--- On Sun, 11/2/08, Olivier Simon <[log in to unmask]> wrote:

> That's the problem with having a regular orthograph and
> a predictable 
> grammar: Some forms may look daunting (though,
> "dauntingness" is a matter 
> of relativity too; "coughing" or
> "swallowed" look daunting for a speaker of 
> French). 

Another argument against using English.

> The major sources of vocabulary for sambahsa, besides the
> IE core, are 
> Romance, other European families and
> "Muslim"languages  (Sinitic words are 
> less common, such as "nowngmin" or
> "gwirlay"). The broader the distribution, 
> the more forbidding will be the outlook of the auxlang. 

> It is surely easier to build a vocabulary out of a single
> existing language, but 
> the risk is that the auxlang gets too much influenced by
> the source language, 
> so that a code-switching may appear and blur the apparent
> primeval simplicity 
> of the auxlang. 

Also true. One of the problems of Interlingua is that it simplifies the learning and even usage task for people with the right background by assuming a lot of information. If you're familiar with a Romance language in particular, it works really well. Otherwise not. And it also is constantly being pulled toward Romance expectations.

But consider Occ as a counterexample: because it has a strong Germanic component, especially in the grammar, it seems to resist such tendencies, yet it is very readable. The lesson seems to be that introducing elements that distance the project from its apparent lexical sources can keep those sources from taking over. This can be as simple as spelling (using <k> for [k]) or tweaking some obvious point of grammar/syntax (Occ is an adj-noun language, and it tends to put object pronouns after verbs).

This also argues for having a basic vocabulary, because then the words can be defined properly and their true meanings be absorbed by repetition. In fact, a limited basic vocabulary practically guarantees distance from lexical sources, because most "natural" languages have overspecified vocabularies.
> While German follows the same model as English
> "spielende Kinder" = "playing 
> children"
> Luxembourg Frankish acts the same as French: "Kanner,
> déi spillen" = "des 
> enfants qui jouent". 

Actually, both constructions are fairly common in English.

> But, some may say "spillend Kanner" by imitation
> of German, though "-end" 
> normally does not exist in Luxembourg Frankish. 
> Such inconveniences may occur with an auxlang too much
> linked to a natlang. 

Again, it happens in Ia and seems to be happening in LFN. But both languages copy Romance a little too much with noun-adjective ordering and other details. I admit I'm surprised that LFN has overcome its creolish appearance to drift in a Romance direction. If only they had used <k>...