For Neto, et al: Bibles

[MaxPC]

This Bible translation comparison chart raised a question for me.
Which type of translation do you feel is optimal?

[Neto]

Was there a link to the chart you saw?

Word-for-Word
Thought-for-Thought?
Somewhere in between?

There is a lot of misunderstanding about these different terms, and as a result, a lot of misuse of them as well.

Generally, people will say that their favorite translation is described by one of these, and then they will go on to explain why the others are lacking, and maybe even prone to errors.

A story circulated in Wycliffe Bible Translators circles about a well-meaning and enthusiastic person who wrote in to the mission head-quarters, offering to help with Bible Translation. They wrote, "Please send the dictionary for a language where a translation is needed, and I will start working on it." (In Bible college, I toyed with the idea of putting up a sign, on a box labeled "Bibles to be translated", and then announce that I wanted to collect Bibles that people were no longer using, in order to send them in to be translated for Bibleless people groups. - But I didn't.)

Here is a quick example of a familiar passage, John 3:16-17, "translated word-for-word":
16 Thus for loved the God the world, so as the son the only-begotten [MASC SINGULAR] gave that all the believing into him not may-perish but may-have life eternal. 17 For not sent the God the son into the world in-order-that [MASC SINGULAR] judge the world but in-order-that might-be-save the world through him.

I actually already supplied additional information not in the text, because the Greek text was written without any meaning-based capitalization. I capitalized the word "God", which in English implies that it is either the Deity, or a personal name.

The word 'world' occurs 4 times in these two verses, but with at least two completely different meanings, unless one interprets it in all cases as "the created people", instead of seeing this verse as saying that God sent his son [down] to earth. (That is, "God sent his son into [the midst of] his created people..." Perhaps this is the real meaning in this verse, not a reference to the physical earth or world at all.)

In WBT, we aimed at what is called "dynamic equivalence".

As an example (I didn't get as far as translating the Gospel of John, so I'll pick another passage), Luke 16:31, where Jesus is giving the "punch line" to the Rich Man & Lazarus story, here as literally word-for-word as I can translate the Banawa into English:

Abraham's voice started up again, "God's thing marking only is good.
(Literally, for the meaning 'good', the Banawa is "not bad", but I didn't render it that way because that means something quite different in English, and I didn't want to go that far with a word-for-word translation. But that is their way of saying 'good', and I suppose one could gather something of their perspective on life from this, sort of tilted to the negative.)
Continuing:
A man died, awoke again be if, that man's voice they like do also not would be certainly. Lazarus arrive be if, they laugh [PASSIVE] he at he repeatedly be if certainly.
(This is a Passive construction, which is used in Banawa to maintain the discourse topic as the subject of every sentence in which the topic is a participant. A free translation would read something like "If Lazarus were to come back, he would certainly be constantly laughed at by them.
Continuing:
Not going to go. Me to only at Lazarus live will," Abraham's voice did long ago.
"OK, Finished certainly." Jesus' voice did. Jesus' story be certainly. Story told long ago non-eye-witnessed, them taught long ago non-eye-witnessed.

Banawa does not have a verb meaning 'say', or 'talk'. The voice of a person is an inalienably possessed noun, the same as 'skin', 'face', and even droplets of water that are on a person's body, or a trail that a person uses constantly. These nouns can "stand in" in a sentence for the person him or her self. So when Jesus was hit in the face during the 'trial', the Banawa would say Face was hand-hit (by) them. Another difference in Banawa is that this specifically says that it was 3 or more men who hit him. Their word for third person plural is different for men and for women, and this pronoun doesn't even occur in the same place in a sentence. The masculine plural pronoun is a prefix, and the feminine plural is a suffix. And as I alluded to above, the 'plural' means 3 or more. They have a dual marker if the meaning is two.

