Page 1 of 1 (19 items)
This post has 18 Replies | 7 Followers

Posts 688
Robert Peters | Forum Activity | Posted: Thu, Apr 9 2015 6:35 AM

Is there the plan to have a Hebrew Grammatical Constructions dataset? 

Posts 1419
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Thu, Apr 9 2015 6:44 AM

It is something we're considering. Is it something that y'all would find useful? Let us know (please respond on this thread).

Thanks!

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 688
Robert Peters | Forum Activity | Replied: Thu, Apr 9 2015 6:46 AM

Yes! It will help me when I am Preaching and Teaching through Old testament passages. It will help me immediately as I write me exegetical paper for Seminary. 

Posts 912
David Knoll | Forum Activity | Replied: Thu, Apr 9 2015 7:03 AM

What constructions are you considering?

Posts 1586
John Kight | Forum Activity | Replied: Thu, Apr 9 2015 7:33 AM

I agree this would be helpful!

For book reviews and more visit sojotheo.com 

Posts 1419
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Thu, Apr 9 2015 7:54 AM

David Knoll:

What constructions are you considering?

What constructions should we consider?

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 912
David Knoll | Forum Activity | Replied: Thu, Apr 9 2015 7:58 AM

I really don't know. It has to be something that we cannot achieve using morph or syntax searches. I don't see anything which could be of interest to me in the current offering (mostly Greek NT) but I use Hebrew and I am looking for examples constantly so if it would save me time...

The Septuagint also comes to mind (or is that too small a market?).

Posts 1586
John Kight | Forum Activity | Replied: Thu, Apr 9 2015 8:04 AM

I just entered into the my first Biblical Hebrew class four weeks ago, and know little more at this point than the Construct Chain. However, like the Greek counterpart, I see this functioning as a learning aid for those of us entering into formal training in the language--especially if accompanied by an explanation/example resource. Not to mention the search capabilities that would assist in exegetical research.

All that to say, I don't know what specific constructs should be considered but I do think it will be beneficial all around when finished.  

For book reviews and more visit sojotheo.com 

Posts 688
Robert Peters | Forum Activity | Replied: Thu, Apr 9 2015 9:59 AM

I would say the different verb construct chains

Also on the noun side something that identifies the various construct relationships. I am just diving into Hebrew so I would like to see one more qualified then myself to make some suggestions 

Posts 1526
Forum MVP
Fr Devin Roza | Forum Activity | Replied: Thu, Apr 9 2015 10:06 AM

Rick Brannan (Faithlife):

What constructions should we consider?

One dataset I'd like to see in Hebrew (and in Greek) is the classification of clauses into their types (main, conditional, causal, comparative, etc.). I remember a while back a question came up on the Faithlife Syntax Search group about how to search for X within main clauses in Hebrew... no one has answered so far (if you know, please share!), using Andersen-Forbes. Being able to identify main clauses might be interesting for the study of modality and the imperfect. And, of course, being able to search within types of subordinate clauses could be incredibly useful.

Along the same lines, maybe a classification of asher-relative clauses. I wonder if it would be interesting to include a distinction between resumptive and non-resumptive.

Verbless clauses might be interesting as well. Then, within those, clauses of identification vs classification.

As well, I am of the idea that while constructions that can't be found by Syntax Search maybe should be prioritized, I do think that Faithlife should strive to create datasets for the main categories that can be found using ONLY Syntax Search. The reason is that Syntax Search is very difficult, and also because of how useful it is to be able to combine all the different types of datasets / search extensions / search fields, something oftentimes not possible in Syntax Search.

Posts 816
LogosEmployee
Eli Evans (Faithlife) | Forum Activity | Replied: Fri, Apr 10 2015 1:35 PM

I'll bite. Clauses in Andersen-Forbes have an "Is Embedded" property, which measures whether the clause is descended from another clause. (You won't find this in their write-ups of the database because as it happens, I added it myself, back in the day. "Is Embedded" is calculated by examining each clause's lineage upwards in the graph. If no clause nodes are encountered, it is not embedded, and if any clauses are, it is embedded.)

So you can search for something that looks like this:

And you get all of the non-embedded clauses that are predicated with a finite verb. That's preeeeeeeeeetty close to main clauses. (Depending on what you mean by "main," of course.)

Posts 1526
Forum MVP
Fr Devin Roza | Forum Activity | Replied: Mon, Apr 13 2015 5:34 AM

Eli Evans:

And you get all of the non-embedded clauses that are predicated with a finite verb. That's preeeeeeeeeetty close to main clauses. (Depending on what you mean by "main," of course.)

Thanks Eli. "Not embedded" looks promising, but I tried this out now and am getting mixed results.

Most of the clauses I'm getting in the search results are clearly main clauses. Some maybe are arguably main clauses (Hebrew Syntax is soooo undefined and unclear sometimes...), but others seem to me to be clearly subordinate clauses. For example, I'm getting multiple hits with this search in Deut 4:40 that seem pretty clearly to be subordinate clauses in Hebrew. The first is an "asher" clause that is translated as "keep his statutes... that it may go well with you", and then following that with a "lemaan" phrase translated as "and that you may prolong your days". Both of these are final clauses, and are even classified as such in Andersen-Forbes ("Aim/Purpose").

