4. Accurate and Consistent
CODE PROTECTED GRAMMAR AND SEMANTICS
The TENT is not just "a new translation" of the New Testament for two main reasons. It can not be compared with any previous English translations because, if you will forgive the anthropomorphisms, it does some things no other English translation does, by doing all of the following ten things:
It takes each word-form in the NT Greek and checks its meaning in every NT context where it occurs.
It checks the known etymology of each word.
It translates using only current English usage and avoids archaic, obsolete or Christian-only vocabulary.
It adheres to the Greek grammar but also is put into English word order.
It maintains semantic connections between verses.
It elevates punctuation above versification.
It makes readable many Greek words left untranslated in English translations.
It adds in brackets or italics only the most basic words needed for English readability , namely a, an, it, is/are/be.
It restores the contextually understood name of YHWH.
It solves many textual problems well known to NT scholars.
The second reason the TENT is not just a new translation of the NT is because it is a new translation of the NT Greek itself and required a new Greek-English lexicon. To create a new lexicon the following questions had to be answered for each Strong's numbered Greek word:
1) Do variations in a Greek word's grammar signal for other meanings consistently -- middle, passive, intensive, substantival infinitives, paraphrastics, plurals, etc.? If not, what rule applies that makes it different? For example, is it a non-initial imperative? Is it an aorist subjunctive after a conditional conjunction?
2) Are all NT grammatical variations of a Greek word translated consistently into English?
3) What is the etymology of the word?
4) Have the constituent words of a compound been translated accurately and consistently?
5) Does a new meaning contradict any traditional theology? Validly or invalidly?
6) Does a word's meaning show up consistently in a Strong's number search of the lexicon and source text?
7) If one meaning does not fit every occurrence of a grammatical variation, does the meaning need refining or more research?
To ensure that English word translations were accurately reflecting the Greek grammar, a rigorous means of protecting that grammar information in the translation process was vital. What could be better than digital coding?
Warning!
If grammar makes you crazy, you might want to skip this page of explanation, or just read the summary paragraph under the heading: Coding Steps (about a fifth of the way up from the page bottom). Or maybe even just read here about The 2nd Miracle: the Codes or click the "2nd Miracle" button at the page bottom.
Before inserting any N2LR-translated verbs into a source text, it was necessary to have a straightforward process that would accurately and consistently import the correct verb form along with its available grammar information. This required digitally coding every verb's variation of person, number, case, tense, voice and mood/mode into all its specific occurrences in the text.
I first made a cheat-sheet* listing all the combinations of tense/voice/mood –called TVM numbering. Each TVM number from 5627 to 5773 (as a continuum after Strong's numbers) described one of these combinations, e.g. NT5712 indicated imperfect tense / passive voice / indicative mood. Under that same number and description on the cheatsheet, I wrote an example translation using the verb "to ring", an English verb with distinct forms for present (ring(s), past (rang) and perfect (rung) tenses. The example translation of ‘to-ring’ with a TVM number of 5712 (imperfect / passive / indicative) would then be, "was/were-being-rung."
Cheat sheets are very handy, but doing each verb individually was so time consuming, that I became frustrated. There had to be a better way - one at a time was way too slow. And so I went to the Lord about it. The outcome of that is the Second Miracle that came as an answer to prayer, the First Miracle having been the answer to setting out a fleece.
All of the Greek verb TVMs had to be "decoded" (translated) using standard English forms. To easily translate Greek verbs into one of the possible forms of English verbs, I gave each TVM number its own short code in two parts. For the first part of the code which would be inserted directly before the TVM number in the text, the code was one of the following:
verb form: ring(s) code: - i
verb form: rang code: - a
verb form: rung code: - u
verb form: ringing code: - ing
verb form: ring (imperative) code: - [!]
For the second part of the code, I made a separate list for the forms of possible auxiliary words which are commonly used before, and as part of, English verbs. Some combinations are left out if not found in the NT (e.g. imperfect/middle deponent/indicative, or future/passive/perfect) or that are found in the NT only as periphrastic forms (substitute Greek pronounceable forms).
For accurately translating the NT, I needed English auxiliary word combinations such as those below:
am/is/are-
be-
being-
has/have-
has/have-been-
having-
having-been-
was/were-
was/were-being-
had-
had-been-
will-
will-be-
to-
to-be-
to-have-
to-have-been-
might-
might-be-
might-have-
might-have-been-
This second part of the code would then be inserted wherever any TVM number which uses it occurs. I put it right after the Strong's number, and before the N2LR code for the verb form.
