out the meaning was correct; it was Jonah who went overboard. That was my first hint that the language allowed long-distance reflexives.
I told this story to K.P. Mohanan during one of my visits to Singapore, and Mo said something like, “Of course, what else would he say?” It seemed perfectly natural; just like Malayalam. But at that time I had not heard of long- distance reflexives in any closely related language, so I was not expecting it.
Now suppose that this had been a translation instead of a spontaneous story, and that the translator (following the English source text) had written: “Jonah told the sailor to throw him into the sea.” I believe that this would have been interpreted in Land Dayak as meaning that some third person was to be thrown overboard, which of course is incorrect. But if I had not accidentally learned this fact about reflexive binding in Land Dayak, I would never have thought to check it, because it looks like a perfectly accurate and natural translation. This is an example of a “blind error”, an error that no one would have caught without specific knowledge about that aspect of Land Dayak grammar.
Let me mention one final grammatical issue in translation. Malay/Indonesian shows a strong preference for the passive voice (specifically the di- passive) to encode main-line events in a narrative, especially where there is a series of actions by the same actor. Now in these contexts, it is clear that the actor is highly topical. In some functionalist approaches, such clauses cannot be analyzed as passives because the passive by definition is a construction that topicalizes patients. But syntactically the di- construction is clearly a passive: the patient has all the syntactic properties of a subject, and the agent has all the syntactic properties of an oblique argument. LFG takes both the syntactic relations and the pragmatic functions very seriously, but recognizes them as being distinct and logically independent of each other. Thus it is possible to ask, “What is the pragmatic function of the passive in language X” in a meaningful way, because the construction is not defined in terms of its pragmatic functions.
This is just one instance of a broader principle: languages can use the same syntactic constructions for quite different purposes. In a number of mainland Southeast Asian languages, the passive is used only for unfortunate events — the so-called ADVERSATIVE PASSIVE. In Biblical Hebrew, as in the Greek New Testament (probably due to Semitic influences), an agentless passive is often used as a way of describing God’s actions without using any name to refer to God. Clearly the functions of the passive in Hebrew, Vietnamese, Malay and English are quite different from each other. I will not spell out the details here, but with a bit of imagination you can see that a translation from any one of these languages to any of the others which mechanically preserves active for active and passive for passive can lead to confusing (and sometimes hilarious) results.