[hist-analytic] The Fundamental Triple-Dichotomy

Roger Bishop Jones rbj at rbjones.com
Sun Feb 22 12:15:18 EST 2009

I have in mind, at present, writing a short monograph entitled:

  The Fundamental Triple-Dichotomy

an overblown name for what may first have appeared as "Hume's Fork"
but whose predecessors can be traced through the entire history
of Western philosophy.

The monograph (as presently conceived) will hang around a presentation
of Hume's fork, and a contemporary re-presentation of the "triple-dichotomy".
To this will be added some lightweight historical material covering
related themes in the period before Hume as far back as early Greek 
and also tracing the evolution of the dichtomies since Hume.
Finally I intend to add to my re-presentation of the dichotomies
various refinements which will include discussing the problem of
regress in semantics, the problem of defining the division with
maximal precision, and the question of whether a language of
sets can provide a "universal" foundation for abstract semantics.

I hope to discuss the various positions which I will present in this
monograph here on hist-analytic, as they develop.
I am especially out-on-a-limb (or out of my depth) when attempting
to write about the history of philosophy, since I am a poor scholar,
so I'm hoping for lots of criticism in that area, though perhaps it may
be more vigorous in relation to the central proposal.

To begin this process I present here in its most concise
form the epicenter of my re-presentation of the dichotomies, which
are of course the Necessary/Contingent Analytic/Synthetic and
a priori/a posteriori dichotomies. 

For the purposes of this concise presentation I will assume understood
the notion of logical necessity, taking this to be a characteristic
of propositions consisting in their being true in every possible world
(which idea I leave unexplained for now).

I next propose that a sentence, together with sufficient context
to disambiguate its truth conditions, should be said to be analytic
iff it expresses a necessary proposition.

I further propose that the term "a priori" should be used to talk
of the epistemic status of a proposition, in particular of the kind
of evidence we would expect to find in the justification of a claim
to knowledge of the proposition, noting the distinction between this
and any usage of the term to refer to the manner in which the truth
of the proposition may have been discovered (rather than established).
I propose in this that we should accept a priori justification for
a propositon iff it is a necessary proposition and an a posteriori
justification iff the proposition is contingent.

In summary I have made certain proposals about how we should use
the words in question, and about what we should demand in justification
of claims to knowledge the effect of which is to make the three
dichotomies in question co-extensive, the first two being so
of necessity and analytically.

This position is described in greater detail at the following URL:


and is supported by a mathematical model at:


both written more than a decade ago.

The position is of course, not novel.
It is Hume's fork, and it is very close to the position in these
matters of the logical positivists, held by many philsophers to
have been decisively "refuted" in recent times.

I invite criticism.

Roger Jones

More information about the hist-analytic mailing list