Quote:
Originally Posted by Casiopea  Exactly my point.  The CGEL's solution to 3b and 4b has everything to do with syntax, when in reality, the distributional facts admit quite clearly that the reason speakers use I as an object and me as a subject has nothing to do with syntax proper.
The solution resides in processes already at play elsewhere in the grammar. In phonology, for example, segments are weakend or strengthened according to their relative position; in morphology, pronouns could be constrained by similar processes: objects are strengthed (me > I) or weakened (I > me) according to their relative position, irrespective of government (CGEL's argument) or case assignment (Prescriptivists' argument). That's not something I came up with; it's what the data admits. |
As I see it, the CGEL is arguing is that coordinated pronouns are treated differently. This seems completely consistent with your argument that they are strengthened or weakened (altho I'm not sure exactly that means) according to their position. It is because they are treated differently that they are free to change form according to their position.
Assuming that these kinds of utterances are not ungrammatical, then all parts of the grammar have to allow them, including the syntax. I do agree, though, that CGEL doesn't explain them. At most, they try to demonstrate how they might be represented in the syntax.