In the Land of Invented Languages - Arika Okrent [82]
Fortunately, you can visit jboski, the online Lojban-to-English translator, and at least see if your Lojban sentence parses. If you've made any major errors, or left out a crucial structural element, you'll get an error message. If you managed to create a valid Lojban sentence, you will get something like this, the product of my first (after several tries) successfully parsed translation of “It is clear that there is no classification of the universe not being arbitrary and full of conjectures”:
I got a little thrill when my sentence returned this parse. It was the same thrill I would get during grad school, when, after a long night of beating my head against the keyboard trying to write a data-crunching program, a beautiful stream of output would finally pour down the screen like a light-dappled waterfall of celebratory champagne.
But was this parse cause for celebration? I wasn't so sure. I knew that I had composed a grammatical Lojban sentence, but I couldn't be certain it meant what I wanted it to mean. When I presented it to the Lojbanists at Logfest, I discovered that it didn't. I had actually said that all classifications of the universe were random and full of people who guess. Smadi means “x guesses y is true about subject z.” According to the syntax of my sentence, I was making a statement about the x argument—the guesser. What I wanted was they argument—the guess. I should have used sesmadi instead.
I didn't feel too bad, though. Lojbanists are always making this kind of mistake. They are always making all kinds of mistakes. I know this because on the message boards where Lojban is used, hardly a sentence goes by that is not questioned or corrected—often by the very person who wrote the sentence. In fact, the main topic of Lojban conversation is Lojban itself. When one heated exchange (in English) led a commenter to write “Go fuck yourself!” in Lojban, it turned into a lengthy discussion of why he hadn't said what he meant to say, and what the proper Lojban expression for the sentiment might be.
I didn't see much live conversation at Logfest, but I did see a little. It goes very, very slowly. It's like watching people do long division in their heads. Of course, the types of people who are attracted to Lojban are precisely the types who are good at doing long division in their heads. Almost everyone there had some kind of engineering or math background (except for one enthusiast who, being fifteen years old, couldn't properly be said to even have a background). For dedicated Lojbanists, only part of the difficulty of speaking Lojban comes from the mental effort involved in keeping track of functions and variables. The rest of the difficulty comes from having to hyper-vigilantly guard their Lojban against the influence of English.
The temptation is there, for example, to use the word gunka (work) in a sentence like “This phone doesn't work.” But gunka means “x labors/works on y with goal/objective z.” It doesn't cover the English sense of “work” meaning “to function.” It would likewise be inappropriate to use dizlo (low) to say you're feeling low, because dizlo only means low “as compared with baseline/standard height z.” The metaphorical extension of lowness to emotions doesn't hold in Lojban. There is a Lojban word for these kinds of mistakes—malglico (damned English!). Malglico is what happens when you let the assumptions of English creep into your Lojban.
And this must be avoided in Lojban, because to remain valid in a test of the Whorfian hypothesis, it must remain culturally neutral. In terms of vocabulary, this means that definitions should be unclouded by connotations and metaphorical extensions that may not be shared from culture to culture. In terms