March is machine translation month


Well, five years ago I returned to the issue of machine translation, having addressed it five years earlier, when a firm offered on the fly translation for my website. So it is almost time for a reprise.1

In 2005 I tested the service I was offered and found it almost, but not quite, usable. The paragraph I set was chosen (almost at random) from the Hypertext Bible: Amos commentary, it concerned city gates.

In 2010 I tested Google Translate and found that it’s results for rendering my paragraph into French were not as good as the effort from five years previously, though mysteriously after rendering it again into from French German this time a back translation into English was actually better than the 2005 effort.

So how has Google improved its translation service over the last half decade?

Here are the results of translating into French side by side:

Dans la ville antique du proche Orient les portes n’étaient ni simplement des entrées, ni seulement utilisé pour la protection militaire. Comme point potentiellement faible dans les défenses, les portes d’Israelite ont muré des villes ont typiquement eu trois chambres donner quatre ensembles d'”portes” et d’espaces défendus entre. Les portes de Hazor (laissé – projetez ci-dessus) et de Gezer dès l’exposition de Solomon cette construction triple.Dans l’ancienne ville près des portes-Orient ne sont ni seulement les entrées, ni seulement utilisés pour la protection militaire. Comme un point de faiblesse potentielle dans la défense, aux portes des villes d’Israël paroi généralement avait trois chambres donnant quatre séries de “portes” et défendu les espaces entre les deux. Les portes de Hatsor (à gauche – plan ci-dessus) et Gezer du temps de Salomon montrer cette construction triple.Dans les portes de la ville antique du Proche-Orient ne étaient ni Entrées de simple, ni seulement utilisé pour la protection militaire. Comme un point potentiellement faible dans les défenses, les portes de villes fortifiées israélites avaient généralement trois chambres donnant quatre séries de «portes» et défendus espaces entre. Les portes de Hazor (à gauche ci-dessus ) – Plan et Gezer du temps de Salomon montrent cette triple construction.

Perversely, or more likely because of the complex interactions between translation decisions, that must (I assume) be made by machines as they are by human translators, Google has got both better and worse in the last five years. The opening phrase is now almost (though confusingly not in terms of syntax) as good as the ten year old effort. Some of the changes are bizzare, in 2010 Google cleverly gave Hazor its French spelling Hatsor, but in 2015 it has returned to its more usual English rendering. On the whole in 2015 Google is still less good than the other (2005) effort at syntax.

In 2010 I remarked that oddly though Google’s French was much less good than the 2005 the back translation from German was better. Here then are the 2015 results


In for the military protection antiken of the Near East the doors were neither simple from the entrances nor only use city. When potentially weak point in the defenses the doors have typically three chambers immured of Israelite of cities had to give four whole of “doors” and from areas to which were defended, between. The doors of Hazor (calmly – above you plan) and of Gezer of the exhibition of Solomon to this three-way construction.In the old city near the East doors are not only inputs or only military for protection. As a point of potential weakness in the defense at the gates of the walled cities of Israel usually had three rooms are four categories of “doors” and defended the gaps. The gates of Hazor (left – show on plan) and Gezer time of Solomon, that tripled the building.In the doors of the old city of the Middle East neither simple one- or just for the military protection were used . As a potential weakness in the defense of the doors of Jewish fortress towns usually had three rooms with four series of “gates” and defended spaces between . The gates of Hazor (top left) – Plan and Gezer Solomon’s show time for the triple structure .

