On Sat, 3 Oct 2009 02:52:41 GMT, [log in to unmask] <[log in to unmask]>
wrote:

>> To me the presence of /&/ (ash) in a system like this
>> suggests English influence as well; the English [&] is
>> kinda characteristic.
>
>I probably mistook the sound for Greek because it is in English
pronunciation of Greek words. Although I can see why it might mark the
conlang's creator as English and possibly not very linguistically savvy, I'm
still not understanding how or why it might be better to exclude the sound.

If not for that reason, there is the matter of the vowel space being crowded
in the low front region, and such things do tend to be unstable.  Given /a &
E e/ in a seven-vowel system one might expect a merger -- /E e/ as someone
else said is likely for English speakers, but any of the adjacent pairs
might -- or if not that some shifts: one possibility is /a/ > /O/, /&/ >
/a/, giving the more balanced /i e E a O o u/.  

>> consult Greenberg's universals to make sure you have
>> something that functions within normal natlang parameters.
>
>Aha! OK, thank you.

If universals are what you're after, the Konstanz univerals archive
  http://typo.uni-konstanz.de/archive/intro/
has loads of 'em.  

>> > I don't know, because the language treats progressive
>> > as a tense. The tendency of the language to pile parts
>> > and pieces together to make words would probably
>> > encourage speakers to say "I goeding home, when..."
>>
>> So you can stack tense markers? What do you get when you
>> combine the past and future markers?
>
>I didn't say you could stack them, only that I imagine its character would
encourage speakers to stack things. I probably wouldn't myself, and, in
fact, I quail in terror upon contemplation of the horrible cosmic doom which
could descend upon the unwary fools who dare combine past and future markers
as did the Mad Arab Al Azif.

Right, given that stacking tense markers doesn't happen, the solution I'd
expect for the past progressive (if it's ever distinguished at all) is some
kind of auxiliary construction.  One obvious construction for "I was going
home" is
  I home.LAT go.SUBORD do-continuously.PAST
where ".SUBORD" is whatever form subordinated (coordinated?) verbs like this
would take in your language (possibly even a finite form?). 
"Do-continuously" is a placeholder and kind of a clunky gloss, but perhaps
if you're going for logicality you might just want to have a verb of that
meaning; natural languages that use this strategy TTBOMK tend to make their
continuous forms from verbs of location ('be at') or body position ('sit',
'stand') or some light verb (English of course just uses its copula 'be', "I
_was_ going").  

>> A clitic is like the English possessive 's, which
>> attaches phonologically like an affix but modifies
>> a phrase and attaches to the first or last word in
>> that phrase regardless of what sort of word it is.
>
>I don't have any clitics, or at least not yet. Are they very common
features of natural languages? What advantages do they offer?

Yeah, they're pretty common (although IME deployment of the word "clitic" is
more common with some types of markers than with others, on ?historical
grounds).  Most affixes ultimately come from independent words
grammaticalising onto other words, and most of these pass through the clitic
stage between being independent words and simple affixes.  

From an engelangy perspective I don't really see any advantages they offer
that one of affixes and independent words don't offer as well.  

Alex