Re: Various weirdnesses in natlangs, plus obconlang questions following them

On 5/18/06, Eldin Raigmore <eldin_raigmore@...> wrote:

> [TRITRANSITIVE]
>
> Has anyone else ever heard of these?
> Has anyone got them in his/her conlang?
> Does anyone know of them in a natlang, or someone else's conlang?
>
If by "tritransitive" you mean a construction by which four arguments
are licensed by the verb, then yes.  (On the other hand, if we're
talking syntactically, and you mean a verb phrase with one external
argument and three complements, then no.  In what follows the licensed
arguments are probably just adjuncts to the verb, syntactically
speaking.)

Anyway, that aside, it's not difficult to construct a verb stem that
takes four arguments in a Totonacan language.  (You do have to derive
one; verb roots are at maximum ditransitive, and there are only a few
of these.)  Let's start with 'i'xki'.  (I'll use a variant of the
practical orthography; for reference, x = S, lh = K, V' =
laryngealized vowel.)

ki(n)-  'i'xki'
1obj- give X to Y
s/he gives it to me (also has the meaning "s/he hits me")

ki(n)-  'i'xki'       -ni  -lh
1obj- giveXtoY  -appl -perf
s/he gave it to me for him/her

ki(n)- lii-      'i'xki'         -lh           tun-kiw
1obj instr-  giveXtoY   -perf        one-stick
s/he gave it to me (hit me) with a stick

kim-  puu-  'i'xki'      -yaa     -'na   hon baso
1obj- loc-  giveXtoY -imperf -2obj the  glass
s/he give it to us in the glass

kim-   maa-   'i'xki'        -ni       hon-cha
1obj-  caus-  giveXtoY  -appl   the-tortilla
s/he makes me give him/her the tortillas
or
s/he makes him/her give me the tortillas

And so on.  You don't have to start with a ditransitive either, since
the valence-increasing affixes can co-occur.  Here's an intransitive
stative /wi'la/ ("standing") bumped up to tritransitivity with a
transitivizer, and applicative/benefactive, and a locative:

kim-  puu- wa'l         -ii       -ni     -lh      hon mesa
1obj- loc- -standing  -trans - appl -perf   the table
s/he placed X on the table for me

Anyway, four isn't the limit, although I don't have an example with five.

> [CYCLIC LANGUAGES]
> Does anyone know anything about these languages?  Or about any
> other "cyclic" natlangs?
Kwakw'ala/Kwakiutl isn't quite so "crazy" as it may seem.  Or, let me
qualify that.  It's quite crazy, but this "cyclic" property isn't the
craziest of it.  Being "cyclic" isn't really a crazy new scheme for
marking argument structure... it's just that Northern Wakashan clitics
attach to the preceding word in the sentence, rather than the
following one with which they form a syntactic constituent.  Here's a
Heiltsuk (Bella Bella) example (minus a whole lot of diacritics):

Daduqvla  wismaxi  waciaxi-his  dugvayuaxi-la  uxcthiasaxi-qen  hirhasaxi
watch       man        dog     -with binoculars -on  roof          -for    chief
"A man watched a dog with binoculars on the roof for the chief."

/his/, /la/, and /qen/ are clitics that are syntactically their own
words but cannot phonologically stand alone for some reason.  Clitics
can either attach themselves phonologically to the previous word or
the following word -- here, it's to the previous word.  Since they
form a syntactic phrase with the following word, however, we end up
with a mismatch, and thus it appears a bit crazy.

It's not a phenomenon of a different kind, however, from the English
copular clitic 's.  In "He's going to the store", the 's is attached
to the preceding word, but in terms of syntactic constituency it's
part of what follows.  The difference is just one of degree.

Here's Kwakw'ala:

nanaqesil-ida i'gel'wat-i  eliwinuxwa-s  -is  mestuwi     la     -xa
    migwat-i
guides-sub/art expert-dem hunter-    instr -his harpoon prep-obj/art seal-dem
"An expert hunter guides the seal with his harpoon"

The sequence "hunter with his harpoon" wouldn't be out of place in
English -- the difference is just that the Kwakw'ala equivalents of
"with" and "his" don't have the status of full words, and instead have
the same sort of status as 's.  Kwakw'ala's just the extreme case of
cliticization -- it happens that all of the markers for case, deictic
status, and possessor happen to be enclitics.

Anyway, saying that something's a "cyclic" language isn't to say that
it has a certain exotic way of marking arguments unlike ordinary
head-marking or dependent-marking.  In terms of argument marking,
Northern Wakashan appears basically dependent-marking; it's just that
this is obscured by the enclitic nature of these markings.

Hope this makes things a little clearer.

-- Pat