So, five more years on Google’s back translation is much improved, but the translation is arguably not improved at all! How can we explain this? I think by noticing the strengths and weaknesses of Google’s approach. It is weak on syntax, often offering a pedestrian word for word translation, however it is good at spotting contextual cues. (As the French rendering of Hatsor in 2010 – mysteriously dropped in 2015, have more Francophone websites taken to using the more literal and less phonetic rendering in this period?) This combination is perfect for providing renderings in successive langauges that will not produce hilarious “mistakes” when back translated into the original language. That is the good news for the Google programmers, malicious reviewers (like me) will get little fodder from Google. BUT it is bad news for users, because what we actually need is not brilliant results from an artificial translate/back translate excercise (no matter how many or few intermediate languages we use), but rather a decent understandable translation. That goal is at least as far away in 2015 as it was in 2010 or 2005 :(

  1. Being impatient, and forgetful, I will not wait till the 30th or 31st, but will jump in today :) []

Camouflage Equivalence: another example

Back in April I somehow missed Bryan Bibb’s interesting post Camouflage Equivalence1 it focuses on places where translators:

…seek to obscure rather than reveal the meaning of the original. He [Robinson] defines the term as “rearranging the semantic elements of the original… in a plausible way that disguises their dynamic meaning” (p. 6).

The idea, like the term used to describe it is really helpful. It neatly describes those places where translators soften the offense inherent in Scripture. The NIV regularly does this when a more “literal” translation leads to theological difficulties. One example is the rendering of ha’almah in Is 7:14 as “virgin”. Whether ‘almah can carry this meaning is at least debatable. As far as I can see the logic of Isaiah’s speech however demands a present focus and a translation like “young woman”. NIV has exercised camouflage equivalence.

I am less convinced by Bryan’s example. He claims that the ambiguous language (full of sexual double entendres) in Ruth 3 contains at least one such camouflage equivalence translation in almost all English Bibles. “Uncover his feet” in Ruth 3:4 is (Bryan thinks evidently, I’d say possibly) a euphemism. While most translations diminish the sexual tension in Ruth 3, where there are a string of words and phrases like this one that might carry sexual connotations, sometimes a foot is just a foot! The whole point (I think) of using that concatenation of ambiguous, possibly sexual, terms in Ruth 3 is surely to remain ambiguous. To uncover what the text deliberately leaves veiled but suggested is as “bad” as to cover what the text reveals…. So, “uncover his feet” (NIV, NAS, NRSV) gets it right neither camouflage, nor sex for the sake of shocking the horses, but a good serviceable translation.

On the other hand in Psalm 90:2 common translations are split, some opt for camouflage equivalence:

NET Psalm 90:2 Even before the mountains came into existence, or you brought the world into being, you were the eternal God.
NRS Psalm 90:2 Before the mountains were brought forth, or ever you had formed the earth and the world, from everlasting to everlasting you are God.

While others dare to reveal the clear implication of the Hebrew:

NAS Psalm 90:2 Before the mountains were born, Or Thou didst give birth to the earth and the world, Even from everlasting to everlasting, Thou art God.
NIV Psalm 90:2 Before the mountains were born or you brought forth the whole world, from everlasting to everlasting you are God.

For both verbs yalad and hul speak of procreation and birthing, and though yalad might refer to the father’s role hul cannot, but clearly refers to birthing.

  1. I had also missed Douglas Robinson’s book, Translation and the Problem of Sway, from which he apparently got the fine phrase. []

How English-speakers misunderstand the gender of God

English uses natural gender, inanimate objects are neuter “it” while animals and humans are gendered along the lines of the individual’s sex (except for some dialects)1 where the sex of the individual is unknown a guess is made (with e.g. cats being often assumed female and dogs male) rather than the neuter used. This usage means that many English-speakers have difficulty understanding languages that use grammatical gender. Since Hebrew is such a language, with only two genders “masculine” and “feminine”, this creates problems for discussions of the gender of God.

English-speakers, based on how their language treats sexual identity, and surprised by the novel idea of this being attributed to inanimate objects,2 assume that, where language refers to people, “gender” in some way correlates with identity. And, of course, if we are talking correlations, it does. Men are usually “he” and women “she”. Except, not always. In French3 if one speaks of a woman (f)4 who is a government minister (m)5 then the grammatical gender one uses, and hence the pronouns used to refer to her, will depend on whether the reference is to her name:Marie would be referred to later as elle (she), or her function: le ministre will be referred to later as il (he).

