And God Said

3 BRIDGING THE GAP:
WRITING HEBREW IN ENGLISH

BLIND IDIOTS

“Out of sight, out of mind” sounds as if it might be paraphrased as “blind idiot,” but of course it cannot. Yet many translations of the Bible make this sort of basic mistake when they render ancient Hebrew into modern English.

That’s because knowing what the Hebrew words mean is only one half of translating the Bible. The second and more difficult half is finding English words that do the same thing as the original Hebrew. More generally, translation consists of two parts: decoding the original language (Hebrew, in our case), and finding a translation in a new language (English, for us) that does the same thing as the original.

In Chapter 2 we used examples from modern languages to get a sense of how the first half of translation works. We’ll use the same approach now to understand the second half. And, as before, some examples will help pave the way for a discussion of the underlying theory.

In 1992, Laura Esquivel wrote a book in Spanish called Como agua para chocolate. In this case, the Spanish words are so familiar and easy to translate that we don’t even need the techniques of the last chapter to figure out what they mean: Como is “like,” agua is “water,” para is “for,” and chocolate is “chocolate.” Accordingly, the English-language version of the book (and then movie) was called Like Water for Chocolate.

Esquivel’s quirky story centers around a woman who cannot marry the man she wants because she is the youngest daughter, and a family tradition insists that she therefore be the one to care for her mother. The book is about the tension that results from unrequited love, family dynamics, and a generation gap.

But here’s the problem. The English phrase “like water for chocolate” doesn’t conjure up any particular image among English speakers. The heroine of the story is a cook, so people who read the book in English sometimes think that the “chocolate” in the title may refer to that. Other English speakers think that “water” instead of “chocolate” may be an image of poverty. But nothing clear comes to mind.

In Spanish, by contrast, “like water for chocolate” is a common expression based on the Spanish culinary tradition maintaining that hot chocolate is best prepared with water that is almost but not quite boiling. Therefore, “water for [making hot] chocolate” is water that is about to boil. “At the boiling point,” we might say in English. The Spanish phrase has nothing to do with poverty and, actually, little to do with food.

So while “like water for chocolate” gets all of the words right in English, in the end it completely misses the point. Ms. Esquivel’s title foreshadows the internal tensions of her characters. The English translation foreshadows nothing.

But the French version of the film is called Les hommes du president—literally, “the men of the president.” Not only isn’t that part of a nursery rhyme, it barely has anything to do with the book. Even disregarding the lack of the word “all” in the French, did the translators at least get it almost right? No.

The core of the issue is what it means to translate. We saw in the last chapter that it’s hard just to figure out what the words and phrases in a foreign language mean. Many people, perhaps because of that difficulty, stop with the meaning of the words. In so doing, they miss the point of translation.

For example, “obtain” means the same thing as “get,” but “How do you obtain from here to the airport?” is not English. “Sincerely” means “with sincerity,” but only the first one is used to sign a business letter.

In the previous chapter, we saw that blau in German sometimes means “blue,” but sometimes it has nothing to do with blue. “Blue” in English, similarly, sometimes has nothing to do with the color blue; rather, it connotes sorrow. But even though blau in German and “blue” in English have metaphoric meanings, they are not the same metaphoric meaning.

Jonathan Safran Foer’s popular book (and then movie) Everything Is Illuminated features a native Russian speaker trying to express himself in English. The hero tells the reader that his clumsy brother is “always promenading into things,” rather than “walking into things.” Similarly, he has a dog: “If you’re wondering what my bitch’s name is . . . ,” he says. Both of these demonstrate common translation mistakes. Foer humorously uses pretend translation mistakes to give the impression of a Russian writing English.

Unfortunately, we see these sorts of things—and much worse—in mainstream Bible translations. So people who read the Bible in English frequently end up missing the whole point of the original Hebrew.

As a demonstration of how difficult translation is, we might consider two lines from the Joseph narrative in Genesis 40. Joseph is Jacob’s favorite son, but Joseph’s brothers can’t stand him. They sell him to the Ishmaelites (their cousins!), who, in turn, sell him to a powerful man named Potiphar.

After Potiphar—who has misunderstood a situation between his wife and their new Hebrew slave—throws Joseph into jail, Joseph attempts to interpret the dreams of Pharaoh’s butler and baker, who, like Joseph, find themselves locked up. (Probably “butler,” “baker,” and even “jail” aren’t quite the right words, but they’re close enough for now.) Joseph first turns to the butler’s dream about a triple-branched vine whose blossoms produce grapes for wine. He has good news for the butler (Genesis 40:13): “Yet within three days shall Pharaoh lift up thine head” (KJV).

What does “lift up [your] head” mean? The rest of the line makes it a little clearer: “. . . and restore thee unto thy place.” Apparently, “lift up your head” was a positive thing in Hebrew, while in English it doesn’t make much sense at all. Even though we have similar English expressions, like “hold your head up” or “hold your head high,” people can generally hold only their own heads high, not other people’s.

The Hebrew phrase is used elsewhere, as, for example, in Jeremiah 52:31, where Evilmerodach (that’s the guy’s name) “lifted up the head” of King Jehoiachin and “brought him forth out of prison” (KJV). In Psalm 24:7, the Psalmist commands the very gates of the Temple to “lift up your heads.” In Judges 8:28, Midian is “subdued . . . so that they lifted up their heads no more” (KJV). Job laments (in Job 10:15) that even if he is in the right, he will not “lift up [his] head.”

So it looks as if to “lift up one’s head” in Hebrew, like the Spanish and German examples we just saw, should not be translated word for word, lest the idiom lose its meaning. In English we do not use “lift up his head” to mean “take him out of jail”—in fact, we don’t use that exact phrase for anything. (Heads are not the only body parts that are “lifted up” in Hebrew. Eyes are, too, leading to the perfectly ridiculous translation, “lifted up his eyes.” The Hebrew idiom means “look around,” while the English suggests torture or perhaps plastic surgery.)

Recognizing the idiomatic nature of the Hebrew, the New Living Translation translates Genesis 40:13 as “Pharaoh will take you out of prison” instead of the KJV’s “Pharoah will lift up thine head.” Similarly, the popular JPS translation offers, “Pharaoh will pardon you.” At first glance, it looks like these two modern translations are better.

But the issue is trickier than that, because in Genesis 40:19, Joseph uses almost the same expression for the doomed baker. Joseph starts off with the same four Hebrew words that mean “Pharaoh will lift up thine head.” (Surprisingly, though, the KJV uses “thy” here, not “thine”—the two English words meant essentially the same thing, but generally when two identical words are used in Hebrew, identical words should be used in their English translation.) In a beautiful linguistic play that distinguishes 40:19 from 40:13, Joseph adds the single Hebrew word mei’alecha—that is, “from upon you.” Pharaoh will pardon the butler but will behead the baker!

We can almost imagine Joseph drawing out the words. Having already promised the butler, “Pharaoh will lift up your head [and save you],” Joseph turns to the baker and promises, “Pharoah will lift up your head . . .”

“. . . off your body by hanging you.”

But the NLT and JPS, having done so well with the idiomatic Hebrew, miss the entire wordplay here and translate Genesis 40:19 as “Pharaoh will cut off your head and impale your body on a pole” and “Pharaoh will lift off your head and impale you upon a pole,” respectively. They got the meaning right but missed the wordplay.

The KJV, by contrast, translates Genesis 40:19 as “Pharaoh [shall] lift up thy head from off thee, and shall hang thee on a tree.” Except for the minor mistake of changing “thine” into “thy,” the KJV gets the play on words right—but at the expense of the meaning.

What we need is an English phrase that means “save you” but that, with an additional word, can mean “hang you.” Do we have such an option in English?

What if Joseph said, in English: “Pharaoh thinks you [the butler] have a head on your shoulders and he’ll bring you back to the court. . . . Pharaoh thinks you [the baker] should have your head taken off your shoulders and he’ll hang you from a tree. . . .” It has a play on words, and it even has the word “head,” but it doesn’t mean the same thing. Which is better? Capturing the meaning and missing the point, or capturing the point and missing the meaning?

As another option, what if Joseph said: “Pharaoh will befriend you. . . . Pharaoh will behead you. . . .” Or do we need to keep heads involved at all? Maybe, “Pharaoh will have you brought to the palace. . . . Pharaoh will have you brought to the palace gallows . . .”?

As yet another option, what about, “Pharaoh wants you to hang around the palace,” which seems to work for both the butler and the baker?

We have many options to choose from. We still don’t have a final answer, but we’ve come closer to the tone and meaning of the original. More generally, though, we need an approach to translation that will help us put this, and many other issues, into a clearer framework.

FOREIGN LANGUAGES

The computer scientist Alan J. Perlis is reported to have quipped, “Learning French is trivial: the word for ‘horse’ is cheval, and everything else follows in the same way.” He was joking, of course, but he put his finger on the two important parts of translation.

What do we mean when we say that the French word for “horse” is cheval? And how does everything else follow?

The Horse

We’ll start with the first question. What does it mean to say that the English word “horse” and the French word cheval are the same?

Anyone who speaks French and English, or anyone who has access to a French/English dictionary, “knows” that “horse” and cheval mean the same thing. But precisely because this “fact” is so widely known, few people stop to ask what exactly we mean by it.

The tendency is to think that the two are interchangeable, but of course they are not completely interchangeable. For one thing, “horse” rhymes with “of course,” while cheval does not rhyme with the French translation bien sûr. (Nor does cheval rhyme with the English “of course.”) Obviously, in some contexts rhyming may not matter, but, equally obviously, it may be crucially important in other contexts.

For that matter, and this is a little more subtle, the two are not interchangeable precisely because they are in different languages. There is something about speaking English that goes beyond the English words, and, equally, something about French that transcends the French. This is why Leo Tolstoy wrote the opening of his Russian classic, War and Peace, not in Russian but mostly in French. The opening line—the very first words of the book—is in French, not Russian: Eh bien, mon prince (“Well, my prince”). Douglas Hofstadter therefore wonders how best to translate War and Peace into French. Should the opening passage stay in French? Or does that destroy the effect? If not French, perhaps English? What about translating the book into English?

