Here is what appears to be a duplicated search hit for a single verb:
There is only one instance of this verb in 1 Corinthians 7:21 yet it is reported twice.
I don't think this will be fixed as resources with GRAMCORD morphology are not supported in L4. See http://community.logos.com/forums/p/6026/51285.aspx#51285
I don't think this will be fixed as resources with GRAMCORD morphology are not supported in L4.
I see where Bradley mentions the Septuagint with GRAMCORD morphology but not the NA27. ??
Edit: BTW I checked the morph search on the UBS4 w/Logos morphology and it does not return this error.
Problem #2. A new search in UBS4 w/Logos morphology. The hit counts don't add up in Analysis view. This is the book of Romans. Reports 1179 hits but the breakdown of hits by mood totals 1194 (if the two at the bottom are to be counted in). Which is correct?
#3. Same search, new book: 1 Corinthians. Says 1322 results, total of individual Mood sections (plus the two uncategorized ones) is 1331.
I'll quit for now until I find out what is going on.
Mark, I've reported similar things with Morph search
Here: http://community.logos.com/forums/t/25290.aspx
and
Here: http://community.logos.com/forums/t/25271.aspx
I don't do much morph searching so don't pay too much attention to other posts on the subject. I am curious about this, however, and would like to know whether to trust the search results and which number to trust.
There are 1194 analysis lines because of multiple assignments to a result ie. some of the 1179 results double-up to give 1194 lines
Mark, to add to what Dave said: When a morph scheme gives multiple possibilities of morphology 2 (or more) lines are generated in the Analysis view. This is because Analysis is all about showing all related information to your original search. Line doubling like this is very common in morph searching, especially if you are searching a reverse interlinear.
It would be nice if the search engine could intuit when the extra lines are helpful and get rid of them, but then we might miss stuff we wanted to notice.
Dave and Kevin, thanks for the analysis.
I think whatever Logos does the numbers should add up. The items should be counted the same way in the individual analysis views that they were counted for the overall view, that is, once for each verb.
You've pointed out reasons Logos has trouble doing that. Those seem to be tagging issues. I understand the problem with always being able to assign the 'correct' morphology for a word or the possibility of two L-N numbers being reported. I don't want the data stripped out, but I do want the numbers to match.
Perhaps the way around this is for Logos to devote a single line per word when data is grouped by a category. Do this my expanding the individual columns to allow for a second (or third) piece of data where needed. So for example in this report, instead of two lines for assigning two LN numbers which is misleading, assign one line that has two L-N numbers in it (since the L-N numbers are not my interest in such a grouping).
When I group by L-N number, combine two (or more) mood possibilities for one verb in one line by adding an extra column (or combining the morphs in a single column) so I don't think there are more of a specific L-N number than there are.
It would be a little more complicated to program, but would give consistent numerical results, which I think is what those numbers should be. It seems right now that the group numbers are just report line counts, not meaningful analysis numbers.
I agree, Mark. Something has to give here.
One thing is certain: We must be very cautious with how we cite data from morph searches.
When I group by L-N number, combine two (or more) mood possibilities for one verb in one line by adding an extra column (or combining the morphs in a single column)
I oversimplified the verb counting, an issue that was discussed with Brian. The 1179 result is all occurrences of a verb ie. if you search Ro 13:1 you will get 7 results for 6 highlighted words in Verse view! The 7 results can be seen in Aligned view as well as Analysis view, but the difference will only be seen in Analysis if you display and look at the Voice column!
There will be only 3 search results for 3 words in Ro 11:18 but Analysis will show 5 because of the doubling of LN numbers in two of those words.
I agree with counting the number of verbs for the search result because that is what your @V term is asking! So that makes it difficult to adopt your suggestion because you will have fewer lines than results in Analysis view! Overall then, Analysis does the best job for its purpose.
if you search Ro 13:1 you will get 7 results for 6 highlighted words in Verse view! The 7 results can be seen in Aligned view as well as Analysis view, but the difference will only be seen in Analysis if you display and look at the Voice column!
