Re: Irrealis mood and non-finite verbs

There's a couple different claims and questions here, so let me
see if I've understood everything right...

Eric:
<<
I am wondering if there is (cross-linguistically) a connection
between irrealis mood and non-finite verbs (or adjectives or nouns).
I know David Peterson calls a certain marker in Sidaan "irrealis/non-
finite marker", which seems to justify the vague impression I have
that they are related somehow. Maybe David could help me out with the
nuances of that marker in Sidaan (and in natlangs that may have
inspired it)?
 >>

The first claim is the connection between irrealis and non-finite
verbs (let's stick to verbs).  In order to make sense of this, we
need to define "irrealis".  *I* defined it for myself as something
that has not (or is not) actually happening.  This would include
something happening in the future (even if it will definitely happen),
something that might have happened, something one wants to
happen, something that would have happened, etc.  I lumped
all those together and labeled them "irrealis" for the purposes
of Sidaan.  Not all languages do the same, and I don't believe
this is exactly how the term is used by linguists.  If it's used within
linguistics, then that would be the definition to use if we want
to compare natural languages and what they do.

Off the top of my head, compare Spanish.  Take "I want to go":

quiero ir
/want-1.sg.pres. go-INF./

One might just have easily said:

quiero voy
/want-1sg.pres. go-1sg.pres./

That's ungrammatical Spanish, but if you just want to talk about
conveying the notion, there's no reason that that shouldn't work,
just as in English one might say, "I want that I go".  It's not
grammatical, but it works.

It seems, though, that a lot of languages prefer to use an uninflected--
or non-finite--form in these constructions.  Why?  Could be a
number of reasons--the first of which is that you already inflected
one verb, so doing it again, when the subject is the same, and
the verbs go together, is redundant.  From there, though, we
have two "actions", shall we say: "wanting", which is a part of
the real world, and is "happening" to the speaker, and "going",
which is not a part of the real world.  It doesn't ever seem to
be the other way around, e.g.:

querer voy = "I want to go"

I'd be interested to see a natlang example like that.  So even if
it was for other reasons, there does seem to be a connection
between the non-real and non-finite verbs--at least in these
IE languages where verbs agree with subjects, etc.  A true
cross-linguistic study would be interesting to see, and I wouldn't
want to make any claims about anything, or even any guesses,
having not read or undertaken one.

When it comes to Sidaan...

Eric:
<<
and in natlangs that may have inspired it
 >>

No natlangs inspired it.  I knew what I had (a perfect/imperfect
system, spired in part by Russian), and knew what I wanted
(a verb system similar to languages of the Philippine type),
but didn't know how to get there.  So what I did was used the
pieces available.

The old system looks like this (using "I" for a subject):

longuno
/eat-PRF./
"I ate", "I have eaten", "I just ate", "I've finished eating", etc.

longunE
/eat-IMP./
"I'm eating", "I was eating", "I'm still eating", etc.

This doesn't quite cover the range of linguistic expression
necessary, so since it was inspired by Russian, I used the four-way
Russian system as inspiration.  This is where /san-/, the "irrealis"
marker ("irrealis" as I defined it above) comes into play:

sanlonguno
/IRR.-eat-PRF./
"I would have eaten", "I should have eaten", etc.

sanlongunE
/IRR.-eat-IMP./
"I will eat", "I might eat", "I may eat", "I would eat", etc.

It simply adds some irreality to the already established perfect
and imperfect.

Now here's where the irrealis+non-finite part comes.  All these
verbs are finite, because they bear the perfective/imperfective
tense (that's what counts as "finite" in Sidaan).  What I didn't
have a strategy for was to do something like, "I tried to eat", or
"I want to eat", or "I kept eating", etc.  You can always create
affixes, but I didn't want to.  So what I needed was some way
to say "This thing is a verb, but it's not a finite verb".  So I just
stripped off the (im)perfective suffixes, and used /san-/.  The
irrealis prefix, after all, could only attach to verbs (so a speaker
knows if it has /san-/, it's a verb), but anything without a
perfective suffix is not a finite verb.  Thus, this produced a non-
finite verb.  I went on to use this non-finite verb in various
ways, but that part isn't relevant.

So the reason that the non-finiteness and irreality are tied
together in Sidaan is really something of an accident.  Imagine
if the verb system looked like this:

longuno = realis, perfective
longunE = realis, imperfective
longuni = irrealis, perfective
longunu = irrealis, imperfective

Now it's all the same, and the irrealis part of it is tied directly
into the aspectual morphology.  If this were what the system
was like, I'd probably have to go to a two word strategy, like
English, since most verbs in Sidaan, without /san-/ and any
perfective suffix, can double as nouns.  It's simply the fact that
the /san-/ prefix is, in a way, a kind of overlay that allows me
to use it as a kind of non-finite marker.  So if this were a
natlang, I'd say it shouldn't be used as evidence of their being
a connection between non-finiteness and irreality.  Not that
there isn't, but here it's kind of a happy accident.

-David
*******************************************************************
"A male love inevivi i'ala'i oku i ue pokulu'ume o heki a."
"No eternal reward will forgive us now for wasting the dawn."

-Jim Morrison

http://dedalvs.free.fr/

Replies