Biblical Hebrew Forum

[mjoseph <mjoseph AT terminal.cz>]

I've been following this thread with great interest, though I haven't 
been posting as the number of messages has been so great. I would like to 
summarize some of the conclusions I've reached while musing over what has 
been written from both sides.

I do not believe that there is such a thing as a "literal translation," 
or to be more precise, the only literal translation is an interlinear. 
When Jehovah's Witnesses state that the NWT is the most literal version, 
I might agree only in the restricted sense that it may possibly be true 
that it most nearly approaches being an interlinear. I, however, would 
consider that to be nothing more than a sign of a particularly poor 
translation (much as the NASB, done by evangelical Christians, is a poor 
translation), *even without* considerations of theological bias, as I 
hope to show.

In short, I believe these "literal" translations to be poor because they 
betray a lack of understanding of the nature of different languages. The 
wider point (details follow) is that, with a few exceptions such as 
concrete nouns (e.g., hippopotamus, telephone), the meaning of terms in 
one language does not completely overlap with the meaning of terms in 
another language, the quantitative difference being lesser (between two 
languages in the same family) or greater (between two languages in 
different families). The attempt, therefore, always to translate a given 
term in the original by the same word in the receptor language, rather 
than giving the reader a chance to see how the word is used in the 
original language, actually takes him *away* from the meaning, not only 
of the text as a whole, but even from the meaning of the words 
themselves. The promised details (some of which have already been 
mentioned by others; all of these arguments against "literal" translation 
can and should be taken as arguments in favor of a dynamic equivalence 
translation):

1) Cases in which the "literal" translation makes no sense.
  (a) Idioms are the most obvious case. When I was in Haiti the head of 
the Bible school, a competent linguist and one of the translators of the 
Bible into Creole, used to regale us by translating Creole proverbs 
word-for-word into English.
  (b) Back near the beginning of this thread, I wrote: "the NWT has 'slow 
to anger' in Ex. 34:6, 'sparks' in Job 5:7 (for B.:N"Y-RE$EP) and 'houses 
of the soul', in quotation marks, in Is. 3:20 (for B.FT"Y-HAN.EPE$). Dr. 
Furuli's phrase 'a strictly literal translation (though with some 
idiomatic renderings)' is self-contradictory. A 'strictly literal' 
translation would translate all three of these (as well as other) idioms 
literally; one done according to to Nida's principles would give the 
English equivalent for each, based on research, etymology, comparative 
philology, historical factors governing and/or affecting interpretation, 
etc." I see no reason to revise that statement now. Yet it should be 
clear that the "literal" translations "long of nose," "sons of Resheph," 
and "houses of the soul" make no sense in English. Not too long ago, 
though, in a first-year Hebrew course, I had two JW students insist that 
Is. 3:20 should be translated literally "houses of the soul," so as to 
let the reader decide what it means. Not a peep, though, about the 
idiomatic renderings in the NWT of Ex. 34:6 and Job 5:7, for reasons that 
I hope to make evident near the end of this post.
  (c) The first idiom every Hebrew student learns is KFRAT B.:RIYT which 
I hardly have to explain would be "literally" translated "to cut a 
covenant." In English, though, we say "to make a covenant." Looks fine, 
and is fine, but the point is, this is *not* a "literal" translation! It 
is strictly idiomatic, and properly so. The NWT has "to conclude a 
covenant." Sounds a little stiff to me, but still, it's an acceptable 
translation, which does convey the proper idea. Again, this is *not* a 
literal translation (neither should it be; my point, though, is that the 
JWs are always priding themselves on how literal their version is). To 
take it a step further, the Czech expression is "uc^init smlouvu," "to do 
a covenant" (in the 1613 version, similar to our KJV) or "uzavr^it 
smlouvu," "to close a covenant" (in the 1985 version, similar to our 
NEB). But wait--the LXX uses DIATIQEMI in the middle voice. OK, Arndt and 
Gingrich give "decree, ordain; assign, confer; make a will," to which 
Liddell and Scott add (for the middle only), "to set out for sale, 
dispose of; to settle mutually; to make a covenant; to make an agreement 
with, promise." So, what is the "literal" translation? To cut a covenant, 
to make a covenant, to do a covenant, to close a covenant, to decree a 
covenant, to ordain a covenant, to assign a covenant, to confer a 
covenant, to dispose of a covenant, to settle mutually a covenant, or to 
promise a covenant? By the way, I have no idea what verb is used in 
Norwegian, Chinese, or Hungarian.

