Hebrew transliteration source; qamets vs. qamets chatuph principles used

Did I miss the post on here for the source of the Hebrew transliteration used in Logos, for morph-tagged English versions of OT (kjv, esv, leb, etc.)? (I saw some posts where some said they did not know source. I even got the same answer myself from logos help personnel.)

In particular, transliteration for qamets or qamets chatuph often seems incorrect and it would be interesting to see what proposed principles led to such decisions. (I saw a post where someone indicated that chatuphs are often transliterated incorrectly as reg. qamets, and also a mention of the ex. of Gn 1:21 [2x transliterated in logos as reg. qamets, but clearly should be chatuph, reduced from cholem in kOl when attached to following noun by maqqeph]. But it seems even more frequent in logos that reg. qamets is transliterated as chatuph.

As in Qal pf 3fs if no metheq - but even some rabbis suggest Massoretes were inconsistent with metheg (Or maybe often assumed familiarity with basic verb patterns. So I saw a pericope that had the same Qal pf 3fs within a couple verses, one with and one without metheg, and logos transliteration rendered them differently, apparently there based on presence and absence of metheg!).

Or Niph impf which I was taught had i-class, daghesh, a-class (so reg. qamets) pattern, but these are often transliterated as chatuph instead in logos.

And many of these verbal examples are transliterated as having chatuph followed by a silent/closing-syllable shewa, even when the next syllable begins with a begadkefat consonant without daghesh (indicating that preceding shewa should be a vocal shewa (or at least medial) and indicating a long, not short, vowel should proceed that.

At any rate, if there is a source, I'd like to be able to examine the underlying reason(s) for those kind of transliterations which seem odd if not plainly incorrect.

[BTW. None of this is to deny or refer to a few cases where even famed grammarians disagree on distinquishing qamets vs. chatuph.]

Find more posts tagged with

Comments

bump for attention 6

Have you checked the internal documentation of the transliteration tool? It has some information.

https://community.logos.com/discussion/comment/1200706#Comment_1200706

Had checked Transliteration under Help and Help for Text Converter interactive. How do I access any other "internal documentation?" (With apologies for my ignorance.)

TY for any and all help.

https://community.logos.com/discussion/comment/1200896#Comment_1200896

PS Unless you are referring to manuals in documentation and didn't see specific one there?

https://community.logos.com/discussion/comment/1200899#Comment_1200899

If you look in Logos Help, under Program Settings you will see a description of Hebrew transliteration formats. The (default) Scientific is a Logos scheme. It is likely the one used in their RI's.

https://community.logos.com/discussion/comment/1200958#Comment_1200958

TY for response, but apparently I am as untransparent in my question as logos is when it comes to the source of its Hebrew transliteration.

I am not asking about the format/style of transliteration, but the source/author of the actual transliteration given, for ex., in the reverse interlinear ribbon.

So for "all-of" in Gen 1:21, transliteration is kal, with flat line over a, which I take for regular qamets. If you put the same BH into the text converter you get kǒl in the scientific format (and simply "o" in the other formats; but not "a" with a line over it for any of the formats.) It seems to me that the transliterator for the interlinear is reading the vowel as a long "a" qamets, instead of short "o," qamets chatuph, as Christo van der Merwe (and note the transparency there!), does in the text converter interactive. Open to correction on that.

But my question is Who is the Source of the transliteration; who did the transliteration? not What is the format/scheme? I do not think the style of format accounts for differences in such instances. Such distinctions can and are made within any transliteration format.

https://community.logos.com/discussion/comment/1200988#Comment_1200988

It depends on the resource - the most common reverse interlinear transliteration is by Faithlife staff. If you provide a specific resource, I may be able to provide a more specific answer.

https://community.logos.com/discussion/comment/1200998#Comment_1200998

Muchas gracias! So, ESV reverse interlinear ribbon, as just one ex. That same transliteration is actually quite pervasive in many places in logos and I would assume is sort of an Ur source.

Would you know of any documentation Faithlife has provided for that transliteration (that might be more useful than mere names of perhaps unpublished biblical Hebrew scholars among Faithlife staffers)? Perhaps especially any feedback they might have produced to justify their transliterations as opposed to transliterations offered by other resources in logos?

