[b-greek] Re: B-GREEK POLL: Smarts and Sharps both depend on context? Y/N

From: Kimmo Huovila (kimmo@kaamas.kielikone.fi)
Date: Tue Oct 31 2000 - 12:05:19 EST

Someone said that it is difficult to get clear yes or no
answers from scholars, because they want to carefully qualify what they
say. So I am not rambling, I am just trying to be like a scholar ;-)

The trouble with some of these grammatical rules is that by adding
enough qualifications, the corpus becomes so small and statistical
correlations have such a big say that it is possible to make almost any
rule "work", even if it is not a genuine grammatical rule and
would not correspond to native intuitions. I think that
grammarians should go deeper than just
look at a few instances. These rules (Smart's, Sharp's) should be
described in terms of their place in the totality of the language. Thus
looking at analogical structures etc. is relevant. It is not useless to
ponder why a proposed rule would work in a language and is it likely
that such a thing would ever have grammaticalized.

I liked Mark's approach of constituency analysis to the question of
Smart's rule. Perhaps unfortunately for his idea to hit home in this
forum, he used UG to support it. While I share the concern Wayne
expressed about that methodologically, I think his analysis is
nevertheless pertinent from the Greek point of view.

I think that also Sharp's rule should be analyzed in terms of
constituency. The one article governing two nouns links the two nouns
together closer than the repetition of the article would. This is
irrespective of the further constraints of Sharp's rule (singularity
etc). Now, this linking can be used when the referent(s) is/are
identical. But it does not mean that there may not be other pragmatic
reasons to want to group them together, such as them sharing a
contextually pertinent feature together. So, we have one structure that
can be used when there is identity of reference, but does not have to

If we exclude plural reference, we decrease the probability that the
referent is not identical. The same thing happens if we include only
references. Thus the rule is likely to be more binding with these
constraints. We could set a scale of likelihood for the construction
(with different constraints) to be used
with identical reference, Sharp's rule with the usual constraints being
somewhat close to the identical reference end. For the exegete to
determine if the reference is really identical, context still must be
considered, because the rule is not as likely to be applicable if there
are strong contextual reasons for associating the two different
referents as practically one for the purpose at hand. And the reverse
side is that Sharp's rule (well, actually no longer Sharp's rule) with
plural nouns is not completely irrelevant, as the two nouns still have a
closer association than with the repetition of the article. One just has
to remember that this association is not necessarily the same as
identical reference, and it is even more seldom identical with plural
nouns than with singular nouns. This illustrates scalarity nicely, by
the way.

In the case of Smart's constructions, I guess the relevant question is
why is the personal pronoun repeated. I think the repetition makes sure
that the pronoun is understood to qualify both nouns. It may also serve
emphatic function. This makes perfect sense in the context of John
20:28, where two separate referents is pragmatically problematic. And
the possessive pronoun could also be repeated in the case of
non-identical reference (a real
example of Smart's rule). Thus Smart's rule is context dependent.
Furthermore, it seems to catch very
little of how Greek grammar actually works: we can explain the data with
other principles that we already need to account for linguistic
phenomena in Greek (such as constituency and pragmatics) with no need to
create a separate rule. The constituency analysis is better motivated
and makes more correct
predictions (e.g. in John 20:28, as pointed out in several posts in this

I could propose a rule that when one person is addressed, and you have
Smart's construction used as an address (vocative function), only one
person is meant: it
works much better with actual data than Smart's rule. (Could anyone find
an exception?) However, this kind
of rule does not really tell us anything insightful about Greek
grammar, in accordance with Clay's insightful comment on this kind of
rule generation. The more useful thing than to learn this rule would be
to understand why it works (which really has very little to do with
grammar: if you address one person with two titles, you mean to
use them to describe the one person. Circular, isn't it?)

I think that Smart's rule is not very likely to contribute much to
our understanding of Greek grammar. The repetition of the pronoun may
bring an element of emphasis that may account for the possible relative
infrequency of violations of the rule (if indeed they are infrequent, I
have not counted any sizable corpus of the construction). Yet, in many
contexts the emphasis may be felicitous (it is not so rare to describe a
personal relationship to something in two different way). There are
probably more contexts in which the violation of Smart's rule is
felicitous than contexts in which the violation of Sharp's rule is
felicitous, though I have not made a statistical study.

Sharp's rule does point out an interesting principle, though I would not
regard it as an absolute rule.
Yet, to deviate from Sharp's rule (with the usual qualifications of the
rule) would be fairly marked and should cause the exegete to look for a
reason for the deviation in the context. (One thing that you learn in
linguistics classes is that very often it is possible to conjecture a
context in which the weirdest clauses are felicitous. Therefore to make
a rule like Sharp's absolute needs great caution.)

So, yes, I think the approach to both rules should in principle be the

And finally, YES, both are context dependent, though Smart's probably
much more so.

Just my vote in the poll,


Dan Parker wrote:
> I have just had a private dialogue with Mark Wilson with regards the
> question I posed in an earlier post where I said:
> "Mark,
> I would assume that you also will accept the "context" clause
> in Uncle Bob's rule to be appended to the end of Sharps, is
> that correct?"
> Mark answered, "Yes. That would be correct."
> Does anyone on this list agree with Mark on this? I would like to hear
> particularly with those who are oposed to Smart's rule.
> Thanks in advance,
> Dan Parker
> ---

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 - 15:36:40 EDT