Worldlangs and schematicism
We have noted many times that large scale word recognition doesn't work in
worldlangs. People learning worldlangs can expect to recognize only a
small portion of the words. There are several reasons for this: Worldlangs
incorporate divergent source languages, for example Chinese, Arabic and
English, which have few words in common. Worldlangs use new spelling.
Despite of Olivier's brave attempt, it's not sensible to incorporate
etymological spelling traditions and several writing systems.
Possibly because of these observations Steve Rice claimed once that
worldlangs are new a priori languages. Seemingly it is so. But current
worldlangs miss the schematicism of a priori languages, which is their
most useful feature. Nonschematic a priori languages are normally just
useless (but potentially fun) artlangs. So while worldlangs lose at-sight
readability of eurolangs, they don't compensate it with schematicism or
large scale compounding. (I dare to claim that worldlangs are less
compounding than Chinese.)
Example: In my esperantoid "kula" means "to open". This word is virtually
a priori, because most people don't have a clue where it comes from. They
just have to memorize it. Word with the opposite meaning is "to close".
Another virtually a priori word could be selected from some language, but
what's the use? Schematicism, such as reversing the root (kula <-> luka),
would be more efficient. It would add some predictability to otherwise
unpredictable word stock. Moreover, it would add predictability that is
independent of prior knowledge (thus bettering at-sight languages, which
always rely on prior knowledge).
I write this mainly as self-criticism, but hopefully this also raises some
discussion in this silent winter. I admit that some languages, such as
Sasxsek, are not touched by this critique.
- 11.2.2011 17:16, <deinx nxtxr> kirjoitti:
> On 2/10/11 12:56 PM, cafaristeir wrote:Actually toneless Chinese word can have more than four meanings because
>>> Monosyllabic words are not a problem. You can mix mono- and
>>> polysyllabic words, just like for example English mixes Germanic
>>> monosyllables and Latinate polysyllables. The tones are a problem,
>>> but I suppose that they would rather have toneless Chinese words
>>> than no Chinese words at all!
>> As you know, a toneless Chinese word can have up to 4 different
>> meanings. Better is to take polysyllabic Chinese words, less prone to
>> such confusions.
there are homophones with identical tones too.
> There's a big difference between *recognition* and mnemonics. In aThat's true. There are homophones and near homophones across languages.
> worldlang recognition is not practical nor realistic. The possibility
> of encountering false friends will ruin any attempt to make it
Normally in learning a foreign language you know whether familiar
sounding words are related to your language or not. (In the latter case
you can just ignore all accidental homophones.) But with worldlangs you
just can't be sure, at least not nearly as often as in case of
eurolangs, for example.
> It's also unfair to say you are taking *words* fromI agree with that.
> another language. We are not. These new languages have their own words
> with their own semantics. All we are doing is assigning labels to those
> semantics, and using something familiar as a memory aid. To most people
> starting to use a worldlang, the vocabulary will be unfamiliar but there
> will be an occasional resemblance to aid in memorization.