God Didn't Say That

Bible Translations and Mistranslations

Why the Debate between Formal Equivalence and Functional Equivalence is Deceptive

The debate between “formal equivalence” and “functional equivalence” has come up again at BBB, this time in the comment thread to a post about David Ker’s The Bible Wasn’t Written To You. (It’s a free e-book. Take a look.)

Dannii started the debate with a reference to his post “In which I ask if there’s any value to conveying morphosyntax.”

John Hobbins countered that “mimicking syntactical patterns of the source text in translation is […] a reasonable default strategy.”

That is, essentially, the crux of the debate: whether or not the grammatical details of the original should be mimicked in translation or not. The formal equivalence camp thinks yes. Functional-equivalence translators disagree.

My take is that mimicking the grammar is as foolish as mimicking the sounds. We don’t translate the Greek ho (which means “the”) as “hoe” just to mimick the sounds. And we shouldn’t translate, say, a passive verb in Greek or Hebrew as a passive one in English just to mimic the grammar. Failure to realize this basic point, it seems to me, is to misunderstand what translation is.

So why is this basic theoretical point nonetheless so hard to grasp?

I think part of the answer lies in the practice of Bible translation. By and large, published English versions of the Bible are either formally equivalent or flawed in other ways, so the debate ends up, in practice, pitting formal equivalence not against functional equivalence but instead against other kinds of mistranslations.

The non-formally equivalent CEB can help us understand how this plays out.

Among that translation’s aims is that it should be written at a 7th-grade reading level. But I think that that goal is a mistake, because the Bible is not written at a 7th-grade reading level, so from the outset, the CEB has made a decision to abandon accuracy in some regards. And as part of pursuing that goal, the CEB’s editors make other mistakes. For instance, the CEB recasts Hebrews 12:1, turning it into a statement about going in a different direction in life, while the original is about going unburdened in the same direction.

Similarly, from the CEB translation comparison, we see that Genesis 2:7 now reads, “the Lord God formed the human from the topsoil of the fertile land” instead of the NRSV, “then the Lord God formed man from the dust of the ground” (their italics, to highlight the comparison). But the original doesn’t have any notion of “fertile,” and “topsoil” is almost certainly wrong for what should be “dust.”

My point is not to pick on the CEB, but rather to use it to highlight what I think goes wrong in the formal equivalence versus functional equivalence debate.

The formal equivalence crowd looks at the kinds of mistakes we just saw in the CEB, and, rightly in my opinion, notes that these versions miss essential aspects of the original. Then they compare, say, the NRSV renditions of these verses. The NRSV correctly has “lay aside every weight” in Hebrews 12:1. It correctly has “dust” in Genesis 2:7. And it doesn’t introduce the notion of “fertile” there.

In these cases, the NRSV is more accurate that the CEB. But I don’t think that the NRSV’s accuracy here comes from its philosophy. Rather, I think it comes, in this case, in spite of its philosophy.

After all, it’s this same philosophy that leads the NRSV to translate Mark 12:18 as, “The Sadducees … asked him a question, saying:” even though we don’t “say” questions in English; we ask them. The NRSV makes the same mistake in Genesis 44:19.

The supporters of functional equivalence use mistakes like these in the NRSV to attack formal equivalence.

And what follows is a debate where both sides are right — because both the CEB and the NRSV have mistakes — but where neither side is really talking about translation theory. They are talking about practice.

So instead of asking which version is better, I think the right questions are:

1. Can functional-equivalence translations be fixed without abandoning their translation philosophy?

2. Can formal-equivalence translations be fixed without abandoning their translation philosophy?


April 21, 2011 Posted by | translation practice, translation theory, using Bible translations | , , , , , , , | 14 Comments

The Ten Commandments Don’t Forbid Killing

The Pope’s latest comments about condoms have again brought up the Ten Commandments, and, in particular, “thou shalt not kill,” which Catholics and some others number as the fifth commandment, while Jews and most Protestants call it the sixth.

