Once it became the interlanguage, it start evolving very rapidly as
new technical terms were added. It has a huge vocabulary compared
with many other languages. That helps keep it as the interlanguage.
English's rich vocabulary has more to due with its invasions in the
Middle Ages than changes since the 19th century, although the lack of a
central authority that tries to control the language probably helps
absorption of new terminology more readily. Technical terms tend to be
exchanged between languages largely as direct transliterations, so they
don't really count for a rich vocabulary.
It also likes to be specific. Not leaving much up to interpretation
is great for science and business but not so great for Zen poetry.
Yet Lobjan never caught on for some reason.
One of the things I think about every once in a while is what would be
the characteristics of a ideal interlanguage. I explored Esperanto. It
has had lots of time to catch on, but has not. I explore why in my
essay on it.
http://mindprod.com/esperanto/esperanto.html
Looking at your essay on it, there are several inaccuracies:
1. Chinese characters are not mutually intelligible between Japanese,
Korean, and Chinese, or even the various dialects of Chinese (Cantonese
and Mandarin in particular). A better relationship is like the Greek and
Latin alphabets: the Latin alphabet directly descended from the Greek
one, and you can sometimes change the two sets and not notice, but there
are definitely cases where you can't. This is why the CJK Han
unification in Unicode was very controversial.
2. Discussing "words" in the context of multiple languages is
inaccurate, so comparing wpm of Chinese IME versus English isn't useful,
especially when you claim later that you could type Chinese characters
via their English equivalents for an IME.
3. The point about radicals letting you sometime infer meaning is like
the claim that sign language is intuitive because it's visual: it works
rather less well than most supporters claim when you go and measure it
(confirmation bias).
4. English is also moderately agglutinative. Consider the word
"antidisestablishmentarianism" and how many roots are in that word. Or
more useful words in modern discourse like nanotechnology.
5. Your point about "unifying Asian languages" flies rather greatly in
the face of what I know about East Asian history and smacks more of
Chinese propaganda than truth (see my point about number 1).
6. You imply a ranking of languages by total fluent speakers but give a
listing of what appears to be languages by native, first-language speakers.
7. It is probably not a matter of time until the US is predominantly
Spanish speaking. The most recent demographic trends, for example, show
greater increase in Asian-descent populations than Hispanic.
I'm too tired to give a full explanation of why I think Esperanto
failed, but the salient points I believe are the following:
1. It's a constructed language, so every fault that it has is less
excusable than native languages. These faults include, but are not
limited to, grammatical gender, agreement, phonological complexity,
orthographic strictures, inflection-versus-agglutination, and choice of
roots for word.
2. Internal warfare about degree of reform of the language (cf., Ido
reforms).
3. The general decline of Western Europe in the early 20th century,
relative to the United States and the USSR changed the position of
languages in Europe from being a family of roughly balanced powers (UK,
France, Germany) whose citizenry were apt to be polylingual to one of
two poles dominated by very powerful monolingual large countries (the US
and the USSR). Polylingual creoles just didn't have as much utility at
that point.
I am also interested in how you might use teams of computer
programmers who do not all speak English. I envisage some SCIDs that
greatly tighten up the ability of programmers to track what others are
doing and to decouple how the program is displayed from how it is
stored.
As someone who tries not to be monolingual, I still see programming as
being basically monolingually English, for the simple reason that there
needs to be a common language to specify, design, and implement APIs in.
That language is English for historical reasons. And before you suggest
machine translation, I will point out that APIs are places where
precision in language is necessary, and machine translation has yet to
be able to translate text (especially the kind of noisy text that you'd
find in programs and their API documentation) to anywhere near that kind
of precision and is unlikely to do so in the next quarter-century.
