Syntax Search questions - differing results
I am going through the videos at
http://www.logos.com/products/details/3135
and trying them in Logos 4 to get the feel for ver. 4 syntax searching.
I have had no trouble duplicating them in ver 4., except for two of the searches which had substantially different results:
In the video
A-F Searching: Infinitive Absolute as Predicator—Learn how to find clauses where an infinitive absolute verb form substitutes for a finite verb. (I am not sure if these are downloadable somewhere or not), in both Logos 3 and 4 I get :
Infinitive absolute alone no finite search either side (his last search), he gets 205 hits I get 40. In the next to last search, Infinitive absolute as Predicator, he gets 753 results I get 195.
I get the same results in ver 3 and 4.
Does anyone have access to these videos, do you know why the results differ so much? I know his ver 3 Logos looks different than mine, has the AF database changed a lot?
Comments
-
Dominick Sela said:
I am going through the videos at
http://www.logos.com/products/details/3135
and trying them in Logos 4 to get the feel for ver. 4 syntax searching.
I have had no trouble duplicating them in ver 4., except for two of the searches which had substantially different results:
In the video
A-F Searching: Infinitive Absolute as Predicator—Learn how to find clauses where an infinitive absolute verb form substitutes for a finite verb. (I am not sure if these are downloadable somewhere or not), in both Logos 3 and 4 I get :
Infinitive absolute alone no finite search either side (his last search), he gets 205 hits I get 40. In the next to last search, Infinitive absolute as Predicator, he gets 753 results I get 195.
I get the same results in ver 3 and 4.
Does anyone have access to these videos, do you know why the results differ so much? I know his ver 3 Logos looks different than mine, has the AF database changed a lot?
I am at work right now and do not have access to L4 or the videos. I did the same thing you're doing a while back and found it very beneficial in learning to use the syntax databases. Can you post a screen shot of the L4 search you're duplicating from the video and tell me the L3 video number and I'll play with it in the morning when I get home.
0 -
Here are the two screen shots. You can see by the titles which one got the 40 results (when in the video he got 195), and which one I got the 195 results (he got 753), as the number is in parens at the end of the title.
On the table of contents screen that comes up when I insert the CD, these searches are both in video 6D (A-F Searching: Infinitive Absolute as Predicator). They are pretty simple searches, I am wondering why the results are so different. Thanks for checking it out!
0 -
Dominick,
I played with this for a while this morning and I get the same results you get. I cannot figure out why the results are so much different in L4 vs. L3. The interesting thing to me is when you run the first search in video 6d (just search for infinitive absolute segments L4 produces 860 hits while L3 only produces 847 in the video. I cannot check L3 on my computer because I uninstalled it some time ago.
Maybe someone from Logos or one of the forum syntax guru's can chime in on this.
EDIT: I created another thread in the L4 forum and titled it "Possible BUG"
0 -
I got the 847 vs. 860 also, but since it appears that's an older version of Logos 3, that there might have been some changes in the AF database since it is being revised.
Just to be clear, the numbers I report above I got in BOTH Logos 3 and 4. Those numbers of course differ WIDELY from the Logos 3 results in the video, as I also reported above.
Since the video didn't show constructing the query we have no way of knowing what if anything could have been set (that does not show in the graphical picture of the query). I hope someone from Logos responds to this, it's quite a difference and undermines the points the speaker made in the video.
Thanks Fred for making a report about the possible bug...
0 -
Dominick Sela said:
I got the 847 vs. 860 also, but since it appears that's an older version of Logos 3, that there might have been some changes in the AF database since it is being revised.
Libronix 3, *I think* had the version .80 of AF at the time -- though even if it wasn't .80, I do know for sure that it was an earlier version number than what we have now. Logos 4 has version .90 -- which is dated August 2009.
0 -
Michael
I thought there was a different version in L4; but that still seems like a huge difference between versions.[*-)]
0 -
Fred Chapman said:
I thought there was a different version in L4; but that still seems like a huge difference between versions.
Indeed!
I don't know for sure if that is the reason. Let me get back to you.
0 -
Fred Chapman said:
The interesting thing to me is when you run the first search in video 6d (just search for infinitive absolute segments L4 produces 860 hits while L3 only produces 847 in the video. I cannot check L3 on my computer because I uninstalled it some time ago.