Just the fact that we have to wonder whether keeping the French in French is the best way to translate French into French shows how tricky translation can be.

There’s a third, even more subtle way in which “horse” and cheval are different. By and large, “horse” refers to a horse in America or some other English-speaking country, while cheval refers to a horse in France or some other French-speaking country. It’s not that that’s what the words necessarily mean, but certainly that’s how they tend to be used. With “horse” and cheval we probably don’t care, but what about with, say, “tree” in English and the obvious translation eitz in Modern Hebrew?

In the Northeast of the United States, for example, a “tree” almost always means a deciduous tree or a certain kind of pine tree. Unqualified, the word almost never means “palm tree.” English speakers who plant a maple tree are likely to say, “We planted a tree,” whereas if they plant a palm tree, they are likely to say, “We planted a palm tree.” They qualify “tree” in this second case precisely because it is not the usual kind of tree.

By contrast, in modern Israel, palm trees are plentiful and common. Maple trees are not. And the pine trees are of a different variety than those in the United States. An Israeli who plants a palm tree may very well say, “I’m planting an eitz.”

What is the better translation of the Hebrew? “I’m planting a tree” or “I’m planting a pine tree”? One better matches the Hebrew. The other makes it clear what’s going on.

Consider another example. The Modern Hebrew word for “blue” is kachol. We’ve already seen how complicated color words can be when they are used metaphorically, but for now we can focus just on the color “blue” itself, because an interesting pattern emerges.

In English, the primary colors are “red,” “green,” and “blue.” (Technically, these are the “additive primary colors,” because when added one to another they produce white; shine a red, green, and blue spotlight on something and it’s the same as shining white light. If colored filters or inks are used instead of lights, a combination of secondary colors are needed to filter all of the white light into black. These colors are cyan, magenta, and yellow, or, approximately, blue, red, and yellow. This is why painters know that the primary colors are red, blue, and yellow [RBY], while scientists know that they are red, green, and blue [RGB].) As it happens, the human eye decodes colors according to the RGB scheme, which is probably why red, green, and blue are usually the most basic color words. Languages do not have words for more exotic colors if they do not have those three.

In English, of course, “green” can refer to a variety of shades of green, including “dark green” and “light green.” The same is true for blue. We have “dark blue” and “light blue.” But “red” is different in English. While we have “dark red,” there is no “light red,” because instead we say “pink.” In other words, we use four color names (green, blue, red, and pink) to describe the primary colors in English.

In Hebrew, however, there are not four but five words. Like in English, there is “green” (yarok), which comes in “light” and “dark” varieties. And like in English, there is “red” (adom), which comes in “dark” but not “light,” because there’s a separate word (varod) for what would otherwise be “light red.” But, unlike in English, there’s also a separate word for “light blue”: t’chelet. So while there is a phrase “dark blue” in Hebrew, there is no phrase “light blue.” Furthermore, this means that the Hebrew word for “blue,” kachol, excludes “light blue,” just as “red” (in English and Hebrew) excludes “pink.”

Here’s the question: What’s the best English translation for kachol? Is it “blue”? That’s not quite right, because “blue” includes light blue, while kachol does not. What about “blue but not light blue”? That’s more accurate, but it seems like something else has gone wrong. Surely so short a word shouldn’t have such a long translation. Or should it?

The way the color words work has broader ramifications. Consider four children’s blocks. One is pink. One is red. And two are different shades of blue. In English, “Which two are the same color?” has an answer. The blue block and the light-blue block are both blue. But in Hebrew, the question doesn’t have an answer. The kachol block is as different from the t’chelet block as the red one is from the pink.

In terms of translation, how could one possibly translate the English “Give me the two blue blocks” into Hebrew? No sentence that starts “Give me the two . . .” seems reasonable. “Give me the two blocks that are either kachol or t’chelet”? Clearly not.

And what happens if “two” is a central part of the content of the sentence? A poem might include something like, “and two blue blocks, one light, one dark. . . .” How does that get translated into Hebrew?

These are issues that arise even when the colors are used literally. As we have seen, the problems get compounded when we take into account the metaphoric meanings of words—in this case, colors. A (slightly) deeper poem might include, “. . . and two blue blocks to match my two blue moods. . . .” How would that get translated?

Everything Follows

The reason Perlis’s statement is so jarring is that, as anyone who has ever tried to learn a foreign language knows, “everything else” does not “follow in the same way.” In addition to vocabulary (which, even from the glimpse we just saw, can be very complicated), there’s grammar.

More specifically, in addition to vocabulary, each language has its own syntax, its own morphology, its own phonology, its own orthography, and more.

Syntax is the way words are put together—that is, “word order.” The first three words of the Bible—breshit,bara, and elohim—mean, roughly, “in the beginning” (that’s the first word), “created” (the second word), and “God” (third). We’ll come back to “in the beginning” later. For now, we note that the Hebrew reads “. . . created God . . .” even though God was the one who did the creating. (A thirteenth-century Jewish work that would become a cornerstone of the Jewish mystical movement known as Kabbalah offers a fanciful wordplay on this line, reinterpreting it as “something created God,” not the other way around.)

In English, subjects—by and large the doers of actions—must usually appear before verbs. That’s why we say “God created . . .” in English, not “created God. . . .” Biblical Hebrew works differently, preferring verb-subject (“created,” the verb, followed by “God,” the subject, or bara elohim in Hebrew), but allowing subject-verb as well. This is a matter of syntax. The syntax of Hebrew is different from the syntax of English.

(There are actually a few times when English not only allows but actually requires the subject to come after the verb. One such instance is after the word “nor.” For example: “Abraham wasn’t all good nor was he completely evil.” Normally in English, subjects like “Abraham” come before verbs like “was,” but after “nor” the order is reversed. That’s also a matter of syntax.)

Morphology is how words get put together. So, what syntax is to phrases, morphology is to words. English morphology is fairly simple compared with that of other languages. In English, we add “-s” at the end of a word to mark a plural noun (“dog” becomes “dogs”) and also a singular verb (“walk” becomes “walks”). We have other suffixes, like “-ly,” “-ness,” etc., and some prefixes, such as “re-.”

Hebrew morphology is considerably more complex. The single word vichuneka has five parts and means “and he will favor you.” (The reader needn’t worry about the details, but for the curious: The first part is the prefix v-, which means “and.” The second part is y-, which marks the future tense. The third part is the root Ch.N in combination with matters that are even more complex yet; together they express the verb “favor.” The fourth part is something called an “infix”—that is, a word part that sits not before or after but in the middle of a word; the infix is -n-, whose meaning we do not know for sure, though some scholars think it emphasizes something. The fifth part is the suffix -ka, which means “you.” A rule of Hebrew morphology changes the three letters V-Y-Ch . . . at the beginning of the word into VIYCh. . . . Another rule deletes the -Y- because it comes after a vowel and before a consonant.)

As a simpler example of Hebrew morphology, we might consider the first word of the Bible, breshit, which has two parts: the prefix b- (“in”) and the word reshit (“beginning”).

Phonology is the sounds of a language. In English, the “s” sounds different at the end of “dogs” than it does at the end of “cats.” (In “dogs,” the “s” sounds like a “z.”) That’s a matter of phonology.

Finally, “orthography” is a fancy word for “spelling.”

When we’re translating Hebrew into English, we usually pay less attention to orthography and phonology than to other aspects, because orthography and phonology tend to contribute less to the meaning than morphology and syntax. For example, we don’t usually care what the Hebrew word breshit sounds like when we translate the word into English, and we don’t care how it is spelled. We want to know what it means (a difficult task, as we saw in the last chapter), and we want to know how it contributes to the meaning of the sentence.

As we will see next, however, meaning is not the only part of translation, so sometimes the phonology will be important after all. And every so often we might even care about orthography.

TRANSLATION

On one hand, it’s important to isolate the various aspects that comprise a language. We have no way to understand Hebrew, for example, if we don’t understand its syntax.

On the other hand, fixating on any one part of a foreign language is one of the biggest traps in translation. In general, it’s a mistake to mimic the syntax or morphology of a foreign language, just as it’s a mistake to mimic the sounds.

In the book of Numbers, for example, King Balak asks the pagan soothsayer Balaam to curse the people of Israel. Balaam responds with three poetic blessings, the third of which begins with the Hebrew word ma. That word means “what.” The next Hebrew word is tovu, and it means “were good.” The rest of the line means “your tents, Jacob.” The KJV translates, “How goodly are your tents, O Jacob . . .” The wrong translation is the literal “What good were your tents . . . ?” The translator has to know that the Hebrew word ma (“what”) sometimes functions the same way “how” does in English, just as the translator has to know that the past tense in Hebrew poetry sometimes has the force of the present tense in English.

Unfortunately, translators are frequently tempted to mimic as much as possible. Usually, they can’t blindly mimic the sounds. For example, in translating the Hebrew word bara (“created”), translators can’t use the English word “bara” because there is no such word. So they don’t mimic the sounds.

Sometimes it’s just as hard to mimic the syntax. The part of Genesis 1:1 that reads bara elohim cannot become “created God” in English (“. . . created God the sky and the earth”), because it doesn’t make any sense. So even novice translators know they have to rearrange the words.

But even advanced translators blindly put “in the beginning,” the translation of breshit, right at the beginning of the English translation, because that’s where it occurs in the Hebrew. (We’ll look at breshit a bit more later.)

The reasoning here is faulty (and usually subconscious). Just because English happens to allow both “in the beginning God created . . .” and “God created . . . in the beginning . . .” doesn’t mean that the two word orders, even though they are identical, mean the same thing in the different languages. In other words, even though both word orders are possible, the syntaxes of the two languages are still different.

