In my enthusiasm to report a naturally occurring instance of a rare but highly studied syntactic phenomenon, I made a mistake: it turns out that the example in my recent posting on parasitic gaps is not a parasitic gap construction at all. Fortunately, the posting generated some interesting further discussion on naturally occurring parasitic gaps and double right node raising, so some good has come of it.
Jon Nissenbaum was kind enough to point out my error, and even kinder to have done it by private e-mail. :-)
He explains that the example I gave
...is an instance of across-the-board
movement, coupled with right node raising out of a coordinate
structure, rather than a parasitic gap:
The coordinate structure is the smoking gun, since
...parasitic gaps are gaps that appear, not in a coordinate structure
as above, but in an island (as you cite Peggy Speas as
mentioning). The "main" gaps (the ones upon which the PGs are
parasitic) are perfectly acceptable as stand-alone extractions, unlike
ATB movement out of a coordination. So, while both versions of (1) are
fine, the same is not true of (2):
(1) a sweater that I wore _ [without returning _]
a sweater that I wore _ [without returning your phone call]
but,
(2) a sweater that I wore _ [and had to return _ ]
*a sweater that I wore _ [and had to return your phone call]
Many thanks to Jon for setting me straight! Again, see the followup posting by Chris Potts for some true parasitic gap examples that have been properly caged.
Posted by Philip Resnik at March 11, 2004 08:12 AM