Something that calls out for research

Another puzzling thing, and one that you'd think might repay research, is the fact that when S is having her conversations with imagined people (we could call them hallucinations though I don't think she usually sees them) or just talking to herself her language is often lucid, coherent and complex. Yet when she is more 'herself' and trying to join in a conversation she often struggles to express herself and people listening usually struggle to make much sense of it.

Of course, some of it may be to do with familiarity and repetition - many of her conversations are constantly recycled - but new variations and new topics do come along and she is equally sure-footed with them. It may also be that in some way she feels more in control - she is after all she is making up both sides of the conversation. Another strange and interesting thing is that sometimes there's a silence whilst she's listening to the 'other person' but sometimes she does both sides of the conversation.

I'm wondering if it's ever occurred to people to study this strange phenomenon?