Verbs with Two Complements

[Peter Bekins]

Two questions:

 

First, I put together a simple search for verbs in the Hifil Binyan that are governing two complements. When I simply ask for two complements within the predicate phrase it works fine:

 

[Predicate Phrase]

[Predicate]      [Complement Phrase]      [Complement Phrase]

[Verb: Hifil]

 

Next I wanted to exclude all the complements marked by prepositions, so in each complement phrase I specified that the first item should be a noun, definite article, or object marker:

 

[Complement Phrase]

[Particle (Object, Article)]

[Noun]

[Place = 1] 

 

After I run it like this, I am also getting hits where the verb governs a single complement but that complement is itself composed of a compound phrase, for instance

 

Gen 3:16 ‏הַרְבָּ֤ה אַרְבֶּה֙ עִצְּבוֹנֵ֣ךְ וְהֵֽרֹנֵ֔ךְ 

'I will in fact increase your pain and your conception'

(which BTW is a proper example of hendiadys, I forget who brought that up on a different post) 

 

I cannot figure out any way to exclude these. The problem seems to be that the NPs inside the coordinated phrase are treated as being at depth 0 themselves.

 

Second, when there are three items (A, B, C) in a construct, does search both directions literally mean both directions so that you only hit on A + B + C or C + B + A? Or does it look for every possible combination, like B + C + A or A + C + B?

 

Thanks!

Pete

[Λύχνις Δαν]

Something rather similar came up a couple of weeks back in Greek I think. Not the same but similar.

 

I rebuilt your search and I see 3:16. I can get rid of it by adding a WITHIN 1 attached to both the Complement Phrases. The problem is that it excludes tons of your other hits also. I suspect it throws out a lot you might want to keep. My hits dropped from 818 to 114 with the WITHIN.

 

So next I tried putting a negated conjunction between the two phrases and set the WITHIN to 1-5. Interestingly 2:9 shows up because the WITHIN distance limit is too short. But again a dramatic reduction in the hit count, more so than before.

 

Gen 5:32 shows another interesting case in that the compound complement is of 3 parts, two connected asyndetically and two with a conjunction. Not sure whether you want that one or not.

 

Ok I'll give you what I have before disappearing down too many rabbit holes and see if it helps at all.

 

Thx

D
 

 

PeterBekinsPredDblCompPhrs.accord.zip

[Peter Bekins]

Thanks Daniel, I will look at your ideas. I was hoping to do it without adding anything fuzzy (like WITHIN). 

 

Pete 

[Λύχνις Δαν]

One other thing that came up in the other thread was the idea that a compound phrase could be excluded directly by use of a new option on the phrase. It exists on the Complement itself but not on the Complement Phrase

 

Thx

D

[Peter Bekins]

One other thing that came up in the other thread was the idea that a compound phrase could be excluded directly by use of a new option on the phrase. It exists on the Complement itself but not on the Complement Phrase

 

Thx

D

 

"Could be" as in if they added this functionality in the future? This is actually what I was getting at originally. We need some precise way to interact with these compounds. Unfortunately, adding another check box adds to the complexity of the construct dialog.

 

Pete

[Robert Holmstedt]

I admit that I also don't like adding fuzzy items, like "within," in syntax searches. 

 

Pete, for future queries, please try attaching workspaces as screenshots or zipped workspaces. It saves me the time of reconstructing that you've done.

 

My search is attached. Rather than specifying "noun," etc., I simply negated PPs.

 

The problem seems to be that the phrase boundary is not being processed correctly, so that two items in the same compound phrase are registering as two comp phrases; this is also the problem behind the hits with compounds, as in Gen 3.16. I've turned this in to the programmers. 

 

 

Hifil_NPCompNPComp.accord.zip

[Λύχνις Δαν]

"Could be" as in if they added this functionality in the future? This is actually what I was getting at originally. We need some precise way to interact with these compounds. Unfortunately, adding another check box adds to the complexity of the construct dialog.

 

Pete

 

I asked for this feature in https://www.accordancebible.com/forums/topic/22110-double-accusative-object-complement-in-matthew/page-2?do=findComment&comment=108263and Robert thought it a reasonable proposal but it went no further in that thread because it turned out for that case that the compound selection on the Complement was adequate. In this case I don't think it is.

 

But I don't know if this yet required if the phrase boundary issue Robert mentions above is dealt with.

 

Thx

D

Edited November 25, 2017 by דָנִיאֶל

[Peter Bekins]

Sorry Rob, I guess I have to learn how to export a workspace.

 

Did you run that search in 12.2? When I run it in 12.1.4 the first two hits are bivalent rather than trivalent:

 

Gen 2:6 ‏וְהִשְׁקָ֖ה אֶֽת־כָּל־פְּנֵֽי־הָֽאֲדָמָֽה

 

Gen 2:10 ‏לְהַשְׁק֖וֹת אֶת־הַגָּ֑ן

 

 

EDIT: figured it out, you used complement instead of place=1 so it seems that since את and גן are at the same level in the phrase they are counted like two complements. I see the phrase boundary problem clearly now.

 

I didn't negate the prepositions because that never works right for me, but I am glad to know the larger problem is being looked at.

 

Pete

 

Edited November 25, 2017 by Peter Bekins

[Robert Holmstedt]

Pete, just use “save as” in Accordance and name the Workspace. To upload to the Forum you have to compress it to a zip file, though. Screenshot might be simpler.

[Peter Bekins]

Has any progress been made on the phrase boundary problem? I have tried to do some double-complement searches again but I am having the same issues with two items in the same complement counting as a hit when I really want two separate complements.

 

Pete

[Robert Holmstedt]

Not that I'm aware of. Let's nudge it up.

[Peter Bekins]

Thanks!