Norms and accuracy

I’ve been absent for quite a while here. I’m on sabbatical and trying to finish not one, but two, books. Since I last posted I’ve been to India on a spur of the moment trip.

One of my wife’s work colleagues got married and invited us to the wedding. So we went. I have for years been telling my wife that I wanted to visit India some day. Seeing the Taj Mahal was on my bucket list. So she called my bluff.

Anyway, we also got to visit some longtime friends in Tamil Nadu as well. Mary met them in Ethiopa in the 60’s. They are a family of polyglots. They speak excellent English in addition to their native Tamil. From their years in Ethiopia, they speak Amharic. Their son, who has had a most interesting work history, went back to Africa to work so he speaks Swahili as well as Hindi and all the major Dravidian languages (Kannada, Malayalam, and Telegu). This is an example of something we linguists say over and over. Much of the world is multilingual. People who speak only one language are the exception, not the rule. And Vinod didn’t learn his languages by studying them in school for years. Rather he picked them up mostly in the context of living and working in places where he needed to have them. Needless to say, he thinks about language and translation in a very different way from you and me. To him language is the tool you use to communicate with.

That’s a position I’ve been arguing for in this blog for years.

If you think it’s the words of the original that are important and that wording must be preserved up to the limits of intelligibility, then you have to be willing to distort the meaning because no two languages work the same way — even if they are closely related.

Let’s look at a subtle example where Koine and English match in categorial distinctions but the where the norms of usage are different, and see what the distortion of meaning is.

The words in question are man, woman, and person on the English side and ἀνήρ, γυνή, and ἄνθρωπος on the Koine side. The categories match.

man = [adult male human]
ἀνήρ = [adult male human]

woman = [adult female human]
γυνή = [adult female human]

person = [human being]
ἄνθρωπος = [human being]

The difference I want to focus on is a subtle one.

In English one normally includes the gender of the referent unless there is reason not to. As a result man is about four times as frequent as person in running text, and woman is about three times as frequent.

But in Koine, it’s the other way around. You don’t use the gender based term unless there’s a reason to.  So ἄνθρωπος is a little more than twice as common as ἀνήρ, and in one in eight of those cases, ἀνήρ means ‘husband’, not ‘man’. The patterns are similar for woman.  Ἄνθρωπος is a more than four times as common as γυνή, and in half of the cases, γυνή means ‘wife’, not ‘woman’.

If you saw a man standing on the corner and you say (1), it is not just a simple report. You imply something more.

(1)  I saw a person standing on the corner.

Because (2) is what we normally say, unless there’s a reason to withhold the gender of the referent.

(2)  I saw a man standing on the corner.

For that reason alone, translations that try to push the gender neutrality of  ἄνθρωπος often sound odd in English. Here are some examples.

καὶ γὰρ ἐγὼ ἄνθρωπός εἰμι ὑπὸ ἐξουσίαν, … (Matt. 8:9a)
The Source: I, too, am a person under authority, …
Stylistically better: I, too, am a man under authority, …

Ἄνθρωπός τις ἦν πλούσιος ὃς εἶχεν οἰκονόμον, καὶ οὗτος διεβλήθη αὐτῷ ὡς διασκορπίζων τὰ ὑπάρχοντα αὐτοῦ. (Luke 16:1)
The Source: There was a certain rich person whose manager was accused of wasting money.
Stylistically better: There was a certain rich man whose manager was accused of wasting money.

Δεῦτε ἴδετε ἄνθρωπον ὃς εἶπέ μοι πάντα ὅσα ἐποίησα· … (John 4:29)
The Source: Come see a person who told me everything I ever did!
Stylistically better: Come see a man who told me everything I ever did!

In these cases ἄνθρωπος is best translated man. That choice isn’t driven by sexism, but by the norms of English usage.

But then that knife cuts two ways.

There are places where translations in the King James line say man (for ἄνθρωπος) where person, someone, or human or some kind of indefinite is a more accurate translation, both referentially and stylistically.

τί γὰρ ὠφελεῖται ἄνθρωπος κερδήσας τὸν κόσμον ὅλον ἑαυτὸν δὲ ἀπολέσας ἢ ζημιωθείς; (Lk 9:25)
ESV:For what does it profit a man if he gains the whole world and loses or forfeits himself?
Stylistically better: What will you gain, if you own the whole world but ruin yourself or waste your life?

Of course, we’ve heard enough sermons now to know what Matt. 15:9 means, but apart from Biblish (or in fixed phrases) we don’t use a nominal construction with man when we mean to highlight humanness.

Human nature does not mean the same as the nature of man.

This is especially when we want to highlight the distinction between human and divine.

To err is of man, to forgive is of God.


