God Didn't Say That

Bible Translations and Mistranslations

How Similar Words Lead Bible Translators Astray

“Nuclear families” have nothing to do with “nuclear energy,” in spite of the word “nuclear” in both phrases.

Most people know that two unrelated words can look the same: the “bank” in “river bank” and in “money bank,” for example. Such words usually mean completely different things.

It’s less commonly appreciated that closely related words can also mean completely different things. In this case, the “nuclear” in “nuclear family” and in “nuclear energy” comes directly from the word “nucleus.” But even so, knowing what “nuclear families” are doesn’t help understand the phrase “nuclear energy.” (This kind of mistake is so common that “nuclear magnetic resonance imaging,” which measures the interaction between magnetic fields and atomic nuclei, was renamed just “magnetic resonance imaging” because “nuclear” falsely suggested that the process had something to do with radioactivity.)

This basic fact about languages has important implications for Bible translation.

One example comes from the Hebrew word hikriv, which means both “draw near” and “sacrifice.” It’s possible that these two meanings, as with “nuclear” in English,” have common ancestry. But that doesn’t mean that the two meanings are related. Nonetheless, it’s a common mistake to assume that “sacrifices” in the Bible had more to do with “drawing near” than the English translation suggests. They did not.

A second example is the Greek work sarx, literally “flesh,” but — as is widely known and often discussed — the word meant something different for Paul than it did for the authors of, say, Genesis.

If identical words can mean different things, certainly related words can, too. Yet many Bible translators ignore this fact.

An example comes from the two related words chamad and nechmad in Hebrew. They are both from the root Ch.M.D. The initial “n” in Hebrew essentially marks passive voice. And the vowel differences are a direct result of the lengths of the words. So it looks like chamad and nechmad should be related just like any other active/passive pair.

But they are not. The verb nechmad means “desirable” while the active verb chamad means “take.” This confusion led to a mistranslation of the last commandment, which should read “do not take,” not “do not covet.” (I have lots more here: “The Ten Commandments Don’t Forbid Coveting” and in this video: “Thou shalt not covet?.)

Returning to the English “nuclear,” it would be a mistake to try to use “nuclear energy” to understand what “nuclear family” means, and it would almost always be a mistake for a translator from English to another language to try to use the same foreign word for “nuclear” in both cases.

Similarly, it seems to me, the Bible translation challenge in this regard is twofold: First, to differentiate between similar or even identical words, so that the meaning of one doesn’t wrongly shade the meaning of the other. And secondly, only to try to use identical English words for identical Hebrew or Greek ones when the original words mean the same thing.


October 26, 2012 Posted by | translation theory, Translation Traps | , , , , , | 6 Comments

Words that Mean More than One Thing

I think it can helpful to look at familiar English words as a way of understanding ancient words and how best to translate them.

In this case, we’ll look at one way words can mean more than one thing.

“Cash” in English

One meaning of the English word “cash” is actual physical money — dollar bills, 5’s, 10’s, etc.

As credit cards became more widespread, “cash” was used in opposition to “credit.” Two ways to pay for something were by credit card or with actual bills. The second one was called “cash.”

Because there was something more immediate about “cash,” it came to represent anything that wasn’t credit. So paying by check was one way of paying cash (even though “cash” could also be the opposite of “check”). Paying with a debit card is also different than credit, so that’s also “cash.” In this sense, too, “cash” (debit card) is the opposite of “cash” (physical money).

“Cash on delivery” (“C.O.D.”) now just means payment on delivery, and is the opposite of pre-paid. You can sometimes use a credit card to pay for something C.O.D.

Most English speakers don’t have any difficulty understanding “cash” in these various contexts.

Sarx in Greek

The same sort of thing happened with the Greek word sarx, variously translated “flesh,” “human nature,” “sinful nature,” “humanity,” etc.

The trick is that the word means more than one thing.

In this regard, it’s like “cash” in English, because sarx does mean “flesh,” but also one particular aspect of “flesh,” and also “flesh” in contrast to various other things.

Even though the English word “flesh” also means more than one thing — think of “human flesh,” “the flesh trade,” “flesh and bones,” etc. — we don’t have anything that migrated in meaning the way sarx did. That’s why it’s difficult to translate.

I think that understanding this basic fact about words is essential to formulating a plan to translate them accurately.

What other polysemantic (“having more than one meaning”) words can you think of?