Do you know if there is any kind of effort to update/coordinate transliteration among different logos resources, ideally based on clearly declared principles; or to at least offer somewhere a cavat re. any common alternative transliteration within logos?

https://community.logos.com/discussion/comment/1201003#Comment_1201003

If you open the information panel in the library or from the resource panel menu, under Reverse Interlinear it will tell you precisely what text was used:

This also usually tells you at least the project lead for the alignment. As for the transliteration, my request years ago for more detailed information did not generate a satisfactory answer ... but then I may not have reached the right person.

https://community.logos.com/discussion/comment/1200988#Comment_1200988

So for "all-of" in Gen 1:21, transliteration is kal, with flat line over a, which I take for regular qamets. If you put the same BH into the text converter you get kǒl in the scientific format (and simply "o" in the other formats

Transliteration for "every"/"all"--> kōl with a flat line over o (in the four RI's that I tried). That corresponds to Scientific and, more importantly, to SBL Academic which is my preferred format in Program Settings. The TC shows kol in the other formats!

Now look at "sea creatures" --> tannîn in SBL Academic, which changes to tǎn·nîn when I change my preference to Scientific (as per TC).

(You have to re-generate the RI pane when you make a change in Settings).

https://community.logos.com/discussion/comment/1201004#Comment_1201004

Well, that does show the lead as Chip McDaniel. (Not so concerned with the text as my questions regard transliterations of generally reliable readings across texts, but that tidbit may be helpful in the future.)

Maybe I will try to contact Dr. McDaniel. I only fear I will get the same response Mark Smith has posted in thiis forum re. NT Greek transliteration:

"Then let me be more specific. I have an ESV English-Greek Reverse Interlinear New Testament. I contacted John Schwandt, the general editor, to ask this question. He said that Logos had supplied the transliteration and he did not know the source. So, what is the source of the transliteration used in the ESV interlinear?"

TY for your effort and responses!

https://community.logos.com/discussion/comment/1201037#Comment_1201037

I believe you at looking at the transliteration of the lemma in the RI, which has cholem-waw long "o" vowel, rather than the MSS transliteration -- "all-of," more literally put, (a construct noun form, enclitic to following noun by means of maqqeph, with subsequent vowel reduction). If you have the MSS transliteration enabled in the RI ribbon and click on the word "every" in Gn 1.21 in ESV or LEB or KJV, you will see a transliteration as kal, with line over the a --which differs from results you will receive in the text converter (as mentioned in earlier post) IF you enter the mss, not lemma, form there.

I will make one last attempt to clarify. If speaking your name to someone on the phone they transcribed it as Dove rather than Dave, the problem would lie with the transcriber, not with the English alphabet. The issue I raised has nothing directly to do with the style/format of transliteration used (each are essentially separate alphabets, but each can distinguish vowels of different class and length appropriately with their own sigla. Actually, they show very minor differences across the styles re. vowels).

But we are certainly talking past each other and probably should desist our correspondence on this Q, Dove, er, Dave.

https://community.logos.com/discussion/comment/1201051#Comment_1201051

Dr. McDaniel (who turns out was once a guest instructor, as TA Ferris McDaniel, in a class I took) graciously confirmed that he personally had nothing to do with the transliteration supplied in the ESV RI and that he did not even see it while he coordinated the actual interlinear (text-to-translation linkage). He suspects (as I did) that older sources like Strong's, TWOT, etc. may have been used in compiling a transliteration database that was then inserted into the RI.

If anyone on this forum has any pull with logos, maybe suggest that the MSS transliteration needs a thorough inspection/correction/revisions?

Andrew Batishko

https://community.logos.com/discussion/comment/1201068#Comment_1201068

Just to be clear... neither the ESV nor the Reverse Interlinear contains any transliterated text. The Reverse Interlinear contains Hebrew text, and this is converted to transliterated text on the fly based on the style you have selected in your program settings. This is exactly as is done in the Text Converter tool. I don't see any difference between what shows in that tool and what shows in the reverse interlinear ribbon.

