Double Accusative | Object-Complement in Matthew

[rwrobinson88]

Greetings.

 

I have spent some time today trying to nail this search down to learn a bit more about the Greek Database. I'd love some feedback on this search. 

 

I have other syntax databases which make this search significantly easier. I did that search and then was checking my searches in this database to see the results. I wanted to keep the range large enough yet not too much data. So, I chose Matthew. 

 

So, I found that I needed to create a total of three construct searches in order to get good results. Here they are:

 

 

 

 

The only difference between the second and third search is that I had to change the first complement phrase's phrase structure to one segment rather than all segments. This was to get Matthew 20:12 back as a result. 

 

There are still false hits. They are all because they are prepositional phrases marked as complements:

Matt 6:13 - Complement as a Prep. Phrase

Matt 10:17 - Complement as a Prep. Phrase

Matt 10:41 - Complement as a Prep. Phrase

Matt 13:33 - Complement as a Prep. Phrase

Matt 13:48 - Complement as a Prep. Phrase

Matt 18:8 - Complement as a Prep. Phrase

Matt 18:9 - Complement as a Prep. Phrase

Matt 21:46 -  Complement as a Prep. Phrase

Matt 26:50 - Complement as a Prep. Phrase

 

The is one that should have come back as a hit but didn't and I couldn't figure out why:

Matthew 27:13

 

Here are all the hits:

Matt 1:21, 23, 25; 3:3, 9; 4:19; 5:36; 6:13; 7:9; 10:17, 25, 41-42; 12:16, 33; 13:33, 48; 14:5, 30; 15:32; 16:26; 18:8-9; 19:4; 20:12, 28; 21:13, 24, 26, 46; 22:43, 45; 23:15; 26:50, 73; 27:31; 28:14

 

Three things:

1. If anyone could help me to get ride of the false hits above, that would be great. 

2. If anyone could show me how I could of made this more simple, that would be great.

3. If someone could should me how/if I'm missing something with Matthew 27:13, that would be great.

Edited October 29, 2017 by RyanWRobinson

[mgvh]

I don't have the Syntax database in Accordance, so I tried to see what I could find using the Cascadia Syntax Graphs of the NT in Logos.

Here are the results it returned:

Mt 1:20, 21, 23, 25; 3:3, 9; 4:19; 5:36; 7:9; 10:25, 42; 12:10, 16, 33; 14:5, 30; 15:9; 16:26; 19:4; 20:12, 28; 21:13, 24, 26; 22:43–44, 45; 23:9, 15; 26:73; 27:13, 31; 28:14

 

I think those are all correct results, and it picks up some your search missed. What's nice about the Cascadia system is that they have all double accusatives tagged as Object and Secondary Object. All I needed to do was search for verses with a Secondary Object.
 

[Λύχνις Δαν]

Hi Ryan,

 

  I haven't yet tried the searches fully but I am not understanding where the double acc is in Mt 27:13 ?

 

  All I see is this text : Μαθθαῖον 27·13 τότε λέγει αὐτῷ ὁ Πιλᾶτος· οὐκ ἀκούεις πόσα σου καταμαρτυροῦσιν;

 

  I've set your searches up now and I am not getting the same hits exactly. I assume you are ORing all three together right ?

 

Thx

D

Edited October 29, 2017 by דָנִיאֶל

[rwrobinson88]

mgvh - That is the other syntax database I used and referenced above. Significantly more simple. The point of this search is to be learning the database. One thing I did catch was there was one hit in the accordance database that the Cascadia search didn't catch because of Cascadia's annoy "gap." That feature is so hard to figure out that I just left that be. Also, the hits that come back from the Cascadia search that are not in the hits from accordance's database is a matter of disagreement. The disagreement is AccSyntax marks some as appositive rather than another complement.

 

 דָנִיאֶל - Matthew 27:13 is a double complement in the database. But, I was doing this past by bedtime and had a brain fart on σου. It's a genitive. Simple as can be. I just wasn't thinking clearly or hard about it.

 

In light of this, two requests remain:

 

1. If anyone could help me to get rid of the false hits above, that would be great. 

2. If anyone could show me how I could have made this more simple, that would be great.

[Λύχνις Δαν]

Ok so that suggests that Cascadia has has tagged some non-acc as a secondary object. I don't have their db so cannot check.

 

Here are my hits with your queries:

 

Μαθθαῖον 1·21, 23, 25; 3·9; 4·19; 5·36; 7·9; 10·25, 42; 12·16, 33; 14·5, 30; 15·32; 16·26; 18·8; 19·4; 20·12, 28; 21·13, 24, 26, 46; 22·43, 45; 23·15; 26·73; 27·31; 28·14

 

(It would be really nice to have a Jupyter notebook style sharing option in Acc., just in passing, but I've attached my ws.)

