Many Biblical Hebrew transliteration errors in logos
There are literally thousands of transliteration errors for Biblical Hebrew in the MSS transliteration [MSSt] in Logos (similarly also in parts of the Word-by-Word section, as in the Exegetical or Passage Guides) and users need to beware. (Some sample types given below.) Of course, in one regard, this warning is merely a reminder that "A little knowledge is a dangerous thing."
Additionally, this is a call for complete revision/update of such transliterations in Logos, especially in the MSSt (and in the Word-by-Word section). (Also, to a lesser extent, elements within the text converter tool [tc].)
The very process of breaking up components to transliterate in the MSSt is not even an actual “manuscript” transliteration (full text-form) and that process has contributed to such abundant errors.
Biblical Hebrew [BH] is frequently a componential/compositional-type language. The ubiquitous addition of prefixes and suffixes, and even close linkages of words, create sound changes to the finished MSS amalgamation/compound word-forms within BH (re. the vocalized Masoretic text). This should be accurately reflected in transliteration.
To a beginning BH student, such attendant changes can seem odd (since similar changes in English, for example, occur almost unconsciously) and generate frustration. Soon, however, deconstructing a componential form becomes an intriguing and enjoyable experience, like puzzle-solving or detective work. This work is done for an English reader in parsings and by providing the lemmas/dictionary words. Later on, for the BH learner, the very sound changes – if accurately portrayed in transliteration -- make accurate parsings more automatic.
However, tearing apart the components within a MSS form and then representing them in isolation can ignore -- completely or partially or misleadingly-- the phonological processes which were involved in the creation of the MSS composition. The result produced may then be an incorrect transliteration, mongrel forms that do not exist as BH words, or a Frankenstein-type transliteration that represents neither the individual components nor the MSS compound-form, nor even any single lemma/dictionary form.
Three common analogies may demonstrate, even to someone without any knowledge of BH, the potential danger of using a "slice-and-dice" computer-generated method to produce a transliteration, such as is apparently reflected in the MSSt ribbon of the interlinears in Logos and, to a lesser extent, can be generated in the tc.
1) Anecdotally at least, there was an early Russian computer Bible translation program used for "The spirit is willing, but the flesh is weak." The program spit out a word-for-word "translation" as "The vodka is strong, but the meat is rotten." For the individual elements, not so far off, but for the whole, unacceptable to most readers. In fact, the result is something of a knee-slapper to anyone familiar with the original and the intended meaning of both the whole and the parts.
2) Another analogy might be drawn to the fallacy of some etymological-type deconstructions. If the term "butterfly" is broken into its components, it results in something like "levitating dairy product!" Simply break apart the components, "butter" and "fly," no?
3) Finally, consider analogies from chemistry: If a friend said he had presented you with a container of hydrogen and another of oxygen – referring to a bottle of water --would he be taken seriously? Or, a misleading chemical formula listing elements of a compound while ignoring an essential catalyst?
Something of the sort is frequently misrepresented by logos in MSSt (and sub-divisions of compounded forms in the Word-by-Word sections) and can at times be replicated via the tc.
Now one of the main reasons any preacher/teacher/writer would speak/write a transliteration is to add credence that their message is based solidly on the Hebrew text (whichever text is the basis of the exposition). As a general guideline, such references should be quite limited. But, if such a citation is offered, an improper transliteration can actually undercut the credibility of the speaker/writer, being a public display of an actual lack of BH knowledge.
[To clarify, this has nothing to do with individual or dialectical or historically-influenced pronunciations, but rather any Hebrew MSS form under discussion. Transliteration of a text-form itself can be distinguished from, say, Ashkenazi vs. Sephardic Jewish pronunciation or Germanic-influenced vav vs. waw pronunciation. And a long A differs from a short o, no matter how these are represented in any transliteration system.]
Previously I pointed to multitudinous places of kol-, "all-of" [kaph-qamets chatuph-lamed in construct with maqqeph linkage; cf. GKC #9r and #9u1(c)] incorrectly in MSSt as kAl, with long A, Gn 1.21, etc; similarly in tc, unless the entire BH construct phrase is entered, then it is generally correctly transliterated with a short o.
Below are some additional types of constructions to give an inkling of the many misleading transliterations, perhaps most easily found in the MSSt.
[I use a simple transliteration system for the vowels, without diacritics, to represent vowels at issue: long a/qamets > A, vs. short o/qamets chatuph > o. Transliteration of actual full manuscript forms from tc are copied using the Simplified transliteration system there. However, where there is another distinct, long o vowel/cholem is represented as O, so as to distinguish that from short o which is at issue in many of the exx.]
