> [mailto:[log in to unmask]] On Behalf Of steve rice > --- Risto Kupsala <[log in to unmask]> wrote: > Yes. For example, in addition to obviously non-Western > items of dress, food, etc. there are simply different > concepts. In Linu I made a point of having spare color > words in addition to the usual European > ones--blue-green, green-gray, etc. There are probably > a number of concepts that are common in some Asian > languages (even unrelated ones that have been in the > same area of diffusion) that a Western language > wouldn't have. I've address colors by going to a more neutral approach making words for the three main light colors red, green and blue. Then I also uses their inverse versions of magenta, yellow and cyan that are common in photography and printing along with names for white, black and grey. There are two ways to reference other colors. One is to blend them like "red+blue" = "purple", or you can just refer to something of the same color like "orange+color", or "sky+blue" I've found clothing isn't too tough. For Sasxsek I have "rob" (< "robe") which can indicate not just a robe, but also a women's "dress" or a "gown". "cemiz" for "shirt" but also for "blouse" or any other garment that is worn on the upper body which is something fairly common across cultures. > > So we could start off with a purely English/Western > > IAL with very > > limited vocabulary consisting of 118 words (as in > > Toki Pona) or even up > > to 850 words (as in Basic English). If the language > > was cool enough, it > > might attract a number of early adopters or even an > > entire > > self-supporting movement. Since the initial > > vocabulary would be so tiny, > > there would be plenty of room to complement it with > > non-Western words > > later. > > I'd suggest the usual 800-1000 range. That still > leaves a lot of room. What about Eur dio vs J kami, > for example? They are not the same thing, except in a > cheap phrasebook. The important thing would be a > grammar that could accommodate very diverse forms > (presumably no POS endings) and relegate problem > concepts (tense, plurality, etc.) to the lexicon. That's why I now like to start with Lojban's gismu as a semantic base for conlangs, especially auxlangs. The ~1300 concepts in the gismu list are fairly neutral, and they can easily be relexed (relabeled) to fit different morphologies and phonologies. I have found also that there are ways to reduce it to less numbers. For example, you may not need distinct words for "man" and "woman" when a word for "person" could be used with a sex marker.