I get 857 hits in L3 vs. 860 in L4.
For the other searches shown by Dom I get same results in L3 and L4. I don't have the video.
Dave
===Windows 11 & Android 13
0 -
Michael Aubrey said:
Libronix 3, *I think* had the version .80 of AF at the time
L3 is v 0.80 (Nov. 08).
Dave
===Windows 11 & Android 13
0 -
Dominick Sela said:
Here are the two screen shots. You can see by the titles which one got the 40 results (when in the video he got 195), and which one I got the 195 results (he got 753), as the number is in parens at the end of the title.
Dominick Sela said:The above query doesn't show unambiguously what it claims. Make IC3 the same as IC1 and you get 195 results ie. Zero Finite Verb IC appears in a clause with an Inf Abs IC (you can prove this positively by seeing how many Finite Verb IC occur with an Inf Abs IC).
Dave
===Windows 11 & Android 13
0 -
Dave Hooton said:
The above query doesn't show unambiguously what it claims. Make IC3 the same as IC1 and you get 195 results ie. Zero Finite Verb IC appears in a clause with an Inf Abs IC (you can prove this positively by seeing how many Finite Verb IC occur with an Inf Abs IC).
Ah Dave, you found the error in my query construction, THANKS! You always seem to figure this stuff out, that's why I am learning!
So one left - I went back and re-researched this, but I still don't get the problem. Here is a complete description of the problem as it stands today:
In video 6D, at 2:27 in the video, he brings up a query "Infinitive Absolute as Predicator". He runs the query and gets 753 hits. I run it in BOTH Logos 3 (ver .80 of the AF PMA database) and Logos 4 (ver .90 of AFPMA database) today, and I get 195 hits - way too few for it just to be a database revision issue. Here is the query from my Logos 4, it is VERY simple!
He does not review the whole query, I just have benefit of seeing the graphical picture of it in the video. Unless there is a modifier that does not show up in the graphical picture it seems like this is the query he is speaking of.
Now I noticed in the video, he gets a hit on Genesis 2:16, and I do not. If I bring up AFPMA, there IS an infinitive absolute there:
HOWEVER, note this - the CIC is ACTUALLY in Gen 2:17, as the clause goes across verses, from Gen 2:16 to Gen 2:17! Could this be the issue, that Logos now is not catching all the ones that go across verses, which would explain only 195 hits vs. 753?
That's all I know - if not this it must be something way more grammatically advanced than I can figure out! Thanks for the help all...
0 -
Dominick Sela said:
That's all I know - if not this it must be something way more grammatically advanced than I can figure out!
I haven't the time to look further but explore a Segment with Verb - Inf Abs, this is where Gen 2:16 comes in. There are 860 in total
Dave
===Windows 11 & Android 13
0 -
Yeah that captures it, I found out too I could get it with a Segment. But why does the Clause-CIC query fail when it used to work in the video? That's very worrisome if queries can change that much between versions of database or Logos, it just doesn't seem like that would happen...
0 -
Dominick Sela said:
Yeah that captures it, I found out too I could get it with a Segment. But why does the Clause-CIC query fail when it used to work in the video? That's very worrisome if queries can change that much between versions of database or Logos, it just doesn't seem like that would happen...
Looks like you figured it out before I got back to you.
In terms of database changes, it needs to be kept in mind that this is still a pre1.0 release.
The best place for questions about the changes might very well be: http://andersen-forbes.org/
They state:
Status of the A-F Database: The A-F database of August 2009 (v0.9) is part of the Logos 4.0 release. The improvements in this release are many and substantial.
I would *guess* that the A-F Database will stabilize significantly when it gets to version 1.0.
Their e-mail address is there. I'm sure it wouldn't hurt to ask about the changes directly.
0 -
Thanks Michael, I sent them an email asking about this. I'll post here if I get a response back! I feel better that the database may be drastically changing, although I do still wonder why a CIC does not catch the hit in Gen 2.16-17.
0 -
I got a reply already, from Dean Forbes himself!
