Missing PREP differing by dialect

[Arnold M. Zwicky]

On Apr 2, 2008, at 8:58 AM, Larry Horn wrote:

> At 11:14 AM -0400 4/2/08, Wilson Gray wrote:
>> WE:
>> "He shit/shat himself"
>> "He pissed himself"
>>
>> vs. BE:
>> "He shitted _on_ himself"
>> "He pissed _on_ himself."
>>

> Actually, the latter two forms are perfectly acceptable in varieties
> of WE I'm familiar with, and there's a slight
> semantic/pragmatic/register difference between the two versions.  If
> I inadvertently allow a couple of drops to hit my shoe, I pissed on
> myself, but I didn't piss myself.  This actually follows from the
> general association with direct objects and "affectedness"...

i thought the shit/piss cases were in the list of P~zero alternations
from charlie doyle's 1977 Am Sp paper, which he mentioned here back on
11 february -- but apparently not.  in any case, a number of those
variants differ subtly in conveyed meaning, and for some
"affectedness" seems to be at issue (beat up (on) a person, play (on)
a piano).

the connection between direct objects and affectedness is a nice
(though subtle) example of iconicity in grammar: closely linked
objects (i.e., direct objects) tend to be understood as denoting more
affected referents, and less closely linked objects (i.e., oblique
objects, marked by prepositions) tend to be understood as denoting
less affected, more tangentially connected, referents.  tighter
syntactic connection, more direct connection in meaning.

(this is not a novel observation of mine, by the way.  for
"functionalist" linguists, it's a commonplace.)

bonus observation: all this means that "direct object" is not a half-
bad name for this syntactic function.  not exactly transparent, and
certainly not a definition, but suggestive.

arnold

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org