2) Multiple meanings of a word/multiple words for one meaning.
  (a) In the original language.
    (a1) I once mentioned the example of NFTAN, which can mean either 
"give" or "permit" in Hebrew. A garden variety example, but one which 
puts the "literalists" in a quandry--either always to translate by the 
word "give" (and make hash of the sense), or use both words (properly, in 
my opinion--that is what context and linguistics are for, after all), and 
give up pretensions to making a "literal" translation. These examples, of 
course, could be multiplied indefinitely. What, exactly, is the "literal" 
translation of the preposition lamed from Hebrew into English?
    (a2) Polysemy. Some Czech prefixes have two distinct meanings, such 
as "vy-" ("up" or "out") and "roz-" ("begin to" or "in all directions"). 
When attached to the same verbal root, they produce two different verbs 
(vyjit, "go up" or "go out"; rozejit, "start to go" or "go in all 
directions"). What is the literal translation (and there are other Czech 
prefixes with three or four different meanings)? The correct answer is 
that it depends on the context--how the word is used in a given sentence. 
Now, if I understand correctly, Dr. Furuli's main point in his criticism 
of Nida's translation theory has been (pardon me if I don't word it 
precisely) that "a word brings to mind one specific concept in the mind 
of a native speaker, from which all the related meanings of the word 
derive."  Dr. Furuli's words were:

>A word, in
>my view, is void of meaning, but it is a semantic signal of a concept in
>the minds of those speaking the same language and having the same
>presupposition pool. This concept usually has a quite clear core of meaning
>elements but becomes more fuzzy when the edges are approached.

About Nida et. al. he adds:

>Nida's two central principles were: (1) There is only one target group: the
>general readers (who do not want to work with the text themselves). (2) The
>individual words have little meaning, what is to be translated are not
>words, but "kernels", which are found by a semantic (interpretative)
>analysis. (J de Waard, E.A. Nida)

>A central principle for "semantic domain" is thus expressed: "a word does
>not have a meaning without a context, it only has possibilities of
>meaning." J. P. Louw

Pardon my brusqueness, but for all his linguistic acumen, Dr. Furuli is 
simply wrong. He may be very close in the case of concrete nouns (except 
for the metaphorical overtones of grandeur, the word "lion" pretty much 
means the same thing to all English speakers), and perhaps not too far 
off with less concrete nouns and verbs referring to more or less specific 
actions, but the idea fails completely with a verb of indefinitely wide 
range (the most famous English example is "get") or with abstract 
nouns--I assure you that the word "love" signals quite different concepts 
in the minds of different native English speakers! The idea that NEPE$ 
signals one specific concept to the Biblical Hebrew speaking person may 
or may not be true (I suspect it's not; the word "run" certainly doesn't 
signal one specific concept to an English speaker), but even if it were 
true, that does *not* imply that it should always be translated by the 
same word in English, any more than the word "run" in an English book 
should always be translated by the same word in a translation into 
Norwegian.

  (b) In the target language.
    (b1) Synonyms. How can one pretend that this or that is *the* literal 
translation, when a number of equally "literal" possibilities present 
themselves? Rich languages (English and Czech both qualify) both have 
many synonyms, some exact (in which case they can be used interchangeably 
for literary variety), and some almost exact (which express nuances, 
which sometimes accord with the nuances of the original language, and 
sometimes do not--as I said, the whole point is that the semantic value 
of words *do not* overlap exactly between languages. If they did, 
computers could translate).
    (b2) Multiple meanings of the same word. "Soul" (in English) is a 
perfect example. To dichotomists, it is the immaterial part of man, 
everything that isn't body, and synonymous with "spirit." To 
trichotomists it means part of the immaterial part of man--often defined 
as that which contains the emotions, volition and intellect, or the part 
that makes me me, and not someone else. To some it conjures up the idea 
of the immortality of the soul, to others a perishable entity, to 
materialists it's at most a symbol (though they'd probably say it's just 
a function of brain chemistry), to some it is a tangible object that can 
be captured or controlled. Now, *even if* NEPE$ has a unitary meaning in 
Hebrew, what is the literal translation into English that will convey the 
true Hebraic content of the word without distorting it for an English 
speaker? Is it not obvious that a one-word-for-one-word translation is 
actually *farther* from the meaning of the text than a dynamic 
equivalence translation in which each usage of NEPE$ has been analyzed in 
its context, and been given the English gloss which most nearly 
represents its meaning in the particular passage in which it is used?