8 Ὁ λαὸς οὗτος τοῖς χείλεσίν με τιμᾷ,
ἡ δὲ καρδία αὐτῶν πόρρω ἀπέχει ἀπ’ ἐμοῦ·
μάτην δὲ σέβονταί με,
διδάσκοντες διδασκαλίας ἐντάλματα ἀνθρώπων. (Matt. 15:8-9)
ESV: 8 “‘This people honors me with their lips,
but their heart is far from me;
in vain do they worship me,
teaching as doctrines the commandments of men.’”
Stylistically better: “‘These people honor me with their words,
but their hearts are far from me;
9 their worship is useless,
they teach human commandments as doctrine.’”

None of our translations even begin to grapple with questions of style beyond the frequently cited argument that Biblish sounds like better English. (A point I roundly dispute.)

But, I fear, that by pointing out the stylistic problems associated with the use of gendered terms, I may have stepped into a mine field.

So let me affirm that I’m saying there are specific passages (and not a few of them) where using man makes for better English than using person or human (being). In asserting that I’m responding to stylistic concerns, not pushing an anti-feminist or complementarian agenda. I’m certainly not a feminist, but I am egalitarian. So if I say there are places where ἄνθρωπος is better glossed man, it’s not because I have a theological ax to grind.

(WARNING: If the comments start to wander off into a debate, theological or otherwise, about men’s and women’s roles, I’ll moderate with a heavy hand.)

8 thoughts on “Norms and accuracy

  1. dkmt says:

    Rich, I think you’re right that, in English, if you say “I saw a person standing on the corner” I assume that you weren’t close enough to them to be able to see if it was a man or a woman (or that it was hazy or raining or something). If it clearly refers to a man, then it’s normal in English to say “man” not “person.”

    I note that Matt 8:9 refers to a man, so it’s not incorrect to say “Man”, but stylistically, I think here it’s more natural to leave out this word altogether and say simply “I am under authority”.

    I agree with your translation in Luke 16 if it can be shown that the person referred to in the story was male (and I assume that since kurios later on in the passage is masculine, that the master was in fact a male? (Can you have a female lord, kuria? I assume so?)

  2. Rich Rhodes says:

    The point was not to exhaust all the options or settle on the absolute arguably best. The point was to raise the question.

    BTW, Donna, I think you’re right about Mat. 8:9. It is more natural your way.

    One of the problems with translation — something that professional literary translators are well aware of — is that translations, even the very best ones, all come out stylistically flat even when the original may have several “voices”.

    Consider Huckleberry Finn. The narrator and the characters have quite distinct voices.

    One of the world’s leading classicists, Anna Morpurgo-Davies, once told me that Homer’s characters all speak the appropriate dialects, something you’d never know from any of the translations.

    I just want to get to the point of being able to talk about how to make Matthew, Mark, Luke, John, Paul, and the writer of Hebrews sound as different in English as they do in Greek. And step one is to talk about subtle differences in usage. In this case, what linguists call markedness.

  3. John Hobbins says:

    Hi Rich,

    Welcome back to BBB! It’s great to have your shiny two denarii. Here are mine, worn and ferruginous though they are.

    I like your last comment better than your post. You say:

    “I just want to get to the point of being able to talk about how to make Matthew, Mark, Luke, John, Paul, and the writer of Hebrews sound as different in English as they do in Greek. And step one is to talk about subtle differences in usage. In this case, what linguists call markedness.”

    Fine. So here is a feature of the usage of Matthew. He likes to conform to Biblish diction and cite in Biblish for the sake of his LXX literate readership. He also likes to calque non-Biblish phraseology, like “kingdom of heaven,” phraseology Jesus shared with the Pharisees. Hence the relatively large number of Hebrew and Jewish Aramaic > Jewish Hellenistic Greek turns of phrase found in Matthew.

    In Matthew 15, that’s why we find the wording we do, per LXX Isaiah 29:13, for example: ἐντάλματα ἀνθρώπων (not an exact calque of the Hebrew; the collective singular “command” is pluralized).

    How then to translate? It seems best to calque and paraphrase at the same time in both cases:

    Isa 29:13:

    “its worship of me has become a commandment of men learned by rote.” So NJPSV (slightly modified).

    Matt 15:19

    “in vain do they worship me, teaching as doctrine the commandments of men.” So ESV (slightly modified).

    I’m not sure there is a better way to capture the Biblishness and Rabbinicness of the style of Matthew than to preserve agreement in matters of detail across registers Matthew wished to conjoin: biblical Greek understood within, and acting as an indictment of, Pharisaic Judaism (a love-hate relationship with the Pharisees is unlikely to be an innovation of Matthew, BTW; it is attributable to Jesus himself). In the case at hand, “commandment(s) of men” is the salient phrase. One rephrases it as considerable peril (though Mark does, in a different context: “traditions of men”).

