Question for Faithlife regarding algorithm behind the corresponding words Visual Filter

There is a discussion at https://community.logos.com/forums/t/211877.aspx regarding how the corresponding words Visual Filter works when there are multiple possible roots associated with a word.

It looks as though the algorithm does not indicate all correspondences in these cases and it would be good to get a comment from someone at Faithlife regarding this

Thanks, Graham

Find more posts tagged with

Desktop App

Comments

MJ. Smith

weekly bumping process in hopes of generating a response for this thread

Don Parker

https://community.logos.com/discussion/comment/1235924#Comment_1235924

Another related homonymic example re. Parsings and Visual filtering for programmers or curious Bible students to consider can be seen in Is 41:2. Different parsings/roots are represented in different English translations, some identifying qr) I, “call” and others qr) II, “meet.” This is parsed in the reverse interlinear consistently as qr) II, “meet,” which has widespread acceptance since Gesenius, but unlike early versions (LXX, Vulg, Targ.; see for ex. John N. Oswalt, The Book of Isaiah. Chapters 40-66 (NICOT) (Wm. B. Eerdmans Publishing Co., 1998) and still some English versions (as a sampling: kjv and csb “called;” leb and nrsv “summoned” -and leb note a:or “called”).

So what a logos user of LEB finds is a translation of root I, call/summon, but a parsing of root II, “meet,” which might legitimately cause some confusion.

And the visual filter accordingly will not link this occurrence to uses of qr) I, “call” in this same chapter. In v. 4 and v. 9 qr) certainly seems to mean “call” and not the much rarer root II, “meet.” And the use of such a leitwort/catch-phrase/linking word/theme is common in biblical Hebrew and significant for Bible interpretation.

Whichever root one deems represented in this verse, a really good software program might want to show a parsing consistent with the English translation being used by a Bible student and a GREAT software program might somehow point to the homonyms that explain the differences among the various translations.

Not an easy task, admittedly, but that is why Logos gets the big bucks, no?

Don Parker

https://community.logos.com/discussion/comment/1237424#Comment_1237424

One last textual example, Is 41.23, which is similar in regards to the reverse interlinear parsing mismatching a translation. It is distinct in not being related to homonyms, but rather different potential readings for the same consonantal text, the kethiv-qere. So this is another call for at least consistency between a translation and a parsing found in the reverse interlinear.

I think any teacher can recall a student offering two disparate answers to a test Q. This is essentially playing with the teacher by offering an incorrect answer along with a correct answer, challenging the teacher to choose which answer the student has in his mind and daring the teacher to call him out one way or another. Or a language instructor can recall asking a student to translate and parse a root, receiving two different answers thereby; and the student wants full credit, rather than 50% either way – a failing grade.

In Is 41.23 English versions are quite divided in translating as from either yr) “fear” or r)h “see.” [As it happens, some forms can look similar in having only two root consonants remaining (here resh and aleph) , because these are both “weak” verbs (I-yod and III-heh sometimes respectively losing either initial root consonant yod or final heh).]

The ESV, ERV, NASB/1995, NET, and NIV all render the last verb in Is 41.23 as “fear/be terrified/in awe,” - reading the root yr). On the other hand, KJV, CSB/HCSB, and LEB translate as “see/behold” - reading the root r)h.

Yet the reverse interlinear consistently parses as r)h even in English Bibles that translate the verb as coming from yr).

Unlike a smart-aleck student, the Massoretes were trying to accomplish two specific tasks: preserve the traditional/received Hebrew consonantal text and vocalize the text as they understood it; and the Kethiv-Qere does so. Where an English Bible has made a choice between two roots by way of translation, the parsing of a different root in the interlinear does not clearly achieve two different goals; but rather, for most logos users, will only create confusion.

[Of course, some English Bibles add a helpful note at the translation gloss: LEB n.w; ESV n.2; NET n.45]

MJ. Smith

https://community.logos.com/discussion/comment/1237424#Comment_1237424

I think you misunderstand the structure of the reverse interlinears in the sense of ideal vs. actual. When I was first using Logos, they tried to create reverse interlinears that reflected the text that the translators actually used. Unfortunately, this is rarely, if ever, available -- rather one is inferring the choice of the translators from among the textual variants. Faithlife later switched to tagging a few standard original language texts and aligning each translation to one of them (well up to three when OT, NT, and deuterocanonicals are based on separate original language texts). This means that the reverse interlinear represents a base text from which the translation was made NOT the precise textual choices made by the translators. We can identify the base texts from the resource information panel.

This at least gives us a defined base rather than some FL editor's guess. To address cases where the translator appears to have chosen an alternative text, there is the Textual Variations section of the Exegetical Guide and the various apparatuses. The parsing depends upon the linguistic theory that the translator is using - even the parts of speech are not fixed. The result is that the user must be aware of the assumptions and theories that the various elements of FL tagging apply, just as they must understand the assumptions and theories of the various authors of the books FL offers. Yes, FL could improve (and probably will over time) as computational linguistics/artificial intelligence expands what is possible and what is practical. But for now, it is important to understand what one is looking at before using it.

Don Parker

https://community.logos.com/discussion/comment/1237700#Comment_1237700

Thank you. That does explain the last example and doubtless other K-Q and countless text critical decisions made by translators that are at odds with the ri.

Your more accurate portrayal clarifies the potentially misleading statements at Logos website and in the brief description in the basic Explorer about the reverse interlinears ("work backward from a translation to the original language").

Still I wonder how many students using logos for basic word study will attempt to construct ludicrous associations between an unrelated divergent root in a particular base text (and a particular reading of that base text as kethiv or qere for the MT) with a specific translation. So perhaps the call for improvement, even in this last example, is not without some merit.

MJ. Smith

https://community.logos.com/discussion/comment/1237991#Comment_1237991

Still I wonder how many students using logos for basic word study will attempt to

I have found Logos users to range from (a) the naive who accept all FaithLife tagging as definitive (b) to the sophisticated who use just enough salt to always double check the results by coming at the same material from multiple directions (c) to the paranoid who assume all the tagging is so contaminated with denominational bias that the forums should be filled with their unique take on all things Biblical (d) to the pragmatic who think its not perfect but it doesn't have to be for their purposes. In short, I've given up worrying about how others take Logos; I just plug away at reminding them of the details of the data.