Perhaps you can post a screenshot of your Text Converter tool, plus indicate which transliteration you have set in your settings.

https://community.logos.com/discussion/comment/1201057#Comment_1201057

I believe you at looking at the transliteration of the lemma in the RI, which has cholem-waw long "o" vowel, rather than the MSS transliteration -- "all-of,"

Correct. I realised that too late to correct my comment. However, whilst it is kāl with my preferred format in Program Settings, TC indicates kāl or kal for the other formats.

But we are certainly talking past each other and probably should desist our correspondence on this Q, Dove, er, Dave.

I think so, Dan[:)]. I will follow your response to Andrew's observations, though.

https://community.logos.com/discussion/comment/1201070#Comment_1201070

Just to be clear- if you re-read earlier posts - the question was about the transliteration in the Reverse Interlinear ribbon. (ONCE it is enabled for MSS transliteration there!)

1) Enable MSS transliteration in RI ribbon for EAV/LEB/KJV, (2) click or hold cursor over "every" in Gn 1:21. Tell me you do not get kal, with line over the "a" in the MSS transl. row and column for the Hebrew text form, and you post that screenshot.

Otherwise, I have moved on.

https://community.logos.com/discussion/comment/1201079#Comment_1201079

Well, good to see you agree that maybe vowels might sometimes have significance, Dave --dOn thought.

And I am pretty glad to see you want to move on as well, but you couldn't resist challenging me again, incorrectly again, and still with irrelevant referencing of transliteration styles:

"However, whilst it is kāl with my preferred format in Program Settings, TC indicates kāl or kal for the other formats." - from your post.

You still betray a limited understanding of the issue - Where did you get your Biblical Hebrew training, btw?

Select the entire Hebrew construct phrase in Gn 1.21 = kol-nephesh (in crude transliteration), and paste that into tc.

You will not get kal, either with or without a line over it, in any format.

Sorry I can't let a misrepresentation pass unanswered.

But I really don't understand some of the replies on here, to the effect that "your eye-sight must be bad; you are WRONG; you don't know about different transliteration systems or what an interlinear is.

I asked a very specific Q: the source of the transliteration. If you don't know the answer, why the compulsion to jump in and start throwing punches??

https://community.logos.com/discussion/comment/1201088#Comment_1201088

I asked a very specific Q: the source of the transliteration.

according to Andrew Batishko (a Faithlife employee) it happens "on the fly" based upon your chosen transliteration format. What option do you have selected in the program settings?

https://community.logos.com/discussion/comment/1201085#Comment_1201085

Just to be clear- if you re-read earlier posts - the question was about the transliteration in the Reverse Interlinear ribbon. (ONCE it is enabled for MSS transliteration there!)

I believe this is a response to the post by the Faithlife employee who is the best available information. Whether you view the RVI in the ribbon or interleaved in the text, the data is the same.

I do not get a diacritical mark on the a.

https://community.logos.com/discussion/comment/1201093#Comment_1201093

TY, but we are dancing around the periphery IMHO. And that screeenshot there is where Dave's comments about the transliteration system used would come into actual relevance. This screenshot with the result kal, instead of kal, with diacritic is INDEED based simply on the different transliteration systems used.

The transliteration system used with this example is a is red herring. (And this is only one example of the kind of problems I find in the transliteration in logos.) It is unrelated to whatever system is used; rather it pertains to the identification of the vowel represented by a single Massoretic sign in Hebrew- alternatively, short o or long a. Most beginning Hebrew students will understand the difference and difficulty.

Your screenshot - and even Dave finally acknowledged that kall [his system with diacritic, line over a] shows the MSS transliteration as a long a- however that is represented in any system.. I insist it should be a short o instead, as the tc will show if you select the Hebrew text (easiest to be sure you don't misrepresent the Hebrew text is to select the entire Hebrew construct phrase from LHI or other strictly Hebrew text).

As I suggested to Dave, and I think you are a more disimpassioned observer on this Q:

Keep the same transliteration system used for that screenshot,

Select the entire Hebrew construct phrase in Gn 1.21 = kol-nephesh (in crude transliteration), and paste that into tc.

You will not get kal, either with or without a diacritc/line over it, in any format.