In Hebrew the word for God ‘elohim is masculine, as is God’s name yhwh. But what does this imply about God’s sexual identity (or indeed gender in whatever non-grammatical sense)?

Tony Reinke in an article Our Mother Who Art In Heaven? on John Piper’s Desiring God website argues that this data implies that God is in some sense masculine.

His argument is based on work by John Cooper whose book Our Father in Heaven: Christian Faith and Inclusive Language for God, [amtap book:isbn=080102188X] whom he quotes:

Linguistically, all the clear and plausible instances of feminine reference to God are imagery or figures of speech: similes, analogies, metaphors, and personification. . . . there are no cases in which feminine terms are used as names, titles, or invocations of God, and thus there are no feminine pronouns for God. There are no instances where God is directly identified by a feminine term, even a metaphorical predicate noun. In other words, God is never directly said to be a mother, mistress, or female bird in the way he is said to be a father, king, judge, or shepherd.6

Notice that Cooper’s argument is not merely the crude misunderstanding I have outlined above. Though he seems to give weight to this misunderstanding :( “There are no instances where God is directly identified by a feminine term,” his more significant claim is that: “God is never directly said to be a mother, mistress, or female bird in the way he is said to be a father, king, judge, or shepherd.”

I challenged this argument in chapter 5 section 4 of my book Not Only a Father. There I engaged with Elizabeth Achtemeier as a strong representative of the claim that masculine and feminine imagery for God works in different ways. She had used more traditional language to express the argument, father language about God is metaphorical while mother language is merely comparison (simile).

Basically I have two problems with Achtemeier’s and Cooper’s argument:

  1. I am not convinced that the neat distinction of how metaphor and simile (direct identification and comparison)  operate in Biblical Hebrew can be sustained. When God is described as “being” a father or the rock (m) of our salvation (Ps 95:1) always only some aspects of rocks and of fathers are in view in any place. Just as is the case also when God is described as like a mother, or indeed as “being” the rock (f) of Israel (Gen 49:24).7
  2. I am however convinced that to call God father in ways which are significantly different from the ways one refers to “him” as mother is idolatry. Such talk (whether indulged in by Achtemeier, a biblical scholar, or Cooper, a philosophical theologian) makes God a member of one class of beings (male or masculine) and not a member of another (female or feminine). Such a partial8 god, one who is precisely a god and not a goddess, is not the God of Scripture.

For more on this see my book Not Only a Father.

  1. In some dialects of English, especially in the westcountry, inanimate objects have gender, a wardrobe is “he”, a fork “she” and so on. []
  2. With only two grammatical genders every noun must be “gendered” []
  3. Another language with two grammatical genders. []
  4. I’ll identify feminine words as (f) and masculine ones as (m). Since she is a woman this ascription of feminine gender to her does seem natural. []
  5. The word ministre is masculine. []
  6. John Cooper, Our Father in Heaven: Christian Faith and Inclusive Language for God (Baker, 1998), 89. []
  7. Incidentally, I must sometime check Cooper’s book to see how he deals with that last case… []
  8. Pun intended []

Jeremiah 4:23-27 (translation and notes)

We looked at Jeremiah 4:23-27 in class this week and I plan a podcast on the text over at 5 Minute Bible so, since Ill use my own very literal translation there I though I’d publish it here wirth a few notes to explain it.It is intended to be as near word for word as I could get and still be English. So the repetitions stand out, it is laid out to show the terse almost staccato feel. I have noted some of my translation choices with footnotes.
23I looked at the earth.

