Greek Construct Search Issues
I was trying to set up some simple Greek construct searches for my students looking for Present Active Indicative verbs with their Nominative subject.
Here’s the construct I set up, and I simply left all the defaults in place. (I.e, default depth of 2, etc.) I ran the search and allowed it to look in both directions.
Many of the results that are returned are accurate, but many are not. E.g., look below at Matthew 4.11. The verb ἀφίησιν is highlighted but its subject (διάβολος) is not. Instead ἄγγελοι is highlighted as the subject, but the two verbs it actually controls (προσῆλθον καὶ διηκόνουν) are not. I played with all the various depths and check boxes, added subject and predicate parameters, and I either eliminated Mat 4.11 from the results or got the same results you see.
The problem may be in the Syntax model it is using. You can see in the syntax tree that it thinks the whole verse is a single independent clause (N). Is that confusing the construct search? Further in the second independent clause with angels as the subject, it looks to me like only προσῆλθον is connected directly to angels.
SO: am I doing something wrong with the search or is the issue with the syntax diagramming?
This is not an isolated incident. I included Matt 5.14 in the graphic above. You will see there that πόλις is correctly highlighted as a subject, but the verb that goes with it, δύναται, is not. Instead the three verbs that are highlighted do not have specified subjects and are not part of the same clause. (And if I toggle off Hits with sub-clauses and leave everything else the same, it will not find either Mat 4.11 or 5.14.)
I would really like to be able to trust the construct results, so I’m hoping I’m just doing something wrong. Thanks for any help!
-
Thanks, and had a quick look at this - I am not a great expert, and so others may be more helpful than I. I too had a look at the depths and subject and predicate options.
I think the simple answer to part of this question is that διάβολος is classified as an adjective, and so at the depth of 2, the search will find the ἄγγελοι that is within a depth of 2 of the verb ἀφίησιν. You can get around this by setting the depth to 1 and using the syntax term “Subject” rather than “noun”.
the other option ot consider would be using the “Agree” connecting Item - which is helpful if you want to incease the depthe for any reson and elimiate findingi ἄγγελοι in Matt 4:11.
I am having difficulty witht the seaches I am doing identifying the other example of Matt 5:14 - But I would note that δύναται is not picked up as it is classified as passive and not active here. The finding of the other 3 indicative verbs must be to do with the depth settings .
1 -
Tim Planche - Thank you! Excellent! I didn't notice that διαβολος was analyzed as an adjective, so that was the issue there. And I should have realized that δυναται in 5.14 was not active. So when I run this construct search, I do get closer to the results I was hoping for:
This search (with depth set at 1) does eliminate those hits in 5.14f with verbs without explicit subjects, so that's good. In 4.11, however, ἄγγελοι is highlighted as the subject, but only first of the two verbs it actually controls (προσῆλθον καὶ διηκόνουν) is highlighted. I have to set depth to 2 to pick up the second verb, but then that picks up many false hits. (E.g., all the verbs without the explicit subject in 5.14-15) >> I think that must be an error in the syntax coding.
I see that you included the diagramming module which does accurately depict how I think 4.11 should be displayed, but the Greek construct is using the syntax module which I think is wrong. (I tried changing quite a few more parameters in the search, and I am more convinced that 4.11 is just analyzed incorrectly syntactically.)
Thanks again!
0 -
Hi Mark, once you get a firm handle on how the syntax module works, maybe you should try giving a webinar on it! :-) My take on your original post is that the highlighted words in Matt 4:11 are picked up because they are included under “N+”. I.e., in the same independent clause, you have a present active verb (ἀφίησιν) and a nominative subject (ἄγγελοι). Also, if you haven't checked “search in both directions” you specifically get a nominative after the verb, as here.
I'm wondering if there's any way of connecting the subject to the corresponding verb and not just a verb that happens to be in the same clause.
But regardless, your original query shouldn't pick up the verbs connected with ἄγγελοι, as they are an aorist and an imperfect, respectively.
So in this case, the syntax module seems correct.
1 -
Mark, you can probably do better, but this eliminates some of the false positives (I put a NOT on λέγω because it shows up in several hits in Matthew, but I'm sure there are other things you would have to exclude. I didn't spend much time on this). There are several hits which are correct but without the nominative noun being highlighted. I think this is a bug:
0 -
Thanks, Donald COBB . My later post did get me closer just looking for indicative verbs (regardless of tense or voice) that are the predicates in combination with a “subject” (which would include nouns and pronouns and substantives). Note that I did a CLAUSE search, not a PHRASE search, to avoid a bunch of unwanted hits. And I do search both directions.
Still, even using that latter construct, it really depends on depth settings. E.g., look at Mat 4.5f. The result correctly highlights διαβολος and παραλαμβανει but misses εστησεν and λεγει (in the next verse). You can see that the diagram displays it just as I would think, but the syntax has a much more complex arrangement.
I can get the verb hits highlighted in 4.5f if I set the CLAUSE depth to 4 and the predicate and subject depth to 2, but then I get lots more incorrect hits. (Mainly predicate verbs without subjects.)
Yeah, I'll have to keep working at this…
0 -
Mark, you have much more practice on this than I do. I haven't put too much time into learning the syntax modules. I would really love to see you do something on it once you feel you've got it down. And there's no doubt some tweaking to do on the resource itself.
1
Please sign in to leave a comment.
Comments
6 comments