Instead you will get, lo and behold, a short o, in whatever format, with or without diacritic [because some systems choose to leave the actual length of the vowel, whether o or a , less clearly distinguished].

If you can't do so, let me know how I can send a MS Word doc/convert if needed, a screenshot showing that, contrary to Dave and Andrew”s assertion, the transliteration is NOT the same in MSS transliteration and in tc.

[As to the suggestion that all the transliterations in logos are done "on the fly" by some program at Faithlife, I am dubious. But IF that is the case, who programed the program and by what principles are outputs produced? As in my example, it is NOT the same as the tc program.]

https://community.logos.com/discussion/comment/1201096#Comment_1201096

[As to the suggestion that all the transliterations in logos are done "on the fly" by some program at Faithlife, I am dubious. But IF that is the case, who programed the program and by what principles are outputs produced? As in my example, it is NOT the same as the tc program.]

I am trained in Sanskrit and Pali not Hebrew and Greek so I was surprised that transliterations were not more carefully defined in Logos. They are a bit more defined now that when I asked. I would expect that the conversion is done automatically using some sort of a graph structure (think a form of AI) but I would not hazard a guess on how since Natural Language Processing is a field that is changing rapidly.

I'm getting the results that Andrew indicated that I would.

I think you are talking about this:

I would assume that the difference is the word standing alone vs. the word in context as part of a compound. In Sanskrit, I would refer to this as sandhi -- fast speech phonetic changes. I don't know how this kind of phonetic change is talked about in Hebrew. Have I correctly understood your concern?

Note also that in the RVI there is an untranslated word.

https://community.logos.com/discussion/comment/1201097#Comment_1201097

Bingo! You and Andrew got the transliteration correct using tc. kol, NOT kal, with or without diacritic.

And you are correct to some extent about sound change due to connection with following word. (But if you look at the regular/dictionary/absolute BH noun form you will see kOl, a long "o" - cholem. [PS. I actually mis-spoke, off-the-cuff, saying cholem-waw when side-tracked by Dave's mistaken reference to the lemma and not the text form.. Reduction typically occurs within its class [a/i/u], barring historical anomolies, so the noun form in construct attached by maqqeph reduces to a short "o," with or without diacritic. And this is NOT what is represented in the MSS transliteration, which is kal, with or without diacritic, long "a."

So even if anyone were to argue in favor of one reading [wrongly as long a, vs. correctly as short o] over the other, no one can insist, as Dave and Andrew did, that there is no difference in these transliteration resources.

TY for restoring my sanity after all the rabbit trails about transliteration systems.

But this one example will perhaps show also some of my hesitancy in accepting the hypothesis that there was any kind of "automatic" conversion and, if so, sufficient principles involved in converting the many instances of some similar Massoretic signs for alternative vowels, reduced sounds, and accents.

https://community.logos.com/discussion/comment/1201098#Comment_1201098

perhaps show also some of my hesitancy in accepting the hypothesis that there was any kind of "automatic" conversion

To this I can only say attend a language processing professional convention and/or a Bible language and technology convention. As to "sufficient principles" -- I'm a Sanskritist -- which computers are finding very difficult to deal with.

https://community.logos.com/discussion/comment/1201099#Comment_1201099

I admit I am no computer programmer, but I have some idea of some of the complexities of Masoretic Biblical Hebrew.

[Maybe indeed there is some incredible processing happening "on the fly" at various logos servers crunching, importing, uploading and filling all the various data fields in the scrollable! interlinear ribbons for the English-Hebrew interlinear text via calculations at the speed of light, even as I and thousands of users jump with a mouse-click from chapter to chapter or book to book. Seems like it would be easier, safer, and swifter to use hyper-links to specific databases, links like so many other logos resources use without special conversion processing by uniquely designed programs. And all that was programmed and packaged, what, 20+ years ago for these interlinears? I choose to remain an agnostic on that.]

At any rate, that "program" for MSS transliteration produces output different from the tc. And the tc has a fine Hebrew grammarian's name attached and the output is more what I would anticipate based on basic principles of biblical Hebrew. So will you forgive me if I don't bow before that particular unseen and somewhat questionable "creator?"

