Grice on "seem" and "may be" (Ex)◊(ϕ x)

by J. L. Speranza
for the Grice Club

ONE THING to consider when discussing Grice's "Causal Theory of Perception" (perhaps his clearest case for 'conversational implicature') is what he calls the D-or-D (denial or doubt) 'implicature' in "seems" statements. His funny example. In broad daylight, of an object at a few feet from me, I say:

--- That British pillar box seems red to me.

(versus the stronger, 'it is red').

--- We have discussed the dictum elsewhere. What I want to consider is the connection, a loose one, between


x seems phi.

(as Grice symbolises the thing)

and,

not

x IS phi

but

x MAY be phi.

-- a bother to symbolise.

Suppose I use the diamond.

◊

--- The thing now is to provide an extensionalist (in terms of possible states of affairs or possible 'worlds') of

"the pillar box" -- call it "Box" -- SEEMING 'red' (call it phi)

The structure of

"Box is red"

(let us simplify)

would be

ix.(Bx & ϕx)

What about

"Box SEEMS red".

What I am submitting is that there IS indeed a doubt-or-denial implication -- but to what extent does this not overlap or connect with the 'entailment' of 'seems'.

For, what is to 'seem'. Surely there is a 'conceptual' connection between 'seems' and 'is'. Grice is claiming that 'seems' meaning "is not" or "may not" -- i.e. respectively denial or doubt -- we have to distinguish from an utterer implying that "seems" triggers "is not" or "may not".

---

His point is well taken. Surely there is nothing incompatible about something BEING perfectly phi for utterer and yet Utterer saying, "seems phi". I know some guarded speakers who think it's rude and ill-mannered to say things like 'know', 'certain', and 'is'. Things ALWAYS "seem" to them. (I think I am one of those persons).

----

The 'denial' is exaggerated. "x seems phi" implying "x is not phi". We do know of journalists who use 'seem' like that. "She seemed a bit intoxicated"? No. Not that case. Perhaps a journalist will be guarded enough and add, "only". "She only SEEMS to have forgotten the lines". No. That does not work.

What about 'doubt'? Doubt about what? The point of bringing in the 'diamond',

◊p

is to connect the 'seems' WITH the 'is'. "Seems" seems to be always short for "seems to be".

"The pillar box seems to BE red".

Now, -- the implicature point by Grice seems obvious now. He is surely saying that there is no easy entailment or analysis of 'seems' that will get rid of the 'is'. What he is claiming is that an analysis of the 'social' implication (i.e. in a context where we are constrained by making the strongest 'conversational' move, caeteris paribus) CONNECTS the 'seems' with a qualification of the 'copula'.

The case of 'denial' is too gross: 'S seems P' becomes "It is not the case that S is P" (at the level of the cancellable implicature only). But it also becomes "S MAY BE P" (at the same level).

Perhaps Grice, alas, took senses for granted all too often. There we have this delightful philosopher expanding on what things seem to us, and what we imply by thus saying, without giving us a detailed truth-condition of what the case actually IS when we do make such a guarded claim. Or not!