By comparison, we can look at the Modern Hebrew word dog. It’s an animal in Hebrew. But it’s not a dog. As we saw briefly in Chapter 2, it’s a fish. No one is too deeply surprised by this coincidence. Hebrew and English have different vocabularies, and, surely, such happenstance similarities are to be expected. Similarly, again from Chapter 2, we know that the Hebrew word kontzert means “classical music concert.” We expect things like that.

(Still, it’s sometimes confusing. As we saw in Chapter 1, we get the phrase “jubilee year” only because the Hebrew yovel in its Latinized form, iobileus, sounds like the Latin iubileus, which means “jubilation.” That mistake, though it involves longer words, is essentially no different than assuming that the Hebrew dog means “dog.”)

In addition to coincidences like the Hebrew and English words for “concert” meaning almost the same things, we also expect happenstance similarities in other areas of language—in syntax, for example. In English, adjectives generally precede the nouns they modify. A block that is blue is a “blue block,” not a “block blue.” Hebrew works the other way around. The famous red heifer from Numbers 19:2 is, in Hebrew, “heifer red”—that is para aduma. Para means “heifer” (or, better, “cow”; a heifer is generally a cow that has not yet given birth, and there’s no reason to think that the Hebrew word is similarly limited), and aduma means “red.”

Translators all know that the literal “cow red” should become “red cow” in English, but they know it for the wrong reason. They know it because English doesn’t allow the possibility of “cow red.” There are a handful of situations in which English does allow the adjective to come after the noun. For example, “Chris is an expert in all things Greek.” While it’s not entirely easy to know how that sentence is different from “Chris is an expert in all Greek things,” the two English sentences are different. And only one of them is the right translation of the equivalent sentence in Hebrew.

Unfortunately, most translators blindly jump at the chance to make the English superficially the same as the Hebrew, so they would choose “things Greek” for comparable Hebrew whenever they could, never stopping to realize that doing so is the same as blindly translating dog in Hebrew as “dog.”

In other words, just because English has something that looks like the original Hebrew doesn’t mean that superficially identical English works the same way as the Hebrew. Frequently it does not, as with dog that doesn’t mean “dog,” in spite of the similarity.

Another example, this time from Modern Hebrew and modern Russian, will further demonstrate the point. (It’s convenient to use modern languages because it’s easier to see if we are right about the exact nuances of the words. We can just ask people who speak the languages.)

Both Hebrew and Russian offer more flexibility in terms of word order than English does. For example, the simple English sentence, “David saw Sarah yesterday,” has only one grammatical permutation in English: “Yesterday David saw Sarah.” By contrast, in both Hebrew and Russian, the words can be scrambled, so the equivalent of “Yesterday saw David Sarah” is grammatical in both languages. (The verb forms and other clues make it clear that David saw Sarah, but even without these clues the sentences would be grammatical, though they would also be confusing.)

Furthermore, Hebrew and Russian both have spoken and written dialects, and, as in English, the written language differs from the spoken one. The writing in The New York Times is different from the speech of the people who write the articles. The same is true in Russian. And the same is true in Hebrew.

One difference between the written and spoken dialects of Russian is that the spoken dialect allows for more word-order scrambling than the written dialect. Newspapers in Moscow frequently follow what we English speakers think of as “normal” word order. By contrast, casual Russian speech allows considerably more variation. So in most circumstances a newspaper would be unlikely to print the equivalent of “Yesterday saw David Sarah” (vchera videl david saru) to mean “Yesterday David saw Sarah,” but it’s a perfectly common way for Russians to speak. In other words, the word order in the Russian “Yesterday saw David Sarah” is a sign of colloquial, everyday speech. (The Russian word-order facts as a whole are actually much more complicated than what we see here.)

Modern Hebrew also has different sets of word-order possibilities in spoken and written dialects. However, in Hebrew, it’s the formal, written dialect that allows for more variation. So the Hebrew equivalent of “Yesterday saw David Sarah” (etmol ra’ah david et sarah) is, in fact, the most reasonable way to write the sentence in a newspaper, while it’s a fairly odd way to speak.

Here’s the question: How should the Russian “Yesterday saw David Sarah” (vchera videl david saru) be translated into Hebrew? Should the translation simply copy the words, giving us etmol ra’ah david et sarah? No. For that would be to take an informal, colloquial phrase and turn it into a formal one. Rather, the informal Russian deserves informal Hebrew. And in this case, the informal Hebrew has a different word order. (This distinction really refers to something called register, which we address in more detail later.)

TABLE 1. TRANSLATION

The details are complex, but the point is not. Hebrew and Russian have different rules of syntax, so a syntactic construction in Russian doesn’t mean the same thing that it does in Hebrew. This is exactly like the same set of sounds meaning one thing in Hebrew and meaning something else in Russian.

Obviously, the same principle applies to translating any language into any other, including ancient Hebrew into modern English. It is a mistake to simply mirror Hebrew in English.

Rather, the goal is to understand not just the vocabulary but also the grammar of the source language (Hebrew, in our case), and then try to do the same thing in the target language (English). The diagram in Table 1 depicts this process graphically.

A trivial example makes the chart clear. Hebrew has a word dog. We first decode its function, using the techniques in Chapter 2. We learn that the word is used to refer to swimmy things. Then we ask ourselves what performs the same function in English. And the answer is the word “fish.”

Earlier, when we looked at the Hebrew ma tovu . . . , we contrasted the meaning of ma (“what”) with its function (“how”) to give us the correct “how good are your tents!”

Our Modern Hebrew/Russian example works the same way. We ask what the function is of a particular word order in Russian, learn that it’s used for colloquial speech, and then ask how we do the same thing in Hebrew.

Blind mimicry involves skipping the most important steps of translation. Rarely, either because of luck or because languages often have common grammatical elements, the blind mimicry works well. That’s one extreme on a spectrum of translation. At the other extreme, sometimes the blind mimicry works so poorly that even untrained translators realize they have to fix the situation. But most cases fall in the middle. The mimicry is just wrong enough that the translation misses something, but not wrong enough that untrained translators realize that the process has gone awry.

No one learning Hebrew is tempted to think that bara means “bara,” because “bara” isn’t an English word. That’s a case of mimicry working really badly, at one extreme. At the other extreme, people who think that amerika means “America” are right. But it’s the broad middle region that causes trouble. Most people learning Modern Hebrew wrongly think that kontzert means “concert,” because that translation is close enough to seem reasonable.

And things get worse when it comes to grammar. Because there’s a better chance that a grammatical construction in one language will at least make sense in another, there’s a better chance of coming up with a translation that seems right but is actually wrong.

In addition to these sorts of issues about correctly understanding vocabulary and syntax, etc., there’s a second, even less appreciated aspect to translation. Sometimes what one language does with syntax is accomplished not with syntax but with another part of another language. The same goes for morphology and vocabulary.

We’ve already seen one example of this. Hebrew has a word (vocabulary) to indicate direct quotation: leimor. As we saw on page 37, that word is usually blindly translated into “saying,” even though in English we don’t use a separate word to indicate direct quotation. We use punctuation. (In colloquial speech, we also sometimes use vocabulary, but in a very different way. We have a verb that means, “said, and I quote. . . .” The verb is “goes.” It’s slang, but when people use it, they use it only for direct quotation. Of the two sentences, “Chris said, ‘Hello’ ” and “Chris said he was doing fine,” only in the first can “said” become “goes”: “[So I see Chris and] Chris goes, ‘Hello. . . .’ ”)

Another example comes from questions, which in English are formed through syntax—specifically, by putting a verb before the subject. So the question form of “Moses parted the Red Sea” is “Did Moses part the Red Sea?” The verb “did” gets put before the subject “Moses,” and “parted” becomes “part” so that together with “did” it can form “did part.” In Hebrew, by contrast, questions are formed through vocabulary. There’s a prefix, ha-, that introduces a question. It goes at the beginning of a sentence.

When God asks Cain what happened to his brother, Cain’s answer is, “Am I my brother’s keeper?” In English, we change the order of the words in “I am my brother’s keeper” to make it a question. Or, to think of it another way, the different word order marks the sentence as a question. Syntax is what differentiates a question from a statement in English. In Hebrew, the difference between “I am my brother’s keeper” (shomer achi anochi) and “Am I my brother’s keeper?” (hashomer achi anochi) has nothing to do with word order. Rather, the prefixed word ha- indicates that the sentence is a question. Where English uses syntax, Hebrew uses morphology.

This is but one example of the sort of cross-language mismatch that is typical of translation. So when we use Table 1, we have to remember that sometimes the function of one part of Hebrew is expressed through another part of English.

One particular way of merely mimicking Hebrew grammar rather than translating it is especially common. Translators have commonly but wrongly assumed that parts of speech have to be preserved in translation, so verbs stay verbs, nouns stay nouns, etc. But there’s no reason to think that this is so.

We can demonstrate this important fact just by looking at English, because frequently in English, as in all languages, we see various ways of expressing the same thing or almost the same thing, but the different ways involve different parts of speech. For example, “Chris is the boss,” “Chris is in charge,” and “Chris runs the company” all mean almost the same thing, but the first sentence uses a noun (“the boss”), the second a prepositional phrase (“in charge”), and the third a verb (“runs”). Certainly there are subtle differences among these three sentences, but the differences do not stem from the differences in parts of speech.

Modern French further demonstrates the point. In English, a verb like “swim” can be modified by the phrase “across the river” to create what seems like a simple English sentence: “Chris swam across the river.” But a quirk of French grammar prevents the combination of “swim” (nager) and “across the river” (à travers la rivière). The only way to express the same simple idea in French is to use the verb “cross,” so the French sentence is, “Chris crossed the river by swimming” (Chris a traversé la rivière à la nage). In other words, the English verb (“swim”) and the French prepositional phrase (“by swimming”) function identically. The English prepositional phrase (“across the river”) matches the French verb and object (“crossed the river”). It would simply be a mistake to translate the basic French sentence as “Chris crossed the river by swimming.” But this is exactly the sort of mistake that plagues most Bible translations.

