[hist-analytic] Roger's Questions about Anayticity
danny.frederick at tiscali.co.uk
Sun Jan 25 17:57:59 EST 2009
I think I can see the source of our disagreement. You say:
'If I speak in English about the analyticity of some French sentence, e.g.
"...." is an analytic sentence of the French language,
this claim will be synthetic (whatever the sentence in quotes), supposing
the phrase "the French language" to mean something like "the language spoken
in France". This is because the language spoken in France might have been
quite other than it in fact is (or at least, that explains it being
If you disagree with my proviso about the meaning of "The French Language"
then chose some other phrase which refers to that language without connoting
However, the name
either refers to an interpreted string or it does not. If it does not, it is
impossible that the string is analytic. If it does, then the string is
either analytic or it is not. If it is analytic, then the statement that it
is analytic is necessarily true and the statement that it is not analytic is
necessarily false. If it is not analytic, then the statement that it is
analytic is necessarily false and the statement that it is not analytic is
necessarily true. In every possibility the attribution of analyticity is
either necessarily true or necessarily false.
'"...." is an analytic sentence of the French language.'
I interpret this as a sentence which begins with a name, which is a rigid
designator. If the name refers to an uninterpreted string, then it does not
refer to something that can be analytic; so let's assume that it refers to
an interpreted string. In this case, it is context-dependent, just like many
other names. Thus, in one context I use 'John' to refer to my brother. In
another context, I use 'John' to refer to a friend. Your phrase, 'of the
French language,' supplies the context which determines the reference. So we
now know we a talking about the sentence interpreted as a French sentence.
The statement that that sentence is analytic is thus either necessarily true
or necessarily false.
But you go on:
'supposing the phrase "the French language" to mean something like "the
language spoken in France"'
Now, this seems to involve a different meaning in that a rigid designator is
replaced with a non-rigid designator. Thus in place of the name
and a disambiguating context, we get a definite description: 'The
interpretation of "...." as a sentence of the language (that happens to be)
spoken in France.' Within this definite description,
is the name of an uninterpreted string. So the definite description refers
to different interpretations in different possible worlds, depending upon
which particular language is spoken by the French in those worlds.
Analyticity still attaches de re to interpreted sentences (or to
interpretations of sentences), but the definite descriptions (non-rigid
designators) pick out different interpretations (or different interpreted
sentences) in different possible worlds, and thus they pick out an analytic
interpreted sentence in one world but a non-analytic interpreted sentence in
You then say:
'the implication of supposing that the name of a language connotes its
semantics is that someone who does not have a complete knowledge of a
language does not know the meaning of its name either.'
Nothing that I have said implies that the name of a language CONNOTES its
semantics. For example, let me just call the (interpreted, actual) French
language 'Peter.' Then the name 'Peter' as I use it REFERS to French with
its semantics (of which I am ignorant, not being a French speaker). And the
'"...." is an analytic sentence of Peter,'
is either necessarily true or necessarily false (again, here, 'of Peter'
provides a context which determines the reference of the quotation name,
which reference is an interpreted sentence).
'It may be argued that for the purposes of delineating analyticity, it is
not in fact essential that semantic assent preserve meaning, or that it
I do not know what Carnap's position is (I have not read him for a quarter
of a century). But it seems to me that semantic ascent must refer to
expressions taken with their meanings, because nothing relevant could be
said about meaningless strings.
I am a bit out of practice with the logic-chopping, so if you or anyone else
can point out mistakes in what I have said, I will be pleased to hear about
More information about the hist-analytic