Unfortunately, “kill” is a mistranslation of the original Hebrew, which does not say, “you shall not kill.”

The Hebrew verb here is ratsach, and it only refers to illegal killing.

We see this pretty clearly from Numbers 35, which deals with different kinds of killing — somewhat like modern murder vs. manslaughter laws.

For example, in Numbers 35:16, we learn that one person who kills another with an iron instrument has ratsached. Verses 17-18 expand ratsach to include killing by hitting someone with a deadly stone object or a deadly wooden object. The reasoning seems to be that iron is assumed to be a deadly weapon, while stones and wood come in both deadly and non-deadly varieties. Hitting someone with a deadly instrument is a case of ratsaching.

The point of these clauses is that there are lots of kinds of killing, and only some of them are instances of ratsaching.

Other kinds of killing — for example, killing the assailant from verses 16-18 — is not only allowed but required. That sort of required killing (capital punishment, as we call it now) is not ratsaching, and is not forbidden by the Ten Commandments.

Similarly, many other kinds of killing are not addressed in the Ten Commandments.

I go through much more evidence in Chapter 7 of And God Said, so I won’t repeat it here.

Much of this information, though, is not new.
Continue reading

November 24, 2010 Posted by | Bible versions, translation practice | , , , , , , , , , , , , , | 13 Comments

Q&A: The Original Baptism

From the About page comes a question about baptism, the essence of which is the observation that the words we now translate “baptize,” “baptism,” “[John the] Baptist,” etc. were actually ordinary words in Greek, like our “wash” in English. They were not technical religious terms like the English “baptize,” and the Greek words did not mean what the modern English “baptize” does.

So perhaps instead of “baptism” we should translate “washing.”

But it’s a little more complicated than that.

Greek Baptism

The Greek word for “baptize” is baptizo.

We know from passages like Mark 7:4 that the word can mean simply “wash”: “[The Pharisees and Jews] do not eat after returning from the marketplace unless they have washed [baptizo] … [Other traditions include] the washing [baptismos] of [various eating vessels].”

We see similar evidence in Luke 11:38: “The Pharisee was amazed to see that [Jesus] didn’t wash [baptizo] before the meal.”

We also see the verb in the OT, once in II Kings 5:14, where it’s the Greek translation of the Hebrew taval (“dip” or “immerse”), and once in Isaiah 21:4, where the word seems out of context.

Equally, we find the verb baptizo in non-Biblical Greek texts — more on this below. In those contexts, too, the verb seems to be a general one.

From all of these sources, it’s clear that baptizo is a common verb, and the specialized “baptize” in English misrepresents the original Greek.
Continue reading

August 24, 2010 Posted by | Q&A, translation practice | , , , , , , , , , , , , , , , | 28 Comments

On James 2:23-24: Why Faith Without Works is Dead

James 2:23-24 uses the same root twice to highlight the point that Faith requires Works. But that important rhetorical device — duplication of the root — is lost in most translations. For example (NRSV):

(23) …”Abraham believed God, and it was reckoned to him as righteousness” [Genesis 15:6] … (24) You see that a person is justified by works and not by faith alone.

That translation, like most others, is ambiguous regarding the exact connection between Abraham’s belief (in James 2:23, which quotes Genesis 15:6) and faith (in James 2:24).

But in Greek, “believed” is pisteuo and “faith” is pistis. The text connects Abraham’s pistis with the general nature of pistis. It’s essentially a grammatical accident that we see a verb in Genesis 15:6 — so also in James 2:23 — and a noun in James 2:24.

Why do translations have such a hard time capturing this basic effect? The KJV, ESV, NAB, NIV, NLT, and NRSV all have “Abraham believed” here, instead of the obvious other choice: “Abraham had faith.”

(The NAB’s lapse is particularly surprising. In Genesis itself that translation reads, “put his faith.” The CEV opts for “had faith” in James 2:23, but then goes with “what we believe” in verse 24.)

