Search help using Greek Construct

[mgvh]

Can someone tell me what I'm doing wrong here...

I'm trying to find all the instances of πιστευω that have an accusative direct object. When I first ran the search, most of the results were ones where the verb is followed by the preposition εις and its accusative object. So I went to the Greek construct (attached) and set it up so that the preposition should NOT appear between the verb and the object / complement.

Why do I still end up getting results with the preposition? (BTW, it doesn't matter if I use the Complement label or not. Also, I can confirm that I used the preposition εις and not the number one εις.)

[mgvh]

So I thought I'd try another way. See attached. This time I used the INTER command and negated it to eliminate any preposition or οτι. Still the results come back with those very things showing up. (BTW, in this and the previous, I unchecked the search both directions.)

Thanks for any guidance!

[David Lang]

Using the negative INTER is the right way to exclude items, but I don't understand why having both οτι and the preposition in the same INTER is not working. It should exclude both items, so I'm guessing this is a bug.

 

Until we get it fixed, you can use two separate negative INTER items and the search works correctly.

[Joel Brown]

I need to double check, but I’m pretty sure the single INTER is saying to exclude all words that are *both* oti and a preposition, rather than *either*.

[David Lang]

You're right, Joel. Normally, if two items in a column or INTER are mutually exclusive, like the NOUN and PRONOUN in the second column, it's treated as an OR. If they are compatible, then it's as if the two items are joined by an at symbol (@). In this case, a lexical form which happens to be a conjunction was joined with a preposition tag, and since a lexical form can be joined to a tag, Accordance treated it like an @ search.

 

I was thinking that they were mutually exclusive because I knew oti was not a preposition, but of course, Accordance doesn't factor in a lexical form's part of speech.

[mgvh]

Super! That does it. Thanks so much!

[mgvh]

Ok, now a follow up question... Using the right construct, I see that an accusative pro/noun occurs 65 times after ακουω without an intervening preposition or οτι.

If I use the Analysis, I can see that αγαπην is one of the pro/noun hits.

How do I find in which verse this hit occurs?

If I use a Concordance tab, it just provides a list of all the times that αγαπη occurs in the NT.

If I triple click on αγαπην, I get all the instances of the inflected form, not just the result with ακουω.

Again, I appreciate your help.

[David Lang]

Drag a LEX item into the same column as the NOUN and PRONOUN tags and specify αγαπη. That will constrain the pro/noun to αγαπη.

[mgvh]

Thanks, David. Yes, that's the way to find that instance with that particular object.

I should have been clearer. My question is actually bigger. I'm trying to find a way to get all the results grouped according to the object with references.

I can indeed do them one at a time, but I was hoping there was another way.

[Λύχνις Δαν]

Hi Mark,

 

  I was wondering if what you really wanted was an amplify back from word in the analysis that would amplify back to the cases in the source text, subject also to the constraint of the original search. But now it seems like that would only get one example for what you want. So I'm interested, what do you want your final output to show ? I usually take this sort of approach when I reach the limit of what Acc can do for me and then I work out if/how to script the rest from there. But it would also help to understand what you want if you need a new feature or extension of one to get it directly.

 

Thx

D

[David Lang]

Mark,

 

The only way to do what you are asking would be to have some kind of REFERENCE item that could be dragged into a column in the Analysis Display, which would list all the references where each item was found. It's a feature I've suggested in the past, but alas, it hasn't happened yet.

Edited October 13, 2019 by David Lang

[mgvh]

David: Yes, a REFERENCE item is what I was thinking would do it.

Ruby: Thank you! I have tried another way of accomplishing what I was trying to do, and I'm getting closer, but maybe you can point me straight.

What I am trying to study are instances with a verb like πιστευω that takes either an accusative or dative direct object. (My other example is ακουω which takes either accusative or genitive direct object.)

I want to see how often and in what contexts which case is used.

 

What I finally figured out I could do is set things up in the Greek Construct and use the CLAUSE option and use my verb (using LEX) and then establish the COMPLEMENT and add NOUN and PRONOUN with the desired cases specified. Further tweaking, and I seem to get better results setting the Clause Depth to infinity, but choosing "limit depth to clause boundary." See attached graphic.

I'm still getting hits with the object as part of a prepositional phrase. (I was trying to weed those out before. Maybe I'm misunderstanding what a "complement" is?) E.g., I get 1Tim 1.16: πιστεύειν ἐπ᾿ αὐτῷ when looking for πιστευω with a dative object.

 

Or, I'm not getting hits like Eph 1.15 or Phlm 5 (ἀκούων σου τὴν ἀγάπην) when searching for ακουω with an accusative object.

For this latter, I tried yet another approach going through the command line directly using:  [PREDICATE] @ἀκούω <AND> [COMPLEMENT] @ ([NOUN accusative], [PRONOUN accusative])

That picks up all the verses I think I want, but a lot of other hits that do not apply.

 

Again, thanks for your guidance!

Edited October 13, 2019 by mgvh

[Λύχνις Δαν]

Hey Mark,

 

  This is a little tricky. There is a bit going on here. Here is a search which gets both Phil 5 and Eph 1:15 but it cannot get rid of 1Tim 1:16. I can explain why below. Alas there is strange highlighting in Eph 1:15 though and this points to another difficulty which concerns me because it probably indicates other hits are missed with my search. So here's the search with the Eph 1:15 hit on screen:

 

 

  The reason I concerned is that πίστιν is not highlighted. The reason is because of the not prep element in the search. I was hoping that because the prep καθ᾽ was part of the subordinate adjunct phrase that it would not be considered a hit. As it is that suggests that other like constructions would also be excluded. The only reason that Eph 1:15 is still present is because the complement phrase in which this adjunct occurs is a compound and that the other half of the compound is of slight different structure - not beginning with a prep. Now, you may care or you may not depending upon what precisely you are trying to do. If you just want some examples to study then being comprehensive may not matter. If you really do want all cases then this is a potential issue and some cross-check will be required.

 

  Now, turning to 1Tim 1:16 here is a screenshot with that hit visible:

 

 

  Now here the problem is that because of the depth setting the hit is included because there is an adjunct phrase containing a complement phrase beginning without a prep. So the hit is included. Ooops. However, if we reduce the depth to 1 we miss interesting cases I think you want. So can we model this differently ?

 

  Alright, so I have attached two workspaces for you. The first just gives you what I discussed above. The second combines a search like the above and one that specifically targets adjunct constructions. It does comp <NOT> adj_comp to get the hits in Philem. and Ephesians (correcting the missed πίστιν) you want and exclude things like 1Tim 1:15. Now, to be fair this probably isn't the end of this either but perhaps it's a decent enough start.

 

Thx

D

MarkPisteuvwAkouwComplement1.accord

MarkPisteuvwAkouwComplement2.accord