 

Matt 1:21, 23, 25; 3:3, 9; 4:19; 5:36; 6:13; 7:9; 10:17, 25, 41-42; 12:16, 33; 13:33, 48; 14:5, 30; 15:32; 16:26; 18:8-9; 19:4; 20:12, 28; 21:13, 24, 26, 46; 22:43, 45; 23:15; 26:50, 73; 27:31; 28:14

 

The extras in yours are : 3:3, 6:13, 10:17, 10:41, 13:33, 13:48, 14:15, 14:30, 18:9, 26:50.

 

More later once I've studied it a bit more but I attached my workspace in case you can see any discrepancy in my implementation of your searches.

 

Thx

D

 

RyanDoubleAccusative.accord.zip
 

[rwrobinson88]

That's super weird. I just downloaded your workspace and searched. The extras you speak of are coming up as hits. See below:

 

[Λύχνις Δαν]

Ok I think I see the difference. I had a depth 0 on the segmented complement where you have a 2.

Now I get 44 total hits in these vs:

 

Μαθθαῖον 1·21, 23, 25; 3·3, 9; 4·19; 5·36; 6·13; 7·9; 10·17, 25, 41-42; 12·16, 33; 13·33, 48; 14·5, 30; 15·32; 16·26; 18·8-9; 19·4; 20·12, 28; 21·13, 24, 26, 46; 22·43, 45; 23·15; 26·10, 50, 73; 27·31; 28·14

 

You original report :

 

Matt 1:21, 23, 25; 3:3, 9; 4:19; 5:36; 6:13; 7:9; 10:17, 25, 41-42; 12:16, 33; 13:33, 48; 14:5, 30; 15:32; 16:26; 18:8-9; 19:4; 20:12, 28; 21:13, 24, 26, 46; 22:43, 45; 23:15; 26:50, 73; 27:31; 28:14

 

Ok so I now have one extra you did not originally report which is 26:10.

 

Thx

D

[rwrobinson88]

That's weird. That's how it came up when I opened it (in other words, I didn't change that). My original search didn't have the depth at 2 on the segmented.

 

Again, in my original search, I was doing three searches in one with the OR function. That is likely why 26:10 didn't come up (which is a false hit).

[Λύχνις Δαν]

Ok so silly me I saved a version that I didn't run - or at least wasn't using for comparison. I must have been experimenting. But in that case I don't understand how the "extra" hits appeared in your example. I only get 33 hits with depth = 0, 44 with depth 2.

 

Again, in my original search, I was doing three searches in one with the OR function. That is likely why 26:10 didn't come up (which is a false hit).

 

As to how you did the search I assume you did what my workspace shows, OR'ing constructs together in the search tab. Is that what you mean ?

 

I am doing this on 12.1.4. Is that what you are using ?

 

Hang on ! I've just noticed that what I sent you is not even remotely complete - I had three construct OR'd together. Try this new version.

 

RyanDoubleAccusative.accord.zip

 

thx

D

[rwrobinson88]

I pulled it up. There must be a problem with sharing workspaces because it's only coming with one construct.

[rwrobinson88]

Also, with the three searches in the original post, I'm getting 43 hits. They are all done with 0 depth.

[Λύχνις Δαν]

I was able to get rid of a bunch of hits on your prepositional phrase list of false positives but at the cost of some real hits I think also. I need to look more at this now that I'm getting close enough. But what I did was alter the 3rd construct to this :

 

 

Thx

D

[rwrobinson88]

I was hoping that I could make that search another link in this search and tell it not to include those hits. But, I can't figure out how to do that. 

[Λύχνις Δαν]

I'm going over this now more slowly re-deriving the searches from scratch. I suspect that there may be a case to made that Matt 1:20 could be re-tagged as double complement rather than as apposition. Most translations seem to handle it that way. I suspect you would have to translate as "do not be afraid to receive [ or accept rather than take  ] Mary, your wife, ...." if it was appositional, which I don't recall ever seeing. Of course arguing from translation is dodgey if overplayed so similar constructions and more research would be required.

 

EDIT: In any case it's something of an outlier in the false negatives, if indeed it is one, because it's really excluded only because of the appositional tagging.

 

EDIT: Not sure that Mt 12:10 should be considered double accusative or noun with attributive adjective. It might be interesting to see why Cascadia thinks it should be included. I don't have the chart so cannot even guess. Likewise Mt 22:44 doesn't look like double acc to me. I could see how 23:9 might be considered double object I guess - more or less appositional but it's nominative in a S clause so we'll never find it in Accordance with these queries. Also 27:13 - not sure what the Cascadia classification is here. As mentioned before this will not be found as an accusative but I'm not sure it should be a double object in any case.

