phrases whose literal meaning...

[Laurence Horn]

At 7:12 PM -0500 11/3/02, Mark A Mandel wrote:
>On Sun, 3 Nov 2002, James A. Landau wrote:
>
>#My daughter needs to know the term to be applied to phrases whose literal
>#meaning is one thing but which are universally interpreted as something else.
># Her examples are "How do you do?" which literally asks "How are you?" but is
>#used merely as "Hello"; and one from Hebrew, "mazel tov" which literally
>#means "good luck" but is universally used as "congratulations".
>
>Actually each of these is one remove further from literality. "How do
>you do?", taken literally, is the same construction as "How do you sew?"
>or "How do you S?", and is almost ungrammatical, since "do" as main verb
>with no explicit object is almost obsolete. It's a frozen form,
>supplanted in current syntax by "How are you doing?" And "mazel tov" is
>literally 'a good star', which by an astrological metaphor means 'good
>luck'.
>
>-- Mark A. Mandel

cf. also "God be with ye'" > Goodbye.  Besides the indirect speech
act, politeness, and indirection factors Dennis mentioned, there's a
process whereby expressions may implicate something they don't
literally express, but do so in a way that is partially incorporated
into their form as well as their meaning.  This is discussed in the
literature under the labels "standardized nonliterality" (Kent Bach)
and "short-circuited implicature" (Jerry Morgan).  The latter, in his
classic paper "Two types of convention in indirect speech acts"
(Syntax and Semantics 9: Pragmatics, Academic Press, 1978), has lots
of nice examples that he tracks through changes over time, including
the aforementioned "Break a leg!" (vs. #"Fracture a tibia!"), "This
is Larry Horn"/"Larry Horn is speaking" (when getting a phone call;
cf. the impossible "Larry Horn is speaking", "I am Larry Horn",
etc.), "You can say THAT again!" (vs. #"You can repeat that!").  In
each case, he argues, a speaker unaware of the relevant convention of
usage (not a meaning convention on his account) might be able to
reconstruct the intended non-literal speech act, but normally these
are understood without having to be calculated--the inference is, in
Morgan's phrase, short-circuited.  Eventually the expression may end
up literally expressing what it used to merely implicate or suggest;
this is typical of euphemisms, and a nice example from the paper (due
originally to Jerry Sadock) is the fact that "go to the bathroom" had
changed its conventional meaning at some point before I could
complain that your dog went to the bathroom on my living room rug.
Another example, of course, is "goodbye".

Larry