[b-greek] On Smart's rule, exceptions and methodology

From: Kimmo Huovila (kimmo@kaamas.kielikone.fi)
Date: Tue Feb 06 2001 - 08:39:10 EST

Dan Parker wrote:

> I can easily apply Sharps words in defence of his rule to the defense
> of Smart's Rule. [If Sharp can capitalize "Rule" so can I!] Of a detractor
> he said
> "he has not been able to produce against the Rules one single
> example from the Greek text of the New Testament, (the only
> true criterion of their truth)"

You have several times referred to Smart's rule as having no exceptions.
In the last round of discussions two specific exceptions were given
(John 20:28 and Mark 3:33-34, Matt 12:49, the last two counted here as
one), and you did not give any satisfactory explanation to either one.
Why should we look for more exceptions? Without dealing with the cotext
(of the NP; cotext means textual context) it may be possible to deny
some counter examples (such as John 20:28) by referring to the rule, but
this is only circular reasoning if the cotext is strongly against the
interpretation as here. Any proposed rule that restricts the semantics
of the expression needs to pass the test that the proposed semantics fit
the context. In the gospels example you do not even need to go to any
more remote cotext for the NP than just the sentence itself to see that
the rule is does not work there. If you seriously want to get this rule
generally accepted, I think, you need to answer the criticism the rule
has already gotten. Otherwise it is no wonder that some people on the
list really feel the rule has been demolished.

Also, you need to address some methodological problems. These include 1)
too small a database (I think you need to take into account that many
examples come from a similar context, that is the opening of a letter,
which does not do much to give statistical credibility of its use in
different contexts) - you could append this by doing an exhaustive study
of a sizable KOINH corpus 2) lack of statistical comparison with other,
non-Smart NP's joined with KAI. 3) lack of plausible explanations for
the reason this proposed significant anomaly in the Greek nominal
syntax, in other words why the semantics of co-ordination of two NP's
with KAI would be semantically so strongly related to the inner-NP
presence or absence of a personal possessive pronoun, and 4) the
relatively high number of exceptions in the corpus examined so far (even
two is quite significant, because you have so few examples of the
construction, especially if you want to make an absolute claim). Also,
the NT, unfortunately for the linguist, is not a corpus designed by the
use of corpus linguistic criteria, which should caution us for
generalizing too much to the language of the time.

By the way, corpus linguistics is a field of linguistics where much
study has been done. A good working knowledge of it surely helps
avoiding pitfalls formulating grammar rules. 'Corpus linguistics' by
Conrad, Biber, Douglas (hope I remeber the authors correctly) serves as
a general introduction to corpus linguistics, though it does not really
discuss parsed corpora, which would be more relevant when formulating
grammar rules. Some experience with parsed corpora of a modern, well
known language will help one avoid some methodological pitfalls.

If you want to give statistical credibility to the effect of the
personal pronoun, study the proportion of identical references of NPs
linked with KAI without possessives to non-identical references, and
compare that to equivalent figures with both non-repeated and repeated
possessives. And make sure that the sample is big enough to be
statistically significant, and also that the corpus is a relatively
representative sample of the language (e.g. the frequency it appears in
epistolary introductions in the NT distorts the picture statistically if
applied to other uses). That would show how much of a heuristic tool
Smart's rule would be for determining reference.

I do not promise to enter into a lot of further discussion of Smart's
rule. I guess last round was enough for most arguments to be heard. I
just hope there is something here that you, Dan, find worth pondering or
at least interesting. If not, just use the delete button.

Kimmo Huovila

B-Greek home page: http://metalab.unc.edu/bgreek
You are currently subscribed to b-greek as: [jwrobie@mindspring.com]
To unsubscribe, forward this message to leave-b-greek-327Q@franklin.oit.unc.edu
To subscribe, send a message to subscribe-b-greek@franklin.oit.unc.edu

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:36:50 EDT