On 9/19/06, Patrick Littell <[log in to unmask]> wrote:
> Hi, Leigh.
> I speak a little Quechua, and the most difficult syntactic thing for
> me is probably placement of the evidential clitics -- the "migratory
> suffixes" that indicate whether you know the information by witness,
> hearsay, or conjecture.  I know the basic rule, but there are
> apparently complications when the topic marker -qa is introduced.
> For a learner more advanced than I, I've gathered that a more
> difficult problem is that certain suffix combinations have
> idiosyncratic meanings that wouldn't be deduced from their parts.
> (I've found the morphology to be pretty regular, but I'm a beginner
> and won't have come in contact with the more advanced stuff.)  This
> goes along with your goal "that guessing something new from what you
> already know will rarely work."

I have several online references for Quechua and Aymara (thank
goodness I can read Spanish, or I'd have a much harder time), though I
don't speak either. Do you have good resources for Quechua, online or
print (not that my library will have them, but it's worth asking

I'm thinking that the normal languages will be fairly regular, which
will help in making a deliberately irregular language difficult to
learn. My current idea is that they all came from a common empire-wide
language which fragmented when then the empire fell, and since then
many of them have creolized (is that a word?), which has eliminated a
lot of the irregularities. Is that practical?

> ----------
> Other difficult stuff: Search the archives for Suffixaufnahme and
> Suffixhäufung, both of which could add significant complication to a
> very suffixing language of the Andean sort.

Oh, cool. Search keys.

> I'm putting together a presentation in which one of the examples is
> Sumerian, so I can give an example of Suffixhäufung off the top of my
> head:
> é         shesh lugal-ak-ak-a
> house brother king-of-of-in
> "In the house of the brother of the king."
> The genitive suffix -ak (really a clitic, in my analysis) is
> "postponed" until the final word of the phrase... and when you have
> nested genitives they will all "stack up" on the end.
> Also, Eldin and I, and some others, had a discussion awhile back about
> the Kwak'wala (Kwakiutl) and Heiltsuk, both Northern Wakashan
> languages, in which case and possession aren't marked on the word they
> modify, but on the *previous* word of the sentence, due again to
> clitic phenomena.

That's weird. I'll have to look at that.

(snip more interesting suffixing examples...)

> Anyway, any combination of these phenomena would lead to some
> fantastically complicated grammar.  You could also search here for
> noun incorporation, which we discuss fairly frequently; it's not
> characteristic of Andean languages but appears in many neighboring
> regions.

I'd thought of compounding in general, especially to create compounds
that aren't always obvious from their components. I'm not too familiar
with noun incorporation. I'll have to look into that, too.