From: Randy LEEDY (Rleedy@bju.edu)
Date: Tue Dec 08 1998 - 15:14:36 EST
Daniel,
Thanks for your thoughtful post on this topic. I am sorry that just
now, with heavy schedule pressures associated with the end of a
semester, I cannot respond at any length. I am glad you invited a
response before writing more just now; you probably saved yourself
some time.
You have developed your thinking further than I have done, but along
largely the same lines I have been considering. Perhaps the biggest
difference is that I have not been thinking about trying to group
words into larger constituents such as the various kinds of phrases
but rather indicating only connections between individual words. Where
connections exist between phrasal constituents, I would indicate those
connections only between the heads of the respective constituents. One
kind of connection I would indicate is that between pronoun and
antecedent (where expressed), thus a relative pronoun would be tagged
both with its function within ints clause and also with its connection
to its antecedent. I would not group the relative clause and then tag
it as a constituent modifying the pronoun's antecedent. The database I
envision would by no means be a complete accounting of every level of
syntax that might be cataloged. The higher-level information related
to phrasing seems to me to be the most subjective and therefore the
least widely agreed upon. I would like to set up a tagging scheme that
is open to further development of this higher-level information but I
do not feel inclined to tackle that part of the job myself.
Since I began posting on this topic, Dale Wheeler has said that he has
already done work very much like what I have in mind. So I have cooled
my enthusiasm for the moment, until I can get a chance to find out
from him a little more detail on what he has done.
I am saving your post for later consideration. If it turns out that I
do want to pursue this project further, I will be back in touch with
you to discuss ideas a greater length. The next step for me would be
to develop a scheme of exactly what syntactical connections to catalog
and what to leave untouched for now. I have not yet developed even a
working list on paper, so I really have nothing concrete to share with
you now.
Again, thank you very much for the time you invested in your post. I
appreciate it and hope that it will not have been entirely in vain.
Randy Leedy
This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:40:09 EDT