AAP 2026 CONFERENCE

In philosophical literature on dementia, a key question is how to determine what is in the best interest of people with dementia. Two opposing views exist: one appeals to a person's former values, while the other suggests that past values matter little, focusing instead on current perspectives. Franklin Hall recently proposed a third alternative: the "revision model”. This model holds that we only consider past values if they have been revised and the person is answerable for why. I make a case for modifying Hall’s answerability requirement. I draw a distinction between a direct answer and a demonstrable answer. I argue that even when the requirements of answerability are not directly met, requirements can still be met indirectly: we may identify answers through epistemic resources and contextual clues available to us. In some cases, where epistemic access is limited and the person cannot offer a direct response, all we have is the absence of evidence that the answerability requirement has been met, not evidence that it has not been met. The upshot is that, in a wider range of cases, people with dementia may still meet the requirements of answerability, or at least, it may remain undetermined whether those requirements are unmet.