For the record, the way H. P. Grice phrases the thing in "Prolegomena" goes:
"Indeed, the precept that one should be careful NOT to confuse meaning and use
on the way toward being as handy a philosophical vade-mecum as once was the
that one should be careful to identify them."
This re: McEvoy's quoting Witters:
"For a large [emphasis Witters's] class of cases—though not for all—in which we
employ the word "meaning" it can be defined thus: the meaning of a word is its
use in the language."
McEvoy was wondering about various sides to this. McEvoy's main tenet is that
Witters SHOWS (and successfully so) rather than SAY or IMPLICATE anything.
Note that in the Witters quote we do have the philosophical term of art
'define'. He is saying that 'it' (i.e. the word "meaning" -- this is a verbal,
not a real, definition -- vide Robinson, "Definition," Clarendon) CAN (does he
mean 'may'?) be defined.
And he goes on to provide this definition. Familiar as he was with Whitehead's
and Russell's "Principia Mathematica," he knows that 'define' becomes the symbol
i. The meaning of a word is its use in the language.
Isn't it curious how 'though' and 'if' are sometimes interchangeable.
ii. It looks as though it might rain tomorrow.
iii. It looks as IF it might rain tomorrow.
Yet, in Witters's idiolect, 'though' and 'if' have OPPOSITE implicatures:
iv. For a large class of cases -- though not for all --, the meaning of a word
is its use in the language.
v. For a large class of cases -- if not all --, the meaning of a word is its
use in the language.
Be what it may, it is THIS Grice is thinking when he speaks, rhetorically, of
the precept that
vi. One should be careful to identify meaning and use.
Grice's use of 'vade mecum' ('go with me') is delightful.
For Grice is contrasting this 'old' precept (vi) -- i.e. Witters's precept if
he claims it does not apply to ALL cases -- but to a 'large class of cases in
which we employ 'meaning' -- even though or even if he never ventures to
provide one single COUNTER-EXAMPLE, Popperian almost -- with a new one. The
vii. One should be careful NOT to confuse meaning and use.
Grice is being rhetorical in using 'handy a philosophical vade-mecum,' too. He
does not define the degree of 'care', admittedly.
The bottom line, metaphorically, is that Witters is into 'definining' "meaning"
and providing a definition that he accepts may not apply to ALL cases -- but
failing to provide one simple illustration on which he bases his guardedness.
Note too that in the first part of the quote
For a large class of cases—though not for all—in which we
employ the word "meaning" it can be defined thus: the meaning of a
word is its use in the language.
Witters speaks of the cases 'in which we [the majestic again] employ 'meaning'
_simpliciter_ and turns to provide a 'definition' of a more specific
collocation, "meaning of a word". This reminds me of J. L. Austin's essay, "The
meaning of a word" in Philosophical Papers, where he distinguishes:
viii. The meaning of 'rat'.
ix. The meaning of the word 'rat'.
and so on.
If Witters seems to sure that his definition:
x. The meaning of word W =df the use of W.
"In the language" seems otiose. Surely when Humpty Dumpty (admittedly not in a
poem by Lewis Carroll, to echo Witters) uses 'glory' to mean 'a nice knockdown
argument', the addition of "in the language" renders the whole thing
anti-Popperian, unfalsifiable and vacuous. So I get rid of it.
Unless we speak of "in the language of Humpty Dumpty".
Grice, compared to Witters's vague wordings, is so careful that it hurts. He
considers procedures of idiosyncrasies of meaning, as when we restrict the use
of W to one particular 'idiolect, for a specific utterer. It is true that
Witters was lecturing years earlier and in a different shire (Cambridgeshire),
And so on.
This post just to provide the quote by Grice where he narrows down Witters's
thing to a 'precept' once a "handy philosophical vade-mecum", but no more. The
point now, in authors like Strawson, Quine, and Grice himself, is to be careful
NOT to 'identify' meaning and use.
Grice's use of 'identify' gives support to this idea that, in Witters, the use
of 'define' should be understood as "=df" which nicely combines "identity" (as
in Grice's 'identify') with Witters's proposal of a partial definition of what
the meaning of a word amounts to.