I just posted a new paper in which I argue against the total evidence requirement on knowledge. The abstract reads as follows:
- A requirement on rational belief frequently invoked in epistemology and inductive logic is the total evidence requirement (TER). This requirement asks one to consult all evidence when making a determination about what one believes or the degree of confirmation to assign to a hypothesis. Despite the wide-spread use of the requirement there are many problems with it. After explaining the requirement in section 1 of this paper I motivate the requirement in section 2. In section 3, I highlight problems with successive interpretations of the requirement. This applies pressure to abandon TER or revise it. In section 4, I create the proportional evidence requirement (PER). This requirement revises the notion of what constitutes relevant evidence by making the notion proportional to the weight of evidence for a given hypothesis. After formulating two key principles behind PER I realize that one of the principles may not be an improvement over the commitments of TER. So, I revise one of the principles in PER to avoid such problems and create a requirement on evidence that is truly an upgrade over TER. I conclude this paper in section 5 by summarizing and indicating directions for future research.
Update: The paper is now down for revision.