Re: Most developed conlang

Jim Henry wrote:

> Here's an impractical method that seems theoretically valid to me:
>
> 1. Give 100 people who aren't speakers of the language in question
> a set of definitions of root morphemes.
>
> 2. Give them a list of compounds using those morphemes.  Ask
> them to guess what the compound words mean.
>
> 3. The opacity score of each compound word is =
> (100 - number of people who correctly guessed its meaning) / 100
Another thing along those lines you could do would be to ask these
non-speakers to produce forms that they haven't seen before. E.g., what
adjective would you use to describe a bird that has a green tail? Hint:
a hawk with a red tail is a "red-tailed" hawk, a deer with a white tail
is a "white-tailed" deer. Since all of these forms are regularly derived
from the same pattern, they don't all need to be separately listed. (One
problem is that words like "tailed" aren't always found as separate
words unattached to an adjective. So it makes sense to treat
"red-tailed" as a single word for some purposes. The Minza word
"kylřunřuchtsi" on the other hand is a more typical compound of two
words "kylřu" and "nřuchtsi", which can each occur as independent words).