There are a lot of other results similar to these (in the verses I reviewed, I especially noted it with with "pen" and "asher"... although maybe some would consider "pen" clauses to be main clauses? Without having studied the issue in depth, I would classify "pen" clauses as negative purpose clauses). Maybe a problem with some missed structures when you were applying the "Not Embedded" classification?

Posts 94
Rokas | Forum Activity | Replied: Mon, Apr 13 2015 6:05 AM

I would love to see it!

Posts 816
LogosEmployee
Eli Evans (Faithlife) | Forum Activity | Replied: Mon, Apr 13 2015 11:06 AM

Right. This is where you run into the boundaries of the Andersen-Forbes grammar/discourse theory. All the clauses in Deuteronomy 4:40 are embedded within a higher-level discourse structure (the sentence) rather than being embedded within a clause. They're subordinate but to a sentence rather than another clause. This is because Andersen and Forbes see a smooth gradation between phrase, clause, and discourse levels, which sounds simple in theory, but in practice means you'll be searching for a lot of permutations, because a "slot" in your query can be fulfilled by a supra-clausal (discourse) unit, or a clause, or a phrase — or a single segment, because A-F doesn't annotate one-word "phrases" as such.

So, putting my foot syntax searching abilities where my mouth is: I just shared a document to the Logos Syntax Searching group entitled "Hebrew main clauses?"

https://documents.logos.com/documents/4b1a1db6183b49819c7bf63282ec8603

The permutations this query matches are:

1) A non-embedded clause with a parent-count of zero and a finite verb.

2) Non-embedded finite-verb clauses directly descended from discourse units that:

  1. appear first in the discourse unit without an introductory cue phrase; OR
  2. immediately follow another clause without an intervening cue phrase; OR
  3. are introduced with a single conjunction that is anything but a subordinating conjunction; OR
  4. are introduced with a single preposition that is anything but a subordinating preposition; OR
  5. are introduced with a cue phrase that does not contain a subordinating conjunction or preposition.

I made this by trial and error. I know am pretty sure this correctly selects all of the main clauses in Genesis 1 and Deuteronomy 4:40. I think it would perform well on other texts, but I didn't have time to check it. There may be a few more variants I haven't accounted for.

Posts 524
Sarel Slabbert | Forum Activity | Replied: Mon, Apr 13 2015 12:31 PM

Hi Eli, please repost the link. I get a "page does not exist" error when following the link.

Posts 1526
Forum MVP
Fr Devin Roza | Forum Activity | Replied: Mon, Apr 13 2015 12:39 PM

Sarel Slabbert:

Hi Eli, please repost the link. I get a "page does not exist" error when following the link.

Sarel, you may need to join the Syntax Search Faithlife group - https://faithlife.com/logos-syntax-searching. It is in the Documents tab there, and Eli's link should work if you are a member of that group as well.

Posts 570
Schumitinu | Forum Activity | Replied: Mon, Apr 13 2015 6:55 PM

They are probably not what we would classify as Hebrew grammatical constructions. But what about case frames / semantic roles for the Old Testament? I'd like to see that developed as well.

Posts 234
Joel Madasu | Forum Activity | Replied: Tue, Apr 14 2015 7:17 AM

Fr Devin Roza:
I remember a while back a question came up on the Faithlife Syntax Search group about how to search for X within main clauses in Hebrew

I remember asking a question like this. It was/is hard to find different clauses. I was trying to find יש in independent, dependent and embedded clauses, but I couldn't. AF database is so complicated. I know they took more then 30 years to produce such a great work, and used a theory: "Enhanced" Phrase Structure grammar/rules... but I wish at times that this is less complicated. 

What I would like to see is to break down or at least indicate the Independent/Main, dependent, embedded, relative clauses, or classes that indicate polarity - yes/no questions, etc., and then indicate the use/function of a word/item/constituent in that clause. Also, if a word/item is a discourse marker, then it will be good to indicate that as well. 

Posts 11
John de Hoog | Forum Activity | Replied: Thu, Aug 27 2015 4:07 PM

I think it would be great to be able to save a syntax search as a "clause list" rather than a passage list or a word list. For example, I ran the "subject" search that comes "standard" in the syntax search window for AFPMA, and then i thought it would be great to be able to save that as a clause list which I could then further interrogate - for example, by asking of that clause list which clauses also contained a predicative participle. If THAT search could then be saved as a clause list then further interrogation would be possible - e.g. search on segment within the constraints governing the first two searches. Something like that could make the AFPMA a lot more accessible for people who have syntax questions to ask because syntax questions could be built up gradually rather then trying to fit all the constraints into one huge syntax diagram. Any possibility that this could be considered?

Page 1 of 1 (19 items) | RSS