Anyway, always appreciate your input. God bless. Hasta la vista.

https://community.logos.com/discussion/comment/1201100#Comment_1201100

Actually, I'm retired from a career in IT. When one uses AI one normally uses a blank canvas pre-programmed package for the AI method you are using. You provide some rules and turn it loose to train itself ... yes, one has to verify the output. One design constraint on Logos, especially in the era the RVI were originally developed was to insure that the package worked without internet for those in remote missions (or politically sensitive missions). I don't look at it as producing different output rather that single word and phrases produce different output. The RVI because the words are out of order, by necessity uses the single word form.

https://community.logos.com/discussion/comment/1201101#Comment_1201101

I've had the time of my life...

But since you are so congenial [and a programmer and Sanskritist!], and

since you attempted to answer my initial Q, and

since at least you finally came to somewhat see what was behind my original Q about the source of the MSS transliteration, and

since I have fully expressed my opinion that that resource bears a thorough review,

I will extend this dialogue.

Now, despite your very astute observation that some transliterations are context-dependent, you are still missing some of the point - but I take it only because you are not quite so fluent in BH. This same minimal BH background was initially displayed in some of the posts, as an apparent lack of awareness that the same single/identical Masoretic sigla/vowel/point can be used for both long A and short o. But it is further revealed - again assuming just because of lack of BH training - when you suggest that the rvi "by necessity uses the single word form."

You are absolutely correct saying there is some constraint made when words in the rvi are out of order, but that does not mean, IMHO, that the MSS transl. gets to make up words. kal (w/without diacritic), with long a, is not "the single word form" for this BH text-form. The single word form is the lemma, kOl, with a long o vowel, a totally different sigla/vowel/sign/pointing - that is, a superscript "dot," vs. the subscript T-sign for, alternatively long a or short o. If the MSS transl. had put kol, with or without diacritic, to represent the "single word form (with long o)" -- rather than purporting to represent the form in the text as kal (long a; aside from any consideration of the merit of such a vocalic representation) -- that would have been less egregious and perhaps attributable to the mere constraint of one transliteration word for a Hebrew word ripped from its multi-component form in context. [Even if that itself would have been something akin to "a text without a context is a pretext."]

But honestly, the Mss transliteration in Gn 1.21 is the type of egregious error that a beginning Hebrew student would not make or would often be called to task and further study. [Actually the distinction in this verse is only significant for accuracy of representation of the Hebrew form and has zero impact on any meaning, i.e., no one suggests the meaning there is anything other than something like "all-of/every." But this distinction of short o vs. long a can have semantic value elsewhere, pointing to different words. That is why a BH teacher would instruct re. accurate principles for making the distinction.]

Such an egregious mistransliteration is, however, quite like the type of transliterations that one might have found over a century ago, somewhat similar, say, to the type of transliterations found in Strong's concordance. There, along with other Q-able representations, seems no basic principles of BH syllable structure [which, btw, apply in the Gn 1.21 ex: "A closed, unaccented syllable will have a short vowel." not a long a].

My suspicion is that at the start of the transliteration project which is now found in the MSS transl, some such older body of data was used as an initial input and then some revisions proceeded from there, often with such dubious renderings continuing unaltered. So -- unlike the computer program development process which you describe as normal now. Just a guess/an impression from the stunning MSS transliteration and from looking at different BH resources over the years, as well as a fleeting discussion with Dr. McDaniel. (But I do not here intend to represent his view or speak for him.) Unless someone who was involved in the production of the MSS transliteration surfaces to explain what was actually done. Why, if unique principles were discovered and applied in the process, by all means, let's get those into the light of day and force retractions from, at a minimum, half of the BH grammars issued in the last 50 years.

https://community.logos.com/discussion/comment/1201108#Comment_1201108

Thank you for taking the time to educate me on Hebrew. Perhaps, Dr. Heiser will hear of the thread and weigh in.

Andrew Batishko

https://community.logos.com/discussion/comment/1201108#Comment_1201108

IMHO, that the MSS transl. gets to make up words. ... My suspicion is that at the start of the transliteration project which is now found in the MSS transl

You dismissed this earlier, but this is not pre-generated and part of the resource. It's done on the fly by rather simple code that is running on your machine that transliterates using table lookup and a set of simple rules.