Sorry, I really get carried away on things like this. So basically, if I have to choose one of the three options on that chart you saw, it would be 'Thought-for-thought'. But also bear in mind that a translation for a pre-literate people group is going to be different than one done in a language with a long-standing written tradition. Ever read the transcript from a spoken message? Oral styles involve a lot more repetition, because in a live address, the listener cannot roll back the tape to hear it again.

Obviously, Greek and English, both being in the Indo-European language family, are much more closely related than either is with a language like Banawa, but I used it to demonstrate why a word-for-word translation does not work in a situation like that.

[barnhart]

Thanks for the demonstration Neto. Translation is not simple or subject to easy formulas. In college I took Hebrew (modern) for the language requirement with hopes it might prove useful in reading or understanding the Bible. What I learned was it is far more difficult than that.

[MaxPC]

Neto, your input is valuable, and your thoughts regarding this query are both illuminating and edifying.

My apologies. Apparently the copy/paste did not happen. Here is the chart I referenced.
Chart

[Neto]

Neto, your input is valuable, and your thoughts regarding this query are both illuminating and edifying.

My apologies. Apparently the copy/paste did not happen. Here is the chart I referenced.
Chart

I may have seen this chart before, but not sure. No translation fits fully into one place on that spectrum in all passages.
One example: KJV vs NIV on the passages where the KJV says "God forbid", and the NIV has different forms of "May it never be".
Those texts (in the Greek) do not have the word 'God' in them, nor the word 'forbid'. The translators of the KJV chose to use an expression to communicate the idea of the Greek expression, which would literally be something like "Not never may it be". I don't know, but I tend to guess that the expression "God forbid", may well have been a common expression during that era of the English language development. A literal rendering from the Greek into English would mean the opposite of what Paul said, because of the deal with double negatives in English. In Greek, it places more emphasis on the statement; in English it reverses it. NIV has the literal rendering on this particular case, with "May it never be". Yet the NIV is, on the whole, not as "word-for-word" as is the KJV. All I would suggest (or ask) is that people be more fair when discussing different translations, avoiding careless generalizations or criticisms of translations one does not prefer. As someone said, The important thing is to read it, and to study it. Then obey it. I do prefer NIV over KJV, but my reasons are to do with adapting to changing language. 1611 was a long time ago, and we can probably all think of words that have changed meanings in just our (relatively) short lifetimes. And yes, it usually rubs me the wrong way, but it's a characteristic of a living language - only dead languages don't change.

[Neto]

Some more thoughts on this subject:
I later wondered if I was misrepresenting the so called "word-for-word" approach, because I took it to include word order as well as the words themselves. I also didn't remember if Young's Literal Translation was included on the chart, and looking again, I see that it is not. They would have probably placed it very close to the Interlinear. But neither the interlinear or Young's Literal are intended for reading - they are both study aids.

I also see now that the definition they give for "word-for-word" includes the idea of not only "wording", but also "structure". In that case, my translation of the Banawa text I used as an example is following their description of "thought-for-thought", as close as I can tell, without an example text from them. To be fair, they DO add the caveat of "without sacrificing clarity". But clarity is a matter of degrees, in much the same way as is the spread between grammatically literal vs paraphrase. At what point does one conclude that clarity has been sacrificed? At what reading level is the testing done? What percentage of misunderstanding is acceptable?

