OT: What makes a good conlang? (was Re: Super OT: Re: CH AT: JRRT)

David P [in another thread]:
> Joe wrote:
> <<Actually, the 'tt' is [?] :p>>
> Not unless you're Eliza Doolittle (or "li'il").   It's pretty much a
> flap in English, I'd say.   Any takers on this issue? (Note: This is
> NOT YAEPT! It's important for scientific purposes.)
Are you trolling?! Joe is English. There are many different accents
of English, and only a few of them (North American, Aus, ?NZ) flap
/t/ in _little_.

> And. wrote:
> <<
[...]
> (I believe I practise what I preach here, and I delight in
> discovering congruences between English and my conlang. I also
> rely heavily on my English intuitions during the design process.)>>
[...]
> Now, what did you mean by "I also rely heavily on my English intuitions
> during the design process"? I'm not saying this is a bad/good thing: I
> just don't know what you mean.   Could you explain?
I can't remember quite what I had in mind when I was writing that.
There are many congruences with English, and influences from English.
Examples of influence from English intuitions: (1) the Livagian
notion of textual length is based on syllables; (2) discourse markers
(like "well", "though", "yet") are (or will be) derived from analysis
of English discourse markers, since they are more subtle than any
system I am capable of inventing out of thin air (and Livagian must
not be impoverished in comparison with English). Another example of
influence from natlang intuitions: conditions on Livagian
geminability seem to be the same as in Italian, but I haven't yet
worked out how to define those conditions.

--And.