I suspect that the we get "kal" in the ribbon instead of "kol" because כָּל־נֶ֣פֶשׁ is split into two separate columns due to aligning it with individual English words, and the transliteration code ends up just operating on כָּל־ (which comes out as "kal" if you place just that into the Text Converter).

I will write up a case for this problem, but I don't have any idea when or if this will be fixed.

https://community.logos.com/discussion/comment/1201131#Comment_1201131

Wow! You are correct.

A similar error can actually be generated in the tc, when using simply the construct form plus maqqeph standing alone (which, of course, never occurs in biblical Hebrew) or using a made-up Hebrew word which doesn't exist as a stand-alone single word for "all/every" (k-qamets-l, with not even the maqqeph, nor an attached noun which is required to generate such a form, non-word). I myself, with a modicum of knowledge of BH, would never have conceived to attempt such an adventure in Xanadu.

Now I understand how, in the earlier post, you could say the same transliteration appeared in both resources, whereas I was generating a different translation in the tc using the entire construct phrase. (You and MJ ultimately produced the same, by inputting much of the surrounding context as well.)

Guess I need to put a poor database-origin theory back in the closet.

Sadly, it also seems to mean that a similar/same? program functions behind both resources. And that users need to be on the lookout to potentially generate such errors in the tc, apart from basic knowledge of BH, as well as look for those more obviously displayed in the MSS translation.

Unfortunately this entire root canal of a procedure has been enough for me and I will not even attempt to display what I consider errors, at least potential if not actual, elsewhere in the MSS transliteration.

Still, one may dream of a day when the veil is removed to know who the transliteration program creators were, their BH expertise, and what resources and principles were used in its development. But maybe they prefer to remain in a witness protection program. Or don't want their name/s attached to the kinds of errors mentioned above, without caveats or attempting corrections.

https://community.logos.com/discussion/comment/1201155#Comment_1201155

who the transliteration program creators were, their BH expertise, and what resources and principles were used in its development.

Barking up the wrong tree [8-|] The programmers have no need of BH expertise. They use an SME (subject matter expert) to assist in design and testing. I would begin by looking at the transliteration tables and guidelines provided by journals as standards for their publications.

it seems even more frequent in logos that reg. qamets is transliterated as chatuph.

If I am reading the table on this page https://support.logos.com/hc/en-us/articles/360016599611-Hebrew-and-Greek-Input-Options#transliterate correctly, the "qamets" and "qamets hatup" are both rendered the same. I cannot comment about the source of the table, but it may reveal the basis for the transliteration product.

https://community.logos.com/discussion/comment/1201176#Comment_1201176

Congrats Dave, you are slowly coming around to the issue in the ex. given -- you have discovered that the same sign can represent two different Hebrew vowels. (That is a genuine beginning.)

The table shows how to generate the single sigla, which is identical for either, a subscript T essentially. It can represent a long a or a short o, but to determine which, within a syllable of an actual Hebrew word, and how to transliterate it correctly, actually requires some knowledge of biblical Hebrew.

My original Q, for the last time, was: What was the source of the MSS transliteration? The type of responses I have received seem mainly:

-No peaking behind the curtain!

-Programmers at Work!

-It was done in the ancient past by unknown scribes whose tradition must be honored at all costs.

-It all occurs seamlessly and instantaneously generated

-There are no contradictions or errors [manifestly false].

-You are wrong about everything! At any rate, Never question The Logos!

-We don't need no stinkin' Hebrew.

I will obediently slink away, vowing never again to speak of the transliteration program that Must Not Be Named. Still I take away trinkets of mundane and frequently incorrect information about transliteration systems, interlinears, Hebrew morphology and syntax, and the manufacturing of new Hebrew terms - all treasures tossed out as I attempted to run the gauntlet.

https://community.logos.com/discussion/comment/1201196#Comment_1201196

Congrats Dave, you are slowly coming around

Leaving this thread. I (a simple logos user who was attempting to help) do not need to subject myself to disrespect

Quick Links