I also think it’s no small matter that the same root appears twice, a topic I’ll turn to soon.

June 10, 2010 Posted by | Bible versions, translation practice, translation theory | , , , , , , , , , , , | 25 Comments

The Ten Commandments Aren’t Commandments

The Ten Commandments — listed in Exodus 20 and again in Deuteronomy 5 — aren’t called commandments in the original Hebrew or in the Greek LXX.

In Hebrew, they are d’varim in Exodus 20, either “things” or “words.” (This dual use of d’varim is a bit like “things” in English — I can own ten things or tell you ten things.)

To the best of my knowledge, of the major translations only the NAB renders the Hebrew as “commandments” in Exodus 20.

For that matter, the number “10” doesn’t come from Exodus 20 or Deuteronomy 5, but rather from Exodus 34:28 and two other places in Deuteronomy. There, the KJV and other translations (NIV, ESV, NAB, NRSV, and others) translate “ten d’varim” as “ten commandments,” sometimes capitalizing the phrase and sometimes with a note that the Hebrew doesn’t say “commandments.”

(Later Jewish tradition would replace d’varim with dibrot, which also doesn’t mean “commandments.”)

Two questions come to mind: Should we keep calling these the “ten commandments” even though that doesn’t seem to be what they are in the Bible? And is the NAB justified in its translation decision?

May 18, 2010 Posted by | Bible versions, translation practice, translation theory | , , , , , , , , | 23 Comments

What Wine and Wineskins can Teach Us about Text and Context

Bill Mounce notes (also here) that Classical Greek had two words for “new”: neos and kainos.

We see them both in Matthew 9:17 (as well as Mark 2:22 and Luke 5:37), where Jesus relates that people “pour new wine into new wineskins” (NIV). The problem is that this translation (along with the NLT, CEV, and others) wrongly makes it sound as if it is the newsness of the skins that makes them suitable for the new wine. That is, the translation seems to suggest that the wine and the skin should match.

But the Greek uses neos for the wine and kainos for the skins. So in Greek, the wine doesn’t match the skin. Rather, there are two kinds of skins (palaios and kainos) and the question is which is better for wine that is neos.

In other words, the original question is “should neos wine go in to kainos or palaios skins?” Some translations prejudice the issue by asking instead, “should new wine go in to new or old skins.”

Simply as a description of the skin, I’m not sure that “fresh wineskin” — the other common option, from the KJV, NAB, NRSV, etc. — is better than “new.” (This might because I get my wine from bottles, so in truth I’m not really sure what this wineskin [askos] is, and what a fresh one looks like.) But in the context of Matthew 9:17, I think it’s more important to convey the point of the lesson than to describe the exact quality of the skin.

I also think that this is a perfect demonstration of why translating each word is not enough to create a good translation.

April 28, 2010 Posted by | Bible versions, translation practice, translation theory | , , , , , , , , , , , , | 12 Comments

The Problem with Forever

Ancient Hebrew divided “forever” into two parts: forever in the past, and forever in the future. Hebrew used the preposition “from” (mi-) to indicate the former, and “to” (l’-) for the latter.

So Hebrew has three words. “Eternity” is olam. “From the beginning of time up to now” is mei-olam. And “from now to the end of time” is l’olam. Variations include ad olam (“until olam”) and la’ad, both of which mean the same thing as l’olam.

English has a convenient word “forever” that encompasses all time, but we don’t have the equivalent of l’olam or mei’olam. This creates a translation challenge when those two Hebrew words are juxtaposed as a poetic way of indicating “all of eternity.”

One example out of a great many comes from Psalm 90 (which recently popped up here and here). The end of verse 2 in Hebrew reads, “and mei-olam [eternity up to now] ad olam [eternity starting now] you are God.” The point is fairly simple: “You have always been God and you will always be God.”

But the KJV offers the barely coherent, “from everlasting to everlasting.” Here I have to wonder if the translators were even aware of the role that “from,” “to,” and “everlasting” played in Hebrew.