My overall point with respect to the idea that a word can have multiple 
meanings is simply that in light of this fact, a "literal" translation 
can only distort the meaning both of the words and the sense of the text. 
This should be obvious from both synonymy and polysemy. A text without a 
context is still a pretext.

3) Semantic non-overlap
Sorry for the terminology, but I don't know what it's really called. 
Think of a spectrum. The slices that English speakers call "red" and 
"orange" and "yellow" respectively don't necessariIy overlap with the 
words used in a different language. Another language may even have a name 
for a color in between red and orange, or between orange and yellow. So, 
color terms can't be translated "literally." The spectrum can be applied 
metaphorically to other things, too. Many languages have different words 
for family relationships--perhaps one word for "mother's sister" and a 
different word for "father's sister." In this case, translation could be 
accomplished paraphrastically, and could be considered tolerably literal, 
but it doesn't always work that way. I just read that in translating the 
Bible into Chinese the words )AX and )AXOT create problems, because what 
is important in Chinese is not the gender of the sibling, but whether or 
not he/she is older or younger than the speaker. I have no idea whether 
or not this is true, but can easily conceive of it being true, if not in 
Chinese, then is some other language. In this case, what is the "literal" 
translation of )AX if the context doesn't indicate the relative ages of 
the two people? The Biblical names of God are notorious for creating this 
type of problem, as the semantic content of the name of the head "deity" 
in a polytheistic tribe may be very far from the Biblical content of 
ELOHIM or YHWH. I even read once where someone tried to claim that QEOS 
was a poor choice of a Greek word to translate the Hebrew ELOHIM!

4) Figurative language
)AP in Hebrew means "nose" and, figuratively, "anger." I've heard it 
explained that since when someone is angry they sometimes snort, the word 
"nose" developed, through the meaning "snorting through the nose" into 
"anger." Whether or not this is a folk etymology I don't know; it isn't 
important anyway, as there are other similar etymological paths that are 
historically valid. The point is that when a Hebrew used the word )AP to 
mean "anger," he like as not was *not* thinking of "nose," and when he 
used the word )AP to mean "nose," he like as not was *not* thinking of 
"anger." If he was thinking of both meanings, he was probably punning. 
The thought that a word signals one specific concept is seen to be valid 
*only in a given context*. The English uses of "turkey" make another nice 
example of this, but there are thousands of examples in any given 
language. There is no such thing as a "literal" translation, only an 
accurate translation in a given context. J. de Waard, E.A. Nida and J.P. 
Louw are correct.

