Syntax Search-Louw-Nida suggestion...and Granville sharp revisited...

+1

The problem is that, in my mind....the best search would come as an ability to do MULTIPLE LN commands...like LN 12 BUTNOT LN93

LN12 does exclude LN93 (and every other LN besides 12) ==>  not sure what you mean here, UNLESS ...

You can specify multiple LN eg. 12 93  (space between = OR)

Dave,

I was thinking specifically for the Granville sharp rule.

NOT a proper name but YES to "deity"

No on 93 but yes on 12....all in the same search....I don't think that there is a way to do that...right?

In my mind this would narrow the search to more relevant hits for the Granville sharp search..

NOT a proper name but YES to "deity"

A word in the syntax databases is assigned a single context-sensitive LN number. So we only need search for LN 12! The only overlap is g:Satan which can be both a proper name and a title of deity, but we have to trust its interpretation as LN 12.

Dave,

I guess what I'm trying to say is this: wouldn't it be useful if you could say:

word = names BUT NOT names of places

I'm just wondering because on the granville thing; we are looking for supernatural beings but not proper names....

Dave,

I guess what I'm trying to say is this: wouldn't it be useful if you could say:

word = names BUT NOT names of places

I'm just wondering because on the granville thing; we are looking for supernatural beings but not proper names....

LN12 is exactly that ie. supernatural beings but not proper names (check this out in the Louw-Nida lexicon).

What you are saying only makes sense if a word is tagged with both LN12 and LN93

Dave,

I think I get what you're saying; ..hmm...

What if you wanted to NOT use "LN 12 supernatural beings" but you DID want to exclude names of places and only use titles given to people; i.e. "lord"?

What if you wanted to NOT use "LN 12 supernatural beings" but you DID want to exclude names of places and only use titles given to people; i.e. "lord"?

If there is no suitable LN domain for titles then you might have to use specific Greek words in the query. Search LN Lexicon for "lord" "master" etc and those words will give you an LN domain to which they belong.

Dave,

it looks like my dream of an accurate Granville sharp search is going down the drain...

Dave,

I guess what I'm trying to say is this: wouldn't it be useful if you could say:

word = names BUT NOT names of places

I'm just wondering because on the granville thing; we are looking for supernatural beings but not proper names....

Subdomains are your friend. LN93A is person names, LN93B is place names. You can do this by specifying person name subdomain (93A), right?

Granville-Sharp's first rule is not supernatural-being-specific though that is the primary application of folks who use it in the NT. I'd also posit (along with so many other syntactic 'rules') that GS is more of a codification of a trend of a relatively small corpus than a linguistic rule per se. That is, the specification of the semantics of the lexical portion as applied to the NT is soft and relatively subjective.

There are other GS issues — some say it is two nouns, but if you read Wallace you'll see he takes it basically as two items joined by και that agree in case. Thus it could be a participle and a noun, or a noun and an adjective (and Wallace has examples of both of these). If you're really interested in Granville Sharp's rule, then it might be worth reading Wallace's book ("Granville Sharp's Canon and its Kin: Semantics and Significance") which is from what I understand a revision of his doctoral thesis. It would certainly have the best/most authoritative list of instances adhering to Sharp's criteria, as discerned by Wallace.

There is so much non-syntactic sifting that goes on with potential results that adhere to GS that I'd be surprised to have a one-search 100% 'accurate' solution (outside of searching a text tagged with 'Granville Sharp instances' for such instances). Yes, we should try to use these tools to test these sorts of ideas — but testing here means (at least as I see it) gathering data and checking the possibles/probables to see if the rule is really a rule. You can see a relatively worked non-GS example (using LDLS3.x and OpenText.org) in a paper of mine here: http://www.supakoo.com/rick/papers/ETS2007NationalPaper.pdf.