers are directly embedded in these dynamics, which is why reflexive practices are so central to 30,31 and entangled within knowledge systems, prompting our focus on epistemic 9 and ontoepistemological reflexivity. 32 The perceived value of knowledge moulds subjectivities, 10 and then influences what research participants think to disclose and what health researchers think to ask during research encounters. 9 Unasked questions and undisclosed knowledge are key examples of the need for reflexivity and showcase its relationality with epistemic privilege. 33 This is crucial within co-produced health research, where despite its political ambitions, lay participants are less commonly involved in healthcare planning, design, delivery and improvement 34 or the review of systems innovation. 35 Instead, they are more typically represented in technical aspects 36 of research more acceptable in neoliberal settings but still branded as co-production, such as activities like usability testing that can be more passive and defined because they have a preset agenda. 37 Concomitantly, theoretical guidance about how to involve lay populations in co-analysis of data has been obscure 38 with some suggestions that these processes need 'downgrading' 39 or risk being 'dirtied' and 'stretched' during co-production. 6 Lurking doubts about the capabilities of citizens to engage in co-produced research 40 alludes to lingering conflicts about whose knowledge is legitimate and in what settings. How knowledge is valued further influences what is viewed as a valuable output from analytical or interpretive processes-further showcasing why the histories that give rise to contemporary configurations of knowledge are manifestations of power that enfold co-produced research. 16 Broadly, the potential role-and value-of involving lay populations at macrostructural levels of in/equality remains undertheorised and underachieved. 16 We suggest that there are two layers to this problem. First, the value of lay expertise might be misunderstood (a reasoning error) and therefore efforts are not invested in developing researcher critical consciousness towards epistemic and political contests. The fact that much of what citizen scientists produce is ignored by scientists 41 is likely symptomatic of this, in addition to unwritten rules about whose expertise matters and whose does not. 42 Second, the short-term and scarce nature of health-related funding and implementation cycles contracts the scope for such work (an execution error) because it does not afford the time and care needed for collaboration 43 that is sustained, open-ended, and able to transform civic arenas. 44 Both of these co-production 'errors' engender layers of cruel optimism.
The reasoning error continues to situate lay participants as the problem, in a position of deficiency where they lack knowledge (per Foucault,45 docile bodies that are positioned as vacuous vessels that need filling with new information)-rather than the knowledge they bring being examined by researchers to break cycles of 'circulating reference'. 46 It then follows that they cannot bring air into closed political systems 15 nor disrupt the quest for monophonic patient voice in health settings 47 because the ways in which the knowledge they bring can be valued is already prescribed by dominant cultural systems that privileges some knowledge above others. 10 The execution error occurs when the value of lay participants is understood (at least to some level) but the expectations, needs and practices of co-production research spill over what health researchers can fulfil-resulting in disengagement (and/or moral injury on behalf of the researcher and participants)-as well as cramping of what can be achieved during the process, 15 which turns the gaze again to what lay participants do, think and know that can be made available for critique in a short-term window.
Considering that epistemic precarity frequently coincides with material precarity, 48 the cruel optimism of co-production can be twofold-rhetorically and materially engaging with lay populations who experience disadvantage because of this perceived vulnerability while dissolving potential benefit from their involvement, because their insight is not viewed as 'knowledge' nor enabled to challenge the 'status quo' of knowledge production systems.

| Research context and prelude to women's thought collectives
In thinking about how to transition the snippets of Ariel Gore's council of experts method detailed in her book (as we were unable to connect with her in-person), she talks about travelling to meet different women she knew or contacted because they had expertise in exploring the question of happiness, drawn either from lay or professional experience. We liken her approach to purposive sampling in an emergent research design, 49 in which Kristen's PhD would be located, to explore how social class shapes alcohol consumption for women in midlife using interview methods 52 . To complement these insights, she designed and received funding for a companion study to explore the social context of alcohol consumption and experiences of marketing to women to increase profit margins and market share. Given that so much of targeted marketing is specifically tailored and delivered to particular groups and individuals to increase purchasing and consumption, involving the lay population seemed unavoidable and completely necessary.
The timely and evocative reading of Gore's 11 book, however, sparked creativity about how research participants could be involved in co-analysing and co-interpreting research findings about them-coproducing knowledge about issues that concern them while also placing them in a position to refine the evolving meanings and interpretations made from the research. There were few details, however, about the liberation psychology forum she held and it was unclear how her ideas were refined by the presence of her council of experts (hence our reason for contacting her). This move from what was said to her, her reflecting over what it meant and then pulling into a narrative-makes stark what is done to knowledge between its 'provision' by research participants and subsequent 'development' by researcher(s) during analysis and interpretation, typically without recourse to participants (member-checking being an exception, although still subject to the researcher terms of conduct). Something Gore captures well is her wrestling with these ideas, both in granular detail and how it changes her ideas about what happiness is. She journals about trying to live differently according to the knowledge she adopts. In using her work to spark intellectual creativity, 53 we set out to explore how a council of experts might resonate with ideas in the social science literature about co-producing research regarding structural distributions of consumption, health, risk and resources.

