Citizen sciences are often 'extractive' in that they involve citizens in order to collect data useful to science. We consider citizen science more broadly, with a research-action stance, and aiming to collect data that is useful not only to scientists, but also to policymakers, politicians and citizens themselves. In this perspective, the question of objectivity is placed at the heart of a dialogue between different actors coming from different scientific disciplines, but also from other areas of society. Here, objectivity must guarantee a decision-making process that is 'enlightened' (i.e. informed) and serves the common good. The communication proposed here is based on several experiences of participatory research projects in the field of water led by the authors and pertaining to this paradigm. These experiences rely on associating citizens and other actors in the subsequent decision-making steps of a plan, project or program: from diagnosis to implementation, through the identification of objectives, planning, foresight and the choice of actions (CoOPLAaGE approach). Actors are notably involved in the framing of the process itself as well as its monitoring and evaluation. In other words, they define who will be involved when and with what role and what data they will need when and in what format. In this sense, they are invited to discuss the role of scientific expertise in the decision-making process and the data they need to participate in an informed way. This reflection is anchored in the fields of research on the engineering of participation as well as on policy analytics. Two experiences will be highlighted: the SPARE project in Drôme (2015-2018) and the PACTE project (2018-2022) in Tunisia. In Drôme, the aim was to allow citizens to make proposals ahead of the revision of the Water Management Plan (SAGE). Citizens have thus participated in the engineering of the participatory process itself, with the underlying assumption of their empowerment and increased ownership of the process. In fact, if this hypothesis has been verified, it has also led citizens to question the very expertise of researchers on participation and to require the extreme transparency of all data produced and analyzed. In Tunisia, conversely, the sharing of the expectations of the various actors vis-à-vis the production of knowledge related to the participatory process revealed a refocusing on the scientific authority underpinned by a vision of necessary partition of roles and skills. In conclusion, rather than questioning 'what form the implication of citizens should take in order to optimize the epistemological conditions of scientific objectivity ?', this communication proposes to return this question and ask: 'what form should citizen sciences take in order to optimize the epistemological conditions of an enlightened decision?'.