Testing Epistemic Injustice

Abstrakt

This work builds on the trivial observation that everyone is not trusted equally. One’s gender, ethnic group, occupation etc. will affect how one’s information is believed and interpreted by others. We begin by reviewing past approaches to reliability and epistemic injustice, and the factors which affect how one’s reliability is evaluated by others in discourse. We then discuss recent experimental results which show that the linguistic manipulation of gender seems  to affect the strategies with which the source’s reliability is evaluated. We argue that masculine sources benefit from more charitable assumptions than feminine ones. To support this claim, we present the results of a fine-grained categorization task. The results of this task seem to support our claim about charity, i.e. that a masculine source can more easily claim competence about a topic categorized as feminine, whereas the converse appears less true.