Noun + pronominal suffix [pro suf] causing reduction:
1 Sm 20.30 long A, can be created in tc like MSSt; correctly as short o in full form tc levoshtekha “to your shame”
**1 Sm 22.8 same form 2x in text!; incorrectly as long A for 1st use; but correctly as short o for 2d use in the verse! [So MSSt contradicts itself!] Full form correctly in tc: eth-ozniy, “(opening) my ear”
Infinitive construct [Qal inf cnst]+ pro suf forms incorrectly as qamets/long A for the inf cnst base/theme vowel. [See for correct transliteration of such inf cnst forms + suffix as qamets chatuph/short o in Lambdin # 115 (pg 128).]:
Dt 6.7 incorrect "base" as long A in both; full form correctly as short o in tc: uveshokhbekha , “and in your lying down”
Jr 45.1 kAth; but * tc for full form correctly has short koth-bO, "in his writing"
1 Sm 13:1 mAlk but * tc for full form correctly has short mol-kO, “(in) his reigning”
Lv 7.36 mAsh' VS full form correctly in tc moshchO, “to anoint him”
1 Sm 19.11 long A, vs. correctly in tc for full MSS form leshomrO,”to watch him”
Similarly, when a noun/here the "subject," is attached by maqqeph to an inf cnst:
1 Sm 22.8 correctly short o in tc for full form, bikhroth-beni, “in my son's cutting/making”
Jussive when particle of entreaty is attached:
1 Sm 16.22 incorrectly as long A in MSSt; correctly short o in full form in tc: ya'amod-na*, “Let stand/serve”
"Imperfect"/wayyiqtol/waw consecutive + preterite form affixed with maqqeph to a following prepositional phrase:
1 Sm 17: 51 yikrAt in MSSt; can be created as same incorrect form in tc; correctly entire multi-form in tc: wayyikhroth-bah, “and he cut with it”
Reduction from cholem theme vowel to short o when a pron suffix is added to waw consec + impf/preterite:
1 Sm 15.17 long A in MSSt; correctly in tc wayyimoskha “and He anointed you”
Occasionally qamets chatuph preceding compound patach shewa; here in Qal inf cnst base:
1 Sm 15.1 mehshAch in both VS correctly in tc short o, limshohokha , “to anoint you”[For this and other exx of same as well as different types, cf. GKC#9v, p. 50]
Similarly, qamets chatuph before compound shewa qamets chatuph for a few nouns [cf. Furtato 9.9].
1 Sam 17.54 long A for cut-up form in both; correctly in tc for full MSS form as be'oholO [cf. "the majority of commentators are inclined to emend the text from be’oholo, “his tent,” to be’oholi, “tent [of Yahweh.”- Barber, The Books of Samuel, vol. One, 206n20"]
Directive [locative] suffix -Ah (Lambdin #58 w/many exx; Gkc #90i on the long A; sometimes incorrectly as -oh in both Mss Transl and in tc; other times correctly in both.)
1 Sam 15.12 hakkarmeloh incorrectly as short o in both! “to-Carmel” [But correct in other places:1 Sm 16.13, 19:18 haramathAh correct in both; 17.49 aretsAh correctly in both MSSt and tc.]
Similarly, -Ah ending on an adverb:
1 Sm 19.17 short o in both, incorrectly as kakhoh; yet correctly in Bible Word Study tool, and #3970 DBL ka·ka(h, and correct even in some places in the context menu.
Less consequentially, but certainly abundantly,and further grist showing the need for an update, daghesh forte conjunctivum is not even represented in transliteration for partis/components in both; but correctly doubled in tc when conjoined by maqqeph as the unified form :
1 Chr 29.18 zoth in both when separated; shomrah-zzoth as conjoined in tc
1 Sm 22.3 (conjunctivum 2x in single multi-form!) both represented for full form in tc: mah-yya'aseh-lli
These samples of some amalgamated forms – where transliteration errors-by-division may readily be produced – occur many times in biblical Hebrew and other types could be added. So, as per the Hill Street Blues admonishment, “Let's be careful out there” when using the transliterations in logos.
Until an update is forthcoming: In the Word-by-Word sections of the Guides, prefer the transliteration given for the full MSS text form. If using the MSSt or tc, copying and pasting the full MSS form into the tc is recommended to avoid some such errors.
Comments
-
Which transliteration scheme are you assuming? There are several approved by academic publishers so I can't check the sample errors you note without knowing what rules to compare against.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Hi.
Repeating from the post:
"[I use a simple transliteration system for the vowels, without diacritics, to represent vowels at issue: long a/qamets > A, vs. short o/qamets chatuph > o. Transliteration of actual full manuscript forms from tc are copied using the Simplified transliteration system there. However, where there is another distinct, long o vowel/cholem is represented as O, so as to distinguish that from short o which is at issue in many of the exx.]"
But to paraphrase a great mind, If you think this is about transliteration systems, "you are barking up the wrong tree."
0 -
PS: The hope was not to once again be side-tracked by the irrelevant topic of different transliteration systems. So here is an addendum for anyone else who decides to follow this thread.
Transliteration systems cannot be used as “smoke-and-mirrors” to escape the obvious contradictions and errors as in the examples provided in the original post. (for the most part, simple and non-controversial exx.; many can be corroborated via tc; refrained from so many more and more significant - If I speak of banal BH examples and those are not believed, how may I tell of more complicated ones with those who have minimal language training?)
In fact, there are relatively few differences between transliteration systems and these are easily accommodated, mutatis mutandis, with appropriate changes.
All the systems in tc represent long A, qamets, as a, or a with a macron, horizontal line above.