I also needed consistent auxiliary forms in English for:
perfect/aorist-subjunctives: would/could/should-
interrogatives: do/does/did-
optatives: may-(aorist tense) / might-(present tense)
imperatives: English initial position command form, e.g. "Do it!", or an auxiliary “must.”
After processing all verbs using global search and replace, each verb would be described by this N2LR coding which combined these two English parts for the verb's auxiliary words plus the verb form itself , as in "was/were-being -u." Each source text verb would then have after it the format: Strong’s number – N2LR coding – TVM number. I suppose one might call it an “enhanced parsing,” since, in addition to the number codes, I would now have an immediate knowledge of how those numbers should be translated.
Once all the words were translated in a verse, I could see right away if something was 'off.' After a little investigation and refining or correcting, the problem would be resolved, and I may even have learned a new idiom, a new grammar construction, or a new meaning along the way. This self-correcting process ensured the most accurate and consistent of translations. Even replacing words digitally can be labor intensive, as I soon discovered, but the results were more than encouraging.
Scripture says that Scripture interprets Scripture, and this I found was supremely true of this method which relies on context, accuracy and consistency. It was self-correcting, though at times my flesh wished it wasn't. I can't tell how many times it pointed out my wrong assumptions or perceptions. I found myself, though not enslaved to any one textual tradition, seeming to have become a slave to context, accuracy and consistency. Yet the payback in terms of things which came to light made it not only well worth while, but imperative. Having God's message to us accurately should be fundamentally important for translators.
Two examples of digital coding
Below are two verse references chosen at random from a NT text I downloaded many years ago, a text having both Strong's and TVM numbers. I needed a NT text with both kinds of numbers as a template for The Transparent Experimental New Testament [The TENT] Source Text. It didn't matter what text it was, since I would be replacing it all anyway - even some of the numberings, doing all the checks as I went. It was simply a rough format to save as much typing as possible, such as not having to type a million pairs of angle brackets.
I would be using this downloaded template as the basis for the “experimental” part of that title, to test all the newly translated words I was collecting through the N2LR method for translating large texts. Only after I had translated all the words used in the NT, and made sure they fit into a semantically and grammatically numbered text, could I confidently give this method a name, the Natural Language Learning Redundancy method – N2LR.
Following each of the two verse references below, from left to right are:
1. the original template translation of one of its subject-verb combinations
2. the verb’s Strong's number in brackets
3. the N2LR verb code for the TVM number, single-underlined
4. the TVM number in brackets, single-underlined
verse reference 1. template verb translated as 2. Strong’s # 3. N2LR v. code 4. TVM #
2 Thess. 3:2 we might be delivered <4506> might-be u <5686>
2 Thess. 2:15 you(pl) have been taught <1321> *was/were u <5681> *[see anomaly below]
In The TENT Source Text, I made the above numbered formats the standard for all verbs, that is: template translation / Strong’s# / N2LR verb code / TVM#. Although, after doing this I could roughly translate any particular verse, the accuracy provided by redundancy would be to some degree less than optimum without the next step in the N2LR method.
After these codes were inserted for every Strong's numbered verb in the NT, the next step was to translate all of the NT Strong's numbered words, whether verb, noun or whatever, of most occurrence in order, down to the ones of least occurrence, according to the N2LR method. Then, the template translation could be updated to the final N2LR-derived TENT translation using a digital ‘search and replace’ procedure described in detail below.
Of course if some translator down the ages from the original Greek had changed a verb form for some reason, this would be reflected in a discrepancy between the template's words and the N2LR verb code as in *2 Thess. 2:15 above, where the template translation uses the perfect passive, even though the N2LR verb code indicates the Greek as an aorist passive. During the N2LR coding-dependent 'search-and-replace' steps listed under "Coding Steps " below, all such discrepancies were automatically corrected to reflect the most probable original Greek.
Another consideration was the interaction between words, that is, the grammatical influences that could change many of the verb forms, such as in 2 Th 3:2 above. To translate accurately one must consider every possible kind of redundancy available. This means looking at patterns of word influences on every level, beginning with grammatical context, then sentence/passage context, and finally document/cultural context.
For example, it soon became apparent during the N2LR translation process, that a perfect subjunctive verb or an aorist subjunctive verb, as in 2 Th 3:2, is changed from a "might" to a "would/could" form if it comes in a conditional phrase, often after the Greek word ἵνα 'hina' <2443> which is translated "(so-)that" in English. Grammar induced changes such as this needed to be discovered, given a standard description in The TENT Lexicon, and noted (see orange - three paragraphs down) wherever found in The TENT Source Text, as in 2 Th 3:2 in the "2016 - TENT Source Text."
