> Since conventional physics is sufficient to give (at least in
> principle) a complete description of the human brain, a partial
> ordering on the set of all possible observer moments S can be defined as
> First we choose an arbitrary brain B.
> If x1 and x2 are elements of S, then x1 < x2 iff both x1 and x2 can be
> experienced by B and an initial condition specifying the entire state of
> the (conventional) universe including that of B exists such that x1 is
> experienced by B at some time t1 and x2 is experienced at some time t2
> and t1 < t2.
Your definition of a brain B cannot be a specific atomic configuration,
because of course the brain state changes depending on what it is
experiencing. So B must include an entire set of possible configurations.
But we know from biology that as brains experience things, their fine
structure changes. The synapses where neurons come together change,
the cells themselves change, the blood vessels probably change, etc.
So I think you will have trouble clearly defining B in a biologically
reasonable or plausible way, to mean all the things B could ever
You're also bringing in the notion of time, which raises many problems
of its own.
Even if you could do all this, isn't it possible that x1 < x2 and also x2
< x1. You haven't proven that it isn't. Biologically we could imagine
a brain going into state x1, and then forgetting about it, and going
into state x2; and we could also imagine the reverse. For example,
it might be possible for a brain to have dream D1 followed by dream D2,
or it could have had D2 followed by D1.
All in all I don't think this is a very promising approach.