All of this confusion is what led the translators of the KJV to italicize words that, in their (wrong) opinion, were added in the English even though they were not in the Hebrew. In Genesis 1:4, one example out of literally thousands, the KJV reads, “And God saw the light, that it was good.” The point of the italics is to make it clear that the Hebrew doesn’t have words for “it” and “was,” that the words of the Hebrew literally mean (in this order): “Saw God et the-light that-good.” (The Hebrew et marks “the light” as an object of the verb “saw.” The word et in Hebrew is generally used before definite nouns when they are objects of verbs. Definite nouns that are not names get the word “the” in English. “The light” is definite, so as an object it gets et in Hebrew. The indefinite object “light” would not.

In addition to the unfortunate fact that the KJV’s authors are wildly inconsistent regarding the “missing” italicized words, they missed the point of translation in this regard. Different languages use different words to convey the same thing, and, except for people who are trying to learn Hebrew, the details of Hebrew grammar should be hidden from the English reader, not italicized. (This is why in this book we generally do not mimic the italicization of words when we quote the KJV.)

In this regard, we can consider two examples. Genesis 1:3 reads, in Hebrew, “Said God will-be light and-was light.”

Linguists know that there are two kinds of languages. One kind, like English, almost always requires a subject. When the subject is the one doing the verb, as in “God said,” every language includes a subject. But the question arises of what happens when no one or nothing in particular is the one doing the verb. The answer is that, in languages such as English that require a subject, a pleonastic subject serves as the subject of the sentence. In English, the pleonastic subject is usually the word “it” or “there.”

The clearest example in English is “It is raining.” The “it” is not “raining.” Rather, the pleonastic “it” is included in English only because English always requires a subject. Similarly, we see a pleonastic “there” in “There is no reason to panic.”

French, which behaves like English, similarly requires a subject even when the subject isn’t doing the verb. “It’s raining” in French is il pleut. Il—a pleonastic subject here—means “it” (or “he”), and pleut means “rains.”

Other languages do not require a subject. Spanish is such a language. Accordingly, the Spanish equivalent of “It’s raining” or il pleut is the one word llueve—literally, “rains.”

How should llueve be translated into English? Obviously, the only reasonable answer is “It’s raining.” Italicizing the “it’s,” or marking in some other way the difference between English and Spanish grammar, would do only that: mark a difference in grammar. It would not make the translation more accurate or in any other way better.

Hebrew is a language like Spanish. It does not require a subject. That’s why “There will-be light” in Hebrew (y’hi or) doesn’t have a word for “there” and literally ends up “will-be light.” We will ignore two other aspects of Hebrew grammar—they explain why “there will be” can mean “let there be” and why “will be” is only one word—and focus on the missing “there.” Its absence in Hebrew is purely a consequence of the details of Hebrew grammar. In fact, not having the word “there” in Hebrew is exactly the same as having the word “there” in English. They both have the same force. In other words, the correct translation of a pleonastic subject in English is no subject in Hebrew, and vice versa.

The KJV gets this right and, perhaps surprisingly in light of Genesis 1:4, does not italicize anything in Genesis 1:3, offering as a translation the fairly accurate “And God said, Let there be light: and there was light.”

Unfortunately, in the very next line, the word “it” is italicized: “And God saw the light, that it was good. . . .” Like “there” in the previous line, there is nothing unusual about “it” (or “was,” for that matter). There is nothing missing in the Hebrew. There is nothing extra in the English. And italicizing “it” leaves the English reader with the mistaken impression that the “it” is somehow less a part of the original than other words.

From the point of view of studying Hebrew grammar, there might be some merit to that claim. The italicized “it” might remind students of Hebrew that Hebrew doesn’t require pleonastic subjects, whereas English does. But from the point of view of translation, it’s simply a mistake. It leaves the English reader not with a better understanding of the original but with a misunderstanding of the original.

Similar but more complicated aspects of Hebrew grammar explain why we don’t see a word that means “was” in Hebrew. And, again, similar reasoning shows us that it’s a mistake to italicize the word in English.

As another way of looking at the same issue, English orthography requires apostrophes in certain places. Even though nouns and pronouns generally behave the same way—pronouns like “her” substitute for nouns like “Miriam”—possessive nouns require an apostrophe and S (’s), while possessive pronouns require an -es with no apostrophe. That’s why “Miriam’s” means “belonging to Miriam,” but if we use a pronoun, the only correct way to spell the word is “hers,” not “her’s.” Similarly, “its” means “belonging to it,” while “it’s” means only “it is.”

Should a translation of English into Hebrew reflect this bit of arcane English orthography? Of course not. The only question is what the words mean, not which words have an apostrophe. That’s another application of Table 1. If we were to start putting apostrophes in Hebrew (which doesn’t use them for the possessive), we would be making the mistake of skipping directly from the source language to the target language, rather than translating.

Italicizing “missing” words is the same kind of mistake.

Most translations don’t italicize missing words, but they do put them in square brackets. That’s why this issue is so important. It doesn’t apply just to the KJV and its typeface choices. It applies to any translation that tries to distinguish two kinds of words—those that are “really” in the Hebrew and those that . . . what? Aren’t really there?

We should be clear. There are at least three times square brackets are used in translations, only one of which is a mistake. Some translations, doing roughly the same thing as the KJV, bracket English words that do not literally appear in the Hebrew. That’s the (widespread) mistake.

In addition, square brackets are sometimes used for addtional information that the English reader might need. For example, a translation of Genesis 6:14–15 might read, “Make an ark . . . 300 cubits [about 450 feet] in length. . . .” Most English readers don’t know what a cubit is. The translation offers an explanation, but to make it clear that the explanation is precisely that—an explanation—and not part of the translation, the explanation is set off by square brackets.

The third usage, really a subcase of the second, involves quotations. Genesis 6:15 reads, “This is how you shall make it . . .” Make what? A citation of Genesis 6:15 might read either, “This is how you shall make it [the ark] . . .” or just “This is how you shall make [the ark].”

There’s some question as to when these second and third kinds of square brackets should be used, but they do have a place. The first kind, parallel to the italics in the KJV, does not.

All of this follows naturally from Table 1. and from an understanding of what translation is (and is not).

Language is perhaps the most complicated human creation, and its parts serve to inform, entertain, arouse, rally, etc. Now that we know how important function is in translating the Bible, we should learn a little bit more about how language functions.

LEVELS

It will be helpful to consider five possible levels of translation, in order of increasing accuracy:

Sounds	What are the sounds of the original?
Words	What do the original words mean?
Phrases	What do the original phrases mean?
Concepts	What concepts are involved in the original?
Affect	What does the original do?

Frequently, English translations differ from one another because they focus on different levels of translation. (They also differ in other ways, including the accuracy of the translation.) In addition to providing insight into the task of translation, understanding the levels and knowing which published translations try to do what makes it easier to uncover the original text. The Appendix offers lots of information about the various published translations, so we won’t repeat that here. Rather, we’ll jump right into the levels.

To demonstrate, we’ll look at Song of Songs 1:1, the Hebrew for which consists of four written words: shir hashirim asher lishlomo. How do we translate that into English?

Sounds

At the most basic level, one might want to translate the sounds of one language into another. We’ll see why in a minute, but first let’s see what translating sounds entails.

In keeping with our general notion of translating, translating sounds means figuring out the function of the sounds of one language, and then finding sounds in the new one that do the same thing. Linguists and cognitive scientists agree that, by and large, sounds don’t have any inherent function, so usually there’s nothing to translate.

Returning to our earlier example, the three sounds of the Hebrew dog—“d,” “o,” and “g”—don’t mean anything by themselves. And the sounds together mean something only as a word, not as sounds. So a translation of dog as “fish” is translating the word, not the sounds, and rendering dog as “dog” isn’t translating at all. It’s mimicking.

On the other hand, when words rhyme, the sounds absolutely play a role beyond their contribution to the meaning of the words. “A stitch in time saves nine,” for example, has a nice ring to it precisely because of the (near) rhyme. The saying has a more powerful impact than the otherwise more reasonable “A stitch in time saves ten.” Normally we use either round numbers or symbolic numbers in sayings. “Nine” is neither. And the saying certainly has more impact than the bland “Doing things in time saves more work later.”

Would a translation of this line into Hebrew have to almost rhyme? Could it rhyme perfectly? Does it not matter?

To make matters more complicated, it turns out that what counts as rhyming differs from language to language. In classical Hebrew poetry—long after the Bible, but still many hundreds of years old—two words rhyme only if their entire last syllables are identical. In English, “time” and “nine” (almost) rhyme. “Time” and “chime” do rhyme. That’s because only the ends of two syllables have to be the same for two words to rhyme in English. In classical Hebrew poetry, “time” and “double-time” would rhyme, as would “muse” and “use” (from Shakespeare’s Sonnet 78). But “time” and “chime” would not.

Most scholars agree that in antiquity poetry itself worked differently. (The evidence is unclear, largely because we don’t know how ancient Hebrew sounded.) What if they are right and ancient Hebrew poetry consisted of clever word sounds but not rhyming? Should that typical-of-the-time nonrhyming Hebrew poetry be translated into typical-of-our-time poetry that does rhyme?

All of this brings us back to Song of Songs 1:1—in Hebrew, shir hashirim asher lishlomo. The KJV and NRSV both translate it as “The song of songs, which is Solomon’s.” The Hebrew line, though, contains significant sibilance (“s”-like sounds) and alliteration (repeating sounds). That creates a certain she-sells-seashells effect. If we are to translate the sounds, we have to find English that captures the same effect.