More surprisingly, modern translations generally keep this odd phrasing. The NRSV, NIV, and ESV mimic the KJV here. The NAB offers the equally odd “from eternity to eternity.”

The NLT (which I generally don’t like because of its inaccuracies) has an interesting solution: “you are God, without beginning or end.” I still don’t think the NLT’s translation here is accurate, but it’s certainly better than the others, in that at least it expresses the correct general thought.

The Message (which I usually find to be even less accurate than the NLT) also has an interesting option: “from ‘once upon a time’ to ‘kingdom come’ — you are God.” The story-like “once upon a time” and the theologically-laden “kingdom come” grate on my ear, but at least the English means something akin to the point of the Hebrew.

I think this is a perfect demonstration of what I called slavery to parts of speech. In this case, I think the two prepositional phrases “from everlasting” and “to everlasting” should be translated as past-tense and future-tense verbs, respectively.

So: “you always were and always will be God.”

April 23, 2010 Posted by | Bible versions, translation practice, translation theory | , , , , , , , , , , , , , | 4 Comments

Top Translation Traps: Missing the Point

[Between six appearances in four cities and then having to buy a new car, I haven’t been in front of a computer in nearly two weeks. So I’m playing catch-up, starting with a much-delayed installment of “translation traps.”]

Following up on some thoughts about myopic translations, here’s one way in particular that a translation can focus too closely on the words and not closely enough on the text.

This is a typical translation of a (Modern) Hebrew text into English:

Rain was falling, it was cold and wet. We sat at home, we looked out toward the street.
I sat with Tali. It was very cold. I said, “What a shame. We can’t do anything.”
[I’m] not allowed to go out and play ball. It’s just cold and wet and [I’m] not allowed. [I’m] not allowed.”
We kept sitting. Just, just, just, just [sitting]. It was the most boring [thing] in the world.
And then something moved. Bump. Wow, what a bump. We were so shocked.
We looked, and then he made his way in. We looked, and we saw, a mischievous cat.

For reference, here’s the original Hebrew, with word-for-word translations:







But the English translation above, even though at first glance it may seem pretty good, is wrong in almost every regard. Can you figure out what happened?

Continue reading

April 22, 2010 Posted by | Bible versions, translation theory, Translation Traps | , , , , , , , , , , | 4 Comments

Q&A: How Mistranslation Created Divorce in the Bible

From the About page comes this response to something I wrote in And God Said:

On p. 155 of And God Said you claim that “there is no divorce in the Bible.”


Two great questions follow. I’ll take them in reverse order:

The Case of Two Husbands

Also, you speculate that perhaps the Bible would call both an ex-wife and a current wife, “his wife” but this is not true, in Deuteronomy 24:1-4 we see “former wife.”

I presume you mean “former husband,” and here we find a true translation gaff.

The KJV, ESV, NAB, NLT, and others translate “former husband” for ba’al rishon. But “former” in English usually implies “no longer,” whereas the Hebrew rishon just means “first.” For example, when Esau is born before Jacob, he is called the rishon. Genesis 26:1 mentions a famine, and then clarifies, “not the first [rishon] famine,” but rather a new famine. This doesn’t mean or imply that the first famine is no longer or famine. Similarly, ba’ala harishon doesn’t “her husband who is no longer her husband,” but rather, “her first husband.”

(There’s a related use of “former” in English that’s the opposite of “latter” and that just means “first.” For example: “Consider two people, the former a senator and the latter a judge….”)

By comparision, we might look at “ex-wife” in English. A man in his third marriage can have two ex-wives. Even if we call them “the former ex-wife” and “the latter ex-wife,” both remain his ex-wives, and the clearer way to refer to them in English is “his first ex-wife” and “his second ex-wife.”

The NIV gets rishon right with “first,” but then errs and translates shilach as “divorced” instead of the more accurate “sent away.”

