Traditionally, epistemic rationality has been distinguished from practical rationality. In the view of many thinkers, to be epistemically rational, beliefs should approximate a number of ideals: coherence with the canons of logic, axioms of probability, and rules of evidential support. For, it is only then that beliefs are deserving of our trust. These ideals are often required to play a role in numerous applications: in our confidence in climate research, medical judgments based on randomized control trials, or in everyday risk-taking that accepts uncertainty but should differ from acting under ignorance. However, this conception of epistemic rationality has been critically discussed in many ways among both philosophers and scientists. The conference will address three central and related challenges.
(1) The challenge of fallibility
Our fallibility extends to the most deeply entrenched of our beliefs, such as logical beliefs. Acknowledgement of our fallibility seems to make it rational to be less than certain. Yet, on the standard Bayesian conception of it, probabilistic coherence requires that reasoners be certain about logical truths. How is this conflict of demands to be resolved? In one’s attempt to be epistemically rational, one needs to properly gauge the force of one’s evidence. What effects should fallibilism have for the assignment of levels of credence to propositions in the light of one’s evidence? Should we never assign maximal credence? But if we only ever assign non-extremal credences, this might still mean over- or underrating the force of our evidence. Finally, it is far from clear whether there can always be evidence for the higher-order claim that one has properly assessed what force one’s evidence has: at some point one may have to take it on trust that one has done so. But if so, where should that point be located? What is it that our credences should aim to match – objective chances (where these are defined), relevant actual frequencies, or objective evidential probabilities?
(2) The challenge of intractability
Many cognitive tasks – such as the traveling salesman problem, or distributing assets in the stock market optimally – are so intractable that the ideal of a reasoner with infinite resources who always follows canons of logic or probability looks as if it set the bar too high. This instills doubts about the centrality of these theories in accounts of epistemic rationality. If reasoning tasks are too demanding for optimal information search and precise assignments of credences, does that imply – as defenders of “bounded rationality” maintain – that we must resort to “fast and frugal” heuristics? But what makes such heuristics epistemically rational? How should we assess their rationality if not against e.g. success rates that are measured probabilistically? Also, what are the borders of tractability, and are they fixed once and for all?
(3) The challenge of instrumentalism
It can be argued that truth is not the only goal of rational inquiry: belief systems should also satisfy other epistemic goals (e.g. simplicity, fruitfulness), and even practical needs. Making one’s beliefs depend exclusively on evidential support, or demanding that we ought to follow the principle of total evidence, may impede achievement of these further goals. Thus, many claim that epistemic rationality is not sui generis, but reducible to instrumental or practical rationality. Can an adequate account of epistemic rationality integrate epistemic and non-epistemic goals? What would be the consequences for attitudes of trust and distrust concerning knowledge claims, e.g. in science and science-based policy making, if we accepted instrumentalism?
Can the traditional conception of epistemic rationality be rescued? What modifications are necessary, what alternatives must be considered? There are some answers to these questions in the recent literature. But there is no consensus about them, and more investigation is needed. The conference brings together leading philosophers and cognitive as well as social scientists to discuss anew the concept of epistemic rationality, while keeping in view its implications for science and society.