God Didn't Say That

Bible Translations and Mistranslations

What September 11 Might Have in Common with Translating the Trinity

I imagine a novel written in a remote location, far from western culture. It’s about the last ten days of summer and the nearing autumn. So they call the book the equivalent of “What Happened on September 11” in their local language.

My question is this. Should the American version of the book be called, What Happened on September 11?

I don’t think so, because even though September 11 is ten days before the end of summer in English, too, the phrase “September 11” has local overtones — the terrorist attacks, the wars that followed, etc. — that override the simple meaning of the phrase.

This is one way that a good translation of the words can be a bad translation of the text.

What this has to do with the Trinity is that the claim has surfaced that in Arabic, “father” and “son” wrongly imply sex, so they’re not good translations for what we know in English as the Father and the Son.

Facts to support this claim about Arabic (and other languages of the Middle East) have been frustratingly difficult to come by, but even the theoretical issue, it seems to me, has been misunderstood.

Some people have claimed that getting rid of “Son” in Arabic is pandering, or wrongly changing the Bible to placate an audience, or giving up on theology, etc. Maybe. But maybe not. Maybe “son” in Arabic is like “September 11” in English. It has a plain meaning, but it also has overtones that destroy the original point of the text.

Other people have claimed simply that the job of the translator is to translate the words. In spite of the hugely intuitive appeal of such an approach, it doesn’t work very well, because sometimes the words convey the wrong thing.

So even before we get a good factual answer about Arabic, I think it’s important to understand the fundamental point that it’s certainly possible for the literal equivalent of “son” and “father” to be the wrong way to translate the Trinity.


February 9, 2012 - Posted by | translation theory | , , , ,


  1. The problem with this particular debate is that we’re not translating into a language that has never had a Bible before. There have been Bibles in Arabic since the 8th century.

    Comment by Gordon Tisher | February 9, 2012

    • I agree that that’s a complicating factor.

      On one hand, it often makes sense to preserve familiar translations because they are familiar.

      On the other hand, the original translations could have been wrong in places (as with the KJV in English), or Arabic could have changed enough to make a previously correct translation now sometimes inaccurate (again, as with the KJV in English).

      Comment by Joel H. | February 9, 2012

  2. It seems, Joel, that this problem closely relates to the same kind of close-minded storm arose over the Revised Standard Version’s translation in Isaiah of “young woman” as opposed to “virgin.”

    There are those who are so adamantly set on their own ideological purity that they insist on sacrificing everything on the altar of dogma.

    After all, NO process of translation from one language to another – especially when the two are not rooted in a common ancestry, e.g. from Hebrew to Sanskrit – is an easy process. Even in two related languages, such as Hebrew and Arabic, social connotations have changed the meanings of words radically over hundreds to thousands of years usage.

    Comment by Colleen Harper | February 9, 2012

    • I think you’re right that some people approach the issue already “knowing” that what they are most familiar with is also most accurate.

      I also think there’s a difference between changing “virgin” to “young woman” versus changing “son” to something else. In the first case, the debate is about the original meaning of the text. In the second case, the debate is about how a modern language works.

      Normally it’s harder to figure out ancient languages (though in the case of “young woman” the evidence is pretty clear), which is why it’s so frustrating that I can’t get a clear answer about Arabic.

      Comment by Joel H. | February 9, 2012

  3. Er, one problem that I’m not sure has been adequately addressed in this [and the previous] discussion is the Trinity’s meaning in English: Father, Son, and Holy Spirit. God is not a father, nor was Jesus His son in any literal, which is to say biological, sense. Thus, those who are offended by the sexual innuendo of the Trinity are simply reading the English text mechanically, word-for-word.

    I suggest that the Bible translators’ response to the offended parties should be to remind the Arabic reader that s/he is to read the text with the figurative meaning in mind. Perhaps this is why God invented footnotes.



