[hist-analytic] Frederick's conception of the A Priori

Roger Bishop Jones rbj at rbjones.com
Fri Mar 27 09:55:37 EDT 2009

On Monday 23 March 2009 15:33:55 Danny Frederick wrote:

>The use of Popper's demarcation criterion for empirical knowledge to define
>a notion of a priori knowledge is due to me.


>Objective knowledge is inter-subjectively testable. It is empirical if it is
>testable against public observation statements. It is a priori if it is
>testable in some other way, such as algorithms or derivations. Of course, we
>need something in addition to algorithms and derivations, namely, intuitive
>evidence of necessity.

It looks like your conception of the a priori and a posteriori are
confined to knowledge, and incorporate some criteria of what may
count as knowledge.

In my "triple-dichotomy" monograph I propose to defend the
positivist position that the three dichotomies are in fact
one, and so I need to intepret the a priori/a posteriori
distinction for propositions in general, even those which
are not likely ever to be known, and those which are in
fact false.
This I would do on the basis of what kind of justification
would be acceptable for establishing the proposition.
(which one may have an opinion on whether the proposition
is true or false).  The idea is that there is a connection
between what a proposition says and how it should be

In this context it is essential to define the "a priori" prior
to, and more generally than, consideration of what justifications
are valid.

If your conception of the a priori and a posteriori is limited
to known or knowable propositions, then our differences in
this area will be primarily verbal, we disagree about how
these terms should be used, rather than about the consequences
of our respective conceptions.

From the differences which are so far apparent (without my
having tabled a proper account of my own proposed usage)
one might expect that my use of these terms is more
liberal than, but consistent with, yours.
However that does not seem to be the case, there are
substantive disagreements.

Broadly speaking I count a justification as a posteriori
if it makes any (epistemic) use of contingent propositions,
and otherwise as a priori. Its status in this respect is
independent of any other characteristic of the
justification, e.g. whether the contingent propositions
are used reasonable and do in fact support
the proposition supposedly being justified.

My criteria does not depend upon the notion of sensory
experience (odd perhaps for a positivist).
The use of a proposition about sensory experience
would on my account render a justification a posteriori,
but a proposition describing any other
experience would do likewise.  Under my definition
an a priori justification would have to avoid dependence
on reports of any kind of experience whatever, or
even upon thoughts which someone has had.

It seems to me that your interest, like Poppers is
primarily in the demarcation of science from non-science
and that the two categories "empirical science" and
"deductive science" are what you are addressing.
I wonder therefore what advantage there is in using the
terms "a posteriori" and "a priori" for these categories,
rather than leaving them for broader use.
As it is, you, I and Kripke are all talking at cross
purposes, we use the same words in different ways and
therefore are in difficulty in getting a grip on
whether we have a substantive non-verbal disagreement.

My position in the proposed monograph in relation to
Kripke is likely to be to argue that his philosophy,
which has been supposed to refute various tenets of
logical positivism, in fact fails to do so because
Kripke's conception of the meaning of the relevant
concepts can be seen to be at variance with the
positivists.  I should have to take a similar position
towards some of your own views about the a priori,
e.g. the idea that "I think" can be known "a priori".

Roger Jones


More information about the hist-analytic mailing list