November 3, 2010 Posted by | general linguistics, translation theory | , , , | 12 Comments

Translating and Improving the Bible

Joel Berkowitz (in Shakespeare on the American Yiddish Stage) writes of the hubris of Yiddish theaters that promoted Yiddish productions of Shakespeare that were “translated and improved.”*

Though we mock it now, I often think I see the same thing in Bible translations, in two related ways:

1. “Translators” want to make the general flavor of the text into something it never was, frequently either overly formal (NKJV, for example) or overly informal (GNB / TEV).

2. “Translators” want to explain not just what the text says, but what it “means.” Sometimes this takes the flavor of theological interpretation. Other times it comes from a desire to make an opaque text simple.

The second issue came up recently in a comment by Peter Kirk, who correctly points out that expanding on bara in Genesis 1:1 to specify details of creation that are absent from the original text “go[es] beyond what is necessary for translation […] into theological speculation.”

One criticism of translating sarx as “sinful nature” is that is, too, is a “translation and an improvement” in that it fills in details on which the original text is silent. (Another criticism is that it’s not what the text meant. But my point here is that even if it is what the text meant, it might not be the right translation.)

Similarly, it seems to me that “translators” who take gender-specific texts and make them generic are “translating and improving.” For that matter, taking a generic text and making it gendered is a mistake, though I think this reverse pattern usually happens by error — because the translators don’t understand gender in the original language as well as they think they do — not by design.

A case in point is “ancestors.” Let’s assume I’m right that the Hebrew avot means “ancestors.” How, then, should we translate “to your avot, to Abraham, to Isaac, and to Jacob” (Deut. 1:9)? Even though the ancestors listed are all male, and even if the biblical culture was such that only the men counted (I don’t think it was — but let’s assume), I still don’t think “ancestors” should be changed to “fathers.”

A more radical case makes the reasoning clearer. If patir refers to God, I think it should still be translated as either “father” or “parent,” not as “God.”

The reason I put scare quotes around “translators” so many times here is that in my opinion translation is incompatible with deciding a priori what the content or style of the translation should be. You can (try to) improve the text, or you can translate it, but you can’t do both.

(*) By the way, though the “translated and improved” slogan is widely cited, I’ve been unable to confirm it. If you have a photo of the original, I’ll be most grateful to see it.

November 3, 2009 Posted by | translation theory | , , , , , , , , , | 4 Comments

Translating Terms of Art

The English phrase “term of art” is nicely self-referential, because it is one. A “term of art” is a term — a word or a phrase — that is used technically in a narrow context. It usually has nothing to do with “art,” except in the now antiquated sense in which “law,” “science,” etc. are all “arts.”

In addition to their specific meanings, terms of art are generally frozen phrases. So, for example, “art term” doesn’t mean at all what “term of art” means (even though “art work” is a lot like “work of art”).

Terms of art create a double translation challenge (as, really, does everything that is to be translated). They have to be identified and understood, let’s say in Greek, and then rendered accurately in translation, say, in English.

As is frequently the case, an example from modern languages may help demonstrate the point. (I’ll use American English and Israeli Hebrew only because I happen to speak those two languages.) In English we have a phrase “third party,” as in, for example, “third party liability insurance.” In Hebrew, that’s called tzad gimel, literally, “side gimel,” (gimel is the third letter of the Hebrew alphabet.)

(The “third party” is someone injured by the insured who is not a party to the insurance contract. The insured is the “first party” and the insurer is the “second party.”)

Translating tzad gimel into American English requires knowing something about the insurance industries in Israel and the U.S., in addition to a familiarity with the terms of art in the two languages.

These facts also mean that the Hebrew tzad should usually be translated “side,” but in certain narrow contexts, the only right translation is “party.”

In fact, tzad gimel is an easy case because English has a matching term of art.

What happens when the target language doesn’t have anything that matches the original?

I think that sarx and simeion are two good examples.

November 2, 2009 Posted by | translation theory, Uncategorized | , , , , | 4 Comments

Sarx, Flesh, and Mismatched Metaphors

T.C. Robinson brings up the issue of sarx again. (We went through this some time ago: Peter Kirk on BBB, Doug Chaplin on Clayboy, Mark Goodacre on NT blog, Jason Staples, a short post here, and more.)

The word is a perfect follow up to our discussions earlier this and again today about metaphors. It’s pretty clear that sarx literally means “flesh.” I think the translation challenge is that the metaphoric framework of the NT uses the concept of “flesh” differently than we do now.

