Re: THEORY: What's a "word"? What's "Polysynthetic? Self-Segrating Morphology, and Milewski

Eldin wrote:
<<
He says that a language has "words" if it has a class of syntactic
groups
whose boundaries are established by exactly the same means regardless of
where they appear in the clause.
 >>

This is an interesting definition, though it relies on syntax, which
is problematic because you have to have a theory to go with it...
Not that that's wrong--it's just that theories differ in exactly this
respect, so someone coming from theory X may have a different
conclusion about the nature of words in a given language than
someone coming from theory Y.

For what it's worth, the first time I took the morphology class
at UCSD, most of the class and the readings were devoted to
exactly what a word is.  We came to the unsatisfying conclusion
that there was no possible unifiable definition (though we didn't
read this paper [book?] by Milewski; we focused on modern
WP writings).  One major problem, actually, was the nature of
clitics.  Now, if what Doug Ball has been looking at is on the
right track, perhaps clitics aren't a problem, since they don't
exist (Doug would need to say more about this; I know the end
result, not the arguments).  In which case, perhaps a phonological
(or prosodic) definition of what a word is is the right one.  Who
knows?

Eldin continues:
<<
Does anyone have a conlang that does have "words", but whose
"definition"
of what exactly a "word" is, is "weird"?
 >>

I do.  :)  Kamakawi, in fact, has caused me no end of trouble on
account of its problem with words.

Some words in Kamakawi are obviously "words", like nouns
and verbs.  Even when they have affixes, they have a single
stress pattern, have one stress per word, and act in all ways like
words.  Then, however, there are the (p)articles.  So, in a sentence
like this:

Ka mata ei ie ana.
/past-new.subject see I Obj.-def. duck/
"I saw a duck."

The object marker /i/ takes on the definite article /e/, and
becomes what I've always thought of as a word, even though
both of these can appear on their own:

Ka mata ei i ana.
/past-new.subject see I Obj. duck/
"I saw a duck."

E ana ka mata i'i.
/def. duck past-new.subject see Obj.-I/
"*The duck* saw me."

However, it becomes much more difficult when that definite
article comes after prepositions.  With some, it seems to me to
obviously glom on, but with others, much less so...

Ka mata'u ei tie ana.
/past-new.subject see-Pass. I Inst.-def. duck/
"I was seen by the duck."

Ka mata ei ie ana ae e pale.  (or /ae'e pale/ ?)
/past-new.subject see I Obj.-def. duck in def. house/
"I saw a duck in the house."

Ka ale ei poiu e pale.  (??)
/past-new.subject go I through def. house/
"I go through the house."

However, this is only a problem in the romanization system
and in transcription.  The reason is that Kamakawi is written
with characters that behave orthographically rather like Chinese
characters, but work more like Egyptian hieroglyphs (with
symbols for actual sounds combined with symbols for concepts).
These are called /iku/.  An example can be seen here:

http://dedalvs.free.fr/kamakawi/

This is the first line of the Babel text:

Ka peka i ape kalaka oi oalala poi.
/past-new.subject land Obj. one language with word few/
"The land was one language with few words."

In fact, you can see an image of the whole Babel text here:

http://www.langmaker.com/db/Kamakawi_Babel_Text

And what should be apparent is that since there are no spaces
between the words, and since characters themselves can express,
say, one meaning, half a meaning, or more than one meaning,
when I write in the romanization, and have to think about how
the language is divided up into words, it becomes rather
difficult.  Often I just guess, or am swayed by the fact that the
word /poiue/ looks unbearable.

Eldin continues:
<<
Whose conlangs are polysynthetic?
And which conlangs?
For those conlangs, what exactly did their conlangers mean
by "polysynthetic"?
 >>

I have one, Epiq.  You can take a look at the webpage to see
exactly how it works:

http://dedalvs.free.fr/epiq/

I'm not sure if I refer to it as polysynthetic on the page, but I
don't find that term very useful.  I'm not even sure if what I've
done fits into a theory of polysynthetic language.  What I did
was I looked at what I saw as a neat feature of the Inuit language
I have a grammar and dictionary for (Siglitun), and I borrowed
one of its concepts.

In Siglitun (or Uummarmiut or Inuvialuit--they use all three of
these names for what they seem to be saying is the same language,
the result of which is I have *no* idea what it's called), there's
this dictionary that lists a few nouns and verbs--not many, really,
compared to what one would find in a dictionary of Spanish,
Japanese, etc.  Then comes a listing of the affixes.  It goes on for
pages and pages and pages and pages.  Essentially, you can
start with a noun or a verb, and add affixes that changes its lexical
class dozens of times over and adds meanings left and right.
You start with the central idea (which seems, more often than
not, to be caribou), and then use the affixes to add what you
want to say about it, including...

(This is in the romanization system; I don't know the IPA.)

tuktuliaqatigitqingnapin'ngitkiga
"I will never go caribou hunting with him again."

I can't break that down, but /tuktu/ is the word for "caribou".

I liked this idea, so I tried to do something similar with Epiq.

Eldin continues:
<<
6. Tonkawa is very weird.  In the root -- but only in the root -- the
_odd_-
positioned vowels appear in "full grade", while the _even_-positioned
vowels appear in "weak grade".  However, to determine parity (oddness-
or-
evenness) of the position, one counts from the beginning of the _word_
rather than from the beginning of the root.  So if the prefix(es)
have, or
add up to, an even number of syllables (for instance zero syllables
or two
syllables), the root is pronounced one way; but if it/they have an odd
number of syllables (in sum), the root is pronounced another way.  There
may also be suffixes.  He gives the example of the root "yamaxo" "paint
someone's face", which appears as "yamx-o?" "he paints his face" and
as "ke-
ymax-o?" "he paints my face".
 >>

This is similar to old analyses of Potawatomi, and related languages.
I'd wager to say that such analyses are outdated.  I haven't looked at
any of the data in detail, but suspect LRM could have something
to say about it.  For another account of these disappearing vowels
from a modern perspective, my girlfriend's master's thesis was
an OT account of the disappearing vowels of Potawatomi.

Eldin continues:
<<
For example in Tunica, each group before the "predicate" (I think he
might
mean "verb" in this case) ends on a rising tone; each group after
the "predicate" ends on a falling tone; and in the VP itself, the
verb can
end on any of four tones depending on its mood.  (So "mood" is marked by
a "suprafix" in Tunica, instead of a prefix or suffix or infix.)
 >>

No freaking way...  Looks like I have a language to look up at
the university library.  ;)

-David
*******************************************************************
"sunly eleSkarez ygralleryf ydZZixelje je ox2mejze."
"No eternal reward will forgive us now for wasting the dawn."

-Jim Morrison

http://dedalvs.free.fr/