Need help with Hebrew search results

I recently preformed a search on bara in the LHI and got "45 results in 38 verses”. Since I have another bible application, the name I won't mention, I did the same search and got "46 verses, 27 forms and 54 hits" -- the text was the Codex Leningradensis Hebrew Text.

Further searches in L4 yielded the following results: AFTA -- "45 results in 38 verses", BHS -- "54 results in 46 verses". This last result agrees with the search result from the other unnamed bible application.

Ok, here are my questions. Why the difference? It's obvious that the BHS and the Codex Leningradensis Hebrew Text are the same text but what's the difference between them and the LHI and AFTA. How would I isolate the 9 additional instances of bara?

Thanks, in advance, for your responses.

Find more posts tagged with

Comments

Alan Macgregor

I have found in both Greek and Hebrew the number of results can vary widely depending on the morphology that is attached to the text I am searching. Since Hebrew and Greek searches are morphological, you need to ensure how you search, whether by lemma or not.

It doesn't matter so much if I am just referring to a round number e.g. "over 40 examples" but it becomes more critical if you are doing this for academic research where you are going to compare results with other data. In such a case I tend to specify which Morphology I am using and how I am carrying out my searches, whilst acknowledging that other results can be returned using other search types/morphologies

In other words, it's not clear cut. There is no "right" answer. It really depends on the precise question you are asking of the search engine.

James Thompson

https://community.logos.com/discussion/comment/95253#Comment_95253

I have found in both Greek and Hebrew the number of results can vary widely depending on the morphology that is attached to the text I am searching. Since Hebrew and Greek searches are morphological, you need to ensure how you search, whether by lemma or not.

In each of my searches I'm performing a "lemma" search so, as I understand lemma searches, all forms of the word should show up. I'm still left with diffferences I can't explain, in particular and within L4, between the BHS and the LHI and AFTA. Again, why the differences and how do I isolate them. Perhaps isolating them would provide a clue...

Kevin Becker

https://community.logos.com/discussion/comment/95259#Comment_95259

There are many homophones in Hebrew. They need to be differentiated. Also there are times when there is disagreement as to what the lexical form a word is. A morphology scheme must make a decision in either of these cases. To figure out what's going on you will have to compare the different searches and figure out what the issue is, whether lexical disagreement or error in a database.

Vincent Setterholm

Kevin is correct that the lemma search is sensitive to homograph distinctions, so you won't get hits for words that are spelled the same but have a different meaning. That accounts for most of the differences you might see comparing a search in our Westminster implementation compared to some other piece of software that might default to ignoring the numbers used to distinguish homographs. The smaller hit counts you sometimes see in LHI owe something to what Alan mentioned - different morphologies represent different theories. In this case:

The Andersen-Forbes analysis generally splits participles from other verb forms. It had to do with the tagging necessary to facilitate their syntactical analysis (and perhaps related to their particular theory of lexicography). The result is that the AF analysis splits the text into more lexemes than you might expect, so some verbs will have smaller counts if you're only searching one related lexeme.