Transcending origins: deconstruction and the value of knowing extra languages

In the Economist, I read tonight that Richard Weaver "wrote a book once called Ideas Have Consequences."

At first, I was annoyed by their use of the word "once." Is there some value in clarifying for us that he didn't write the book two, three, or sixty-one times? But the sentence felt wrong without it. Why? I started to think about the role that "once" played in this construction. Once (!) I arrived at the answer, I thought it would be neat to write about this episode, because it's actually a good, accessible story about the difference between a superior translator and a mediocre one.

One is tempted to think of once--which is derived from "one"--as fundamentally meaning "one time." Then, I thought about the Economist sentence in Japanese: The Japanese sentence that came to me automatically--based solely on intent and meaning, not words--was, "wrote a book quite some time ago." So, there it was. "Once" transcends its origin, and turns out sometimes to mean "long ago," with little/anything to do with one-ness.

A native English-speaker quickly realizes the truth of this proposition. "Once, my friend and I were on a trip," and so forth...

Where it gets more interesting, perhaps, is that "one time," a synonym of "once," turns out to convey the same meaning: "some time ago." For instance, "One time, I went to that store, and they were totally rude to me." (In fact, we almost never use "one time" to count occurrences." We do use "once" for that purpose. "One time" is used almost exclusively to place an event in the past.)

So, both "once" and its twin brother "one time" seem to transcend their origins in the counting world.

There are two morals to this story: First, a good translator doesn't translate words; he or she translates ideas. I know (pretty inept) Japanese translators who would have rendered that sentence as, "wrote a book [exactly] one time." Non-translators or poor translators think that translation is just word-for-word replacement. They don't get the fact that you need to understand the concept at hand, internalize it--separate it from its original words--and then choose new words for it from scratch in the target language. (I'm oversimplifying here, but the principle is important. It also explains why translation done by machines is still so shitty: they're basically word-for-worders, not internalizers.)

The other moral is that a second language can also be a useful tool for a writer--a prism through which he can peer at his native tongue. Kind of like pouring a mixed substance back and forth between two glasses to separate it out: idea expressed in English...then Japanese...then back to English.