Difficulties in distinguishing genuine reports of experience from confabulations have implications for theories of consciousness. Through an analysis of out-of-body experiences (OBEs), the author argues that no existing or possible evidence could determine whether variations in reports reflect actual differences in experience or are post-hoc confabulations. How evidence is interpreted depends on prior theoretical commitments. Vehicle theories of consciousness, which cannot accommodate confabulations, allow only the hypothesis that reports reflect elaborate phenomenology. Because it is impossible to know that reports are not confabulations, vehicle theories are insufficient to explain OBEs.
The sense of agency—the feeling of being the cause of one's own actions—is typically explained by identifying the cognitive mechanism that produces it. This paper argues that an unconscious representation of one's own agency can also exist and play a different role in cognition. The author uses the "occasionality problem" to show that distinguishing conscious from unconscious agency representations has theoretical benefits, framing this as an instance of Dennett's "Hard Question": once a representation is elicited, how is it used differently when the subject is aware of it versus when they are not? Two single-function hypotheses—that awareness involves attention or verbal report—are considered and rejected. Multifunction accounts like global workspace theory remain plausible, as does a non-functionalist alternative: the difference may lie in the representation's nature, not its use.