It’s higgledy piggledy.2
To heaven,
but no light there!
24I looked on the mountains.
They are quaking.
All the hills shake themselves.
25I looked.
No human,
and all the birds of heaven have fled.
26I looked.
The field’s a desert,
and all its cities are destroyed
before YHWH,
before the heat of his anger.
27For thus says YHWH:
All the land will be desolation.
But I will not make a full ending.
28Because of this the earth will mourn,
and the heavens will be dark above,
because I have spoken,
I have decided;
and I have not relented
nor will I turn back

Warning, I may update this post, adding notes, or even adjusting the translation. I did this one some years back and need to revisit it when I have time, my son did years ago name my translation the Temporary English Version ;)

  1. הִנֵּה hinneh “look!” can serve a number of functions. In old translations it was often rendered “Lo!” or “Behold!” The important part this construction plays in giving language a “biblical” flavour, illustrates its significance to Hebrew speech. 

    In narrative hinneh often marks a change in view-point:
    Ruth 2:4 where we are invited to “join” Ruth in watching Boaz’ arrival;
    Ruth 3:8, having followed Ruth to Boaz’ feet, we share his surprised awakening.

    It also serves other functions:
    affirmation (translated something like “indeed”) – Ruth 3:2 (where the “look” seems redundant in English);
    explanation “that is…” (which we would put in brackets) Am 7:1;
    call to attention (Ruth 1:15)
    marking events that happen contemporaneously – Ruth 4:1 where וְהִנֵּה suggests that, hardly has Boaz sat down, than the other Goel arrives. []

  2. תֹהוּ וָבֹהוּ tohu vabohu
    This phrase is found most notably in Gen 1:2 (also though split by other words in Is 34:11) translators have to choose a rendering which ideally captures:
    the sense of confusion – rendered in traditional English translations “formless and void” –
    and the echoing sound.

    Various proposals have been tried; Robert Alter’s literary “welter and waste” is good. I have opted in Jer 4:23 for the more homely “higgledy piggeldy”. []

Did Yahweh seduce Jeremiah?

In response to my podcast “The last Confession of Jeremiah: Jeremiah 20: Yahweh seduces his prophetDavid Haslam asked (on Facebook) about the choice of “seduce” here. He noted that most English translations have other words:

“persuaded/denounce” (ASV & WEB),
“deceived/report” (KJV),
“coerced/denounce” (NET)
“deceived/persecute” (DRC & NIV)

From that list you will see that פתה is not easy to translate, like most words, but more than many it carries a meaning that will in other languages be rendered in different ways according to the context. It does indeed suggest persuading, though often in the sense of deceiving, sometimes coercing. In the qal it has the sense of being simple, open minded, or deceived. Its first occurrence in the Bible (Gen 9:27) it just means “enlarge”.

The piel that we have in Jer 20:7 is used 17 times:

  • enlarge (Gen 9:27)
  • seduce virgin girl (Ex 22:15 v.16 in English)
  • coax, entice – of Delilah technique for getting information from her husband (Jud 14:15; 16:5) of tricking Ahab (1 Kgs 22:20,21,22 also 2 Chron 18:19,20,21) or of humans attempting to trick Yahweh as if he were a god (Ps 78:36)
  • deceive (2 Sam 3:25; Pr 24:28)
  • seduce (Hos 2:1) of Yhwh as husband persuading his wife (Israel) to return to him from her lovers
  • Pr 1:10 might be either coax/entice or deceive but Pr 16:29 suggests the use of force
  • Ez 14:9 is perhaps the closest usage at first sight, it involves someone deceiving  a prophet into inquiring of Yhwh on their behalf even though they are an idol worshipper, in which case Yhwh will do the same to the prophet, and even kill him.

So basically most of the usages involve persuading someone to do wrong, often by using sexual wiles. The question that remains concerns Jer 20:7. Does Yhwh here trick a gullible Jeremiah into doing wrong? or Does Yhwh here seduce Jeremiah? Clearly the sexual overtones here cannot be intended literally, but is this the picture being painted? I find it difficult to see Jeremiah in this case claiming that Yhwh has treated him like the prophet in Ezekiel, for Jeremiah is firm that he has spoken the truth. Rather, I suggest that he is claiming to be like an innocent girl (cf. his first confession 11:18ff.) whom Yhwh has persuaded to do as he wishes.