Toward the end of translating all the words for The TENT Lexicon, I was able to quickly translate whole verses and passages. Continuing with the example of 2 Thess. 3:2, note the changes below between the earliest archived version of The TENT Source Text from 2008, and the one from 2016.
2008 - earliest archived TENT Source Text:
2Thess. 3:2 - And/also/then [fig.: (2x) both...and] <2532> so-that <2443> we might be delivered <4506> might-be u <5686> from <575> unreasonable <824> and/also/then [fig.: (2x) both...and] <2532> bad(-ness/) <4190> men/persons/humans(-beings) [fig.: people] <444> * for/as <1063> all/one/every/any(-one/thing) <3956> [men] have not/no <3756> trust/belief <4102>
2016 - TENT Source Text (after eight more years of translating and formatting - finished words are bolded in light green below):
2Thess. 3:2 - and /also/too [fig.: now/then/even / (2x) both...and / neither…nor(w neg.) / idiom#1: w psub/asub v. = then(-could/would(-have/be)-v.)] <2532> so/(so-)that [idiom: psub/asub v. aux. = “would/could”] <2443> we- might/ would /could-be-pulled(-away ) [fig.: rescued] <4506> (pass.) (p/a-sub) <5686> of/from /off/away/out/fully/by <575> the-ones <3588> out-of-place [fig.: having-(some-)thing-wrong ] <824> and /also/too [fig.: now/then/even / (2x) both...and / neither…nor(w neg.) / idiom#1: w psub/asub v. = then(-could/would(-have/be)-v.)] <2532> (geni.) of/from /for -harmful (-one/thing) [fig.: bad(-one/thing)] <4190> men / humans / human-beings [fig.: persons / (pl.) people ] <444> , because /for/since/as [idiom#1: (before contrasted statement) "although"] <1063> the <3588> trust(ing ) [fig.: belief] <4102> [of /from/for/through/in-the-sight-of/by/with/because-of/concerning x -YHWH (implied by def. art.) link] [is] no/not [idiom#1: not? - not? (neg. question with another neg.) = "no? (do/be-)not? / idiom#2: (Eng. style) NT3756 + pres. or aorist v. = do(es)/did-not + pres. v.] <3756> of/from/for/by- all (s./pl.) / (s. w/ artic.?) any/each/every(-one ) /thing) / (neut. pl.) all-things / (m. pl.) all-people [fig.: everybody] <3956> .
Note: the words necessarily bolded in the above quote (due to this website's formatting limitations) are instead double-underlined or italicized single underlined in The TENT Source Text itself.
[*note to self - change texts in green: in final website pg use double underlines and strike-thru instead of green/bold/single underlines then delete this note]
I then timed how many minutes it took to manually extract (by actual typing, not by formatted auto-extraction) a study version verse from the above verse in the 2016 TENT Source Text, and then how many minutes it took to manually extract a devotional version verse from the study version verse:
[note to self: for the coding of the three versions see: zz Formatting code test for three versions.docx]
2016 - This TENT Study Version verse below was manually extracted from The TENT Source Text above in five minutes:
2Thess. 3:2 - and <2532> so-that [idiom: psub/asub v. aux. = “would/could”] <2443> we-would-be-pulled-away <4506> <5686> from <575> the-ones <3588> having-some-thing-wrong <824> and <2532> (geni.) from-harmful <4190> people <444>, because <1063> the <3588> trusting <4102> [of-YHWH (implied by def. art.)] [is] not <3756> by-everyone <3956>.
2016 - The TENT Devotional Version verse was manually extracted from The TENT Study Version in three minutes:
2Thess. 3:2 - and so that we would be pulled away from the ones having something wrong, and from harmful people, because the trusting [of YHWH] [is] not by everyone.
Of course, if I had to spend the amount of time for each verse as I did for the versions above, it would not be possible to translate the whole NT into three versions by the N2LR method, at least not for me at this stage of my life. But what the digital coding also allows in the way of time-saving, is that one can extract only the needed amount of information from version to version by formatted auto-extraction, that is, if it has been encoded/formatted in the right way. Both automatic version extractions would take less than an hour, although some clean up would add to that considerably.
