[YG Conlang Archives] > [ceqli group] > messages [Date Index] [Thread Index] >

[Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: [ceqli] Re: new words

In a message dated 8/18/2007 7:54:24 AM Central Daylight Time, rmay@hidden.email writes:


Re discussion on another board, Lojbab said that Lojban has about 1350 roots.  Ceqli already
has more than that in the glossary, and I'm wondering if I'm maybe not taking advantage of
compounding enough.  True, I have a lot of proper names listed, but I want to end up with
something as clean and simple as possible.


Lojban has lots of gismu for filling out sets of words considered important for an international (non-aux) language, such as metric prefixes, metric units (basic only), lots of nationalities and languages and religions.  There has been a lot of discussion in recent years about how poorly Lojban gismu cover semantic space.  Of course, that's where lujvo come in, but still, there is no official gismu for Italian (unofficial is talno), along with a host of others (including 'rodent').  
Also keep in mind that many basic ideas are not embodied in gismu at all:  numbers, interrogatives, connectives are all cmavo.
So, even though Lojban has 1350 gismu, that isn't necessarily an optimum number of roots for a language to have.

My own Saweli (based on Rick Morneau's Latejami) has 3692 roots so far, divided into 188 semantic classes which can combine with 155 modifiers (along with a few ad hoc creations).  Several of those classes are small and closed, but most are open and not well fleshed out yet.  Each of them can conceivably contain thousands of roots, and each root may take suffixes and endings to change the meaning in specific ways.  188 classifiers + 155 modifiers + 100 suffixes/endings (at a guess) = 443 morphemes.  That's much smaller than Lojban's 1350 gismu and however cmavo there are.

Another example:  The Esperanto-Cxina Vortaro has an appendix called "Baza Radikaro Oficiala" (List of official basic roots).  Out of 10,000,000 words of text, they picked nine frequency groups of morphemes.  The sum total for the list is 2466 morphemes, which includes all the endings, suffixes, prefixes, prepositions, pronouns, conjunctions, as well as other roots.  
These combine to form many thousands of other words.  (The explanation of the list is in Chinese, which I can't read yet.)

It just depends on your language.  I suspect you're right that you're not taking advantage of compounding enough, but it took Esperanto a long time to come up with lots of its compounds.  

Morneau's classifier+modifier approach to building vocabulary has become a favorite of mine.  He discusses other topics in his monograph that most languages (nat and con) don't handle very well (modalities, in particular).  

http://www.eskimo.com/~ram/Latejami/

Good luck.

stevo