[hist-analytic] The "Analytic A Posteriori"
danny.frederick at tiscali.co.uk
Thu Sep 3 11:37:39 EDT 2009
You object to Steve:
'A posteriori knowledge is not just knowledge by experience in your sense;
if it were, only observational knowledge and possibly memory knowledge would
be a posteriori knowledge. But many things we know about the world are
known inferentially; and what is thus known has always been considered a
I don't think you are entitled to say this. I will suppress for the moment
my scruples about the possibility of a priori knowledge. You want to say
that a priori knowledge arises from knowledge of conventions and deductive
inferences from items of knowledge known in this way. Thus you are committed
to saying that things known by inference may be a priori.
Further, given you're view of the a priori nature of logic, the application
of deductive inference to an item of empirical knowledge will yield a priori
additions to what we know empirically. These additions will be relatively a
priori: they depend on empirical knowledge as premises, but they require no
more experience than that (they are a priori given the premises).
I would guess this is Steve's point: we know some stuff by experience; we
know other stuff by a priori inference from experience; and we know some
stuff a priori without inference [but God knows how! - sorry, I couldn't
Further still, I think it is plain false that we can acquire new empirical
knowledge inferentially (by 'new' I mean knowledge not deductively implied
by what we already know). There is no valid method of non-deductive
inference. What we do in acquiring new empirical knowledge is make a guess
and then test it.
More information about the hist-analytic