N suggests - rightly I think - that trying to get from a delusional to a sane experience can be like trying to solve a koan. Let me conflate Wittgenstein and Zen a little (as Canfield and Gudmunsen and Huntingdon have taught us to do (cf also, though, Magliola on Derrida and Nagarjuna)) and suggest that what is needed in relation to the koan is not a solution but a dissolution. Within the terms of reference that the koan sets there is no solution; the dissolution comes from rejecting its tacit premises.
So, to take one koan with which I'm a little familiar: A master says (something a bit like this) to his students: 'If you say this is a stick, I will hit you with it. If you say it is not a stick, I will hit you with it. What do you say?' Predictably, all the students get hit right and proper. There is no solution in the terms offered. But, the further thought goes, why should we buy into this bit of demandingness? What we - the novices - do in so buying is accept the presupposition that we must answer the master's puzzles. We accept the power dynamic, we maintain - by remaining caught within - the bubble of the transference relationship. Better by far to grab the stick and give the old goat a thwack, or just chuck it away. That would show an emancipation from the terms of the argument. (I once had a similar experience on a training week at the Tavistock Clinic: the group leader (twice per day, 90 minutes per session, for 5 days) acted like a complete bastard, sighing and hurrumphing and interrupting and generally being very 'rude', throughout all the sessions. It wasn't until the end of two days of getting furious with him, or profoundly doubting whether my experience could really be veridical, that I realised instead his real lesson, which was: Why on earth would I be giving him, a complete stranger, such power to affect how I am left feeling during and after the session?)
If such a dissolutive rather than solution-based deconstruction of the koan is (in the analogical context) a matter of using thought to reconnect with reality after being caught within the purview of a phantasy/delusion, then I will happily accept this use of 'thought'. (What I meant in the post by (the need to abandon) 'thinking one's way to reality' really only referred to attempts to solve, rather than dissolve, the delusional experience - or, by analogy again, to try to disprove the justifiability of scepticism whilst accepting its tacit premises about the world-independence of the mind. When it comes to a thoughtful dissolution or deconstruction, however, then: bring it on!)
So to put better what I was previously saying: it is not that I want to urge that we would be misguided to provide ourselves with "reasons for returning from the study to the billiards table". The main such reason is the fruitlessness of attempting to arrive at veridical perceptions by supplementing inner experience with (epistemology-provided) causal or justificatory connections to reality. But start where the sceptic starts and remain there (despite all the thinking in the world); question his tacit non-disjunctivist premises, and (only then) abjure his problematic. (N's playful embracing/negating of the epistemological enterprise (toward the end of the 'part ii' post) thus seems to me de trop.) Hume found himself untroubled by his sceptical doubts when engaged in the game. Augustine found himself untroubled about the nature of time when taking it or giving it to someone. What got them into their respective puzzles was not, I submit, anything about time or experience per se, but rather, time or experience already refracted through misbegotten prior assumptions about the way that the respective language games work.
Where does this leave the discussion of schizophrenic experience? Here I want to clarify something I wrote before and to register what I think is a disagreement with N: Whilst I think it is entirely wrong to blithely accept that we know what is meant when someone tells us, or we tell ourselves, that we have had the experience or thought that (e.g.) nothing is real, I really can't accept that what is actually happening is that the someone in question is having a particular experience that they merely happen to thematise (narrativize) in this manner. As N says, the question then remains as to why it is so entirely natural for the someone to thematise it thus. (And here we are back in the territory of my earlier 'talking teapots' post.) What I would rather urge is that the content of the experience is given precisely through that avowal of it which makes use of the notion of 'unreality'.
But yet, or so it seems to me, this need not cause us to become complacent in assuming that we (could) know (that there could be such a thing as knowing) what it would mean to talk of an experience (as) of everything being real or unreal. We know what it is for a dollar or a smile or a problem to be real; do we really - really - understand what it would mean for the world to be real or not? Isn't "the world" precisely that within which discriminations of the reality of this or that are made? Isn't the world a transcendental precondition for necessarily localised talk of 'reality' or 'unreality'? Isn't it to 'sublime the logic of our language' - to ignore the necessary background context within which the concept of 'real' operates and to try to use it about the context itself - to carry on in this way?
One aspect of the model I would, then, use for making sense of the twin utter propriety and yet strict meaninglessness (although is it really right to call it 'meaningless'? I have a feeling that doesn't quite make the point I'm aiming for) of talk of depersonalisation or derealisation is secondary sense. We are drawn to using the terms in the ways we do by the experiences we have. That we are so drawn is not something which could be justified by reference to the content of the experiences; rather, our being-thus-drawn is the 'beginning, and not the end, of this [particular] language-game'. It is the condition of our finding it intelligible. Wednesday is fat and Tuesday is thin; happiness is up and sadness is down; sopranos are high and basses are low; anger is red and envy is green; and so on. We know how to use the relevant terms in their everyday deployments, and then we redeploy them spontaneously and groundlessly to express or report such further experiences and phenomena. I know how to talk about unreal diamonds, and then I find myself using the same word to characterise the very being of others. The logic of my language is sublimed (I cannot - however hard I try, with however much sincerity - mean it in the way I normally mean 'real'). But, damn it, that really is the only way to describe the experience!