All the systems in tc represent short o, qamets chatuph, as o, or o with a breve, short u above.
There is zero difficulty, no matter what system, in distinguishing what is represented as a long A from a short o, in any system of your preference.
Look up some of the verses and word-forms cited, using whatever transliteration system you prefer, and compare. For most of the exx. you will be able to see two different transliterations offered – one for the full MSS form and one for a sliced-up portion cut out from the amalgamated form. Maybe easiest to see in a Word-by-Word section in a Guide (like Exegetical Guide or Passage Guide).
Are both true representations when they differ? Can they be harmonized so that a long A is at the same time a short o? Which will you use and how will you explain the difference? And if one is actually incorrect, how do you know which is right?
[Just a hint: Many of the transliterations of those portions/parts cut out of amalgamated MSS forms are nowhere represented in BH and are not even BH lemmas. But to know the latter requires a bit more BH.]
Take the following different “transliterations” in different schemes, but all equivalent representations, mutatis mutandis::
2= 1 + 1
two equals one plus one
dos es uno y uno
But say you also see: 2 = 1 + 3, etc. Now a choice must be made as to which representation is correct, not between “systems.” Of course, in some bizarro situations, hypothetically, both equations could be correct.
Now suppose someone said 2 = smoggedlygoop y six. Most English speakers would protest there is no such word smoggedlygoop. (Other language speakers might not know this!) And even if treated as a variable, it would at least be misleading to say that smoggedlygoop is universally common in English parlance and represents a specific negative number.
Until Logos recognizes and corrects these types of manufactured BH transliteration errors, use them with some caution.
0 -
Sorry my question annoyed you.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Did not annoy me. But concerned that others in this forum would misunderstand that the topic is in some way simply connected to whatever transliteration system, as in previous thread which went on interminably down that rabbit hole and didn't want that to happen again.
Really, I have no doubt that you will be able to follow the breadcrumbs I dispersed.
0 -
Don Parker said:
as in previous thread which went on interminably down that rabbit hole and didn't want that to happen again.
Is there any new information that takes Andrew's response into consideration? I took his response as a reasonable explanation based on one more piece of information than I had available.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
All I recall from that is 1) OK,- reluctantly and after a lot of hair-pulling - there was acknowledgement of the many Logos mistransliterations (100s, likely 1000s of instances) of kol- as kAl and, meh, maybe someone will look at that sometime in the future (in the millennium?). 2) He himself seemed content with all the BH transliterations as produced by well-functioning computer software. Maybe he is a BH scholar, as well as a programmer, but he certainly did not bring any such BH scholarship to bear in the previous thread.
So now, at this point, I was simply appealing to users of logos to examine some of the evidence themselves and to protect themselves from errors, misinformation, and contradictory transliterations that can be found in Logos. I did so by providing c. 11 other types of constructions, as with kol-, where errors are found or can be generated with tc. And these types of constructions each occur anywhere from several to several thousands of times in BH, with the mistransliterations regenerating themselves in numerous places throughout Logos. I contend that many more constructions besides have similar mistransliterations attached in Logos. Discussing some of those additional mistakes would entail even more technical discussions. And that does not look promising when even basic examples of mistransliterations are pretty much ignored.
Logos told me to give my input here and that they have no one on staff to discuss issues of biblical Hebrew related to the program. So I think I have done my due diligence to inform Logos of some of the errors of its ways in misleading users who have not enjoyed the privilege of learning BH. These are the primary consumers of transliterations.
Users have this warning and may ignore it (as I expect most will) or follow it.
Logos may do as it pleases with its product. As far as many of its BH transliterations, it is misleading users and that doesn't seem ethical, knowledge-based, or even a smart long-term business practice.
Despite enjoying some real BH language guffaws along the way, there are SO MANY mistransliterations in Logos that it can become a bit tedious. If no one cares to correct made-up stuff (under the post-modern umbrella?), what's the point?
If anyone has specific questions about the individual constructions and examples cited in the original post or interest in further discussion on the topic of the BH mistransliterations in Logos, you know where to find me. (Unless one may be banned for speaking ill of the Logos.)
TY again MJ, for giving some consideration to my ramblings and for your always-appreciated responses which show a willingness to get to solutions, even if that means wading into BH. I regret if anything I wrote offended you.
0 -
Don Parker said:
He himself seemed content with all the BH transliterations as produced by well-functioning computer software. Maybe he is a BH scholar, as well as a programmer, but he certainly did not bring any such BH scholarship to bear in the previous thread.
The personal credentials of Adam are irrelevant; he is the face of Logos responding to problems in the code or data. He provided the precise answer to your question either from his own knowledge or from asking someone who knew. The problem appears to be that you misunderstand the contents of the MSS line - which is not to say that there might not be some genuine errors in the MSS line. In a reverse interlinear Bible, the words of the text appear in the sequence of the translation target language and the MSS line is resequenced to correspond. The little numbers on the MSS line provide the original sequence of the words. In Hebrew, this sometimes means that a compound is split into two words. The MSSt line simply takes what is in the MSS line word by word and runs it through the transliteration routine. It looks at NOTHING beyond the single word in the MSS line. As I understand your question, this answers your question. In this context MSS means that the word retains any conjugation or declension; it is not necessarily true that it retains all compounding. You need to check the MSS to see if it has or hasn't.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
“As I understand your question, this answers your question.” Well, yes, in that you understand that there is a slice-and-dice process often at work throughout Logos in presenting BH transliterations.