Words

Words are what people think of when they think of translation. Continuing with Song of Songs 1:1, a word-level translation takes each of the words and translates it. The first word, shir, means either “song” or “poem.” The second word, hashirim, consists of the same word shir with the prefix ha- (“the”) and the plural suffix -im. It means “the songs.” The third word, asher, means “that” or “which.” The last word, lishlomo consists of the prefix l- (“to”) and the name Shlomo, “Solomon.”

So “song the songs which to Solomon” is one word-level translation.

Another similar word-level translation adds in the English words that are “missing,” in the sense we saw earlier, from the Hebrew.

Putting two words side by side is one way that Hebrew expresses possession or, more generally, “of.” Cain’s famous question, “Am I my brother’s keeper,” reads literally “. . . keeper my-brother. . . .” Because two adjacent words indicate possession in Hebrew, in English this becomes “keeper of my brother” or “my brother’s keeper.”

Additionally, when two such words are juxtaposed, the word “the” gets put on the second, not the first, word. So “song the-songs” is how Hebrew expresses “the song of songs.” The word-by-word pattern gives us “of,” and we have to move the word “the” to make it English.

So that gives us “the song of songs.”

Accordingly, another word-level translation offers, “the song of songs which is to Solomon.”

Phrases

But we’re not done translating, because words join together to form phrases and express ideas. Again, Song of Songs demonstrates, because “to Solomon” might mean “Solomon’s.” If so, a phrase-level translation would be “Solomon’s song of songs.” (There are other possible meanings for “to Solomon,” including “for Solomon.”) What makes “Solomon’s song of songs” a phrase-level translation is changing “to Solomon” into “Solomon’s” and moving the word to the beginning of the sentence.

But, in fact, we still might not be done. The construction “X of X’s” that we see in “song of songs” seems to have been used widely in ancient Hebrew to express “best.” It may have functioned like our English “song to beat all songs” or like the spoken English “the song” (as in, “that’s not just a song; it’s the song”). If so, the English “song of songs” misses the point completely, and—depending on the nuances of the expression in Hebrew—it should be “the best song,” “the song to beat all songs,” or even just “The Song.”

In English, as it happens, we have two ways of expressing the comparative and the superlative. Sometimes we have special words, either “good/better/best” or more regular words like “dull/duller/dullest,” and other times we use phrases that involve “more” and “most.” So even though “dull” and “boring” mean the same thing in some contexts, parallel to “duller” and “dullest” we have not “boringer” and “boringest” but rather “more boring” and “most boring.” (Readers interested in why should consult the Appendix for further reading.)

Modern Hebrew has words like “more” and “most,” but it doesn’t have the first pattern, in which words change to form the comparative and superlative degrees. More importantly, though, the words for “more” and “most” in Hebrew are often optional. One way of saying “Indian food is better than Thai food” in Hebrew is, literally, “Indian food is good than Thai food.” Similarly, the same word “good” works for the superlative “Indian food is the good food there is.” Other parts of Hebrew grammar that are too complicated to address here substitute for “more” and “most.”

A word-level translation of a Modern Hebrew sentence might be “Indian food is the good food,” while a phrase-level one might be “Indian food is the best food.” Unfortunately, both translations are grammatical in English, and in most situations when the second one is true so is the first. Taken together, these facts conspire to make the wrong translation sound plausible. It’s grammatical. It even seems to mean almost the right thing. It does sound a little odd, but people who don’t speak Hebrew may think the Hebrew itself is odd. But the translation is wrong.

We see that pattern a lot with word-level translations of the Bible: They are almost right even though they sound a little odd. But they are wrong. That’s why we need phrase-level translations like “Solomon’s Song of Songs” to start off the book of Song of Songs.

Genesis 1:1 (breshit bara elohim) gives us a more difficult challenge. While the word-level translation is fairly straightforward, the phrase-level one is not.

As we saw earlier, the verb in Hebrew normally comes before the subject, as it does here. Putting something even before the verb in ancient Hebrew has the force of emphasizing it or contrasting it with something else, in a process known as contrastive emphasis. In English, we don’t have the exact equivalent of the Hebrew pattern that puts something before a verb to emphasize it, but we have something pretty close. The phrase “it was X that . . .” is nearly equivalent. Accordingly, one fairly accurate phrase-level translation of the first line of the Bible begins, “It was in the beginning that God created . . .”

The point is that Genesis 1:1 answers the question “When?” not “What?” We know that from the order of the words in Hebrew. A phrase-level English translation should reflect that nuance.

Even though we don’t have a perfect way of indicating this sort of contrastive emphasis in written English, we can do it in speaking, by changing our tone of voice.

For example, we might imagine a man on trial for robbing banks. He’s already admitted to robbing First National. When the thief is on the stand, his lawyer asks him, “What bank did you rob in February.” The man has two answers, a safe answer and a dangerous one. The safe answer is simply, “I robbed First National in February,” or, with neutral intonation, “In February I robbed First National.” The thief is simply answering the question. However, in a more dangerous answer, the thief emphasizes the phrase “in February”: “In February I robbed First National.” The most reasonable way to understand this in English is that the man robbed another bank in another month. At the very least, emphasizing “in February” raises the possibility of other months, with which “in February” contrasts.

(This is why airlinese sounds the way it does. Flight attendants tend to emphasize exactly the words that normal speakers do not. For example: “We have arrived at the Atlanta airport. . . .” Most speakers naturally emphasize “arrived” in that sentence. But the emphasis on “arrived” naturally raises other possibilities in the minds of those who hear the sentence. “Crashed,” for example, is one possibility the airlines would rather passengers not think about. By emphasizing “have,” the flight attendants only raise the possibility of “have not [arrived],” which, by comparison, isn’t so bad.)

By way of another example, we return to our observation that parts of speech need not be preserved in translation. We’ve already seen that the pattern noun-noun in Hebrew can represent possession, or, more generally, “of,” as in “song the-songs,” which means “the song of songs.” The same Hebrew construction has another, related meaning, by which the second noun describes the first. In other words, the second noun acts roughly as an adjective.

It may not seem like these two meanings are related, but they are. This is why English has a similar construction using “of ” in which the second noun describes the first. “Bricks of gold” are almost exactly the same as “gold bricks,” though in English the construction is limited to certain words and kinds of description—“bricks of heaviness” doesn’t work, for example.

Hebrew allows this construction more broadly. And when it does—as with the more basic meaning—the second word gets prefixes and suffixes that we might expect to see on the first one. So just as Hebrew grammar requires “song the-songs” to mean “the song of songs,” so too “his song of songs” would become “song songs-his in Hebrew.”

Psalm 47:8 (malach elohim al goyim, elohim yashav al kisei kodsho), numbered 47:9 in Jewish tradition, demonstrates why this is important. The second half of the verse reads, literally (a word-level translation), “God sits on throne his holiness.” The KJV pretty much stops there, adding only enough words to make the sentence sound grammatical: “God sits on the throne of his holiness.” But that’s wrong.

It’s wrong because it’s not the holiness that’s “his” but, rather, the throne. And it’s wrong for another reason. The KJV translation doesn’t reflect the fact that the noun-noun pattern in Hebrew here has adjectival force. The phrase-level translation is, “God sits on his holy throne.” (The question of “throne” versus “chair” brings us back to Chapter 2 and the beginning of this chapter. The Hebrew word that we see here, kisei, can probably mean either one. Because the first half of the verse reads “God reigns,” setting up an image of royalty, we assume that even if the Hebrew word is more general, here it means “throne.”)

Once again, the word-level translation gets things roughly right but ultimately still wrong. The phrase-level translation helps. But, as with words, phrases go only so far. Just as words combine to create phrases, the phrases combine to create concepts, as we see next.

Concepts

Eugène Ionesco remarked that “the French for ‘London’ is Paris.” His point was roughly that “Paris” plays the same role in the lives of French speakers in France that “London” does for English speakers in England. If so, should “London” be the English translation of Paris? The answer, of course, is no, but only because Ionesco’s remark was an exaggeration (though he had a very real point to make).

English speakers know what “London” is and what “Paris” is, and they know that the two are not the same. But another example is trickier. There were no candles in antiquity, not until the melting wax that forms modern candles was invented. Instead, people used oil lamps for small sources of light.

So how are we to translate the Hebrew word ner? It appears to be an oil lamp or a lantern. The word-level translation suggests that if ner means “oil lamp,” then the only possible translation is “oil lamp.” But here’s the problem. When the original text uses ner, the point is something readily at hand, a common object used by default to light up dark spaces (among other purposes). That’s why, in Zephaniah 1:12, translations for ner are “candle,” “lamp,” or “lantern”: “. . . I will search Jerusalem with candles” (KJV), “I will search Jerusalem with lamps” (NRSV), or “I will search with lanterns in Jerusalem’s darkest corners” (NLT).

At a word level all of these translations are wrong. A “lamp” in English is almost always an electric lamp, not an oil lamp. (A decorator who is supposed to “just get any lamp that fits the decor” is not supposed to return with an oil lamp.) Modern lanterns are not of the same technology as ancient ones. And they didn’t have candles in antiquity. But on a concept level, all three translations are better than “oil lamp.”

Another example from modernity comes from the different dining habits in America and Argentina. In America, “dinnertime” is generally between 5:00 P.M. and 8:00 P.M. A late dinner might be 9:00 P.M. In Argentina, 9:00 P.M. is an early dinner, and families regularly bring their school-aged children to restaurants at 10:00 P.M. on school nights. How would we translate the Argentine Spanish equivalent of “He wandered around the city at dinnertime” into American English?

Is “dinnertime” the right translation? At a word level, yes. What else could it be? But at a concept level, the answer might be no. It depends on the point of the sentence. If the point is that it’s summer but it’s been dark for a while because it’s dinnertime, then “dinnertime” doesn’t work in English. If the point is that it’s time to eat dinner, then maybe it does.

The Biblical story of Esther takes place in Shushan “the bira,” from which Ahasuerus reigned over 127 m’dinas. The KJV renders bira as “palace,” even though it uses the same English word for the Hebrew bitan. Other translations offer “citadel” or “fortress,” even though, clearly, Shushan is a city.