To the last question, the WBT Bible translation comprehension testing assumes that we want it to be as near zero as humanly possible. But there is also another consideration. Sometimes the text itself is vague. For the word-for-word approach (Ignoring the question of grammatical sentence structure - that is a questionable goal, and I'll come back to it in a bit.), comprehension is not as much of a concern, because the goal is to "stay true to the text, at nearly all cost". And, here is my point, while INTERPRETATION is a part of any translation work, the real interpretation dilemmas are ideally left for the reader to grapple with. (This was a "fault" of mine, as a translator - I was too reluctant to accept the commentaries' statement that it is not possible to determine the correct interpretation in every case. So I would spend many hours, scouring every commentary for an answer to those questions which no other scholar had been able to solve. It was foolish of me to think that I might be able to solve such questions, but what I was searching for was for convincing evidence from a real scholar, not really to solve it myself.) The authors of some commentaries, the ones I call "devotional commentaries", will nearly always present the different options in such a way that it is obvious which option the author thinks is correct, sometimes going to great lengths to 'defeat' the opposing views. A good exegetical commentary, however, will seek to present the pros and cons for each view without partiality. (Like my Theology II professor in Bible college - students sat around after class, trying to figure out if he was a millenialist or not, and if so, if he was Pre-, mid-, or post-trib. No one knew. And it was the same for other theological questions. His aim was not to teach us WHAT to think, but HOW to think.) So the objective for the Bible translator is to render the text in such as way that if they read it carefully, in depth, they will eventually come face to face with the same questions that the translator struggled with. We do not know how clearly the original readers/hearers understood the text, but Peter gives us some idea of an answer to that, when he states that "Some of Paul's writings are difficult to understand".

I almost forgot this, but I said above that I would come back to the question of literalness of grammatical form. So, in the Banawa example I gave, the grammatical form is very different. In the one case, I substituted a Passive construction for what is an Active construction in the Greek text. But there are levels of accuracy that go beyond the word level: sentence level, and discourse level, to name two. The discourse level is where the Banawa language dictates that a different sentence structure is imperative.

Generally speaking, English uses a Passive when there is a desire to leave the Actor un-named or even unknown. "The dog was badly beaten." The speaker doesn't know who did it, or may wish to leave it unsaid, but make a strong implication that it was a certain person, and the listener will understand, but no charge of a false accusation can be raised, because no perpetrator was named.

The Banawa Passive is used in an almost opposite way - the Actor is known from the context of the story, or is unimportant, because the Actor is not the discourse topic - it's not about them. But (in the English example sentence above) the TOPIC of the discourse is not the Actor, but the dog. So the dog is kept as the Subject of the sentence, and is not named, because the listeners already know what the story is about. If a sentence is included in the story where the discourse topic is not involved as either the Actor or the Undergoer, then an Active sentence structure is used.

There are also elements of discourse structure that indicate things like relative importance, or emphasis. In Banawa, this is akin to when a discourse topic is changing. They will repeat the new discourse topic, just as they did when they introduced the original discourse topic. This is usually done by repetition, as in "The dog, the dog came." It's not 'stuttering', or just repeating the phrase to give time to think what comes next in the story. (The Banawa have a way of preventing long pauses, or allowing interruptions by a listener, something I never completely mastered - ingressive speech. They never stop talking, they just keep talking while they are breathing in to catch another breath.)

Sorry to have go on so long in this post. I realize that the number of interested parties is not large, so thank you to everyone who took the time to read this.

[MaxPC]

Some more thoughts on this subject:
I later wondered if I was misrepresenting the so called "word-for-word" approach, because I took it to include word order as well as the words themselves. I also didn't remember if Young's Literal Translation was included on the chart, and looking again, I see that it is not. They would have probably placed it very close to the Interlinear. But neither the interlinear or Young's Literal are intended for reading - they are both study aids.

I also see now that the definition they give for "word-for-word" includes the idea of not only "wording", but also "structure". In that case, my translation of the Banawa text I used as an example is following their description of "thought-for-thought", as close as I can tell, without an example text from them. To be fair, they DO add the caveat of "without sacrificing clarity". But clarity is a matter of degrees, in much the same way as is the spread between grammatically literal vs paraphrase. At what point does one conclude that clarity has been sacrificed? At what reading level is the testing done? What percentage of misunderstanding is acceptable?

