The process of defining the notion and applying it to different cognitions has just started, and we still have some problems to solve, but I thought I would update you on my own progress with it (and Ema will do the same in posts to follow this one). My initial questions were the following: In what circumstances do delusional beliefs, distorted memories, and confabulatory explanations contribute to the acquisition and preservation of relevant truths? Do delusional beliefs, distorted memories, and confabulatory explanations have benefits that are genuinely epistemic? Are people epistemically blameworthy for having "imperfect cognitions"? What are the consequences of acknowledging that delusional beliefs, distorted memories, and confabulatory explanations can be epistemically advantageous?
One of the main objectives of the project is to develop an account of epistemic innocence for imperfect cognitions. Ideally, we would have cognitions that satisfy norms of truth and accuracy and that are supported by, and responsive to, the evidence available to us, as well as fostering the acquisition and preservation of other relevant true beliefs. But we have limited cognitive capacities, and imperfect cognitions that are false or inaccurate and badly supported by, or irresponsive to, the evidence are a common occurrence (e.g., delusional beliefs, distorted memories, confabulatory explanations).
I want to explore the possibility that at least some imperfect cognitions are epistemically innocent, where the notion of innocence captures the fact that for a given agent at a given time it is epistemically better to have such cognitions than not to have them, even if they fall short of central epistemic norms. The exact formulation of conditions for epistemic innocence will vary depending on one's epistemological commitments and the type of cognition to be considered.
My current sense of the notion is that a cognition is innocent if the following two conditions are met:
1. Epistemic Benefit. The cognition delivers some significant epistemic benefit to a given subject at a given time, that is, it contributes to the acquisition, retention or use of relevant true beliefs.
2. No Relevant Alternatives. Alternative cognitions to the imperfect cognition are either unavailable or fail to deliver the same epistemic benefits as the imperfect cognition to that subject at that time.
The example of potential epistemic benefit I started with was that of a distorted memory in the context of dementia. Suppose a woman remembers that yesterday she went on a trip with her parents, when the trip occurred many years ago, when she was a child, and her parents have been dead for some time. This memory can help her retain a sense of herself is in the context of fading autobiographical memories.
Different notions and degrees of unavailability will apply to alternative cognitions (Ema will expand on this). In general terms, there may be no genuine alternative to an imperfect cognition, because (a) information that would lend support to a different, more accurate, cognition is opaque to introspection, not open to investigation, irretrievable, or blocked for motivational reasons; or (b) the alternative cognition could be strictly speaking available, but it would not carry the same epistemic benefit as the imperfect cognition.
Much more work will need to be done before a proper account of epistemic innocence can be developed and applied, but the contributions of the Imperfect Cognitions network members on this blog have already been a great source of knowledge and inspiration for us.