One might think that, like “New York” and “New York City,” perhaps Shushan refers both to the palace and to the city in which the palace is located, and that the word bira serves to make it clear that it’s the palace, not the city. But we read in Esther 9:12 that the king laments to his wife Esther: “The Jews have slain and destroyed five hundred men in Shushan the bira, and the ten sons of Haman; what have they done in the rest of the king’s m’dinas?” (KJV). So we know not only that Shushan the bira is a city but also that the m’dinas are the same sort of thing as Shushan.

The point was that Shushan was the seat of power, a m’dina from which the king ruled over other m’dinas. We almost have a word for that in English: “capital.” Even though capital cities didn’t exist yet, so bira couldn’t have literally meant “capital,” the translation “capital” gives the American reader a good idea of what Shushan was.

We’ll have much more to say about concepts in Chapter 5, where we look at the very important Biblical concepts of “king” and “shepherd.” But as a final example for now, and as a bridge to affect-level translations, we ask how “kilometers” should be translated into English. Almost everyone agrees that the translator should convert metric units into something the American reader will understand. So a Modern Hebrew sentence that reads “The speed limit in Israel is 120 kilometers per hour” should be translated into miles per hour: “The speed limit in Israel is 75 miles per hour.”

But it’s not as simple as that.

Affect

“The store is about 6.2137 miles away.”

Is that a reasonable modern English translation of the Modern Hebrew sentence that literally means “The store is about ten kilometers away”?

The issue is twofold.

As we just saw, American readers usually don’t know what kilometers are, which is to say they know that kilometers are a measure of distance, but they don’t have a sense of how much distance a kilometer represents. Almost every translation of modern English into Modern Hebrew, in which the situation is even more pronounced, converts miles into kilometers. Usually, translations from Modern Hebrew into English convert kilometers into miles.

It turns out that one kilometer is about 3,281 feet, or roughly 6/10 of a mile. Ten kilometers is 6.213712 miles. The word-and concept-level translations therefore suggest, “The store is about 6.2137 miles away.” The problem is that the phrase “about 6.2137 miles” is absurd. The figure 6.2137 is not “about” anything, but rather “exactly” something.

As a second possibility, we might suggest, “The store is about six miles away,” rounding 6.2137 down to six. But ten is a nice round number, while six is not. Maybe “The store is about five miles away” is better.

As a third possibility, we have to wonder if the hypothetical store is just “far away.” Perhaps the person describing the distance to the store is a child who doesn’t drive and just likes the round number ten. If so, would the translation be, “The store is about ten miles away”? The phrase “thousands of years” has little to do with “one thousand.” For that matter, “one hundred fortnights” is different from “four years” or (more accurately) “3.83 years.”

The fundamental issue is that accuracy isn’t the only criterion of successful translation, because language does more than convey information.

We can make the store situation just a little more contrived and demonstrate this fundamental fact even more convincingly with the following hypothetical Hebrew sentence: “Ten people went to ten stores at least ten kilometers away.” Surely the point is “ten . . . ten . . . ten,” and a translation shouldn’t destroy the pattern.

Let us consider yet another possibility. What if the “cleverness” of having used “ten kilometers” doesn’t come from within the sentence but rather from the culture? For example, suppose we have an American English story about a patriot who, in a demonstration of his love for his country, walks 1,776 miles by foot. How should that be translated into Modern Hebrew? Americans reading the story immediately recognize the figure “1776.” Should the Hebrew translation have a number that, like the English, is immediately recognizable?

We run into this exact problem with numbers in the Bible, which generally refer back to the Babylonian system of base-60 mathematics. The Babylonians couldn’t multiply large numbers—something like 671 times 419 was considered a math problem of immense difficulty, as was, even, 67 times 41. (Readers who don’t understand why might remember that zero hadn’t been invented yet and might try their hand at multiplication using a pencil, paper, and Roman numerals: XLI times LXVII, for example, or even DCLXXI times CDXIX. Good luck.)

But the Babylonians could multiply small integers. Accordingly, in addition to multiples of ten, “round numbers” in antiquity were products of small numbers. Two times three, three times four, etc. That’s why there were originally six days in a workweek (two times three), twelve hours in a day and twelve hours in a night (three times four), sixty seconds in a minute and sixty minutes in an hour (three times four times five), and so forth.

How then should the description of Noah’s ark in Genesis 6:15 be translated? The Hebrew tells us that the dimensions are “300 amas long by 50 amas wide by 30 amas high.” The KJV version, not surprisingly, keeps the numbers and translates ama as “cubit.” By that translation, however, a matter-of-fact statement about the ark has become esoteric. (The English “cubit” comes from the Latin word cubitus, “elbow,” and one cubit is the length from the king’s elbow to the end of his middle finger. So “cubit”—that is, “elbow”—was just like “foot.”)

The U.S. version of the New International Version converts the figures into feet: “The ark is to be 450 feet long, 75 feet wide, and 45 feet high.” That’s much more useful for an American reader, but what about the figures? Did they have some particular significance that they no longer do?

We have the same problem when talking about years. Genesis 5:8 tells us that Seth lived to be 912 years old. Notice the “12” at the end. That was a round number in antiquity. Whether Seth was really that age or not, readers of ancient Hebrew would see such a number as a round number, while we do not. Genesis 14:4 talks about “twelve years” of service. Should the translation make it “ten”?

An even clearer case for translating round numbers (according to antiquity) into round numbers (according to modernity) comes from Genesis 17:20. There the second part of Ishmael’s blessing consists of two parallel parts: “twelve princes shall he beget, and I will make him a great nation.” Clearly, having “twelve princes” is poetically akin to becoming a great nation. Any ancient reader of Hebrew—regardless of their view of the literal truth of the story itself—would know the “twelve” here isn’t meant to be taken literally. Twelve was a round number, similar to the “thousand” in “I’ve told you a thousand times.”

When we translate Modern Hebrew into modern English, sometimes we have to make sure that the number “ten” in the original stays “ten” in the translation, because, in both languages and cultures, ten is a round number. But in ancient Hebrew, “twelve” was also a round number, with perhaps more significance than “ten.” If we follow the same principle, we might be forced to translate “twelve princes” as “ten princes.”

This translation is where most people stop feeling comfortable. After all, the text clearly says twelve, doesn’t it? Actually, it obviously does not. The text clearly says shneim-asar. What we do with that ancient Hebrew is up to us.

We started this chapter with the quite clear observation that “out of sight [out of mind]” doesn’t mean “blind,” even though, literally, “out of sight” sounds like it ought to mean “blind.” Everyone agrees that taking an expression like “out of sight, out of mind” and turning it into a statement about the blind and the stupid is a mistake. There is no justification for it. But that fact is clear only because we fully appreciate the significance of the English, being English speakers ourselves.

An understanding of ancient Israelite culture makes it equally obvious that “twelve”—that is, shneim-asar—was a round number. Translating a round number into a non-round number is a mistake, particularly if, as here, the whole point was to choose a round number.

Fortunately, we have a third option in English. Rather than “twelve” we can translate shneim-asar as “dozen.” Even though “twelve” and “dozen” both refer to the same amount of something (unless they are bagels, in which case a “dozen” is probably a “baker’s dozen” and therefore “thirteen”), a “dozen” is a round number in a way that “twelve” is not.

In fact, the distinction in English between “twelve” and “dozen” demonstrates how a number can carry with it assumptions of exactitude or of approximation. Compare the two sentences, “I have a dozen reasons not to go out with you” and “I have twelve reasons not to go out with you.” English-speaking Americans know that, in the first case, “dozen” is not meant literally. In fact, the statement is almost the same as “I have lots of reasons not to go out with you,” and it leaves open the quite reasonable possibility that the speaker hasn’t even given thought to the actual reasons. In the second case, “twelve” suggests that the speaker has actually made a list of reasons, and they happen to number exactly twelve.

Surprisingly, though, in other contexts “dozen” is precise. “Tell the waiter to set a dozen chairs” means exactly twelve chairs.

So at first glance, “twelve” and “dozen” mean the same thing. But upon closer examination, we find that they are used in different ways, and that very subtle clues let English speakers know what the words mean in context.

And this is an example of the underlying point of affect-level translation. Words do more than simply convey data.

In the case of Genesis 17:20, it turns out that we are lucky. We can translate the verse as “. . . a dozen princes . . . ,” capturing the meaning of the text and the sense of the text. Even more than the particular solution, though, it’s the nature of the problem and the issues involved that are important.

Poetry

The opening line of Song of Songs presents a challenge. We have already seen sound-, word-, phrase-, and concept-level translations. Now we take into account the affect of the line, most importantly its alliterative and therefore poetic impact. As Mark Twain would say, it sounds well.

Each of the four words that form the title of Song of Songs has the sound “sh” in it: SHir haSHirim aSHer liSHlomo. We don’t know for sure how the ancient words sounded, and, in particular, we don’t know what the vowel sounds were. (Our current approximation comes from work done roughly one thousand years ago in Tiberias by a group of people called the Masoretes. It was the Masoretes that recorded the largely vocalic diacritic marks that supplement the mostly consonantal text of the Hebrew Bible. But the Masoretes lived well over a millennium after Song of Songs was written, and nearly that long after Hebrew had ceased to be a spoken language. So we expect that the Hebrew of the Bible sounded at least a little different than the way we currently understand it.)

This is important because it seems like the line has a certain poetic cadence in addition to the alliteration. In fact, with a minor emendation, the line demonstrates a catalectic form of dactylic tetrameter.