    Comment by Michael | February 9, 2012

  4. If “father” and son” in an Arabic Bible really do not work (I don’t knw Arabic, soi am _assuming_ this as as act, _not_ stating it as a fact), then what do the Arabic Bible translators use/recommend instead? — and how did they make that choice, whatever it is?

    For what it’s worth, I read the other day (In Huston Smith’s book THE WORLD’S RELIGIONS) that Muslims do not envision even a metaphorical sense for being a “son of God” or “child of God” — they do not say, or agree with saying, such things as “We are God’s children” or “God is our father”: it just doesn’t strike them as admissible metaphor. (On the other hand, I have been told by a Christian Aran that the Christian Arabs — who were around for centuries before Muhammad — quite normally refer to Jesus as “ibn Allah”: meaning “the son of God.”

    Comment by Kate Gladstone | February 9, 2012

  5. In thenscenarioyou describe, I think the best translation for the book-title would be ONE DAY IN SEPTEMBER.

    Comment by Kate Gladstone | February 9, 2012

    • I think even that title would be construed as being related to September 11. If I saw it in a bookstore today that’s what I would assume. But I admit I don’t connect September with much else now, even though I now have a two year old with his BD in September.

      Comment by Marnie | August 15, 2015

  6. Your more general point, Joel, draws attention to the difference between the linguistic technical terms ‘semantics’ and ‘pragmatics’. I don’t know your blog constituency, but the former word refers to what most people call the word’s meaning, and the later refers to how the word interfaces with the reader’s (hearer’s) context. Sept. 11th is a particular date on the yearly calendar–that’s the semantics. For Americans in particular, Sept. 11th lights up a pile of meaning-chunks within the reader’s mind, that is, the reader’s context. A lot of people think of ‘context’ as the words around the words. It’s not. It’s really the interpretive environment within the reader’s mind. Obviously, as the reader reads, this environment is conditioned by the words around the words. But, there’s more to the context than just what the author has said (or even will say).

    My point is a translator needs to be aware that the words he or she chooses in order to render the original are going to interact with a rather different context than the original audience brought to the text. I think this is the point you are driving toward. So, care must be taken so that the reader will be adequately enabled (note: nothing is perfectly determined here) to arrive at a correct interpretation. I’ll also answer an objection to my just made statement by saying that the coherence of the text is absolutely vital in both the original and translation, or abuse of the reader can happen. That is, the translator is not freed to translate any way he or she feels like it. Also, on the reader side of this same issue, our too common proof-texting “exegesis” trips-up too many readers. As readers, we’re not well trained to think in terms of paragraphs, especially so when reading the Bible. This coherence, as I see it, greatly pushes back the exegetical failures where we inject our context into the text.

    So, on the one side the translator must not require the reader to learn an entirely new context just to simply read the Bible. On the other hand, the reader needs to pursue understanding the flow of thought of the original author.

    I once worked with a Pakistani. He was well read and enjoyed exploring different religious ideas. We talked about “Son of God.” It was interesting to me (in a linguistic sort of way) seeing him struggle with knowing, honestly knowing, the Christian definition of “Son of God,” And yet, he could not tear himself away from the weirdness that God somehow had sex with Mary. The semantics argued with the pragmatics in his head and did so within the confines of his own language intuitions. He understood the theology; it was the language which tripped him up.

    Lastly, to add even more contextual complications, “Son of God” in the original context had a number of pragmatic features not readily held by contemporary English speakers. ‘Son’ was very easily connected to its legal context. A son had rights to his father’s estate. We can see this in the legal adoption activities which even the biological children needed to go through. We also see this in the Galatians 4 passage.

    “Son of God” was also used by the Ceasars. So, as used by the early Christians, there were connections to Jesus’ right to rule the world.

    So, in English, we loose some of this pragmatic meaning when WE read “Son of God.”!!

    Sorry my comment is somewhat long. I really appreciate your point, and I really think it needs more exposure so the average person better appreciates the Arabic as well as the English issues.