“The MSSt line simply takes what is in the MSS line “word by word” and runs it through the transliteration routine. It looks at NOTHING beyond “the single word” in the MSS line. “
[And that does not even suggest any potential to create problematic transliterations? Like, Describe America, but NO LOOKING at its history! Nothing could go wrong there, huh?]
The problem with the process is that the routine is then looking at sliced bits – often at non-words --and so produces errors. The sliced portion is very rarely any “single word” that would be found in any BH dictionary/lexicon. And because it is a sliced bit, now different phonological rules will produce different transliteration results, simply because it has been separated from linked elements activating different phonological rules involved in its original formation. Or, back to an earlier analogy, the mistransliteration is at other times merely a bogus chemical formula that omits essential catalyst(s) and so is less-than-helpful.
More analogies, since no one wants to discuss specific mistransliterations:
The sliced specimen in transliteration is sometimes, in effect, “a fish out of water.” By removing the surrounding elements that gave it life, the dicey transliteration stinks and is ultimately trash, not a living fish. Or, when I slice up an apple, I don't expect to then bite into a slice of onion; I expect it to retain its apple characteristics.
You are correct when you earlier said it comes down to expectations. I would expect BH transliterations in Logos to be accurate and consistent throughout various locations [not only in MSSt, which on the surface might seem to give cover for such mistransliterations, but also in word-by-word sub-sections and in tc]. Others, apparently like the programming community herein, may be fine with distortions, misrepresentations, and ill-begotten mistransliterations plainly contradicted elsewhere in Logos and not found in BH.
But if users have now at least received more of a peak behind the curtain of the processing program for BH transliterations “on-the-fly” in Logos, they may be better equipped to make their own decisions as to proper BH transliterations. In that hope, I bid adieu.
0 -
Don Parker said:
And that does not even suggest any potential to create problematic transliterations?
To me, it simply means that you need to know what you are looking at. It is correct for what it claims to be; it is not correct for a manuscript in original language sequence, but it does not claim to be. I will agree that it is difficult at times to ferret out precisely what a piece of data is and that there are inconsistencies especially where the users have insisted on anachronous tagging in the Old Testament. But for data that has been around long enough for professors and users to run into and report errors, the data has a high reliability when read as intended.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Good admonition to know what one is looking at.
And really, Just when I thought I was out, they pull me back in:
"it is correct for what it claims to be" - We remain at opposite ends on that issue. I feel like if you were to spend a few hours on BH syllable structure and morphology you would not make that statement.
and "data that has been around long enough for professors and users to run into and report errors, the data has a high reliability when read as intended."
Would that be the same degree of correctness and reliability as shown with the previously acknowledged mistransliteration of kol- as kAL-, that occurs hundreds of times in Logos??
0 -
Don Parker said:
I feel like if you were to spend a few hours on BH syllable structure and morphology you would not make that statement.
I doubt it as Sanskrit has issues that makes the Hebrew look simple and I've studied some Central Asian Turkic languages that are agglutinative and written without vowels. The best Sanskrit parser available today on the first sentence of a well-known Sanskrit poet offers 103 possible parsings of the first sentence ... they missed the correct one. Try reading a sentence consisting of a single compound word of 195 letters (428 in transliteration) with declensional and conjugational ending eclipsed by the compounding and with all initial/terminal letters subject to modification by fast speech rules (sandhi). Provide a dictionary organized solely by the root (often very similar to the Proto-Indo-European root if that helps). So sorry, I do believe I am qualified to judge whether Hebrew transliterations following consistent rules are acceptable when the input is divided in a reverse interlinear that is not how it appears in the original manuscript. I would expect an original language interlinear to reflect what you are expecting in a reverse interlinear.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Whoa. Definitely outta my league talking Sanskrit, even if generative linguistics -- if my poor memory serves -- showed many startlingly similar principles across many languages.
Maybe we can find some common ground without spending hours discussing foreign languages (and hopefully without giving any offense):
Help me using an English example. But let's imagine we are non-native English speakers and are looking at the word "comparable" and using a slice-and-dicer program break it into compar and able.
No issue with compar as a "transliteration" for that first "slice," right?
But how about a "transliteration" given as cAmpar ?
Would that raise any kind of objection - like, that is not only not an English "single word," it is not an accurate representation of the English "root" form, nor with regard to the the full word "comparable?"
Would it be deemed disrespectful to suggest that the proper transliteration should be compar, with short o, and not cAmpar with long A? And that if one speaks or writes it as cAmpar, their facility with the English language might be questioned, even mocked?
If you saw such a transliteration as cAmpar would you say "it is correct for what is intended?" Or might you suggest either compar as a slice or compare as a lemma are correct, but not cAmpar?