“Dactylic” means having a strong beat followed by two weak beats, as in the English word “poetry”: PO-e-try. “Tetrameter” means having four sets to a line. Dactylic tetrameter, therefore, is four dactyls to a line: DUH-duh-duh, DUH-duh-duh, DUH-duh-duh, DUH-duh-duh. And “catalectic,” from the Greek for “left off,” refers to leaving off the weak beats at the end of a poetic line, in our case: DUH-duh-duh, DUH-duh-duh, DUH-duh-duh, DUH. (The reader may wish to memorize the phrase “catalectic dactylic tetrameter,” as it may prove useful in resolving unpleasant situations at cocktail parties. Drop the observation, “That sounds like catalectic dactylic tetrameter,” and an annoying interlocutor will almost certainly find someone else to bother.)

The Hebrew might have been read: SHIR ha-shi-RI-im a-SHER li-shlo-MO, with capital letters representing strong beats, either through emphasis (as is common in modern poetry) or length (as may have been common in antiquity). Or perhaps we have understood the pronunciation exactly, which would make the meter an even more complex tetrameter.

Whatever the case, Song of Songs seems, appropriately enough, to start with alliterative metrical poetry. In other words, like all poetry, the point is not just to convey information but to sound well. The function of the words, in this case, is to form poetry.

So we need to find a way of doing the same thing in English. Returning once again to Table 1, we have found the function of the Hebrew, and now we need to find English that achieves the same function.

But what counts as “the same function” here isn’t clear. Does the English poetry need to be the same kind of poetry as the original Hebrew? Do we need catalectic dactylic tetrameter? Would another kind of tetrameter be good enough? Another dactylic form? Is the catalectic ending important?

Or is any poetry good enough? To judge by Greek and Roman works—Homer’s lyric epic poems, for example—dactylic poetry was common in the ancient world, but the most widespread form was dactylic hexameter, that is, six dactylic units to a line, not four. If Song of Songs begins with an unusual amount of a familiar kind of foot, perhaps so too should our translation. Iambic pentameter (five beats of weak-STRONG, like Shakespeare’s “this A-bove ALL to THINE own SELF be TRUE”) is common in English. Perhaps, therefore, Song of Songs in English should begin with iambic tetrameter or even iambic trimeter, so that it, like the Hebrew opening, will be shorter than the common form of poetry.

Rhyming may have been rare in ancient Hebrew poetry (but it’s hard to know for sure, because we don’t know how the Hebrew sounded). It’s common now. Should we translate nonrhyming ancient poetry as nonrhyming modern poetry? Or should we translate common-style ancient poetry as common-style modern poetry (in which case, even if the poetry didn’t rhyme in antiquity, it still should now)?

And what about the alliteration? How important is it to capture that?

Remember the KJV translation of Song of Songs 1:1: “The song of songs, which is Solomon’s.” We’ve seen many problems with that translation, and now we add another. It’s not poetic. In particular, the word “which” breaks things up. The New International Version suggests “Solomon’s Song of Songs,” which captures more of the alliterative feel of the original Hebrew. The New Living Translation, correctly realizing that “song of songs” is a superlative in Hebrew but not in English, translates: “This is Solomon’s song of songs, more wonderful than any other.” It got the phrases but missed the poetry, and therefore the affect, entirely.

What about “the Song of Songs Sung by Solomon”? That captures the alliteration and, probably like the original, it has four beats, making it a tetrameter of some sort, nearly an iambic tetrameter. It means nearly the same thing. By substituting “sung” for the KJV “which is,” we augment the poetry and alliterative affect. But we also lose a bit of the meaning. The original Hebrew is vague about the connection between the Song of Songs and Solomon himself. All we know from the text is that the Song is Solomon’s. We don’t know if he wrote it, liked it, sung it, commissioned it, compiled it, or something else. (Tradition, however, ascribes the authorship of the Song to Solomon.)

What about “the Supreme Song Sung by Solomon”? That translation fixes the mistake of rendering “song of songs” too literally, adhering to the words but missing the meaning of the phrase.

Even though the solution may be elusive, the point is clear by now: The opening line of Song of Songs is poetic. The translation should be, too.

More Poetry

Rhyming and meter don’t seem to have been widespread in Biblical poetry. Rather, as we saw on page 39, poetry in the Bible seems often to have been built around parallel lines; “saying the same thing twice,” we might call it.

By way of example we can return to Balaam and Balak from Numbers 24:5. We saw the first half of Balaam’s third blessing earlier. Here it is again, along with the second half: “How fair are your tents, O Jacob, your encampments, O Israel!” (Numbers 24:5, NRSV). Translations vary widely in the way they render “fair” and “encampments,” but regardless of the nuances of those two words, the verse demonstrates the most common form of Biblical poetry.

The line is poetic because of the parallel structure between “tents” and “encampments” and between “Jacob” and “Israel.”

We see the same form in Isaiah 1:2: “Hear, O heavens, and listen, O earth.” Here the poetic effect comes from a pair of near synonyms in parallel (“hear”/“listen”) coupled with a pair of opposites (“heaven”/“earth”).

This form of poetry is in fact so common that scholars are careful to distinguish two varieties of it: simple parallel structure, in which two ideas are repeated in the same order, and chiasmus (pronounced kye-AZ-mus), in which the order is reversed. (The term comes from the Greek letter chi, which looks like our English letter X. The crossing arms of the X represent the crossed structure of the poetry.)

Numbers 24:5 and Isaiah 1:2, which we just saw, demonstrate straight parallel structure. We see chiasmus in Psalm 1:1, literally: “Happy is the one who does not walk according to the advice of the wicked, nor on the path of sinners stand.” “Stand” is like “walk,” and “advice of the wicked” is like “path of sinners.” Unlike straight parallel structure, in which the two parallel parts appear in the same order both times, in chiasmus, as we have here, the order is reversed. Symbolically, parallel structure is “AB/AB,” while chiasmus is “AB/BA.”

Psalm 2:1 also begins with chiasmus, literally: “Why assemble nations, and peoples plot in vain.” Because, as we have seen, Hebrew allows for more word-order variation than English, the phrase “why assemble nations” means “why do the nations assemble.” In the Hebrew, the word order “assemble nations/peoples plot” is chiasmus. “Assemble” is like “plot,” and “nations” is like “peoples.” Unfortunately, “why assemble nations” doesn’t mean what we need it to in English, which is why the KJV translation destroys the chiasmus, translating instead, “Why do the heathen rage, and the people imagine a vain thing?”

Psalm 2:2 follows up with more chiasmus, also difficult to translate grammatically in English. From the KJV we have, “The kings of the earth set themselves, and the rulers take counsel together,” but the Hebrew reads “set themselves . . . kings/rulers . . . take counsel.”

Following up on two lines of chiasmus, Psalm 2:3 offers straight parallelism: “Let us break their bands asunder, and cast their cords away from us.” So Psalm 2 uses chiasmus for what “they” do and straight parallelism for what “we” do. The whole thing is poetic, and different reference points get different kinds of common poetry.

These are just a few of the numerous examples of chiasmic and straight parallel structure in the Bible.

Chiasmus and parallel structure are rare in English poetry. We do find it sometimes in Shakespeare (as in Othello 3.3: “. . . who dotes, yet doubts; suspects, yet strongly loves”; “dotes” is like “strongly loves” and “doubts” is like “suspects”), but certainly the poetry of Shakespeare comes more from rhythm, rhyme, and lyric beauty. By contrast, much of the poetry of the Bible comes from parallelism and chiasmus.

We should be careful to distinguish chiasmus from antimetabole, in which the exact same words are repeated in inverse order. We see antimetabole in the English “I want what I have and I have what I want.” A chiasmic version of that might be, “I want what I have and I possess what I desire.” Occasionally we find antimetabole in the Bible, but it’s not as common as chiasmus. Genesis 9:6 demonstrates: “Whoever shed the blood of man, by man shall his blood be shed.” The order “shed . . . blood . . . man/man . . . blood . . . shed” is antimetabole. (And let’s add “antimetabole” to “catalectic dactylic tetrameter” as we build our list of cocktail-party words.)

More than the order of the words, though, in the Bible it’s the choice of synonyms, near synonyms, or other words that creates the poetic affect. How is “hear” like “listen” in Isaiah 1:2? How is “heaven” like “earth”? In Numbers 24:5, how are “tents” like “encampments”? (“Tents” and “encampments” are probably only approximately right. Translating the words is difficult for roughly the same reason that translating “tree,” earlier, was difficult. People used to live in “tents” and “encampments,” but they don’t any more. “Houses” and “developments” might be closer to the right modern idea, but those words are wrong for other reasons.) How are “swords” like “spears” and “plowshares” like “pruninghooks” in Isaiah’s famous prediction about the end of war (Isaiah 2:4): “And he shall judge among the nations, and shall rebuke many people: and they shall beat their swords into plowshares, and their spears into pruninghooks: nation shall not lift up sword against nation, neither shall they learn war any more.” How is “lifting up swords” the same as even merely “learning war”?

We have a dilemma. If poetry in the Bible generally involves parallel structure, and if poetry in modernity generally involves meter and rhyme, should the parellel structure be translated as rhyming poetry in English? That is, is the function of the parallel structure in Hebrew the same as the function of, say, rhyming in English?

The problem is compounded by the fact that frequently the Hebrew poetry specifically uses near synonyms, putting two poetic words in parellel. But sometimes we don’t have two poetic words to translate the Hebrew. In addition, translators have frequently, and wrongly, focused on the literal meaning of each of the words, rather than on the effect that they create together. What results in English is a combination of two words, one or both of which is odd and sounds out of context, while the Hebrew had two words that combine to create poetry.

We don’t have much parallelism or chiasmus in English, but we do have word doublets, so we can get an idea of how the Hebrew would have sounded. For example, “no rhyme or reason” is catchy because of the alliteration. The line comes from Shakespeare’s Comedy of Errors: “Was there ever any man thus beaten out of season/When in the why and the wherefore is neither rhyme nor reason?”