5) Miscellaneous
OK, so it's not the most sophisticated linguistic term. There are still a 
few other things to be considered, such as (a) diachronic changes in a 
language. This affects both the original language and the target 
language. The Hebrew Bible was written over a span of many hundreds of 
years (the "Maccabean political tract" school will have to forgive me on 
this one) and the language changed. ):A$ER in 800 BC (or 1000 or 1400 BC! 
:-) is the relative pronoun; in 500 or 400 or 250 BC it can be the 
relative pronoun, but can also be a conjunction meaning "for." What is 
the "literal" translation? In 1611 English "suffer" means "permit." In 
1985 English it means, well, "suffer." (b) Lexical holes (my, I am making 
up a lot of terms today). Czech has no word for "sophisticated." So, how 
does one literally translate an English text using the word 
"sophisticated" into Czech? Beats me; I've been trying to figure it out 
for six years now, without success. French had no word for "leadership" 
(in the sense of vision, not in the sense of administration, 
"direction"), and finally borrowed the English word. Look at the chaos 
surrounding the translation of the Hebrew word XEREM into English. These 
cases of lexical holes are, of course, just extreme cases of semantic 
non-overlap, caused by cultural non-overlap. Czech, for example, didn't 
have most of the words used in business and enterprise, for what I hope 
are obvious reasons. The Czech language has simply taken over words such 
as "management" and "businessman" as loanwords (sometimes with Czechified 
spelling, sometimes not). If some phenomenon does not occur in a given 
culture, that culture's language of course will not have a word for it. 
The difference between American English and Czech is not large, as the 
cultural difference amounts only to the difference between a European and 
an American setting plus 40 years of stagnation under communism, and the 
linguistic differences, while important, are still not immense, as both 
languages are Indo-European. But the difference between English and 
Biblical Hebrew is immense--2,000+ years of cultural difference, and a 
Semitic language vis-a-vis an Indo-European one. It's surprising, not 
that words like XEREM, which make literal translation problematic, exist, 
but rather that there are relatively few of them. (c) Lexical 
proliferation. This is the opposite of lexical holes, and exists for the 
same reason. Like the traditional example of Eskimos having 30 words for 
"snow" (again, whether or not that is true is irrelevant), so all 
languages develop a terminology of almost tropical lushness for concepts 
that are important in their culture. This creates a nightmare for those 
who would translate "literally" into a language that has only one word, 
or a few words, for the concept. Feudal titles in the middle ages are a 
good example. And how do we distinguish between the Hebrew words ):ARIY, 
)ARY"H, K.:PIYR, LFBIY), LAYI$, $AXAL and translate them "literally" into 
English, which has exactly *one* word for "lion"?

All of what I've written is obvious to anyone who speakes more than one 
language which, I presume, is everyone on this list, and I apologize for 
the pedantic nature of the post. So, why did I spend a considerable chunk 
of time writing it? Because in reading the thread entitled "Translations 
and Arian bias" I get the feeling that I'm trying to hit a moving target 
in trying to understand what the Arian/JW/NWT side means by a "literal" 
translation. The NWT is *not* a stupidly wooden translation. It has its 
literary faults, but it is readable. But something is very wrong with a 
theory that insists that every occurence of NEPE$ be translated by the 
English word "soul," that demands that Greek proskunew be translated 
"worship" when the object is God, but "do obeisance" when Christ is the 
object, that translates HFLAK idiomatically and in different ways--"go" 
in Gen. 14:24, "travel" in Gen. 19:2, "come" in Gen. 26:26, "get" in Gen. 
32:2 (Eng. v. 1), "walk" in Ex. 14:29, "go away" in Num. 24:1, etc., and 
then calls itself a "literal" translation. These three examples "just 
happen" to line up with Arian theology. Do Arians have a specific concept 
of the nature of the soul? Yes. Do they wish to draw a distinction 
between God and Christ such that God is divine and Christ is not? Yes. Do 
they have a settled theological conviction concerning the action of 
walking? No. Hence theologically motivated and linguistically 
indefensible "literal" translations of the words NEPE$ and proskunew, but 
an idiomatic rendering of HFLAK. Back to the three idioms that I 
mentioned near the beginning of this post, "slow to anger" in Ex. 34:6, 
"sparks" in Job 5:7 (for B.:N"Y-RE$EP) and "houses of the soul", in 
quotation marks, in Is. 3:20 (for B.FT"Y-HAN.EPE$). Again, the only one 
of these three phrases translated literally is Is. 3:20 
(B.FT"Y-HAN.EPE$). Is it possible that this is not due to linguistic 
considerations, but rather to the fact that Arian/JW theology has an 
important and well-defined role for NEPE$ (and hence it gets translated 
"literally" as "soul"), but none for "long," "nose/anger," "son," or 
"Resheph," and hence feels free to translate them idiomatically? Yes, I'd 
say that's possible. In other words, do I think that the NWT is smuggling 
Arian theology into the Bible under the guise of a seemingly 
sophisticated sounding but thoroughly indefensible linguistic theory, and 
then trying to impress the unknowing with a claim of literalness which 
does not stand up under scrutiny?

Yes, I do.

Mark Joseph