PS: As far as "professors and users running into and reporting" errors to Logos, would they have received the same "ready welcome" as I have experienced previously (present company excepted)?
0 -
I would convert comparable into com + par + able which avoids the problem. In fact, I like to draw my tree diagrams down to the morpheme level which frequently results in "pieces" of words we are not used to seeing independently. I expect the parts to work exactly according to the rules as I have set them for the diagram. In this case, FL set the rules - we have a simple choice: use them under the rules assigned to them by FL OR do it ourselves ignoring the FL input. It is really that simple. The rules are neither right nor wrong - they are simply the rules that generated the data and allows us to understand it.
There is a consistent logic in what FL has done; there is a slightly less consistent logic generating more intuitive data in what you wish FL had done. There is no error.Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Yes, dividing that English word into syllables might allow avoiding answering the questions posed by my example. It also obliterates any English "root" identification, so I don't find that "comparable" to the typical associated grouping of three root consonants for regular verbs in Logos transliterations. Clever evasion though.
Maybe we need to backtrack to square one and a single syllable example. Do you want to revisit and defend what was previously acknowledged as a mistransliteration in Logos of kol- as kAl-. Would you insist it had passed high scrutiny and is "correct for what it intends?"
0 -
Don Parker said:
Yes, dividing that English word into syllables might allow avoiding answering the questions posed by my example
You misunderstand - I divided it into morphemes/sememes com=with par=equal able=ability The Proto-Indo-European root is either pere (grant) or per (sell).
Don Parker said:Do you want to revisit and defend what was previously acknowledged as a mistransliteration in Logos of kol- as kAl-. Would you insist it had passed high scrutiny and is "correct for what it intends?"
Not particularly as it makes no difference either way to me. What I am interested in is understanding the tool sufficiently to explain it to others... I am much more likely to follow up by tracking down the transliteration rules and their sequencing.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Actually, immediately after posting it did occur that some might want to go back to par as the historical root (though that completely obliterates my intended comparison as a parallel to the common regular verb transliteration of C-V-C-C so frequently represented in Logos). That may also border on an etymological fallacy in that few contemporary English speakers think in explicit terms of "granting or selling" or even "being equal with" when using the word compare. But at any rate, it was the varied vocalization of long A vs. short o in the made-up transliteration systems and what one should do when they encounter different transliterations (such as within logos) that I was hoping you would address by that using that example. Clearly it was not a good example and/or its intent was not clear.
Since however you are explicit that "it makes no difference" to you whether kAl is a mistransliteration, then none of the other examples I offered make any difference either. So my original post is really not directed to you or other programmers, only to those who might care about transliteration errors in Logos.
Therefore probably should table this, but I am curious how you reconcile an acknowledgement that kAl is a mistransliteration with the insistence that the slice-and-dice method is a highly reliable process that produces good results with no errors.
0 -
Don Parker said:
but I am curious how you reconcile an acknowledgement that kAl is a mistransliteration with the insistence that the slice-and-dice method is a highly reliable process that produces good results with no errors.
I don't have to - I never said "no errors" ...I would argue only for statistically reliable. In natural language processing perfection is not assumed to be possible ... success is defined by the tolerance for errors. Every report of errors that results in a refinement of the processing narrows that tolerance.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Thank you. That makes sense, in theory. In application it might imply that reports of errors would be welcomed and received with serious consideration to improve refinement.
Hey, I want to make suggestion that might move toward refinement without merely addressing individual errors which disinterest you. Back soon.
0 -
I think I may have gotten a brain-cramp to move us in a productive direction.
You are interested in "tracking down the transliteration rules" and I am interested in correct transliterations appearing in Logos. Maybe we can work together a bit toward our different goals?
Maybe if I explain some related phonetic processes in Hebrew and my opinion of how the current slice-and-dice plus attendant rules might produce some of the wrong transliterations, you can compare my explanation with the current rules and consider how they might be tweaked to produce, IMHO, better results.
I assume your acceptance of this challenge. Otherwise you will simply stop reading and not reply.
I will use the previously acknowledged error of kAl-.
(Deep breath to make it through the boring Hebrew part so as to begin your exciting programming part of tracking the transliteration rules.)
The single word/lemma and absolute state for "all" is kOl, kaph-cholem/long O-lamed. But the Hebrew writer wants to "say all-of/every creature/soul," so he prefixes/attaches kOl to that following word nephesh, "soul." How so? Well by the maqqeph/hypen linkage, but also by putting the original lemma into a slightly altered form: in this instance, the long O reduces, staying within its vowel-class, down to qamets chatuph, short o > kol [not kAL], indicating a construct state word, "all-of."
Now there is another element to bear in mind. In BH there is a close relation between accents and vowel length and syllable structure. kOl', "all," as a monosyllabic word, is necessarily stress-accented. ([We need not talk about any seeming variations which "prove the (BH) rule" here.] But here is where an accent diacritic, or simply a basic knowledge of BH, might be important. [Programmer-type question just for your consideration: Is such a word-form stress accent, at least for monosyllablic words, somehow indicated by some transliteration rule? It does seem to me that the answer is yes, and I will shortly propose that very rule is part of the generation of many mistransliterations.]
