Abstractions and Idealisations in Epistemic Logic Ottilia Kasbergen Abstract: In this thesis we investigate to what extent the many abstractions and idealisations that epistemic logical models make constitute a problem towards the practice of epistemic logic. To this end we look at examples of abstractions and idealisations in epistemic logic, in particular focusing on the problem of logical omniscience. We find that the acceptability of different idealisations depends on the research context of the model in question. However, in epistemic logical literature we rarely find discussions of aims, success criteria and intended applications. Therefore, we asked ten epistemic logicians to share their views on the practice of epistemic logic. As their views diverged greatly, we propose a categorisation of the field of epistemic logic into four research agendas. These agendas will function as a starting point for a discussion of the advantages and disadvantages of idealisations in epistemic logic. As idealisations are inherent to modelling, and as such not an issue to epistemic logic only, we study modelling and idealisations in the sciences more broadly as well. We find that usage of the terms ‘abstraction’ and ‘idealisation’ is not standardised, therefore we redefine these concepts for a discussion of the practice of epistemic logic. We find that it is of importance that epistemic logicians are explicit about both their aims or intended applications, as well as about the idealisations that their models make. Being explicit about the research context may enhance an adequate conception of the extent of use of the epistemic logical models, and may enhance interdisciplinary research.