Asking original language questions in All (smart) search: intriguing finds

Francis · 2025-03-31T18:13:44+00:00

There was an error rendering this rich post.

All search will automatically switch from smart to precise mode if you enter original language syntax such as "lemma.g:". However, it works as smart search if you do not use such syntax and yet use original language words. Note however the difference in the results and synopses between the search that uses greek and the one that uses transliterations:

This was a helpful discovery. It's logical but I had not realized the possibilities this offers.

The possibilities are also quite interesting in Bible Smart Search:

Find more posts tagged with

Comments

Francis

Other interesting use where smart searches approaches the precision of "precise" search but using natural language.

Francis

I have to explore the limitations (if someone has not done it yet) but here is another one that apparently worked, which is wow because the search is launched from the NASB:

Beloved Amodeo

@Francis These findings which you cite are wow! Thanks for sharing. I especially like the final one.

Beloved Amodeo

@Francis Try as I may, I can't get my data to display as yours has. Do I have a prioritization problem or have you employed some other method?

Francis

@Beloved Amodeo

The data has to come from somewhere that you own. The main title I can think of that does that is the "Septuagint with Logos Morphology" which has the Greek tagged to the underlying Hebrew. Do you own it? Or perhaps it pulls from somewhere else, not sure. All I know is that it sure works for me.

Beloved Amodeo

https://community.logos.com/discussion/comment/1364844#Comment_1364844

Yes I do own it. Is there a way to exclude it from operating in this instance?

Francis

@Beloved Amodeo sorry, I misread your first post. Upon looking more carefully at your screenshot, you do have results but they show differently from mine indeed. Do you resources selected under "Add Versions"? If so, try removing them to see if that's causing it.

Beloved Amodeo

I had a dream there was a way to remove all additional references and it manifested. One need hover over any reference and a drop down popup appears, select remove all and one gets the desired results.

Francis

It seems that the results are not completely reliable though. First, for some reason, your results and mine have both verses in common and some that differ. Second, some of the results I got are not right.

As one post had mentioned early with regard to AI searches: if you can't verify visually from the result that it is correct, check it. In this case, this search does help me find relevant results but apparently includes false positives as well. So to be used with caution by those who know how to check.

Graham Criddle

It seems that the results are not completely reliable though

Your screenshot has chased instead of chesed. Changing that gives better results - though they still require checking

Francis

@Graham Criddle yes, I am aware of that, but thank you for pointing it out. I had input chesed but Logos suggested chased instead, so I tried it. However, I just tried again and it did not suggest it. Go figure. Looks like this specific type of search has interesting potential but can be a bit erratic. Upon further reflexion, it dawned on me that perhaps Logos' suggestion was not an alternate spelling for the Hebrew "chesed" but perhaps interpreting what I wrote as a mispelling of the English "chased" 😆. I was too immersed in original languages mode yesterday to notice that.

NichtnurBibelleser

https://community.logos.com/discussion/comment/1364828#Comment_1364828

Thanks for sharing! From my German Einheitsübersetzung, I get similar but somewhat different results:

MJ. Smith

@Francis

First, for some reason, your results and mine have both verses in common and some that differ. Second, some of the results I got I not right.

This is the nature of AI in most contexts as most AI exposed to the public is not exhaustive in its search nor predictable (to humans) in its paths.

Bradley Grainger (Logos)

https://community.logos.com/discussion/comment/1364862#Comment_1364862

These results are not reliable nor comprehensive. The search warns you about this:

For this search, I'd strongly recommend using a precise search. It's not hard:

Find an occurrence of this in the text. (Smart Search can help with this part!)
Right-click the word, choose the Greek Lemma, then Copy Reference: Search

3. Paste that into a Precise Search query, then type INTERSECTS.

4. Right-click again, choose the Hebrew lemma, then Copy Reference: Search

5. Paste that into your search query. You should have lemma.g:ἀγαπάω INTERSECTS lemma.h:אהב

6. Set the search range to Logos LXX (or any LXX with a Hebrew reverse interlinear) and search:

You'll have a complete, accurate list of all Hebrew/Greek alignments between those two lemmas.

Donovan R. Palmer

https://community.logos.com/discussion/comment/1364923#Comment_1364923

@Bradley Grainger (Logos) (and @Mark Barnes (Logos) )

This is how I would have built this search too and I wonder if Logos would consider at some point building a UI to visually build complex original language searches. The copying and pasting around to build searches like this is not difficult to do, but until you know the software really well, it might not even occur to you that this is possible.

Francis

You'll have a complete, accurate list of all Hebrew/Greek alignments between those two lemmas.

Thank you @Bradley Grainger (Logos).

For my part, my posts were primarily about experimenting and making discoveries such as the differences in my first screenshot between the results when using transliterations vs original language letters, which can be fruitful alternatives.

I do recognize and echo your warning about misusing smart searches as substitutes for precise search.

I also echo @Donovan R. Palmer's request because as experienced a user as I am, I did find it pleasant to use natural language and transliterations to quickly find some relevant passages rather than have to use arguments and syntax. It may be not super complicated ("easy" is not the word that comes to my mind) but it remains not that user-friendly.

Bradley Grainger (Logos)

Agreed to both your points, @Donovan R. Palmer and @Francis .

Donovan R. Palmer

💥Boom! 💥

This would be awesome! My use case would be to quickly build searches to find language constructs that are being described in a grammar that I am studying.

Beloved Amodeo

If Logos is motivated by the direction of this thread to make adaptations, it would be nice if the feature resulting would be made available to the Pro subscription tier😊