| Developing insights from contemporary feminist and hermeneutic philosophy
We summarise here the learning from (1) hermeneutic and (2) feminist philosophy, which were read in response to Gore's 11 method to ground it in the contemporary thinking about knowledge, truth and method-and infused a co-production methodology to create space for researcher reflexivity under the guidance of co-analysis and cointerpretation of lay participants.

| Hermeneutic philosophy
Philosophical hermeneutics explores in-depth what it means to 'know', 'understand' and 'interpret'. 54 Centrally, hermeneutic thinkers are convinced that our being and essence as humans is to interpret (following Heidegger 55 ) and that we are what we know, which comes from understanding 56    Interested women were encouraged to contact the researcher.
The following script was used to convey to them why participating for such a long period of time was useful, per Box 1: We distributed this material around university networks (i.e., flyers) and relevant community groups (i.e., notice boards and by contacting specific groups). Snowball sampling was also used, where motivated participants contacted others to suggest participation. Those who participated were provided reimbursement for their time and participation. We aimed for as much heterogeneity in the sample as possible, including in living circumstances (i.e., single/married/divorced, having had children or not, having elderly parents, living alone or with others, in metropolitan or regional areas) and social class (a mixture of economic, social and cultural capitals as measured by a validated tool 85 ; drinking level (light, medium or heavy as measured by the AUDIT-C tool) and age range between 45 and 64.
We did not target participation from any Aboriginal or Torres Strait Islander communities, recognising that Australia's colonial legacy shapes alcohol practices in ways that differ for Caucasian women in midlife. 86 No women communicated identifying as Aboriginal or Torres Strait Islander. We recognise the paradoxical nature of arguing that disadvantage accumulates at layers of intersections and the knowledge they bring is critical to unsettle entrenched hierarchies and power dynamics 27 but not specifically targeting groups likely to have diverse ontoepistemological positionings. 33 Following our logic of widening participation in knowledge development, draft outputs will be circulated to a wide range of audiences, that accounts for groups (i.e., organisations or champions) that may not have been represented in the Collective and could be used to develop future partnerships if desired by these groups. This will create further entry points to co-producing (or contesting) knowledge, and aligns with Erikainen et al. 79 support for widening the way in which 'research engagements' are conceptualised and the structural conditions under which they occur.

| Undertaking the research
The Collectives were shown Figure 2 at the first meeting to try and offset any feelings their knowledge was not valuable and might not influence the trajectory of the project-by way of blank spaces being included in the plan, contingent on forthcoming discussions. We also asserted that the purpose of the Collective was not to come to a consensus about an issue, because such a task was unlikely and intractable-rather, to unearth a range of different perspectives and how these all might help to understand the commercial determinants of alcohol consumption for Australian women in midlife. This openness to dissent and difference is important for co-produced research so that perspectives are not marginalised nor recast through hegemonic values needed to present uniformity or 'one voice', 87 ideas contested by democratic methods which highlight the problematic nature of conflict-free consensus such as agonistic pluralism. 88   nonresponse would be interpreted as them being too busy at the time, and would be re-contacted again in a few weeks-but that they could cease their participation if they asked via return email.
Participant preferences were followed regarding communication, and recognising that some preferred phone contact, emergent thoughts and ideas were discussed with participants via phone rather than in addition to email.

| CRITICAL CONSIDERATIONS FOR HEALTH RESEARCHERS DURING THE CO-PRODUCTION OF RESEARCH IN NEOLIBERAL SETTINGS
Clear synergies exist between hermeneutic and feminist praxis: in both, the researcher is embedded and entangled in knowledge production, to the extent that they cannot be separated from it. 33,78 Health researchers become subjects of inquiry, rather than objects/ objective parties in inquiry, messying the power dynamics of traditional research 89  working knowledge of the philosophy of co-production methodologies as well as the research methods they might use, will be critical during editorial and peer-review processes.
We agree that co-production requires agreement on principles and values rather than definitions, 2 and suggest that when principles that illuminate the democratic and political undercurrents necessary for productive co-production research are outlined, 14 reasoning and execution errors may be lessened. The value of genuinely involving lay participants in research must be coupled with a wariness of coproduction as a new buzzword of neoliberalism 34,91 that can engender complicity in structural inequalities 92 during co-produced research that apparates the researcher and the knowledge systems in which they work within out of the picture. Centring the difficulty of this work may help to further guard against these errors in coproduced research and glossing over the dislocation and disorientation during forms of public deliberation. 93 Our work adds to crucial reflection about co-produced research which can scapegoat lay participants during 'inefficient' co-production research 89 and risk a twofold cruel optimism, which not only impedes flourishing but also diminishes it. 8 By contrast, we present a methodology that situates the researcher as responsible for cohering co-production methods that take form around socio-political structures and the distribution of power and resources, 16