To the last question, the WBT Bible translation comprehension testing assumes that we want it to be as near zero as humanly possible. But there is also another consideration. Sometimes the text itself is vague. For the word-for-word approach (Ignoring the question of grammatical sentence structure - that is a questionable goal, and I'll come back to it in a bit.), comprehension is not as much of a concern, because the goal is to "stay true to the text, at nearly all cost". And, here is my point, while INTERPRETATION is a part of any translation work, the real interpretation dilemmas are ideally left for the reader to grapple with. (This was a "fault" of mine, as a translator - I was too reluctant to accept the commentaries' statement that it is not possible to determine the correct interpretation in every case. So I would spend many hours, scouring every commentary for an answer to those questions which no other scholar had been able to solve. It was foolish of me to think that I might be able to solve such questions, but what I was searching for was for convincing evidence from a real scholar, not really to solve it myself.) The authors of some commentaries, the ones I call "devotional commentaries", will nearly always present the different options in such a way that it is obvious which option the author thinks is correct, sometimes going to great lengths to 'defeat' the opposing views. A good exegetical commentary, however, will seek to present the pros and cons for each view without partiality. (Like my Theology II professor in Bible college - students sat around after class, trying to figure out if he was a millenialist or not, and if so, if he was Pre-, mid-, or post-trib. No one knew. And it was the same for other theological questions. His aim was not to teach us WHAT to think, but HOW to think.) So the objective for the Bible translator is to render the text in such as way that if they read it carefully, in depth, they will eventually come face to face with the same questions that the translator struggled with. We do not know how clearly the original readers/hearers understood the text, but Peter gives us some idea of an answer to that, when he states that "Some of Paul's writings are difficult to understand".

I almost forgot this, but I said above that I would come back to the question of literalness of grammatical form. So, in the Banawa example I gave, the grammatical form is very different. In the one case, I substituted a Passive construction for what is an Active construction in the Greek text. But there are levels of accuracy that go beyond the word level: sentence level, and discourse level, to name two. The discourse level is where the Banawa language dictates that a different sentence structure is imperative.

Generally speaking, English uses a Passive when there is a desire to leave the Actor un-named or even unknown. "The dog was badly beaten." The speaker doesn't know who did it, or may wish to leave it unsaid, but make a strong implication that it was a certain person, and the listener will understand, but no charge of a false accusation can be raised, because no perpetrator was named.

The Banawa Passive is used in an almost opposite way - the Actor is known from the context of the story, or is unimportant, because the Actor is not the discourse topic - it's not about them. But (in the English example sentence above) the TOPIC of the discourse is not the Actor, but the dog. So the dog is kept as the Subject of the sentence, and is not named, because the listeners already know what the story is about. If a sentence is included in the story where the discourse topic is not involved as either the Actor or the Undergoer, then an Active sentence structure is used.

There are also elements of discourse structure that indicate things like relative importance, or emphasis. In Banawa, this is akin to when a discourse topic is changing. They will repeat the new discourse topic, just as they did when they introduced the original discourse topic. This is usually done by repetition, as in "The dog, the dog came." It's not 'stuttering', or just repeating the phrase to give time to think what comes next in the story. (The Banawa have a way of preventing long pauses, or allowing interruptions by a listener, something I never completely mastered - ingressive speech. They never stop talking, they just keep talking while they are breathing in to catch another breath.)

Sorry to have go on so long in this post. I realize that the number of interested parties is not large, so thank you to everyone who took the time to read this.

Please, do not apologise. This is an highly interesting topic, this differential structuring among languages.

Re the bolded above, I am reminded of a similar discussion about the structure of some of the languages among Native American tribes, specifically the Lakota Sioux family of languages. I hope I am recalling that correctly. I can most certainly see your point regarding the necessity of thought-for-thought when the structures are so very different from the syntaxes classical Latin and Greek from which many translations are derived. This type of translation work is difficult and I can honestly say that it is both art and science in my opinion.