But what happens when kOl' morphs into a construct form and is attached to a following noun? It not only reduces its vowel length but, as part and parcel of the process, it loses the stress accent, which moves ahead to the following BH word. > kol-ne'phesh, "every soul" as an example.
Now after these phonetic changes occur to make the compounded amalgamation in the MSS, the Transliteration program [Tp] says, I got it from here, I will cut this puppy up as kol- ne'phesh. Ahh, there are two additional wrinkles when slicing and dicing and these may throw off the Tp, or the rule/s as applied currently: 1) the qamets sign, as in the middle of kol, can be either short o or long A. We put it above as short o, because we know the original lemma was kOl' a long O and that vowels reduce within their same class (and here I add another tidbit which may or may not be represented clearly in Tp rules: a closed [= CVC], unaccented syllable (but not a monosyllablic, accented word) will have a short vowel. So we, but perhaps not Tp, know that the proper transliteration is kol- unaccented.
2) slicer-dicer has to choose what to do with maqqeph, decides to put it with enclitic/prefixed kol-; but at the same time apparently (??) is told (by some rule?) not to consider the maqqeph in deciding whether the single qamets, sub-scribed small T, represents regular long A or chatuph short o.
So now what is Tp to do with sliced bit kaph- qamets or qamets chatuph -lamed- maqqeph? Well it has apparently been told to ignore maqqeph (?), leaving a seemingly isolated monosyllablic "word" - and Tp is apparently smart enough to know (or here applies a rule) that if it is a monosyllabic word it must be accented, and also smart enough to know (or here applies a rule?) that qamets chatuph/short o only occurs in a closed, unaccented syllable. But LOOKING AT NOTHING ELSE (other than sliced bit minus maqqeph) Tp triumphantly pronounces the qamets a regular long A, as if in an accented monosyllabic word [not so, but Tp has apparently been told to ignore the maqqeph? which makes it unaccented syllable and not an accented monosyllabic word] yielding a mistransliteration of kAl.
So, can the programmer see what rules match these phonetic processes and what tweak or additional rule can produce the correct transliteration kol rather than kAl?
0 -
Do you see the correct transliteration when you right-click a word instead of using the interlinear?
Using Logos as a pastor, seminary professor, and Tyndale author
0 -
Sorry, Justin, You will see the same mistransliteration throughout Logos features: context menu, guides, interlinear ribbons, all created by the same computerized slicing-and-dicing program using its currently applied rules. (Try yourself for "every" in Gn 1.21. You will see an incorrect long A, for the "slice" as kAl-, there shown with diacritic macron/superscript horizontal line over a, at the top left of the context menu. Context menu comes up with a right-click). This is precisely why I am advocating for a complete revision re. transliterations in Logos.
But you can also see a correct transliteration for the lemma/dictionary form in the context menu, as well as in the lemma interlinear ribbon and elsewhere. That may be enough for you. And as I mentioned, you can sometimes get the correct transliteration of the actual textual form by looking at a fuller context, as perhaps in the text converter tool. Or maybe consult some separate resources in your library: TLOT, EBC, DBL sometimes contain transliterations besides or in addition to the Hebrew. Lambdin's Introductory Grammar contains many transliterations.
I'm curious though: Wouldn't you like to see the correct transliteration just pop up when you open the context menu, without having to do additional research or not knowing if you were accurately representing the Hebrew text? Unfortunately the transliteration program in Logos is a slice-and-dice method (plus current programming rules) that produces many thousands of these type of errors, not just for the word "all/every-of."
0 -
Don Parker said:
Wouldn't you like to see the correct transliteration just pop up when you open the context menu, without having to do additional research or not knowing if you were accurately representing the Hebrew text? Unfortunately the transliteration program in Logos is a slice-and-dice method (plus current programming rules) that produces many thousands of these type of errors, not just for the word "all/every-of."
Can you explain to me why the Bible college and seminary professors of Hebrews have not made an issue of this in the past nor have them jumped in to support your position? This is the basis of my dropping out of the conversation.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
I can suggest several reasons, though I fear they will not make much of an impression. If I read some of your earlier comments correctly, you primarily look at the issue strictly from a programming perspective.
1) transliteration in general is for people who do not know BH or are beginning Hebrew
2) for many people, the lemma/dictionary form is all they want or need, they are not really attempting to represent the text itself, more interested in word studies
3) many professors would simply laugh off such mistransliterations off and prefer to teach BH so that their students would not replicate such obvious errors (or maybe use them as classroom exx of how not to do transliteration)
4) it is not like every verse contains mistransliterations, sometimes may go two or even three verses to see one
5) some of these are minitiae and most do not effect the semantics, and meaning is where most of us want to get
6) some do not see the beauty and consistency of BH morphology and so have no inclination to point such out (the benefit comes in seeing the principles clearly and being able to apply them elsewhere)
7) not everyone uses Logos and language-enabled packages
8) so much of this is just basic Hebrew and many would rather go on to other matters
9) it does take up some time (I thought I would only have a question or two about how such errors were produced)
10) maybe someone tried and came up against the same stone wall I have experienced
11) a very few, maybe only one ex. I gave, are genuinely controverted transliterations
12) some of the issues involve Masoretic accents which are often ignored in general BH courses
13) qamets/qamets chatuph is an issue which is almost never treated thoroughly, but generally with just a few comments in BH courses
14) many might prefer to actually study the Bible or have a life rather then bother
15) many Christians ignore the OT/BH text altogether
16) many people sleep through their BH classes and few are consistent and diligent and read the BH text
17) I only broached the subject after years of feeling like my head was being punched in by looking at so many mistransliterations (Would I have even bothered when I taught BH 40 years ago? Absolutely not.)