Not a problem as long as the 1179 includes only six verbs in 13:1 (which is the number in the verse). Reporting seven instead of six would be wrong. (The 1179 appears to be the actual verb count and 1194 the number of lines in the report.) My point is that Logos should not be be reporting the number of lines in its reports but the number of verbs. As you point out the extra counts are due to vagaries in the morph tagging and L-N domain assignments. That can be accounted for without simply reporting the number of lines under a section. I think this is what Logos should do. The lines report is simply misleading.
Overall then, Analysis does the best job for its purpose.
I can argue that the way counts are displayed or determined is misleading. I don't want to lose information, just bring all the information together in a way that doesn't raise questions.
Not a problem as long as the 1179 includes only six verbs in 13:1 (which is the number in the verse). Reporting seven instead of six would be wrong. (The 1179 appears to be the actual verb count and 1194 the number of lines in the report.)
The 1179 result includes 7 verbs for Ro 13:1 because you asked how many morphological verbs @V there were (do a search on Ro 13:1)! The issue is that there are only 6 words that are verbs. I agree there are 7 results only because it is consistent with the search term. OTOH I see your point that there are only 6 verbs in the text, therefore 6 results!
I haven't done a search on LN numbers to see if the same logic applies to those results.
I agree it is misleading simply calling them Results. Call them 2 results for 1 word, or 2 lines!
The 1179 result includes 7 verbs for Ro 13:1 because you asked how many morphological verbs @V there were (do a search on Ro 13:1)!
Well maybe we are arguing semantics at this point, but there are only six verbs no matter how you count them and Logos only included 6 in its total of 1179. There are seven lines of data because of the uncertainty in the morph data about voice for one of the verbs. No uncertainty and we have six lines of data and six verbs. Perfect. Ambiguity and we have the situation we are in. Imperfect. But Logos can count the data differently and that would make all the difference to me. I would then consider the report perfect even if the underlying data is not.
As far as I can tell it does.
That would be better.
Oh well, you and I can't change it anyway and I suppose Logos isn't in the 'mood' to do so.
Well maybe we are arguing semantics at this point, but there are only six verbs no matter how you count them and Logos only included 6 in its total of 1179.
No! Logos counts 7 results for Ro 13.1 not 6 (do the search yourself).
I haven't done a search on LN numbers to see if the same logic applies to those results. As far as I can tell it does.
Let me make your day with a search for LN domains in the range 33-88! A Bible Search for <LN 33-88> in Ro 11:18 gives 7 results from 5 words! The Analysis view shows 10 results from the same 5 words.
You're right. I was looking at the search I did for the entire book of Romans. There it reports 1179 results but gives 1194 lines of data. I assumed that it counted just 6 times in Romans 13:1, that the extra line was counted in the 1194, but not counted at the top. Now I'm not so sure what that 1179 number is.
The search for 13:1 should report six verbs since that's what I'm searching for.
That really makes my day.
It also makes me wonder why Logos reports any numbers at all. They don't mean anything. In one case they are lines of the report, in another they are not (7 with 10??).
OK. Logos can chime in any time it wants and explain its philosophy and what its numbers mean.
My head hurts!
Now I'm not so sure what that 1179 number is
Imagine that the Bible is structured as a number of Interlinear lines. The @V search will report the number of matches in the line for Morphology.
The <LN 33-88> search will report the number of matches in the line for Louw-Nida numbers.
Analysis then looks at all the Interlinear lines and produces a grid with only one value per column, whether it be Strong's number, LN number or Morphology (Part of Speech, Mood, Voice etc). So Rom 11:18 will have extra Analysis lines for the @V search because of the multiple LN numbers.
Now my head is beginning to hurt!
Available Now
Build your biblical library with a new trusted commentary or resource every month. Yours to keep forever.