Creating a study version and a devotional version was a matter of pre-planning and the use of text formatting elements in The TENT Source Text according to what kind of information was wanted for the study and devotional versions. You can see this version formatting at work most easily in the different uses for underlining, italics, spaces and hyphens between the 2008 and 2016 TENT Source Text verses above.
Coding Steps
In summary of the above for the non-grammatically minded:
In order to arrive at the latest TENT Source Text, and the first task before beginning to digitally translate the NT was, for each TVM number, to insert the N2LR verb code in between the Strong's and TVM numbers for all verb forms in The TENT Source Text. To easily do this, I used a cheat sheet matching all TVM numbers with their N2LR verb codes (*see "2nd Miracle: The Codes" button at bottom).
Then it was simple to insert the N2LR codes in front of their TVM numbers with digital search and replace operations, one for each different TVM number. One can only see this in the earliest 2008 version of The TENT Source Text in the previous section, "Two Examples of Digital Coding", in the first example of 2Thess 3:2, where it has “might be delivered <4506> might-be u <5686>.”
The second task in translation was to insert each Strong’s numbered word, newly translated by the N2LR method (greatest to least occurrence and adequately fitting all its NT contexts), into The TENT Source Text to replace the original template translation. Since those meanings had to be inserted with grammatical accuracy and consistency according to their contexts, a final digital coding procedure was applied as follows (some people might find this process interesting or you can skip to the last pararaph):
Step 1
Going to the beginning of the Word document (CTRL+Home) and clicking the search icon/button, in the "Find" box I typed in the Strong's number of the N2LR translated word I wanted to insert in all its occurrences in the text. The number must appear as it does in the text, in brackets, e.g. "<4506>". Then I copied it into the “Replace with” box.
Step 2
In the “Replace with” box, I added an “x” after the Strong’s number, e.g. "<4506>" became "<4506>x". Then I hit the "Replace all" button. [As an aside, I could see the number of occurrences for that Strong’s number in the ‘find and replace’ dialog box, which noted the number of replacements just made. However that held only for non-verbs, since verbs usually required many search and replace operations. At this point it is only necessary to say that inserting translated words other than verbs and nouns, such as conjunctions or prepositions, usually required only a single ‘find/replace’ operation. Nouns generally required two, one for singular and one for plural forms. The verbs required the more extensive procedure which is described below, and only after all the N2LR verb codings had been inserted in The TENT Source Text.]
Step 3
Again searching from the beginning of the document (CTRL+Home) for the first occurrence of "<4506>x", and if the word was a verb, I copied from the text that occurrence's TVM number and pasted it one space behind the "x" which was after the Strong's number in the "Find" box. Then I copied whatever word or words the template text used in that occurrence to translate it, pasting the word/words into the "Find" box in front of the Strong's number. In the "Replace with" box, I substituted for the template translation the new TENT Lexicon translation but in the form of that occurrence's N2LR verb code, keeping the Strong's number but removing the "x" after it, and also keeping the TVM number, but deleting its N2LR code. Then I clicked "Replace All".
Step 4
I had to repeat step #3 as many times as I found a new occurrence of that Strong’s number having an "x" until there were no more. When no more "x" numbers were found with that Strong's number, the translation for that Strong's number was finished. With a frequently occurring verb, the search and replace operations could number in the dozens. One such verb could take an hour or two to insert accurately into The TENT Source Text. That may not seem like quick progress, but it often meant that I had just automatically, yet accurately and consistently, translated hundreds and hundreds of words in the NT, an encouragement especially in the beginning with the words of most occurrences. For me, it was well worth the effort.
Step 5
When I had finished a Strong’s number and wanted to go on to another, I had to begin the process again with Step #1 (not step #3!)
As a side note, whenever an N2LR code was deleted along with the “x” after the Strong’s number of a verb, it left behind the Strong’s number followed by the TVM number. At some point however, I found it helpful to insert in that space between those numbers, certain special verb labels such as middle, deponent, passive, or aorist-subjunctive (p/a-sub). One can see an instance of this in the 2016 version of The TENT Source Text above, where 2Thess. 3:2 has, “we- might would /could-be-pulled(-away ) [fig.: rescued] <4506> (pass.) (p/a-sub) <5686> .” The orange coded words signal that not only is the verb in the passive voice, but that it is an aorist subjunctive verb possibly under the influence of a ‘trigger word’, usually hina (NT2443) or a conditional particle occurring somewhere before it, so that the verb should then be understood as a “would/could” form.
*change texts in green: in final website pg use double underlines and strike-thru instead of green/bold/single underlines then delete this note