18) Logos does not make it easy for people to ask Qs (so even here on this forum have to worry about people getting offended, etc.)
But, that's off the top of my head. Maybe you want to show me the record of all the past discussions Logos has had with BH professors and students over the years about the specific issue of textual form transliteration?
0 -
Don Parker said:
If I read some of your earlier comments correctly, you primarily look at the issue strictly from a programming perspective.
You could not be further off -- my interest is primarily linguistic recognizing multiple linguistic theories each have contexts in which they are most effective. Second, I evaluate it in the context of years on the forums, considering both idiosyncratic, vocal users and quiet academic users. Third, I evaluate it in the context of the international standards for interlinears (multi-disciplinary). But my primary reason for bowing out is my belief that I have not figured out how to communicate effectively with you - the responses don't reflect what I attempted to say.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
I apologize and hope you will forgive me if I misrepresented you.
I was going off this comment: "it makes no difference either way to me. What I am interested in is understanding the tool sufficiently to explain it to others... I am much more likely to follow up by tracking down the transliteration rules and their sequencing."
And I am doubly sorry because I thought you might actually take a look at the transliteration rules.
I was also hoping that someone might want to discuss the actual examples I cited in the original post of this thread.
I, too, am out, unless there is anyone who wants to do so.
0 -
1) look in program settings and determine the Hebrew Transliteration Format you have requested
2) Look in the Text Converter under Help which likely will link you to the SBL Handbook of Style and will show you how to adjust it for the precise format you selected.
3) If you find examples where the Hebrew textual unit being converted does not follow the rules found in 2, you will have identified an error that Faithlife will identify as such and adjust accordingly (or at least put it on their to do list).
I suspect that you'll get hung up on the "Hebrew textual unit being converted" portion. At least that is where I felt I was not communicating effectively.
=====
The reason the transliteration makes no difference to me is that I am not a Herew scholar - it's that simple,but, no, I took no offense.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Thank you for taking a stab. I have appreciated all your input.
I'm afraid that merely alters the transliteration system once again, not the representation of the type and length of the Masoretic vowel, as in most of the examples. The other person/Andrew/head programmer? said essentially they would "put that example of kAl- on their list" to correct. Until they get around to it, 100s of uses of that specific construct noun in the Hebrew Bible will continue with that mistransliteration (and it is replicated in multiple places throughout Logos, not merely in the interlinear. Maybe if I had started with the context menu as a location for such mistransliterations - as Justin's Q pointed out - some discussion about interlinears would have been unnecessary.)
However, the other exx. I listed are just a few samples of other types of mistransliterations that should be addressed, in my opinion. [Everyone is free to disagree and defend those transliterations, of course. But since they are transliterations of Hebrew, I would expect someone might be able to give a defense based on principles of Biblical Hebrew. And no one has jumped in to do so.] I have indicated some specific constructions, using tc and in a few specific exx,, another resource or two to support the need for a different transliteration.
I don't have the time or inclination to track and report tens of thousands of mistransliterations.
I'll settle for warning Logos users.
TY again for bearing with me.
0 -
Don Parker said:
The other person/Andrew/head programmer?
No, Bradley is the head of programming. Andrew is one of the staff that handles new releases, testing, maintaining the bug management system, spends some time in the forums . . . I'm not sure of how duties are divided. Andrew, Philana, and Savanna are all somewhere in that area.
Don Parker said:I don't have the time or inclination to track and report tens of thousands of mistransliterations.
No one has suggested that. I have pointed you to the rules. You need to offer a few examples that clearly violate those rules -- not your mental rules. If you can do so, Faithlife will accept it as an error/ug report.
Don Parker said:I'll settle for warning Logos users.
I doubt that these threads will do that - users have to actually seek the threads out..
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Well, this is the place Logos reps directed me to post.
Might you suggest a better place to warn Logos users?
0 -
Don Parker said:
Might you suggest a better place to warn Logos users?
No. Nor do I think it appropriate to warn them before proving that there is a systematic violation of the SBL rules on the textual units being transliterated. Do that work and you win a hero button and serious consideration from the people in FL who help set priorities.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
All I see are 4 different SBL formatting styles
Might you direct me specifically to the "SBL rules on the textual units being transliterated." I do not find even in Google.
Then I will do a couple other exx. of constructions that occur hundreds of times in the Hebrew Bible. Sound fair?
0 -
MJ. Smith said:
Andrew is one of the staff that handles new releases, testing, maintaining the bug management system, spends some time in the forums . .
