Odd Search Question

Last night, in class, someone asked me about "the servant of the Lord," being applied to Joshua at the end of that book, so I popped over to Logos (after tolerating some snide comments about how easy this would be to find in BibleWorks, ugh), and put together a quick search in Logos. I actually built two searches, one for the Hebrew Lemmas, one for the English glosses --and what I found was they return different results.

I am baffled by this result --why is it the search based on the lemma doesn't turn up the Exodus 14:31 result, when it looks like, to me, that the lemmas are actually the same? The pointing is different, but isn't the point of searching with the lemmas that the pointing doesn't matter? Other cases of the pointing being different don't show up.

Can anyone 'splain why this is happening?

Thanks!

Russ

Find more posts tagged with

Comments

Rich DeRuiter

I am baffled by this result --why is it the search based on the lemma doesn't turn up the Exodus 14:31 result, when it looks like, to me, that the lemmas are actually the same?

I'm not sure you are searching for lemmas when you search for non-pointed Hebrew. And I don't know enough about how Hebrew searching works to say why you're getting the results you are.

I can get the results you're looking for in the above search quite quickly by right-clicking on the word "slave" or "servant" in one of the relevant verses and select "lemma" and "search this resource" (limit your search later, if you like). The do the same for the tetragrammaton. Then copy one term to the search box of the other and insert NEAR between them. What you end up searching for then is: <Lemma = af/he/עבד:3> NEAR <Lemma = af/he/יהוה>. This was all done via the ESV, but can be done from the LHI (with the same search terms), or the BHS with Westminster 4.2 with these search terms: <Lemma = lls/he/עֶבֶד:1.n> NEAR <Lemma = lls/he/יהוה>.

Note: I got the search terms by using right click on the term > Lemma > Search this resource for each term then copy/paste one to the other search pane. I find this method both quicker than typing in Hebrew/Greek and means I don't need to remember all the syntax of the search term construction.

Russ White

https://community.logos.com/discussion/comment/390840#Comment_390840

I can get the results you're looking for in the above search quite quickly by right-clicking on the word "slave" or "servant" in one of the relevant verses and select "lemma" and "search this resource" (limit your search later, if you like).

That's actually how I built this search --I can also build the same search using Shibboleth and typing in the words I want, then copying and pasting them in the search box. Either way turns up the same results.

It's not how to build the search that's problematic, it's why the search results using the lemmas doesn't turn up an entry that seems like it should be there.

Russ

Fred Chapman

https://community.logos.com/discussion/comment/390840#Comment_390840

Russ,

I ran it just like you did and I agree it is odd. I understand what Richard is saying also. I am not sure why the program sees these two searches differently.

Here is a screen shot of what Richard did and what you did orginally. To me it does not make sense for these search results being different

I wonder what BibleWorks returns

Todd Phillips

https://community.logos.com/discussion/comment/390842#Comment_390842

It's not how to build the search that's problematic, it's why the search results using the lemmas doesn't turn up an entry that seems like it should be there.

Actually, your search is not a lemma search. You are just searching the surface manuscript text.

To do a lemma in the Bible search you need to specify the lemma field:

Alternatively, using a morph search makes the syntax a little easier:

Notice that the first lemma has a "3" in front of it, that is necessary to differentiate it from other homographs. You can't leave it off or else it won't match, since the text is tagged using specific homographs.

The easiest way to select the proper lemma is to use the Morph search and put the "lemma:" label in front of the Hebrew, and you'll get a selection list to choose from:

Russ White

https://community.logos.com/discussion/comment/390845#Comment_390845

Aha --thanks for the insight.

Russ

Fred Chapman

https://community.logos.com/discussion/comment/390845#Comment_390845

Actually, your search is not a lemma search. You are just searching the surface text.

To do a lemma in the Bible search you need to specify the lemma field:

I'm not the OP and I get what both you and Richard saying about the difference in the searches. My question would be if I right click a word and copy the Lemma(s) into the search bar of a bible search, why does it not return the same results as when I use the specific Lemma search string? To me the fact that it returns some of the results indicates the program has the ability to search the RI Hebrew surface text, but fails to catch all of the occurrences.

Russ White

https://community.logos.com/discussion/comment/390848#Comment_390848

That is a problem --when I copy the lemma to the clipboard, why doesn't it copy the lemma information, as well? Or at least give me the option to do so? And why should a surface text search not return every instance of the homophone, rather than skipping some because of the coding? I.e., I would expect a search on the English surface text for "wash" to return every instance of the word, regardless of the meaning. That Logos assumes I don't want to do the same in a Hebrew manuscript puzzles me...

I'm glad I know how to do this now, but I still don't understand why Logos would return these specific results.

