| |
| (2) |
|
it
has the following S-structure representation: |
| |
|
|
| (4) |
|
 |
|
| |
Here
the anaphor apan's (occurring as apni because
of agreement with the head N kitab) antecedent
is ram. As in (3), here too, ram
does not c - command the anaphor. Yet, contrary
to the predictions of the theory, the sentence
is grammatical with the given coindexing. |
| |
One
way to save the c - command approach could
be to modify the notion of c - command such
that ram in each of (1) and (2) c - commands
the anaphor. The formalization to yield this
result is not difficult, but this alternative
would really solve the problem proposal that
renders the post position irrelervant for
the purposes of C-command, in (2) shyam too
would c - command the anaphor. That is, the
theory would predict that the sentence is
ambiguous but actually it is not. |
| |
The
analysis of (5) and (6) reinforces this point.
|
| (5) |
|
ram ne hari ko uski kitab di
j
j
ram CM hari CM his book give+PAST
(Ram
gave his book to Hari)
|
| |
|
|
| (6) |
|
[ram ke bhai] ko uska makan pasand hai
i j
i,j
ram CM brother CM his house like be+PRES
(Ram's borther does notlike his house)
|
|
| |
