I love Japanese, but it’s a confusing language sometimes. The Japanese word hato is used to refer to both the horrible grey birds that steal your sandwiches and then poo on you, and the lovely white birds that used to fly out of the Olympic flames until the Koreans accidentally barbecued them.

So what does hato mean? Does it mean “pigeon” or does it mean “dove”? Well, neither, because to answer that question is to see the Japanese language through the categories of English, and different languages see the world in different categories. There isn’t a one-to-one mapping between words in different languages. When you start to learn a language you might ask “what does that word mean?”, as if Japanese was actually a coded version of English and if you just make the right substitutions you can decode it; it’s a useful shorthand in the short term but it will mess you up in the long term. If you really want to learn a language well, you need to enter into the system of categories that the language uses, and that means you have to stop thinking in terms of “meanings” and start thinking in terms of “uses”. That’s why I said what hato is used to refer to. Words don’t have meanings, they have uses.

The other great example of this is colour terminology. There’s considerable evidence that people with different languages actually perceive colour differently. When you start learning Japanese you will be told that aoi means “blue” and midori means “green”. And then someone else who’s learning Japanese will tell you “Hey, did you know that the Japanese think that green traffic lights are blue, ha ha ha isn’t that stupid?” But of course they don’t. They don’t say that traffic lights are blue, because “blue” is English; they say that traffic lights are aoi. It’s only English speakers who say that traffic lights are blue. Aoi doesn’t really mean “blue”—because words don’t have meanings, they have uses. Aoi is used to refer to light with wavelengths of between roughly 400 and 500 nanometers, while midori is used for light between about 490 and 550nm. Traffic lights really are aoi, but it’s our broken system of translation-as-symbol-substitution that makes us think that Japanese think they’re blue.

We are learning JSL. One might say that JSL has five words for “to eat”—but one would be wrong, because “to eat” is English and JSL isn’t based on English at all. It actually has five words to refer to five different actions which in English we lazily lump together into the one word “eat”. They’re not the same thing. They are expressing different concepts, and while all those concepts would be rendered in English as “to eat”, to say “they all mean ‘to eat’ ” blurs all that conceptual difference, losing the nuance. How do you say “eat” in JSL? It depends entirely on what you’re eating and how.

Two lessons from this: First, translating out of a language with different conceptual categories requires you to resolve ambiguity, which is a judgement call. What do I mean? Well, knowing what you know about pigeons and doves, if you see the word hato in a text that you are translating into English, how would you translate it? Answer: It depends. It could be either, and before you write any English at all, you have to work out in your own mind which of the possible concepts you need to render. You have to make a call between the ugly grey thing and the pretty white thing. How would you make that call? You can’t just do it by looking at the word alone. You would need more information than just the individual word to resolve that ambiguity. You would need to look at the surrounding context to see if there’s any clues as to whether the bird is grey or white. If that doesn’t help, you may need to approach the topic from an angle; you might have to look at the text as a whole to see if there is some kind of emotional message the author is trying to convey, and translate accordingly. You might still not be able tell, in which case it’s a judgement call. Translation is not an exact science.

Second, translating into a language with different conceptual categories can unintentionally create ambiguity. In my Japanese Bible it says that the Spirit of God descended on Jesus at his baptism like a hato. I giggle every time I read that verse because I can’t help but think:

Well, it could be. I mean, that’s certainly one way to interpret the image portrayed by those words. It may not be the one you intended, though. If you want to make sure that people don’t think about grey birds that steal sandwiches, you have some choices at this point. You could use a circumlocation, explanation or footnotes to resolve the ambiguity in the right way. (“white hato”) Or you could think a bit more carefully about what the text is saying. In English we would say “like a dove”, so the point here is not the physical bird but all the social, cultural and emotional associations implied: images of peace, reconciliation, allusions to the Noah story, and so on. * The thing here is that pigeons are just grey doves and doves are just white pigeons, but because European languages and cultures have been heavily influenced by the Bible, the concept of “dove” now has emotional associations of peace and reconciliation precisely because of the Noah story; so there is some circularity there which is not present in non-European languages.

