On 2017-01-12, Robert South wrote (excerpted for brevity):

> You [I, Paul Bartlett] wrote:

> "[...]  Based on my
> observations, I tend to think that some language designers first choose
> a representational scheme according to some orthographic method and then
> conjure up a system of lexis squashed to fit, whether all presumably
> needful lexical items fit or not. To me, this is backwards, but it seems
> to be a common tactic. Work out a needful vocabulary scheme, and then
> *later* invent an orthography to fit, not the other way around."

>  Suppose your ideal base vocabulary, selected independently of
> "orthographic" limitations, has 2000 words.

Fair enough. There have been various proposals with a controlled 
vocabulary of a relatively fixed number of words (usually apart from 
inflectional forms and proper nouns).

> Now if you have a orthographically limited base vocabulary of 200 words,
> then as long as that limited base vocabulary is equipped to
> synthesize new words and large enough to synthesize all 2000 words of
> that ideal base vocabulary, then the orthographically limited
> base vocabulary is equivalent to the freely selected ideal base
> vocabulary.

This may be, but to be honest I am somewhat skeptical.

> It gets the benefit of having a smaller initial
> vocabular that matches the desired orthographic limitations, but it also
> gets the benefit of giving you the ideal base vocabulary,
> albeit in the form of longer words.

Where is the matter of judgment? In some instances, with an 
oligosynthetic vocabulary some words may become so long and so complex, 
so difficult to decode, that one might as well have individual words to 
be learned. For some individuals, this might be as efficient as decoding 
(or learning) conglomerations. Thus a minor vocabulary might not 
necessarily be the most efficient in actual practice.

> That's exactly what JARRAPUA has.  It can emulate any base vocabulary
> you like.


> The vocabulary list I provide, the several thousand
> compound words, demonstrates that JARRAPUA word synthesis can produce
> all the words of other proposed based vocabularies.

Well, different language divide up semantic space differently, so how 
can JARRAPUA deal with these divisions without major contortions? At 
least some conIALs have their own acknowledged divisions of semantic 
space which any potential learner just has to deal with.

> JARRAPUA isn't really oligosynthetic.

Say what??? It seems to me to be definitely oliogosynthetic, somewhat in 
the manner of Toki Pona, Sona, (the late and now mostly forgotten) 
Ygyde, aUI, and perhaps others. I would say that JARRAPUA is for sure 
oligosynthetic. I am not at all asserting, by any means whatever, that 
no such language can never succeed. I only remain to be convinced.

> It's a set of rules for
> generating vocabulary in the form of compound words made of the base
> vocabulary.  Given a compound word synthesized in JARRAPUA you might
> still have some doubt about what it means if you know just the
> base vocabulary and the rules.  But you have a good inkling, and when
> you learn what that compound word is intended to mean you see
> the sense of it and will have a very good chance of remembering it.

You might remember it, but will you understand what it means 
unambiguously in semantic space for active use in real time?

> Lingua Francas are subject to snowballing.  The more successful they are
> the more advantages there are to learning them, up to limits imposed by
> outside constraints such as the finite number of people who
> want to communicate with foreigners at all.  If Esperanto was going to
> do it, then it would have done it.  Instead it's English that's
> snowballing, and English is so flawed.
> Oh well.

Yes, English is the 400-kilo gorilla in the room. In actual fact today, 
in my estimation, any conIAL must confront the (figurative) juggernaut 
of English in the world today. Even as an educated native speaker of 
English I concede that English has a lot of characteristics that make it 
difficult for others (particularly adults), non-native speakers, to 
learn and use, but that is the way it is. Any conIAL (and I include 
Esperanto and Interlingua) must push against this tide.

Paul Bartlett