    Comment by Mike Sangrey | February 9, 2012

    • Hi Mike.

      Thanks for your detailed comments.

      Lastly, to add even more contextual complications, “Son of God” in the original context had a number of pragmatic features not readily held by contemporary English speakers.

      I’m not convinced that “son of Man” or “son of God” are that successful as translations into English. (And the CEB, for what it’s worth, agrees.) But even if those phrases miss the mark, I don’t think they do the same kind of damage as specifically suggesting that sex is involved.

      Comment by Joel H. | February 10, 2012

  7. The overtones of “son” in Arabic may well destroy the original point of the text, but is it possible to find a better word that will do the original justice?

    Jesus was not merely a man, but God with skin on him. Or, as John put it, the Word made flesh. Obviously, calling God his father didn’t sit right with many of his contemporaries either.

    If “son” is inappropriate, and Jesus was not merely a man, would it not be a greater mistake in going for something like “representative”, despite the absence of any overtones? In any case, how would “representative of God” do any justice to Jesus being the only begotten of the heavenly Father, which “the son of God” actually implies?

    My point is, whatever term one uses, please find one that properly conveys the unique relationship between Jesus and his heavenly Father. Anything less, and you have a lot of explaining to do.

    Comment by Robert Kan | February 10, 2012

    • The overtones of “son” in Arabic may well destroy the original point of the text, but is it possible to find a better word that will do the original justice?

      I don’t know.

      Sometimes there is no good translation, and the best a translator can hope for is a mediocre second. In general, I think a translation that destroys the original point is worse than one that misses the mark more subtly.

      Another way of looking at the general situation is this: if a reader just reads the text, and not the footnotes/explanations/etc., which of the wrong translations does the least damage?

      For example, what if a foreign text has something like, “he used a carefully selected four-letter word.” In English, “four-letter word” means “curse word,” so an English speaker reading that will get the wrong impression that the sentence is about profanity. Is it worth it? Is it worth misleading the reader in this way to convey the nuance of how many letters are in the word? I don’t think so.

      In the case of “son,” he question may come down to (again, depending on the facts about Arabic), “is it worth giving the reader that sex is involved in order to convey the other aspects of ‘son’?”

      And it seems to me that the answer is “no.”

      Comment by Joel H. | February 10, 2012

  8. Can we be reminded of what the perception was when Jesus referred to himself as the “bread of life”? “How can this man give us his flesh to eat?” They thought he was talking of cannibalism. And his response?

    “Truly, truly, I say to you, unless you eat the flesh of the son of man and drink his blood, you have no life in yourselves. He who eats my flesh and drinks my blood has eternal life, and I will raise him up on the last day. For my flesh is true food, and my blood is true drink. He who eats my flesh and drinks my blood abides in me, and I in him.” (John 6:53-56)

    Just as Jesus expects us to eat and drink of his body, the Bible also expects us to read that God got Mary pregnant. Jesus could have arrived here without conception, but the story wouldn’t be the same. To draw the conclusion of sex, one needs to read between the lines. As for eating and drinking the body and blood of Jesus, the Catholics have their doctrine of Transubstantiation.

    However, Michael and Kate, in their responses above, remind us to appreciate the figurative/metaphorical dimension of the biblical literature.

    Comment by RobertKan | February 10, 2012

  9. RobertKan, I think your reference actually underscores Joel’s point. I’m not sure whether that was intended.

    I don’t know Aramaic, but I’ve been told that ‘bread’ stands as an Aramaic metaphor for ‘teaching’. So, all of John 6 is Jesus teaching that his teaching is an authoritative teaching from heaven itself. That’s a difficult position to accept without some solid proof (like a resurrection). In fact, Jesus says he embodies the teaching. We are to make him our model and fully take in who he is. The misunderstanding by the Pharisees simply underscores the nature of a double entendre. They were rejecting Jesus teaching–a common Johanine theme. The Pharisee’s question also comes across as a group trying to “put Jesus down” through a form of ridicule.