Russ

David Thomas

https://community.logos.com/discussion/comment/390845#Comment_390845

To do a lemma in the Bible search you need to specify the lemma field:

Interesting observation - I followed Todd's recommendation about doing a LEMMA search and get all 13 verses in the ESV (the underlying manuscript), but when I do the same search in NASB95, it misses Exodus 9:30

Todd Phillips

https://community.logos.com/discussion/comment/390856#Comment_390856

I followed Todd's recommendation about doing a LEMMA search and get all 13 verses in the ESV (the underlying manuscript), but when I do the same search in NASB95, it misses Exodus 9:30

It is related to how "NEAR" works (NEAR is equivalent to "WITHIN 48 CHARS"), and differences in the translations. If I change the operator to "WITHIN 49 CHARS", then Ex 9:30 shows up in the NASB95 results. (Yes, I know the English words are the same, but maybe it's counting footnote markers?)

Rich DeRuiter

https://community.logos.com/discussion/comment/390850#Comment_390850

I'm glad I know how to do this now, but I still don't understand why Logos would return these specific results.

Russ

As I understand it, when we search, we search the actual data behind the surface text. For English, the results would be the same as the surface text, since no morph info is included in the English data. The Hebrew and Greek texts include a lot more information for each word, so discrepancies are common in the type of search you did above.

We should see this as evidence of a much more powerful and flexible database, rather than a flaw in the search routine. The downside is that we have to search with greater compliance with that data structure. We can overcome the downside, by knowing how that a morph search is easier, or the right-click method I suggested is easier than typing in the search parameters ourselves. Without any testing, the above search might work in a Hebrew text that doesn't have any morph tagging, but I don't have any such version to try it out.

Dave Hooton

https://community.logos.com/discussion/comment/390873#Comment_390873

As I understand it, when we search, we search the actual data behind the surface text.

If you search a Hebrew/Aramaic/Greek word then you are searching the manuscript text, whether it be in a RI or original language bible (with/without morphology). If you use the lemma:word syntax then you search that part of a bible with morphology; not the surface/manuscript text.

So, if I search for עֲבֹד (servant) I get the same result in AFAT, LHI, BHS/WIVU, BHS/WHM 4.2 (3 different morphologies).

Russ White

https://community.logos.com/discussion/comment/390984#Comment_390984

I think this is true for the Hebrew versions, but for the English versions (from what I gather), all searches are lemma searches... At least that's what I'm taking away from this thread (?).

Todd Phillips

https://community.logos.com/discussion/comment/391494#Comment_391494

I think this is true for the Hebrew versions, but for the English versions (from what I gather), all searches are lemma searches... At least that's what I'm taking away from this thread (?).

(I am not totally sure I understand your statement, but...) There is no underlying English "lemma" in any resource (the lemma is the dictionary form of the word.). What you are seeing in your English search is called "stemming", which happens when you have "Match all word forms" checked in the search options. Stemming is done by looking up the search words in an external table to find all possible forms of those words and searching for all of the forms, but that's an inexact method because homographs can't be distinguished and not all possible words and forms are included in the look-up table. Greek and Hebrew lemmas, on the other hand, are not found using an external table, but are manually mapped to the corresponding manuscript word directly in the resource.

Dave Hooton

https://community.logos.com/discussion/comment/391494#Comment_391494

but for the English versions (from what I gather), all searches are lemma searches...

From the Context menu you can Search any bible by Lemma or Manuscript word (=Selection in Hebrew/Greek bibles). Your first search using two Hebrew words will be a Manuscript search even if the words are lemmas. A lemma search always has the word lemma together with the word e.g. lemma:אדין It will produce different results to a (manuscript) search for אדין.

Russ White

https://community.logos.com/discussion/comment/391506#Comment_391506

Okay --so why doesn't the Exodus hit show up when searching based on the manuscript text?

Todd Phillips

https://community.logos.com/discussion/comment/391513#Comment_391513

Okay --so why doesn't the Exodus hit show up when searching based on the manuscript text?

עַבְדּ

Probably because the dagesh (the dot) in the dalet (the letter on the left ) makes it not match.

Russ White

https://community.logos.com/discussion/comment/391518#Comment_391518

If the דּ (the dalet with the dagesh) is considered a different character than ד (the dalet without the dagesh) at the manuscript level, then. if the dagesh is considered part of the phoneme, rather than a diacritic, then the search results make sense --but (I know wikipedia isn't a good source), but Wikipedia does show the dagesh as a diacritic... (?).

David Thomas

https://community.logos.com/discussion/comment/391523#Comment_391523

Could it be the difference between a dagesh forte and a dagesh lene since Begedkephat letters can take either? I think this instance is a forte - in essence doubling the letter which "changes it's spelling" in the manuscript level.

http://www.hebrew4christians.com/Grammar/Unit_Three/Dotted_Letters/dotted_letters.html

I don't presume to be an expert by any means, I am asking as a student trying to learn from those of you who know Hebrew better than I.