From: c stirling bartholomew (email@example.com)
Date: Thu Mar 14 2002 - 22:53:50 EST
on 3/14/02 7:29 PM, Moon-Ryul Jung wrote:
> Dear Clay,
> I am pretty sure that those who work in the field of "natural language
> processing" or "computational linguistics" will claim that they can
> provide a complete specification for the syntactic structure of free word
> order languages.
As I understand this, the specification would be made up of a set of
production rules which meet two limiting criteria:
1 - The rules would produce every possible "well formed clause"
2 - The rules would exclude all clauses which are not "well formed"
So we would need to start out by finding a consensus about what a well
formed clause might be in Hellenistic Greek. Would this be a possible task?
Perhaps it is theoretically possible but software developers building search
engines for grammatically tagged ancient language texts don't travel in the
world of theory. Their stuff has to work.
> They can write down all rules that human readers know about the structure of a
> given language, so that these rules can be processed by computer programs. I
> always thought that it is a matter of time and effort.
This sounds like what people used to call an expert system.
> Sometimes, a given sentence may be analyzed to produce more than one
> tree structures. But that is OK. Even the human reader often encounters
> such cases.
You are right. When I started trying to "read" 2nd Peter for the first time
I got about as far as 1:3-4 and "hit the wall" in terms of syntax. There are
places in the GNT where reading becomes a chore for humans. The
specification rules that would accommodate these passages are likely to
accommodate pure nonsense as well. The big problem is allowing what is real
Greek including texts like 2Pet 1:3-4 but filtering out garbage text. Sure
it can be done, but not on a tight budget by a small software developer.
This is a topic worth kicking around. I think I will let the linguists take
over from here with it.
I appreciate you contribution here.
Clayton Stirling Bartholomew
Three Tree Point
P.O. Box 255 Seahurst WA 98062
B-Greek home page: http://metalab.unc.edu/bgreek
You are currently subscribed to b-greek as: [firstname.lastname@example.org]
To unsubscribe, forward this message to leave-b-greek-327Q@franklin.oit.unc.edu
To subscribe, send a message to email@example.com
This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:37:21 EDT