    The problem with double entendre, and things like puns, is they present translation complexity. Translating the Aramaic ‘bread’ into Greek presents one level of difficulty which can be handled by current mother tongue speakers. However, going further and bringing that over into English, and a contemporary audience largely separated from the original linguistic context, presents the same types of translation issues that ‘Son of God’ does.

    My point is that the meaning, as complex as it was, was a bit obvious in the original. The trick is to enable the translation to approximate the same level of clarity. Is the double entendre important enough that a translator must keep the original’s double entendre at the expense of the original’s clarity? Or is it wiser to prioritize the various meaning elements differently? Unfortunately, given people’s assumptions about what makes a good translation (note the reference to the pragmatics of the contemporary audience!), people assume that prioritizing the elements in a way that brings clarity is somehow compromising accuracy. In my opinion, it is often the other way around.

    Comment by Mike Sangrey | February 10, 2012

    • Mike, thanks for putting forward your position. I take it that part of your theory of Bible translation is that the modern text should reflect what the author would be saying if he were mingling amongst us today.

      In our western culture, do you think Jesus would be telling us to eat his flesh and drink his blood? Do you think he would be telling us he was the son of God? How do you think he would communicate these concepts today?

      Comment by Robert Kan | February 11, 2012

  10. In our western culture, do you think Jesus would be telling us to eat his flesh and drink his blood? Do you think he would be telling us he was the son of God? How do you think he would communicate these concepts today?

    I think it would depend on which language he was speaking today. If he was speaking modern Aramaic (there is such a language) he just might use exactly the same words as he did when he was, as some claim, speaking Aramaic of his own day. In other words, if the literal meanings of the words and their semantic associations within the cultural context of the people he is speaking to fit the meanings he intends, then I think he would use those words. If, however, the semantic and cultural associations of a particular language quickly trigger wrong meaning associations then I suspect he would use other words which would communicate his intended meaning.

    As I understand the translation problem being faced in the current debate, it’s not so much the semantics of the Arabic language itself, but the cultural and theological context of minority languages into which the Bible is being translated. These minority languages are spoken by Muslims for whom any idea that God has sex with a human is repugnant, sacrilegious. Pushback against adopting the metaphorical meaning of sonship to the lexical options available within these minority languages (or Turkish, for which a recent translation has been made by some non-SIL group), comes from Christian leaders whose ministry “fathers” have lived with Muslim neighbors for centuries. The Bible used by these Christians communicates to neighbor Muslims wrong ideas about God having sex to create a son. But it is misunderstanding that this long Christian tradition has been willing to tolerate, partly out of resignation that the bridge of misunderstanding about God’s “son” will ever be bridged. But a newer breed of linguistically trained Bible translators are suggesting that it isn’t always necessary that translations of the Bible create misunderstandings by their translation word choices. So there is a tension between an older way of approaching the misunderstanding and a newer attempt to bridge the gap. Both parties are attempting to be accurate as they deal with the Bible. With the traditionalists the accuracy is to the literal words, with the idea that misunderstanding might be cleared up by teaching. To the newer translators the attempt is for accuracy of meaning, with the idea that initial understanding of words is more important than trying to get to unknown or rare meanings of words.

    It’s a significant dilemma and it is not surprising that it is raising the temperature of a linguistic tension that has existed for centuries.

    In response to a comment that emphasized that Jesus was God’s “begotten,” I would suggest that we need to take the same careful look at that English word “begotten.” To me it means ‘created as a child is normally created’. I realize that theologians at church councils and in textbooks have tried to give the word a different meaning, but I’m not convinced that taking words and trying to change their meaning works very well. Most people understand words with the first meanings that come to mind, unless the context is clear enough to nudge them to consider some other rarer meaning. Notice how the English word “gay” has changed its primary sense of the past to a different primary sense today.