--------------------------------------------------------------------
One locus of frustration for Frank Andersen and me is that our book ("Biblical Hebrew Grammar Visualized") has not yet come out. It is with the publisher awaiting
the tender mercies of a phalanx of copy editors and what-have-you. Anyhow, in it (section 3.2.6.1) we explain our handling of infinitives absolute, but that doesn't help you!
So allow me to explain: we recognize 860 infinitives absolute. We distinguish three clausal functions:
predicator [inf abs prd] -- 190x
intensifier [inf abs int] -- 503x
amplifier [inf abs amp] -- 57xThis leaves 110 instances unaccounted for. These are scattered about. For example, in coordination phrases: inf abs AND inf abs; e.g., Jer 7:25, end of verse.
Hope this clarifies matters.-----------------------------
I replied back and told him I get 195-503-57 in Logos 4 with ver 0.90 of the AF-PMA database, I'll see what he says.
0 -
Dominick Sela said:
I replied back and told him I get 195-503-57 in Logos 4 with ver 0.90 of the AF-PMA database, I'll see what he says.
I get 190-503-57 = 750 or 750 in a single Clause IC with all three attributes ticked.
A Phrase with Inf Abs under Verbal yields 74 (Union/Disjunction or AND). So now we account for 824 of the 860.
The rest are problematic - some are classified under Infinitive Construction & Phrase Inf. Intensifier gives zero results. There is more work tracking these down than I have time - so we need some clarification of the search techniques needed or possible bugs.
Dave
===Windows 11 & Android 13
0 -
OK here is the answer from Dean Forbes, he was gracious enough to figure it out for me!
Statements are his:
I figured out why you get 195 predicative inf. abs.
If you run a search for a CIC that is a predicative inf. abs., specifying nothing as regards the clauses that these CICs appear in, then you'll get 190 hits (what Dean Forbes thought it should be0.
BUT, if you search for: CLAUSE --> CIC: inf abs pred, then you'll get 195 (what I get in Logos 3 and 4 copying the query shown in the training video). How can that be? Well, it's because of multidominance. There are five places where a single inf abs pred serves in TWO clauses:
2 Kgs 19:29, Isa 37:30, Jer 14:19, Prov 17:12, and Neh 7:3.With regard to the version of database in the Logos 3 training video 6D:
Based on the copyright date that is shown at the onset of the video snippet (2006), Mike Heiser was using either version 0.5 or 0.6. Whichever is the case, at that stage of the game we had not yet
split out the intensifying function (e.g., Gen 2:17, "surely die") and the amplifying function (e.g., Deut 2:24, "begin to possess"). They were all lumped together. Hence the large number (in the video, 753, vs. now, with 195).Also, it would appear that in the interval between 0.6 and 0.9 [3 years], 3 items formerly tagged as inf. abs. have been reclassified as something else.
CASE SOLVED! A tip of the hat to Dean Forbes for all his help, what a nice guy to take time out of his busy schedule to help...
0 -
Dominick Sela said:
Also, it would appear that in the interval between 0.6 and 0.9 [3 years], 3 items formerly tagged as inf. abs. have been reclassified as something else.
CASE SOLVED! A tip of the hat to Dean Forbes for all his help, what a nice guy to take time out of his busy schedule to help...
Yeah, he is a great guy. I met him at BibleTech once.
And changes of this magnitude between .6 and .9 make more sense than between .8 and .9.
0 -
Michael Aubrey said:Dominick Sela said:
Also, it would appear that in the interval between 0.6 and 0.9 [3 years], 3 items formerly tagged as inf. abs. have been reclassified as something else.
CASE SOLVED! A tip of the hat to Dean Forbes for all his help, what a nice guy to take time out of his busy schedule to help...
Yeah, he is a great guy. I met him at BibleTech once.
And changes of this magnitude between .6 and .9 make more sense than between .8 and .9.
Good job Dominick in staying with this. Thanks Mike and the rest who contributed to this thread.
0 -
Dominick Sela said:
CASE SOLVED! A tip of the hat to Dean Forbes for all his help, what a nice guy to take time out of his busy schedule to help...
The videos may have been out of date but the detective work has been very fruitful! It's helped my understanding somewhat but Logos still need to improve the tool tips and results tracking in either AFPMA or AFAT.
Dave
===Windows 11 & Android 13
0