[Neto]

....
I almost forgot this, but I said above that I would come back to the question of literalness of grammatical form. So, in the Banawa example I gave, the grammatical form is very different. In the one case, I substituted a Passive construction for what is an Active construction in the Greek text. But there are levels of accuracy that go beyond the word level: sentence level, and discourse level, to name two. The discourse level is where the Banawa language dictates that a different sentence structure is imperative.

Generally speaking, English uses a Passive when there is a desire to leave the Actor un-named or even unknown. "The dog was badly beaten." The speaker doesn't know who did it, or may wish to leave it unsaid, but make a strong implication that it was a certain person, and the listener will understand, but no charge of a false accusation can be raised, because no perpetrator was named.

The Banawa Passive is used in an almost opposite way - the Actor is known from the context of the story, or is unimportant, because the Actor is not the discourse topic - it's not about them. But (in the English example sentence above) the TOPIC of the discourse is not the Actor, but the dog. So the dog is kept as the Subject of the sentence, and is not named, because the listeners already know what the story is about. If a sentence is included in the story where the discourse topic is not involved as either the Actor or the Undergoer, then an Active sentence structure is used.
....

....
Re the bolded above, I am reminded of a similar discussion about the structure of some of the languages among Native American tribes, specifically the Lakota Sioux family of languages. I hope I am recalling that correctly. I can most certainly see your point regarding the necessity of thought-for-thought when the structures are so very different from the syntaxes classical Latin and Greek from which many translations are derived. This type of translation work is difficult and I can honestly say that it is both art and science in my opinion.

The only American Indian languages I studied are Comanche and Cheyenne, and the first only very shortly. I didn't get deep enough in either of them to begin discourse level analysis.

One really interesting thing about Cheyenne is the pronoun affix structure. They have a cultural attitude of respect toward the second person that shapes the way pronouns are used. I cannot chart it all here off of the top of my head, but the 2nd person pronoun is ALWAYS given the place of prominence and respect by being placed in the prefix position. So if you say "I see you.", the order is 2ND_SINGULAR_OBJECT--'see'--1ST_SINGULAR_AGENT/ACTOR. "You see me" is 2ND_SINGULAR_AGENT/ACTOR--'see'--1ST_SINGULAR_OBJECT. Of the singular pronoun affixes, only the 2nd Singular has two Agent or Actor forms, both of which are prefixes. If I recall correctly, the Plural 2nd person is the same.

I said above that the culture shaped the language. However, this is an interesting area of study, and the answer is not as clear as I stated it. Not everyone would agree that language conforms to culture, but that is the direction of influence that makes the most sense to me. So, if someone has managed to reconstruct Proto-Plains Algonquian, it would be interesting to see if they found evidence of these structures far back. Or, even if this pronoun structure is evident in other Algonquian languages. (I looked on-line to see which language grouping they fit in. Then apparently Algonquian is classified at the next level back as a part of the Algic language family.)

Banawa is currently classified as a member of the Arawan language family, which remains unclassified at the next level up. When we started there, it was tentatively classified as an Arawakan language, I think based on a very old short word list from a 19th century explorer, who is probably the one who gave the name to this language family. No one knows what the name was based on. In contrast, Greek, all of the Romance languages (Latin derived), all Teutonic languages (which if I recall correctly includes English); these are all classified as Indo-European. I think that this influences the attitudes that people have toward 'literal' translation - it is simply easier to go from Greek to English than to a language that is possibly related only back at the Tower of Babel. (And God may have so thoroughly confused language at that time that there is actually no relationship at all between such languages. An interesting thing about that, however - The Banawa word for father is 'abi'. Seeing the similarity to Hebrew, I told them that in Jesus' language it was 'aba'. They made a linguistic deduction that 'mother' in Hebrew must then be 'ama', because in Banawa it is 'ami'.)

[MaxPC]

....
I almost forgot this, but I said above that I would come back to the question of literalness of grammatical form. So, in the Banawa example I gave, the grammatical form is very different. In the one case, I substituted a Passive construction for what is an Active construction in the Greek text. But there are levels of accuracy that go beyond the word level: sentence level, and discourse level, to name two. The discourse level is where the Banawa language dictates that a different sentence structure is imperative.