    Secondly, it’s not clear that the English word “begotten” is the most accurate translation of the Greek word monogenes, as in John 3:16. This is still debated today, but I think a majority of Greek scholars today recognize that monogenes has little, if anything, to do with birth or parent-child relationship, and, instead, has to do with, literally, ‘only existing, that is, ‘only one’ or ‘unique. The difference in Greek between ‘only born’ and ‘only one’ is very slight, something like one letter. So the theological concept of “begotten” may actually be erroneous, based on a mistranslation of the Greek.

    Comment by Wayne Leman | February 17, 2012

  11. Would anyone suggest that, in Aramaic, it is highly improbable that sonship normally implies sex? As long as this new breed of translators focus on meaning, rather than on the observation of what is merely misunderstanding the author, they may just be on the right path. To me, arguments that are based on cultural leanings are directed more towards political correctness and being socially sensitive, of which Jesus was neither. Who else on earth would tell his followers to let the dead bury their dead? Where do our cultural associations come from anyway, if not from preconceived ideas? If the first meaning that comes to mind is “sex”, when one hears the word “son”, is this not basically a one-dimensional way of approaching language?

    Is this not the hearer’s problem, rather than the speaker’s? Is it not the translator’s responsibility to accurately render the very words used by the author himself, that is, those words that were commonly used by the majority of the people during that period of time? To me, this discussion doesn’t appear to be about language only, but about how to adjust peoples’ cultural perceptions to our liking.

    Comment by Robert Kan | February 17, 2012

  12. I don’t think we can say, “It’s the hearer’s problem,” at least not without considerable qualification. Communication is a shared experience. As a general rule, some responsibility for miscommunication should be laid at the feet of both parties.

    The Bible is, of course, a rather unique document. In the context of Joel’s post, it is a means for God to communicate with his creatures. On the one hand, I don’t think we can say that God somehow failed in communication when we fail to understand it correctly. However, I think it’s accurate to bring the theological principle to the table that God is gracious. In the light of the great metamorphosis that God became human (the Word became flesh and took up a temple-like residence among us), I think it safe to assume that God goes the extra mile (or even three) in order to meet his creatures where they are. This grand graciousness brings our misunderstanding into graphic bas-relief. That is, if God’s communication condescends to our need, then our failure to understand clearly shows our problem is not in the understanding of the language. God was graciously clear. It has more to do with what we are–broken and sinful. It has little to do with what the language is.

    I hope you see the distinction–the distinction between what the text brings and what the language brings. If I clearly yell, “Stop!!” to a person who is starting to cross the road in front of unseen, speeding traffic, and that person believes me to be wrong, then the consequences for their actions sadly rest with that person. I did the best I could. However, if I yell, “Terminate!!” in the same context, they probably won’t even know I’m talking to them. At a minimum, I share the responsibility for their death. And, it’s an effort at deception–plain and simple–for me to claim the word ‘terminate’ means “to stop.” It’s deception because the conventions of the language dictate the appropriate words. One can’t blame the hearer for not understanding.

    In other words, the meaning is given to me, and the destination language is given to me. I have to choose the appropriate words as dictated by the conventions of the destination language. That language is given to the translator by the people for which the text is translated. The meaning is given to the translator by the original text in its original language. The translator’s task is to take that original meaning and form it with the destination language.

    Comment by Mike Sangrey | February 17, 2012

  13. Thanks Mike, I do see the distinction you describe. My bias has always been towards the importance of consistency in translation. That is, if the original text says that believers are sons of God, and, the original text says that Jesus is the son of God, the phrasing of these statements should be delivered consistently.

    On the other hand, I have no fundamental opposition to a translation that renders the literal phrase “son of man” to “human” or “person”, provided that the rendering is consistently applied wherever the phrase “son of man” occurs. Joel has me convinced that being literal is not always best.

