Your question is one that I have been trying to address for a long time.
Since we have to consider the notion that an observer cannot have itself
directly as an object of experience, it seems to me that we can instead
consider how the observables of one observer are different from another's in
a way to, indirectly, defining and distinguish one observer from another.
My idea is to start with the notion of a class X of all possible
"observables" (in anthropomorphic terms: perceptions) and think about how
they might be partitioned up such that each "observer" would be associated
with some subclass of X.
We notice immediately that the idea of a "light cone structure" (used in
Relativity) is related to this class.
----- Original Message -----
From: "Hal Ruhl" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Wednesday, April 21, 2004 9:08 PM
> I would like to explore just exactly what the various members of the list
> mean by "observer" as in the following from Wei Dai's post.
> >Consider all possible worlds consistent with your memories and current
> >experiences. In other words, all possible worlds that contain at least
> >observer with memories and current experiences exactly identical to
> >Are there more than one such world?