I claim that it is in the limiting case possible for a being to lack any significant sense of itself as an agent, and as something that has a personality, and as somehting that has long-term persistence, and still experience itself as a self or mental subject at a given time.
Many disagree … I won't say more about it now, except to note that there are recognized pathologies that can involve the weakening or loss of all three of these aspects of ordinary human self-experience –aboulia, apraxia, depersonalization, passivity phenomena in schizophrenia, autism, and loss of time sense…. [p.46]
I want now to consider the ontological question:… Do there in fact exist (1) subjects of experience that are (4) single (3) mental (2) things during any gap-free period of experience, whether or not they can persist across gaps in experience?
I think there are. … I will call them SESMETs (Subjects of Experience that are Single MEntal Things). I think that gap-free periods of experience are always short in the human case… So I think that many SESMETs exist in the case of a human being. In all essentials, in fact, I agree with William James… He holds that "the same brain may subserve many conscious selves" that are entirely distinct —numerically distinct— substances. […] On this view the apparent continuity of our conscious experience […] derives from the fact that SESMETs "appropriate" […] the experiential content of their predecessors´ experiences. They do so in a way that is entirely unsurprising in sofar as they arise […] from brain conditions that have considerable similarity from moment to moment even as they change.
I understand that the reference to the brain and to the SESMET as a "thing" could not be endorsed by modern and classical Buddhists. But what about the rest of the claim? Is this more similar than my preceding, Humean proposal, to the Pramāṇavādin conception of the self? (I refer to the Pramāṇavādins since they developed a consistent philosophy of Buddhism).
As far as I am concerned, I share the doubt Strawson attributes to a hypothetical reader (p.48): even if they were true, these SESMETs seem to be useless as they are too far away from our experience of ourselves.