    Truth to be told, most NT authors demonstrate a strong commitment to the Biblish of their day.

    If you naturalize the Biblishly language too much, you remove a component of its markedness.

    To calque or not to calque: that is the question. An exact calque is not necessary; an approximate probably is – in both passages. RSV=ESV, as often, has a great deal in its favor.

    The chief error of RSV=ESV is not getting the tense of Isa 29:13 right: the narrative past tense conveys an essential semantic feature.

    HCSB cannot be recommended across the passages, with its Matthew still speaking in Biblish and its Isaiah speaking in more natural English.

    Given the above discussion, NIV 1984 is superior to NIV 2011. The latter also truncates the text of Matthew; “the teachings they teach” would have been appropriate.

    If you have an alternative method of retaining the stylistic choices of the original to the one suggested here, a method which retains agreement across the texts, with Matthew quoting the KJV of his day to the letter – I would love to hear it.

  4. Rich Rhodes says:

    I’ve been arguing for years that the reason that you can’t translate much of the NT to Biblish is because authors used Biblish Greek as part of their communicative repetoire. You and I probably disagree most about how much of the NT is in Biblish, not whether there is Biblish Greek that should be translated into Biblish English.

  5. Rich Rhodes says:

    A further point about Biblish Greek. I think I can make a good argument that LXX era Greek sounded to Roman era Koine speakers like Biblish. So the place you encounter Biblish in the NT is not book by book, but primarily in the LXX quotes. So, John, I beg to differ, Matt. 8:9 isn’t Koine Biblish, it’s ordinary talk.

    I guess we differ on whether individual writers leaned to Biblish in general. I think not, although, Luke, of all people, uses what is likely a Biblish construction when he starts off sections using ἐγένετο. But I haven’t figured out what to do about him because after starting sections with a tip of the hat to Biblish he lapses back into the Koine of the Roman era.

    What’s worse, when you start digging around in questions of style you come to realize that for the most part the authors of the NT weren’t particularly great writers qua writers. What they said was enormously important, but how they said it may leave the critics yawning. (This kind of criticism has been leveled at Tolkien, for example. And I think I agree. Compare Tolkien with Lewis.)

    One of the things you often hear translators, especially simultaneous interpreters, say is they have trouble not “cleaning up” what they are translating — making the translation sound better than the original. I think we’ve done that in spades in Bible translation, at least in NT translation.

  6. John Hobbins says:


    Here are some points of agreement and disagreement.

    You say:

    I think I can make a good argument that the LXX era Greek sounded to Roman era Koine speakers like Biblish.

    LXX Greek in particular, as opposed to LXX *era* Greek (say, Xenophon,) would have sounded to Roman era Koine speakers like Biblish. Quite simply, it was Biblish. That’s because LXX Greek is translation Greek of a generally wooden, stilted kind, like much of what we call Biblish in English.

    The classic discussion of the generally calque-like quality of LXX Greek is the monograph by James Barr entitled “The Typology of Literalism in Ancient Biblical Translations” (1979). If you haven’t read it yet, you will enjoy every minute of it. Since then a number of LXX scholars have gone so far as to suggest that LXX Greek is by and large “interlinear” Greek. That slavishly imitative of the lexical and syntactic (not pragmatic) details of the source text.

    You say:

    *What* [the NT authors] said was enormously important, but *how* they said may leave the critics yawning.

    Yes and no I think. The very fact that so much of the NT is written in unadorned Greek is of interest to literary critics. The impact of this fact on the history of Western literature has been enormous. The most famous argument in this sense is the second chapter of Erich Auerbach’s Mimesis.

    On the notion that the gospel of Matthew is “ordinary talk.” I’m fine with classifying it as such, so long as, by analogy, one classifies popular hymns and popular sermons in the English language more or less heavily indebted to KJV English (lexis more than taxis) on the one hand, and to “updates” of the same found most commonly in religious language, as “ordinary talk.”

    Because that is what we have in the gospel of Matthew, 15:1-20 for example: a text swarming with excerpts large and small of calquefying translation Greek with updates in Greek literate Pharisaic language the whole of which is framed in unadorned and unremarkable koine Greek of the Roman period current among Greek literate Roman Jews.

    I am still inclined to translate Matthew 15:9, and the biblical text on which it depends by way of the mediation of a less than imitative translation, in a literal-as-possible but free-as-necessary fashion. I know of no other way to retain agreement across text and subtext, with Matthew quoting the KJV of his day with little modification.

    I expand on this here:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s