Enough of pigeons and doves, let’s think about ὁ υἱὸς τοῦ θεοῦ, which means… no, wait, it doesn’t, because words don’t have meanings, they have uses, and υἱὸς had lots of different uses: it was used to mark out a physical, biological son (Romans 9:9); to refer to an adopted son (Heb 12:8); in the wider sense of a descendant (Gal 3:7); symbolically to show a warm social relationship, not necessarily a biological one (1 Peter 5:3); symbolically to mark out people with a particular disposition (Eph 2:2); and so on. Note that four of those very different uses came from the same author! So before we translate υἱὸς τοῦ θεοῦ into any language at all, we have to make a judgement call to resolve that ambiguity, and doing so needs to take into account the social, cultural and emotional associations implied by the phrase.

Once we’ve resolved the ambiguity into a concept in our heads, we then have to render it into a target language, and that language will probably use a completely different set of categories again. In the same way that we want to avoid the image of ugly grey birds pooing on Jesus’ head, we probably also want to avoid the image of God having sex with a woman and bearing a physical, biological son—well, unless of course we’ve resolved the ambiguity and we think that’s exactly what the text means, I suppose; in which case, go right ahead—like I said, it’s a judgement call. Now in English we’re pretty lucky because most of the concepts included in the Greek word υἱὸς do map nicely to the English word “son”. It maps so well that get sloppy and we think that υἱὸς “means” “son”, when of course words don’t have meanings, they have usages.

In Japanese, the mapping is much less simple; we’d use musuko for a biological son, youshi for an adopted son, shison for descendant, ko for some of the symbolic and social uses, bouya for others—although the last two can also carry a nuance of youth and childhood. So we have a choice. We cannot rely on the circularity of English, where, just like Noah’s dove, the phrase “Son of God” has certain social, cultural and emotional associations precisely because we’ve always translated ὁ υἱὸς τοῦ θεοῦ as “the Son of God.” That would be nice, but without several centuries of Christian influence, that option is not available at present. We have not make not only a judgement call but a tradeoff. We’ll go for ko and hope that the “youth” usage of the word does not cause too many problems.

Other language slice and dice the semantic field differently, and we get different tradeoffs. We may have words for “son” that are only used to refer to biological offspring. Assuming that we don’t think that’s what ὁ υἱὸς τοῦ θεοῦ is about, then if we want to avoid the problems of a biological implication, we would need to choose another word which might well have some of the same symbolic usages as υἱὸς but none of the physical usages. Of course because it does not have the biological usages, we might end up choosing a word which when rendered back into English would be rendered by something other than the word “son.”

At which point everyone calls you a heretic.

ὁ υἱὸς τοῦ θεοῦ does not mean “the Son of God”, because words don’t have meanings, they have usages. Most of the time they have more than one usage, and translators have to really think hard to work out whether we’re talking pigeons or doves. And different languages have different usages, which means that rendering a phrase from one language to another is not a simple matter of decoding and re-encoding words, but involves judgement calls, trade-offs, and really, really understanding the usage patterns of both languages well.

If, on the other hand, you merely think in terms of words having meanings which can easily be transported from one language to another, then you end up saying stupid things like “Japanese people think that traffic lights are blue,” which of course they don’t, or “Wycliffe Bible Translators are watering down the Word of God to make it more acceptable to Muslims”, which of course they’re not.

---

*: For another important and even more extreme example, see Amaamalele Tofaeono Siolo, 2005, “Behold, the Pig of God: Mystery of Christ’s Sacrifice in the Context of Melanesia”, Pacific Journal of Theology, II:33)

(PS: I’ve just come across an email - sent a while ago to an account I rarely check - which said that if I comment publicly on this issue I need to point out that I’m only speaking personally and not representing my organisation. I’m sure most readers are intelligent and have come across the concept of personal blogs before, but to fulfill all bureaucracy, there it is: these opinions are mine alone.)