In our culture, “flesh” has at least three main metaphoric uses: physicality (“he’s here in the flesh”), robustness (“flesh out”), and sex (“the flesh trade”).

In the NT, and particularly as Paul uses the word, sarx has a slightly overlapping but very different metaphoric use. In his essay “Flesh” in Romans: A Challenge for the Translator (in The Challenge of Bible Translation), Dr. Douglas Moo observes that one usage out of five of the word sarx is to “designate the human condition in its fallenness.”

And there’s the rub.

The metaphoric use of “flesh” in English relies on a system of metaphor that differs significantly from the NT metaphor of “flesh.” It’s not that sarx in Greek means different things in different places, but rather, I think, that different metaphors are at work in different places, and only some of them are compatible with modern, Western ones.

Here are some questions that come to mind:

Should mastery of a new system of metaphor be required just to read the Bible? (If so, “flesh” is a fine translation of sarx. If not, “flesh” doesn’t work.) Is it possible to ingore our native system of metaphor?

Can the meaning of the text be conveyed independent of the metaphoric system that accompanies it? (If so, “sinful nature” is a fine translation.)

Or is the metaphor part of the meaning? (If so, part of the “sinful nature” concept of sarx is its connection to the flesh.)

October 29, 2009 Posted by | translation practice, translation theory | , , , , | 18 Comments

Translating Words That Mean More Than One Thing

Frequently a Hebrew or Greek word will, in the eyes of English speakers, “mean more than one thing.”

A Foreign Word With Multiple English Translations

A Foreign Word With Multiple English Translations

There are two ways for this to happen. The first is when there are really two foreign words, similar to the situation with “bank” in English (both a financial institution and the side of a river); that’s not what I have in mind here. The trickier case is when the foreign word only has one meaning, but that meaning is more general than any English word, so it takes two (or more) English words to cover the same semantic territory as the one foreign word. This is depicted graphically to the right.

A simple example might be eitz in Hebrew, which means both “tree” and “wood” in English. It’s not that eitz means more than one thing. Rather, the Hebrew term is more encompassing than any English word. So we say that it “means more than one thing,” but really we just have a mismatch between English and Hebrew. (When the situation is reversed, we again generally resort to English-centric terminology, and say that the foreign language has two words “for the same thing.”)

Translating Words With Multiple English Translations

Translating Words With Multiple English Translations

I see three possible translation scenarios, depicted to the left. In the first two, context makes it clear how to translate the foreign word into English. These two cases are usually easy for the translator, and it’s generally only a linguistic curiosity that the foreign language has but one word for the two English ones. Continuing our example, the “eitz of knowing good and evil” is a “tree,” while the eitz of which the ark was built is “wood.”

But sometimes the usage of the foreign word spans both English words, and this is always a true dilemma for the translator. Neither English word suffices as a translation. We don’t see this with eitz, but lots of other words come to mind.

One example seems to be sarx in Greek (as was discussed extensively about a month ago by Peter Kirk, Clayboy, Mark Goodacre, Jason Staples and others, and again in passing yesterday by T. C. Robinson). It’s not exactly that sarx means more than one thing. Rather, its meaning includes “body” in English, but it is broader than that English word. When sarx is used for “body” or “flesh” (say, in Leviticus 13:24), it’s easy to find an English translation. But when it includes “body” and other important denotations as well, a good translation is elusive.

I think another set of examples comes from gender words. The Greek adelphos, for example, includes the English “brother,” but also what we might awkwardly call “co-member of society.” Again, the word doesn’t have more than one meaning, just more than one good translation, depending on context.

What other important words like this present themselves?

October 22, 2009 Posted by | translation theory | , , , , , , | 8 Comments

Weak Flesh and Sarx

Just as people (Peter Kirk on BBB, Doug Chaplin on Clayboy, Mark Goodacre on NT blog, Jason Staples, and more) are talking about sarx again, the New York Times wrote today:

Il Giornale’’s attack expanded on Thursday, with another editorial aimed at the Catholic Church itself, mocking not just the “hypocrisy” of sexually active priests with “weak flesh,”….

So apparently the editors of the Times think that even in secular contexts, English speakers know that “flesh” can have connotations beyond the literal meaning of the word.

September 4, 2009 Posted by | translation practice | , , , | 6 Comments