Compare that original Shakespearian line to what seems like a paraphrase: “Was there ever any man thus beaten out of season/When in the why and the wherefore is neither poetical validity nor rational coherence?” While that ridiculous paraphrase means the same thing as the original, it misses the whole point, namely that Shakespeare’s version sounds well. But most Bible translations are as bad as that paraphrase.

Shakespeare’s As You Like It uses the same expression in an even more clever context. Rosalind asks Orlando, “But are you so much in love as your rhymes speak?” Orlando’s answer is a pun of sorts: “Neither rhyme nor reason can express how much.” His answer cannot be accurately “translated” as “nothing can express how much,” because the whole point is the dual use of the word “rhyme”—once literally, once in an expression.

When we translate poetic or clever Hebrew into mundane English, focusing on the minutiae of each word rather than on the effect, we similarly destroy the effect of the Hebrew.

For example, Psalm 2:5 in the KJV reads, “Then shall he speak unto them in his wrath, and vex them in his sore displeasure.” The biggest problem is “sore displeasure.” The Hebrew chiasmus has two words that refer to speaking and two words that refer to anger. We might prefer in English something like “speak”/“chastise” and “anger”/“ire”: “He will speak to them in anger and in ire chastise them.” It’s not a very good translation, but it’s still better than “. . . sore displeasure.” At least “anger”/“ire” form a nice pair. One reason it’s not a good translation is the oddity of chiasmus in English. This sample translation sounds stranger than “He will speak to them in anger and chastise them in ire.” So our translation sounds odd while the original Hebrew does not.

Going back to Psalm 2:1–2, we might prefer something like, “What purpose have gathering nations and peoples grumbling?” or “What purpose have gathering nations and grumbling peoples?” Again, “gathering”/“grumbling” creates a nice pair in English to match the nice pair in Hebrew. And again, it’s still not very good, but it’s much better than “Why do the heathen rage, and the people imagine a vain thing?” which has no hint of the poetry or chiasmic parallelism of the original Hebrew.

A full treatment of the poetry of the Bible, and possible poetic renderings of it in English, is a book unto itself, and space considerations force us to stop our discussion here. The nature of the issues is clear, even if the details and solutions are not.

It turns out that the lessons we learned about poetry apply to all parts of the Bible, even the prose. That’s because prose, like poetry, always has a certain style to it. We classify poetic styles according to “meter,” “rhyme,” “alliteration,” “imagery,” etc. Prose has imagery, too, and it can even have meter, rhyme, etc. There’s such a thing as poetic prose and, therefore, nonpoetic prose.

A good translation conveys the style of the original, so it’s important to understand one of the biggest ways language can differ and still say the same thing, a topic we turn to next.

Register

The difference between “I don’t care for that food you’re eating” and “Yuck!” is called register. More generally, “register” includes such things as familiarity, politeness, humor, sarcasm, and even poetry. So the difference between “yes, sir” and “you betcha” is likewise a matter of register. Curse words are commonplace in some registers and rare in others. As we saw earlier, newspapers can have their own register, as can other narrow contexts.

We also saw that Hebrew word order in newspapers is more flexible than it is in spoken Hebrew, and that Russian works the other way around. A more systematic way of looking at the situation is this: Both Hebrew and Russian have a newspaper register, and they both have a colloquial spoken register. By chance, one difference between the two registers in each language is the degree of possible word-order variation, and, again by chance, different registers in each language allow greater variation.

We also saw that in airlinese, also a register, different words are emphasized than in other English registers. With airlinese, we also found a reason behind the difference, but usually we are not so fortunate.

A few more examples of register from English will be helpful.

In formal settings, the past tense can be used for the present tense. At the end of a meal, a tuxedoed waiter in a fancy restaurant might ask a diner, “Did you want some more coffee, ma’am?” The wrong answer is, “I did, but now it’s too late.” In restaurantese and, more generally, in very formal registers, the past tense substitutes for the present tense.

In slightly less formal registers, the conditional is used instead of the present: “Would you like some coffee?” Again, the wrong answer is, “If what? Would I like it if what?”

Other registers offer other ways of asking the same question: “Do you want some coffee?” “Want some coffee?” “Coffee?” Or, between a husband and wife, perhaps just a barely intelligible pre-morning-caffeine grunt.

As a final example, we might consider sports. In certain situations sportscasters use the future tense instead of the past. For example, take the often-cited cliché of a baseball grudge match that comes down to the bottom of the ninth inning with the team at bat down one run: There are two outs, the bases are loaded, and the batter has two strikes against him. It all comes down to one pitcher against one batter. The sportscaster watches the pitcher throw a strike and then yells into his microphone, “And that’ll be stike three!”

“That will be”? Why not “that was”? It’s a difference of register.

Different languages have different registers. Most languages have different styles for casual and polite speech. Some have odd things like special language for sports or music or whatnot.

In addition, the particular ways that register is expressed differ from language to language. Many languages, among them French and Spanish and German, have formal and informal pronouns for “you,” with verb forms to match. Some have short words that indicate register. And others, as we have seen, have different word orders in different registers.

We know the drill by now. A good translation preserves register—the function of the language—and not necessarily the parts of the language that indicate the register. Funny things should stay funny in translation. Formality should stay formal. And, as we just saw, poetry should stay poetic.

Most translations do a fairly good job of noticing when the original Hebrew is poetic (even if their success in rendering the poetry is limited). Unfortunately, most translations do a terrible job in preserving other register distinctions.

Instead, they choose a register and force the entire text into that register. Proponents of the KJV frequently like it because it all sounds archaic (even though the Hebrew wasn’t). Partially in response the New Living Translation was created. In the NLT, everything is chatty. In the NRSV and JPS translations, everything is (supposed to be) colloquial.

But the Bible contains a great many registers. The story-laden prose of Genesis is different than the legalistic prose of Leviticus or the narrative prose of Kings. And even within Genesis, the Joseph story is graced with particularly vivid writing.

Reading most Bible translations is like reading The New York Times, Shakespeare, and Beatles lyrics but destroying the differences among the three. When Shakespeare sounds like the daily news, something is wrong. When everything in the Bible sounds identical, the same thing is wrong.

And we know exactly what is wrong. We go back to Table 1 and remind ourselves that the point of translation is not to mimic the original but, rather, to translate it. Getting the words right and failing to preserve register is no different than getting the sounds right and getting the words wrong. They are both mistakes.

An Example

Before we move on, a quick example is in order. Here’s a passage from Numbers 31, as it appears in the KJV:

“And the LORD spake unto Moses, saying, Avenge the children of Israel of the Midianites: afterward shalt thou be gathered unto thy people. And Moses spake unto the people, saying, Arm some of yourselves unto the war, and let them go against the Midianites, and avenge the LORD of Midian. Of every tribe a thousand, throughout all the tribes of Israel, shall ye send to the war. So there were delivered out of the thousands of Israel, a thousand of every tribe, twelve thousand armed for war. And Moses sent them to the war, a thousand of every tribe, them and Phinehas the son of Eleazar the priest, to the war, with the holy instruments, and the trumpets to blow in his hand.”

Here’s the same passage in the usually easier-to-read NRSV:

“The LORD spoke to Moses, saying, ‘Avenge the Israelites on the Midianites; afterward you shall be gathered to your people.’ So Moses said to the people, ‘Arm some of your number for the war, so that they may go against Midian, to execute the LORD’s vengeance on Midian. You shall send a thousand from each of the tribes of Israel to the war.’ So out of the thousands of Israel, a thousand from each tribe were conscripted, twelve thousand armed for battle. Moses sent them to the war, a thousand from each tribe, along with Phinehas son of Eleazar the priest, with the vessels of the sanctuary and the trumpets for sounding the alarm in his hand.”

And this is what the passage should be:

“Adonai said to Moses, ‘Avenge the people of Israel on the Midianites; then you will die.’ And Moses said to the nation, ‘Choose men and form an army, and let them avenge God on Midian. Send one unit from each and every tribe to the army.’ So one unit from each tribe was furnished from among Israel’s units (twelve units of soldiers in all) and Moses sent them (all twelve units) to war, them and Pinchas Son of Eleazar, the army priest, armed with holy instruments and loud trumpets.”

By focusing too closely on the words, both the KJV and the NRSV miss the point of the passage, which is to describe the start of a war and to show how Moses changes God’s command. (God tells Moses, “Avenge the people of Israel,” while Moses tells the people, “Avenge God on Midian.”) The text is not supposed to be flowery, obscure, archaic, or otherwise dense. It is supposed to convey a message, a message that is hidden by most translations.

TRANSLATION, AGAIN

In light of all this, we return to the question of what a good translation should be. And the answer is that in addition to being accurate, the ideal translation will work at every level—from sounds, through words, and up to concepts and effect. But because this is seldom possible, the translator must choose which levels should be given priority. Most people are of the mistaken opinion that the words should always be given priority, that as long as the words are translated correctly, everything else falls into place.

But this is not usually true. In fact, there are very few times when the words themselves are the most important part of a text, just as there are very few times when the sounds are the most important. While it’s hard to understand the text as a whole without knowing what the words mean, just knowing the meaning of the words, as we have seen, is not nearly enough to understand the text.

More generally, a “literal” translation is almost always just a “bad” or “wrong” translation, inasmuch as it fails to give a reader of the translation an accurate understanding or appreciation of the original.

We have now seen how to use context to figure out what the ancient Hebrew words mean and what to look for when we render those words in English. Now it’s time to turn to some more concrete and important examples. In Part II, we’ll apply our knowledge to some of the most commonly quoted parts of the Bible, including Jesus’ most important commandment, Psalm 23, the Ten Commandments, and finally Isaiah’s prophecy of a “virgin birth.”

We’ll see that bad translations have radically misrepresented even these central themes.

3

BRIDGING THE GAP:WRITING HEBREW IN ENGLISH

BLIND IDIOTS

FOREIGN LANGUAGES

TRANSLATION

LEVELS

TRANSLATION, AGAIN

BRIDGING THE GAP:
WRITING HEBREW IN ENGLISH