Generally speaking, English uses a Passive when there is a desire to leave the Actor un-named or even unknown. "The dog was badly beaten." The speaker doesn't know who did it, or may wish to leave it unsaid, but make a strong implication that it was a certain person, and the listener will understand, but no charge of a false accusation can be raised, because no perpetrator was named.

The Banawa Passive is used in an almost opposite way - the Actor is known from the context of the story, or is unimportant, because the Actor is not the discourse topic - it's not about them. But (in the English example sentence above) the TOPIC of the discourse is not the Actor, but the dog. So the dog is kept as the Subject of the sentence, and is not named, because the listeners already know what the story is about. If a sentence is included in the story where the discourse topic is not involved as either the Actor or the Undergoer, then an Active sentence structure is used.
....

....
Re the bolded above, I am reminded of a similar discussion about the structure of some of the languages among Native American tribes, specifically the Lakota Sioux family of languages. I hope I am recalling that correctly. I can most certainly see your point regarding the necessity of thought-for-thought when the structures are so very different from the syntaxes classical Latin and Greek from which many translations are derived. This type of translation work is difficult and I can honestly say that it is both art and science in my opinion.

The only American Indian languages I studied are Comanche and Cheyenne, and the first only very shortly. I didn't get deep enough in either of them to begin discourse level analysis.

One really interesting thing about Cheyenne is the pronoun affix structure. They have a cultural attitude of respect toward the second person that shapes the way pronouns are used. I cannot chart it all here off of the top of my head, but the 2nd person pronoun is ALWAYS given the place of prominence and respect by being placed in the prefix position. So if you say "I see you.", the order is 2ND_SINGULAR_OBJECT--'see'--1ST_SINGULAR_AGENT/ACTOR. "You see me" is 2ND_SINGULAR_AGENT/ACTOR--'see'--1ST_SINGULAR_OBJECT. Of the singular pronoun affixes, only the 2nd Singular has two Agent or Actor forms, both of which are prefixes. If I recall correctly, the Plural 2nd person is the same.

I said above that the culture shaped the language. However, this is an interesting area of study, and the answer is not as clear as I stated it. Not everyone would agree that language conforms to culture, but that is the direction of influence that makes the most sense to me. So, if someone has managed to reconstruct Proto-Plains Algonquian, it would be interesting to see if they found evidence of these structures far back. Or, even if this pronoun structure is evident in other Algonquian languages. (I looked on-line to see which language grouping they fit in. Then apparently Algonquian is classified at the next level back as a part of the Algic language family.)

Banawa is currently classified as a member of the Arawan language family, which remains unclassified at the next level up. When we started there, it was tentatively classified as an Arawakan language, I think based on a very old short word list from a 19th century explorer, who is probably the one who gave the name to this language family. No one knows what the name was based on. In contrast, Greek, all of the Romance languages (Latin derived), all Teutonic languages (which if I recall correctly includes English); these are all classified as Indo-European. I think that this influences the attitudes that people have toward 'literal' translation - it is simply easier to go from Greek to English than to a language that is possibly related only back at the Tower of Babel. (And God may have so thoroughly confused language at that time that there is actually no relationship at all between such languages. An interesting thing about that, however - The Banawa word for father is 'abi'. Seeing the similarity to Hebrew, I told them that in Jesus' language it was 'aba'. They made a linguistic deduction that 'mother' in Hebrew must then be 'ama', because in Banawa it is 'ami'.)

I can certainly see this line of thinking as being the best explanation yet. I suppose that linguistics is a science that is still evolving as well. I remember an article from years past about how the words for mother and father are quite similar in all languages. They hypothesised that the words are based on the natural sounds from infants. It might be. Or it might be something else.