To: Jason Baldridge <firstname.lastname@example.org>
Subject: by-phrase representation
From: Tonia Bleam <email@example.com>
Date: Wed, 14 Oct 1998 16:16:36 -0400 (EDT)
>From firstname.lastname@example.org Mon Feb 15 14: 01:14 1999
So i'm looking at McCawley's syntax text to see what constituent tests he
gives in general and found an argument about where the by-phrase is
specifically for passives, McCawley gives the structure where the
by-phrase is adjoined. (not to be confused with adjoining in TAG).
1. Smith was awarded an honorary degree by Harvard.
the VP that McCawley gives for (1) looks like (2).
2. [VP was [VP [VP awarded an honorary degree] by Harvard]
3. Smith both [was awarded an honorary PhD by Harvard] and was named a
special assistant to the governor.
(3) shows that whole thing in brackets is a constituent.
4. Smith was both [awarded an honorary PhD by Harvard] and named a special
assistant to the governor.
(4) shows that everything after "was" is a constit.
5. Smith was both awarded an honorary PhD and offered an endowed chair by
(5) shows that the V and its object form a constituent without the
by-phrase -- arguing for the "extra" VP node.
So now the question is whether the constituency shown in (5) is the base
constituency -- or whether it's extraposed (right-node raising?)
One way to tell is whether the by-phrase can be extracted from in this
6. Who was Smith both awarded an honorary PhD and offered an endowed chair
and actually, we need to check with speakers about normal extraction
(altho i have at least one judgment on this already).
7. Who was Smith awarded an honorary PhD by?