    However, are we prepared for all the potential outcomes for not being consistent with the “words” of the original? We could tell someone a half-truth, with the view of telling them the full truth later on, but this could come as a huge shock. Is this the best way of teaching people what God really thinks?

    Comment by Robert Kan | February 17, 2012

  14. […] Bible Translation Blogger, Joel Hofman, weighs in on the subject; here and here. […]

    Pingback by Bible Translation Controversy: Background | February 18, 2012

  15. @ Robert Kan,
    I think that “consistency in translation” is not to be applied at the word level, as you suggest, but at what we might call the “context level.” So, for example, dikaios can mean either ‘righteous’ or ‘just’ and might be translated either way depending on context. Or huios ‘son’ might be translated into a particular language as ‘biological son’ in one context and as some metaphorical term for sonship in other contexts.

    I would look for consistency within each context, not consistency across all contexts. The point is to carry across the meaning in each case as determined by the original context, and as required by the receptor language for such a context.

    Comment by Scott Youngman | February 26, 2012

    • Yes, provided we can preserve the intertexuality so that any important connections between different passages using the same “original” word are not destroyed.

      Comment by Robert Kan | February 26, 2012

      • Yes, I agree fully that “context” goes beyond the immediate clause — even up to book and sometimes Biblical level.

        I highly recommend Chapter 10 of Translating the Word of God by Beekman and Callow (1984). They stress the goal of concordance in translation, but also explain legitimate reasons that a translation may not (or even should not) be fully concordant with the original. And they helpfully distinguish between true and pseudo concordance.

        There are dangers at both extremes: either not maintaining true concordance, or forcing pseudo concordance.

        Comment by Scott Youngman | February 27, 2012

  16. There are soooo many somewhat distinct uses of ‘intertextuality’ floating around, that someone will probably contradict me. However, I think we need to be careful to not confuse ‘concordance’ with ‘intertextuality’. Concordance is translating identical original words (and sometimes phrases) with identical destination words and phrases. Technical and key terms benefit immensely by concordance. Intertextuality is often CAUSED by single words, so it is caused at the lexical level. However, it is not the concordance itself.

    I think an excellent example of both terms, though a relatively complex example, is the temptation of Jesus. The concordance would need to be maintained in the New Testament quotes from the Hebrew Testament. Though a more apropos observation is the use of the word translated in many cases as ‘wilderness,’ the reference to Jesus being “led into the wilderness,” and the use of the term ‘forty.’ These are a complex translation choices because the translator is faced with translating from two different languages into one destination language.

    The intertextuality comes into play as one understands that the temptations of Jesus had key similarities to the temptations of Israel. The concordance should TRIGGER this connection. The intertextuality should cause the reader the understand that Jesus is undergoing the same temptation/trial as Israel went through. He passed the test; Israel did not.[1] Unfortunately, many exegetes seem to think Jesus is proof texting, even talking about something that I’ll describe as “Jesus and the Devil getting into a ‘my verse is better than your verse’ battle.” It’s not that at all–intertextuality should be providing the interpretive framework within which one understands the temptation of Jesus.

    Robert is right to ask questions about how the differing texts of “sons of God” and the Jesus referring text of “Son of God” are translated. There’s a concordance consideration there.

    I have not read broadly about the Arabic problem, I just understand the linguistic issues that come to the fore. But, I wonder if there has been discussion around this very thing. I’d be interested whether the association of sex with ‘son’ causes difficulty with the phrase ‘sons of God.’ Does such a translation simply not make sense? And, I’ll note, that “not making sense” is different than making the WRONG sense which many Arabic speakers seem to get from “Son of God.”

    [1] An intertextuality informed analysis gives quite a bit of insight into what Jesus was actually facing.

    Comment by Mike Sangrey | February 27, 2012

  17. […] Bible Translation Blogger, Joel Hofman, weighs in on the subject; here and here. […]

    Pingback by The Politics of Bible Translation « Thinking Out Loud | March 11, 2012

Sorry, the comment form is closed at this time.