It wasn't IME ESP exactly, but...

by Michael S. Kaplan, published on 2011/01/31 07:01 -05:00, original URI: http://blogs.msdn.com/b/michkap/archive/2011/01/31/10122135.aspx


I was flipping through Carl Sagan's Contact the other day, and ran across a part I must have seen before but don't recall:

The austere marble lobby displayed, perhaps incongruously, a real statue--not a holograph--of a nude woman in the style of Praxiteles. they ascended in an Otis-Hitachi elevator, in which the second language was English rather than braille, and she found herself ushered through a large barn of a room in which people were huddled over word processors. A word would be typed in Hiragana, the fifty-one letter Japanese phonetic alphabet, and on the screen would appear the corresponding Chinese ideogram in Kanji. There were hundreds of thousands of such ideograms, or characters, stored in the computer memories, although only three or four thousand were required to read a newspaper. Because many characters of entirely different meanings were expressed by the same spoken word, all possible translations into Kanji were printed out, in order of probability. The word processor had a contextual subroutine in which the candidate characters were also queued according to the computer's estimate of the intended meaning. It was rarely wrong. In a language which until recently never had a typewriter, the word processor was working a communications revolution not fully admired by traditionalists.

Now the book first came out in 1985.

But it was largely imagined starting back in 1980, at that time thinking about a movie.

That movie that didn't happen and eventually became the book a few years later.

And then, it became a movie 12 years after that.

The term IME or Input Method Editor is mostly regarded as having been originally coined by/for Windows, though as far as I know everyone uses it now. It just struck me that this fictional description that predates even the most primitive incarnations of the technology is probably better,more accurate, and to be perfectly frank more engaging than just about every description of an IME I have ever seen in my life, and written in a way that presumed very little prior knowledge of Japanese or for that matter Chinese in doing so.

That's pretty impressive.

Of course perhaps he was just describing a demonstration of something that he had been shown which, even if not yet on computers and not yet as sophisticated as modern IMEs are, was working well enough that the future possibilities were a part of the demonstration. But the words are still immortalized in this tiny corner of a book that was about something else entirely (though it was still along the lines of one of the book's common themes).

Most of that theme was lost in the movie that came out over a decade later, as did several other themes of language and culture that lent a lot of depth to the book. And of course they didn't have this scene either (I watched the movie last night to make sure they didn't slip it in passively, because even in 1997 it might have seemed pretty amazing to people, though only a text description could have done it real justice). Most of these themes were replaced by the throwaway "awful waste of space" imagery that just made me groan.

I've never met an intelligent woman who would fall for that kind of a line the way Jodie Foster's Ellie Arroway did, but maybe I'm not meeting the right kind of intelligent woman yet.

But I digress.

Anyway...no worries, though. The book is still there....


John Cowan on 31 Jan 2011 12:21 PM:

Sorry to burst the bubble, but the Xerox Star, the first commercial window/icon/menu/mouse workstation, was first publicly sold in 1981, and it had built-in Japanese IMEs (though not so called) one using hiragana and one using romaji, as well as the first interactive bidi implementation for Arabic.  What's more, the *Scientific American* article "Multilingual Word Processing" by Joe Becker described the interface in detail, and was published in the July 1984 issue.  That's a bit late, but Sagan very likely either saw a Star demo or had one described to him: I myself played with Arabic typing on the Star around the same time as the article.  

The numbers are a bit inflated: the Star used the Xerox Character Code Standard (XCCS), an enhanced version of JIS X 0201 with the other scripts stuffed into the unused areas.  As a result, there were on the order of 10,000 rather than 100,000 available kanji.  Later editions of the Star's software could handle Chinese or Japanese but not both, as neither Unihan nor the astral planes were yet thought of.  Indeed, it was the decision to break from rigid backward codepoint compatibility with JIS X 0201 that was the first step taken in the creation of Unicode, many of whose characters were taken directly from XCCS, though with new codepoints.

Michael S. Kaplan on 31 Jan 2011 12:43 PM:

I have seen that demo before, and there is certainly a bit of future vision being applied here (plus none of the text descriptions were as good, in my opinion!).

But point taken. :-)

mpz on 31 Jan 2011 11:17 PM:

Back in the DOS days, IME used to be called the FEP (Front End Processor). There were FEPs for MS-DOS as far back as in 1984.

Michael S. Kaplan on 1 Feb 2011 5:07 AM:

By comparion, FEPs were *much* more primitive, and the comparative typographical experience was also not very good (something people really noticed). Even today the CCJK experience is not great....


go to newer or older post, or back to index or month or day