D. E. Over on H. P. Grice

--- I have corresponded with D. E. Over, and I love his style. This is from Over et al at

http://www.cog.brown.edu/~slomanlab/papers/CausalConditionals.pdf

---

On Grice on the implicature of "if p, q"

--

"P(q/p) [could] be high simply because P(q) is high." (p. 31).

They assume a Gricean defense:

It could be argued that

"the use of conditionally pragmatically
suggests, in certain ordinary
contexts, that p raises the probability
of q, or that p causes q."


(i) If you take extra vitamin C, your cold
will be gone in three days.

Over et al write:

"In most contexts, asserting (i) would be
misleading"

-- but not false

"and very bad advice, if extra
vitamin C was not a causal factor
raising the probability that
the cold will be gone in
three days."

"The argument would be that there
is often a pragmatic implicature,
when a conditional like (i) is
asserted: that NOT taking
extra vitamin C will make it
probably that the cold
will last longer than three days."

"This inverse implicature could
cause a second conditional"

(ii) if ~q, ~p

---

"A Ramsey test on this
[second] conditional would
be believable to the
extent that

P(~q/~p)

is high."

---

"A positive effect of this
probability is the same as a
negative effect of its complement
P(q/~p)."