 

EDIT: Mt 15:9 another apposition versus double acc. question.

 

So I think that takes care of or at least accounts in some way for a number of questionable items. I now have a number of hits that Cascadia does not have which I need to look into and perhaps weed out somehow.

 

EDIT: And finally figured out what I'd done wrong on trying to share my workspace. Here is the one I've been working on and the Alt tab is the one where I have been trying to do this in a single search.

 

RyanDoubleAccusative.accord.zip

 

Thx
D

Edited October 29, 2017 by דָנִיאֶל

[rwrobinson88]

Are some of your comments on the proper tagging of the database itself? I wasn't seeing the relevance to the search itself. Please help me understand if I'm being a bit slow here. 

 

For some reason, when I add the prep to the complement phrase, it doesn't change the results as it seems to in your search. Can you help me understand why maybe this isn't working for me?

 

From what I can see, I set my search up identical to Part I-III on your search and I have 44 hits and your query has 32 hits. 

 

I don't know how to attach my ws. I would so that you could check it out.

Edited October 30, 2017 by RyanWRobinson

[Λύχνις Δαν]

Yes many of my comments above were about tagging in Acc or in Logos. I have not really much studied the double accusative construction but some cases that appear in the list Mark provided seem to not be double object. I assume that in some cases the double object tag is being applied where the object is not in the accusative which I'm fine with in principle. But I am not sure if they are double object or attributive adjective. Some of these cases simply have to be thought through and worked out.

 

I'm not quite sure why the not prep doesn't help but it may not do much in the 3-construct search because of other searches interfering. I also turned off search both directions as that was causing me issues. I have one true and obvious false negative as a result, Mt 3:3, which really should be included. I don't think I can see how to include it without either enabling search both dir or adding a second search.

 

Thx

D

[rwrobinson88]

So, I figured out the search that would make them a hit:

 

 

When I search that, I get 12 hits in Matthew. But, if I negate the preposition I don't get any hits. 

 

But, what I found is the second search is getting the hits back with the prepositions:

 

 

So, when I run the three searched linked, I'm only getting the hits for the first two (44 total). I really have no clue how you're getting rid of the hits with prepositions in the complement....

 

edit: I was wrong. It wasn't the picture above that was giving me the complements with the prepositions, it was the other search.

Edited October 30, 2017 by RyanWRobinson

[Λύχνις Δαν]

I no longer recall exactly how I did that except that I added the not prep, but I think I may also have at about that point turned off search both directions. The problem is that the search is almost symmetrical so it's possible that it could match either way. In fact your second search above is symmetrical. So it's possible that it's finding the hits you want to exclude by flipping the query around. Try turning that off and see.

 

Thx

D

[rwrobinson88]

So weird!!! I search your Part I and I get no hits with the with the prepositional phrases. I search mine which seems to me identical, I do get them. The only difference I can see is your noun in the first complement phrase doesn't have an accusative tag. But, I took that off and searched and it did nothing. 

 

I'm lost on this one haha.

Edit: ahhhhh there it is!! it was the searching both directions. Good call!

Alright. Now I am at 33 hits with no prep phrases.

Edited October 30, 2017 by RyanWRobinson

[Λύχνις Δαν]

Yay !

[rwrobinson88]

Hits: Matt 1:21, 23, 25; 3:9; 4:19; 5:36; 7:9; 10:25, 42; 12:16, 33; 14:5, 30; 15:32; 16:26; 18:8; 19:4; 20:12, 28; 21:13, 24, 26, 46; 22:43, 45; 23:15; 26:73; 27:31; 28:14

[rwrobinson88]

Looking at the hits for the whole NT in Cascadia, there is a total of 241. When I do this search in all of the NT in the AccSyntax Database, I get 194 hits...

 

hmmm. Probably need to look into this one a bit more. But, I'm happy with Matthew for now.

[rwrobinson88]

Just peering into Mark has me shuttering. I just don't know if this double complement search is cut out for this database to be quite frank.

[Robert Holmstedt]

The structure of the Accordance syntax databases in general can indeed account for double complement searches. In fact, it was made precisely for such things (and I use the Hebrew database for this kind of complex search all the time). 

 

I'm too busy to look into Greek right now, but why are you constructing these searches with the "within X" feature? Why not a predicate phrase containing two complement phrases that each specify a complement as a noun?

Edited October 30, 2017 by Robert Holmstedt

[Λύχνις Δαν]

I think there was at least one case, perhaps more, where the complement phrases were not adjacent. I would have to go back and check in detail.

 

Thx

D