I'm one of the senior developers on the Logos Desktop team.
Don Parker said:said essentially they would "put that example of kAl- on their list" to correct.
There is no intention of specifically fixing kAI- or any other specific examples.
It is very likely that all of the problems you are seeing are purely due to the fact that we transliterate in isolation just the portion of the word that is represented in the interlinear column.
If a fix was to be made, it would be to transliterate that portion of the word in the context of the entire word.
Transliterating the entire word works correctly, as has been demonstrated in the Text Converter tool. If you find a complete word that isn't transliterated correct in the Text Converter tool, then that would be a separate bug that we would want to consider addressing.
There is nothing wrong with trying to raise the visibility of a bug in order to build support for getting the problem fixed. When prioritizing which bugs to work on, Faithlife often takes into account how many people are affected by a bug.
Andrew Batishko | Logos software developer
0 -
Wow.
I misread your post from Mar 23, 7.36:
"will write up a case for this problem, but I don't have any idea when or if this will be fixed."
I did try to present a bug and build support to get it fixed. Didn't work. OK, then. It's been real.
0 -
"Transliterating the entire word works correctly, as has been demonstrated in the Text Converter tool. If you find a complete word that isn't transliterated correct in the Text Converter tool, then that would be a separate bug that we would want to consider addressing."
I was getting tired of beating my head against a wall.
But here are a four from my exx. in original post that meet your criteria. So I may anticipate they will indeed be addressed?
N + pro suff causing reduction
1 Sm 22:17 incorrectly as long A in both, and **even for full form in tc* eth-aznO [correctlly as short o in TLOT: glh )oznô “to uncover his ear” (1 Sam 20:2, 12f.; 22:8[bis], 17) [same short o if Q is read with pro suf -iy instead, as leb and esv]
Ernst Jenni and Claus Westermann, Theological Lexicon of the Old Testament (Peabody, MA: Hendrickson Publishers, 1997), 71.
waw consec + impf/preterite with post-positive conj. accent telisha qetana [Tp stress accent in wrong place]
1 Sam 23.13 incorrectly in both; also full form in tc wayyaqam
[ct. correctly closely after in 23:16 correct in both, and as full form in tc when no post-positive accent: wayyaqom!]
directive [locative] suffix -Ah (Lambdin #58 w/many exx; Gkc #90i on the long A; sometimes incorrectly as -oh in both Mss Transl and in tc; other times correctly in both.
1 Sam 15.12 hakkarmeloh incorrectly in both
similarly -Ah ending on adverb
1 Sm 19.17 short o in both, incorrectly, kakhoh; yet correctly in Word study, 3970 DBL ka·ka(h, and even some places in the context menu0 -
Don Parker said:
So I may anticipate they will indeed be addressed?
I have no way of knowing when or if any specific case will be addressed. As I indicated above, I can write up a new case for these issues.
I had a hard time following your description of the problem, since I have no skill with Hebrew. Please check my description below and confirm if this sounds correct:
The Text Converter tool incorrectly transliterates the following Hebrew text (Hebrew taken from the Lexham Hebrew Bible, the transliterations are using the SBL General style):
-
1 Sa 22:17 - אֶת־אָזְנוֹ should be eth-ozno instead of eth-azno
-
1 Sa 23:13 - וַיָּקָם֩ should be wayyaqom instead of wayyaqam
-
1 Sa 15:12 - הַכַּרְמֶ֙לָה֙ should be hakkarmelah (?) instead of hakkarmeloh
-
1 Sa 19:17 - כָּ֙כָה֙ should be kakah instead of kakoh
Andrew Batishko | Logos software developer
0 -
-
{Excuse if this is a duplicate post. Thought I posted, don't see it.}
Exactly!!
0 -
Thank you. I've created a case for these problems.
Andrew Batishko | Logos software developer
0 -
Don Parker said:
Might you direct me specifically to the "SBL rules on the textual units being transliterated." I do not find even in Google.
The link is in About in the Text Converter Tool but Andrew has done a good job of assisting you. It may have taken a while but you have achieved your goal - a bug report. Congratulations - and thanks for sticking to it.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Thank you for putting up with me.Actually a bug report wasn't my goal.I grew up in S. FL where there was a saying, "If you see one cockroach/bug, there are hundreds more."
I just hope someone, somewhere (of course, only after confirmation of any or all of these proposed 4 specific mistransliterations), will ask, How did that happen? Does it happen elsewhere? What do we need to do so as not to generate error(s) in similar constructions throughout Logos? And are there even many more situations, albeit in different constructions, where some computer transliteration rule/s are generating many other mistransliterations? Do we need to consult with scholars or even just those who have a good grasp of Biblical Hebrew?
Maybe even get a focus group of Logos users and pastors with no BH training together with Hebraists, ask them (obviously for selected potentially questionable transliterations) what they understand the MSS transliteration (for word-slices) - as found, say, in the context menu - to mean in Logos and see whether any Hebraists balk and suggest a different transliteration.
No discussion needed really, until the 4 individual "bugs" have been inspected.
0
