MorphemeAddict wrote:
> Gary, I grant all of that. I think my concern is simply that English doesn't
> deal with lots of things, and so a list based on English sentences is going
> to be lopsided towards the things that English thinks are important, and
> away from things like evidentials (how do you know what you're
> saying), attitudinals (how do you feel about what you're saying),
> formality/humbleness registers, etc.
> I think the list is a great idea, for a start. Perhaps sentences from other
> languages could be added to the list to make up for what I see as
> shortcomings in a single-language list.
> (As an aside, the list might make a good basis for ESL training.)

Choosing what evidentials to use is always tricky for me when 
translating into Tirelat (probably one of the reasons I stopped using it 
for a while). So I have to make certain assumptions about the speaker. A 
sentence as simple in English as "The sun shines" can be an interesting 
challenge. The idea is not that the sun is shining at any particular 
place or moment in time, but that it shines in general (as far as the 
ordinary human perspective is concerned).

A speaker who wants to be really pedantic might come up with something 
along the lines of "I have personally observed the sun shining, and I 
infer that the sun shines in general."

jtuletin sy kih, të jtulełan sy kih

Both phrases use the perfective aspect suffix -n; the first because it 
describes one or more events in the past, and the second because it 
refers to the sun shining generically, not to any particular instance of 
the sun shining. I briefly had an "aorist" (or gnomic) tense to refer to 
generic statements like this, but it never really fit in well in the 
Tirełat grammar, so I went back to the basic past/non-past distinction 
and use the non-past tense with the perfective aspect for generic 
statements.

The first phrase uses the direct-observation past-tense suffix -ti, 
while the second one uses the inferential non-past suffix -ła. For 
practical use, there's no need to be so explicit; you might just get by 
with "jtulełan sy kih". Although this could be understood as "The sun 
will shine" (non-past perfective is the usual way of expressing future 
tense), Tirełat speakers are more likely to use -žan (predictive nonpast 
perfective) for future events.

When I was translating this, my first thought was to use something like 
"jtuleden sy kih", with the non-past direct-observation suffix -de. But 
as I'm sitting here in the dark, I realized that a speaker might not 
currently be observing the sun shining at the time they are speaking. So 
I think the version with -ła is a better translation.

Maybe I should try a few of these; I might start getting a better feel 
for how to use the Tirelat verbs (that area in particular of Tirelat 
grammar is a bit rusty from the years I wasn't using it much).