Because in Jer 20:7 that seems to me to be the choice we have: either Jeremiah accuses Yhwh of treating him like a prophet who takes payment from idolaters to give an oracle, or Jeremiah is claiming Yhwh charmed him into what he has done, like a girl seduced by a lover.

What do you think?

Articulated texts

Articulated trucks are easier to turn ;) photo by crabchick)

In this post I am NOT thinking of the clear or muffled ar-tic-u-lation that my speech teacher prized, but the other sort. And, teaching “Understanding and Interpreting the Bible” this week the topic of textual articulation came to the fore. First in trying to explain the nature and function of a “conjunction”  to students who have no understanding of grammar (not even those who attended secondary schools with “Grammar” proudly flaunted in their historic names).

Conjunctions, I said are the (often little) words that join and articulate text. They tell us how the parts work together. As such they are very important clues to what a text is doing.

They are. And all1 languages have them. But2 not all languages have them, or use them, equally. And3 they certainly do not use them in the same places. Different languages and different speakers articulate their texts differently.

For this week on spotting the workings of text at a local level, we studied 1 Tim 6:17-19. Most of our students do not learn Greek or Hebrew :(4 so we were working on an English text and with English grammar. 1 Timothy 6:19 provides a nice example:

thus storing up for themselves the treasure of a good foundation for the future, so that they may take hold of the life that really is life.
ἀποθησαυρίζοντας ἑαυτοῖς θεμέλιον καλὸν εἰς τὸ μέλλον, ἵνα ἐπιλάβωνται τῆς ὄντως ζωῆς.

Eduard von Grützner's Falstaff from Wikipedia

Actually the NIV makes the point more dramatically opening the verse “In this way” where the Greek just has a participle. Hebrew texts offer even more of these challenges, since the paratactic constructions favoured by the language use fewer written markers of syntax.

At which point I’ll call back my speech teacher, a grandiloquent old act-tor, for it is only by articulating a written text clearly that we can begin to understand it. For where written grammatical markers of syntax are lacking only clear articulation can “make sense” of the text.

  1. Or at least, all that I have studied so far. []
  2. Yes, I know this is the second time I have started a sentence with a conjunction :) I do hope all prescriptive grammarians are spinning like tops in their graves, or soon will be, since prescriptive grammar is unnecessary and unwanted. Well actually it is not, I need to know that starting sentences with conjunctions is “wrong” for my use of this construction to be chosen for effect, and not mere carelessness. So prescripts you may cease your rotations forthwith :) []
  3. Yes, a third! When you are on a roll it is hard to stop ;) []
  4. No, I don’t know how someone can be a serious Bible student without the languages, either. Though I note that only Greek was compulsory at Oxford, and that I failed to take Hebrew, to my shame. To Oxford’s shame I believe that even Greek is now not required for the Honours School of Theology :( []

The censored Bible: translating Psalm 90

Aristotle’s Feminist Subject has a post in which various translationsof Psalm 90 are compared. As always I’m astounded by the way most treat verse 2:

בְּטֶרֶם׀  הָרִים  יֻלָּדוּ
וַתְּחֹולֵל  אֶרֶץ  וְתֵבֵל וּמֵעֹולָם
עַד־עֹולָם  אַתָּה  אֵל׃

Before the mountains were born
or you gave birth to the earth and the world,
from everlasting to everlasting you are God.

It seems quite clear to me. I cannot see how else to render the words!

The nearest to this explicitly (I think) maternal imagery for the creation of our world (among the translations in front of me here) comes from the NASB:

Before the mountains were born
Or You gave birth to the earth and the world,
Even from everlasting to everlasting,
You are God.

though the NIV comes close:

Before the mountains were born
or you brought forth the earth and the world,
from everlasting to everlasting you are God.

But the rest fudge it. Why? (There is a fuller, though still aimed at non specialist readers version of my take on it in chapter two of my Not Only a Father. Since the format of that work invites, needs, discussion, please go there and discuss either this or one of the other things I say!)