Priority and Prejudice : The Epistemology of Salience and Attention

What is salient to us and what we attend to play a fundamental role in shaping how we perceive, think about, and act in the world. Salience and attention shape our mental lives in ways that have profound epistemic significance, determining how we gather evidence, what sorts of inquiries we undertake, and what we do with the beliefs we form as the result of that. And yet they haven't traditionally fallen within the purview of epistemology. We have a lacuna in our epistemic resources: What should be salient to us? What should we attend to? How should we evaluate how we prioritize and select information? We need a framework for evaluating salience and attention from a distinctively epistemic perspective. This book proposes a novel construct, salience structures, which describe how our informational landscape is contoured by a range of causes, reflecting what is salient to us, and what we are likely to attend to. It offers an evaluative framework for salience structures, describing the epistemic norms which apply to them. It applies that framework to a range of phenomena including ignorance and inquiry, showing how it helps us better understand the nature of prejudice and the role of search engines in our lives. The book develops several themes. Firstly, the framework provides new ways of engaging with 'negative' epistemology: the project of understanding and evaluating the negative space of information we deprioritize and forget, attitudes we don't form, and questions we fail to ask. Secondly, the domain of the epistemic is richer and more various than we have sometimes allowed. But it also needs to be kept in its place: we value many things that do not fall within its aegis. Finally, the epistemic role of salience structures is in competition with the many practical purposes they must serve for us. How deep that tension runs depends on facts about our social organization. We should strive to build just societies that minimize the practical costs of epistemically optim