Re: Isolating languages

On 5/28/05, Caleb Hines <bachmusic1@...> wrote:
> 
> 
> I may have been a bit unclear. It seems people thought I was looking for
> non-tonal, but mono-syllabic isolang. I'm not really planning on it being
> monosyllabic. In fact, I was hoping there were very isolating langs that
> weren't monosyllabic, because I don't know if monosyllabic can be done
> without using tones

You could get non-tonal monosyllabic, if you wanted to. Not assuming you'd 
want to, but I've always liked the idea. There's plenty of "room" in 
possible-syllable space provided there's an ample phonemic inventory or 
consonant clusters are allowed.

A simple demonstration of this: Take some language that you consider 
monosyllabic and without too much homophonic collision. Oh, say Cantonese. 
Now, Cantonese has eight tones or so, but there's also some serious 
restriction of things like final consonants. (And Cantonese is quite 
generous compared to Mandarin.) Final consonants limited to, hmm, voiceless 
stops and nasals, right? What if, in our phonotactic sandbox here, we 
allowed voiced stops and nasal-stop pairs? Now there's only need for four 
tones. Or even final fricatives... now only two tones are needed. Add 
distinctive voicing to initial consonants and no tones are needed at all. 
There, we have a language that fills the same "space" as Cantonese but lacks 
tones.

One of the language sketches I have on my plate (for a sample submission for 
the summer contest) is monosyllabic non-tonal. It allows stop-fricative 
clusters both initially and finally... any of p, t, and k may pair with any 
of s, S, and K (Welsh "ll"). (Maybe a retroflex fricative, instead of or in 
addition to S.) Voice is also distinctive, so that {b,d,g} may pair with 
{z,Z,K\}. The resulting "space" one can fill with words is quite large.

>> My current ideas are:
> >> Tentative name: G'nan
> >> Morphology: Isolating
> >> Typology: SOV/Po/GN/AN, with alternate SVO.
> >> (Seriously consdiering making SVO the primary word
> >> order)
> >
> >How strange! I was working on a conlang called Gi-nàin
> >for a while, until it 'evolved' into my current one,
> >Nèm. Both were isolating, SOV and, so some degree, tonal!
> 
> Interesting indeed! The idea for the name "G'nan" came from my trying to
> remember the typology: GN/AN -> G'nan. At one point I was going to have it
> be the opposite (NG/NA) and was tentatively calling it (you guessed it)
> "Nigna". But NG/NA is so heavily associated with type I (VOS) natlangs, 
> such
> as Semitic and Celtic, that I decided to swap it around.

Well, both of those are VSO... but VSO and VOS are quite similar 
typologically anyway.

How do you handle keeping S and O straight? For me, one of the problems
> 
arises because genitive constructs are consecutive nouns that have no case
> marking. Other problems I haven't quite solved yet arise when I want to
> treat a series of nouns as a list of items (without conjunctions) instead 
> of
> a genitive construct, and when I want to state equality of two items
> (without a copula).

Maybe you could have an optional genitive postposition. Usually left out but 
there when a speaker needs to disambiguate.

Or some sort of rule about, say, articles. Say you have both definite and 
indefinite articles. Art NP Art NP = S O. Art NP NP = genitive phrase. (With 
the definiteness of the possessor "inherited" by the possessed, as in 
Arabic.) NP Art NP = equality sans copula?

A variation of this could probably work with classifiers, as well.

I've discovered that I have much less trouble if I introduce an ergativity
> particle for transitive verbs, although this won't solve everything. I'll
> probably have to either rely on context for disambiguation, or introduce
> more particles.
> 
> ~Caleb
> "Prepositions are great things to end sentences with!"
> 
Anyway, my answers to some of the original questions:

> Are there many isolangs that are not mostly monosylabic?
There are, although the monolangs are probably the most isolating. I suspect 
this is for the simple reason that monolangs are much more resistant to 
adpositions becoming grammaticalized into affixes. If you have a bunch of 
polysyllabic words, what's one more syllable? But if you're a serious 
monolang, one more syllable is one syllable too many. 

> Are there many non-tonal isolangs?
I suspect the bias towards tone-iness among isolangs is a co-symptom of the 
above. Monolangs are quite likely to be isolangs, and also quite likely to 
be tonal. But other than this, I can't see any causal connection between the 
two.

It's like the fact that married people eat less candy than single people. 
It's not like being married causes decreased candy consumption (in general), 
or increased candy consumption decreases the likelihood of marriage (in 
general). It's just that a further factor -- growing older -- tends to cause 
less candy eating and more matrimony.

I figure the arrows go Monolang => Isolang and Monolang => Tonal. It leads 
to a bunch of tonal isolangs, but I don't think there's a causal arrow 
either way between isolangs and tonality.

> What kind of typology distributions (SVO/SOV/etc, Pr/Po, NA/AN, NG/GN) 
> are ther in isolating langs?
As was stated above, strong tendency towards SVO, in order to disambiguate 
the S and the O. Perhaps a compromise between your desire for SOV and your 
serious consideration of SVO is a V2 like German or Dutch?

Another possibility is discourse phenomena that combine to avoid the 
realization of both the subject and the object. Despite our typological 
obsession with classifying things according to the sixfold S,V,& O 
categorization, most sentences actually spoken don't have full noun-phrase 
(non-pronomial) subjects and objects. 

Mayan languages, for example, seem to prefer to string together lots of 
little VS and VO clauses, with previous topics expressed only as pronomial 
affixes. Actually eliciting a full VOS sentence in normal discourse is 
difficult. Not the solution for your isolang, of course, but there are lots 
of possibilities. Like a sort of switch-reference conjunction system... each 
conjoined clause can indicate with a certain word for "and" what role the 
most recent topic plays in this clause. One conjunction could say "the topic 
of the last clause is the (non-surfacing) S of this clause" and another for 
the O. 

"There was this man, see, and he went into a store, where he bought some 
apples. Then a bus ran over him! Finally, the bus hit an apple tree, and it 
dropped some more apples on him."

TOP man was, AND-S store went, AND-S apples bought, AND-O TOP bus ran-over, 
AND-S TOP apple tree hit, AND-S apples him-on dropped.

Anyway, just thinking on paper.

I look forward to seeing what sort of solutions you find!

-- 
Patrick Littell
PHIL205: MWF 2:00-3:00, M 6:00-9:00
Voice Mail: ext 744
Spring 05 Office Hours: M 3:00-6:00