The NJB’s combination of “first husband” and “repudiated her” isn’t bad, except for the fact that the Hebrew shilach is a common verb while the English “repudiate” is not.

The NRSV’s translation is pretty accurate here: “…her first husband, who sent her away…”

So here we see Hebrew that just talks about two husbands, while the English, with the word “former,” wrongly suggests that one of them is no longer a husband.

The alleged divorce only takes place in translation.
Continue reading

April 9, 2010 Posted by | Bible versions, Q&A, translation practice, translation theory | , , , , , , , , , , , , , , | 27 Comments

All in All Not Much of a Conversation

All in All

Dannii at BBB has a post about “all in all” as a translation for panta en pasin in 1 Corinthians 15:28. The full verse is (NRSV):

When all things are subjected to him, then the Son himself will also be subjected to the one who put all things in subjection under him, so that God may be all in all.

I think this is an excellent demonstration both of what can go wrong in Bible translation and how hard it can be to talk about Bible-translation issues.

Dannii read “all in all” and “almost made the mistake of thinking the Bible taught pantheism.” Therefore, for Dannii, the NLT’s “[God] will be utterly supreme over everything everywhere” is “a much better translation.”

In a comment, Marshall Massey replies that, “Personally, I’d rather have a mechanically literal translation of this verse than a translation ‘corrected’ to fit someone’s preconceived notions of what the Bible can and cannot teach. If a verse appears to teach pantheism, then let the readers wrestle with that fact!”

Another comment agrees: “Others are saying it better than me, but I’ll weigh in against this theological filtering.”

Gary Simmons then suggests that modern audiences are “less capable of interpretation” than Paul’s audience and “[i]f our audience is less capable of working through interpretation, then clarity becomes a prominent issue.” So a translator has to take obscure Greek and clarify it in translation.

Peter Kirk notes that “all in all” isn’t a literal translation at all, because “it misses the significant fact that in Greek both ‘all’s are plural, and the first is neuter, whereas the second could be any gender.”

John Hobbins then raises the issue of consistency: “It also makes sense to translate a neologism [such as panta en pasin] in the same way across all of its occurrences,” suggesting that the “ESV among recent translations is the way to go. It’s as simple as that.”

Four Conversations

At this point, we have four conversations going on:

1. What did the Greek mean?

2. How do we judge the success of a translation?

3. Which English best applies the right answer to (2) to the right answer to (1)?

4. Which published translation best represents the right answer to (3)?

(J.K. Gayle rightly notes three of these, in a different order: “1. the Greek; 2. the English; and 3. what translation means and does.”)

I think one common source of frustration is when people appear to be engaged in dialog but in fact they are having different conversations.

For example, Marshall pits literal translations against translations that are corrected. For him, this is a conversation about the second question: he prefers accuracy over emended theology. But for me, this is a combination of the 1st question and the 3rd. I agree that accuracy is important, but I don’t think that “all in all” accurately represents the Greek. So even though I disagree with Marshall’s conclusion that “all in all” is a good translation, fundamentally, at least on this, I agree with him.

Dannii seems primarily to address the third question, without first having answered the first and the second (but with the caveat: “I won’t say the NLT is necessarily correct, but at least they tried.”).

Peter is apparently addressing question 3 (“is this really a literal translation?”), though I don’t believe that he is in favor of literal translations in the first place.

Gary’s point concerns the second question.

John seems to address the second and fourth questions. I agree with John that consistency is important, though I disagree that the ESV achieves it. The same Greek phrase appears in 1 Corinthians 12:6, but there the ESV translates, “all in everyone.” (I don’t know of any translation that offers consistency here.)


I think this expanded view of what’s going on is important for two reasons:

1. This is typical of how Bible translations are debated.

2. When we aren’t clear about which question we’re addressing, fundamental agreement (e.g., “a translation should be accurate”) ends up looking like disagreement.

April 7, 2010 Posted by | Bible versions, translation theory | , , , , , , , | 9 Comments