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

(Some notes, with reference to Mithun's _Languages of Native North
America_.)

Eldin Raigmore wrote at 2006-05-18 20:34:00 (-0400)
 > [CYCLIC LANGUAGES]
 > Milewski mentions a group of languages, all of which, he says, are
 > restricted to various parts of British Columbia, which he calls
 > "cyclic languages".
 >
 > In these languages, a word's "function in the clause" is not marked
 > in/on that word, but in/on the previous word.  To say it a little
 > differently, in these languages, a word does not mark its own
 > "function in the clause", but rather, marks the next word's
 > "function".
 >
 > There is one function reserved for the first word -- it appears to
 > always be the verb, or, at least, is in the examples Milewski
 > actually printed.
 >
 > There is one marking which means "this is the last word in this
 > clause".
 >
 > Has anyone's conlang ever done anything that weird?
 >
 > Milewski's example languages that he mentions are;
 > Kwakiutl (a Mosan (?) language)

Mosan is a proposed macrofamily comprising Salishan, Wakashan, and
Chimakuan.  According to Mithun it's generally been abandoned by
specialists, the similarities being attributable to contact.  Kwakiutl
is a Wakashan language (as Patrick mentions in his reply).

 > Nass
 > Tsimshian

Both members of the Tsimshianic family.

 >
 > Apparently Kwakiutl is what we would call accusative, while Nass and
 > Tsimshian are what we would call ergative.
 >
 > Does anyone know anything about these languages?  Or about any
 > other "cyclic" natlangs?
 >

Like Patrick says, it's (to the best of my knowledge) a matter of case
particles that precede the noun and cliticise to the word they follow.
(Although I have a feeling I've read _something_ more like Milewski's
account, about something, somewhere.)

Here's Mithun on Sm'algyax (which is the same language as Tsimshian).

| (18) SM'ALGYAX ERGATIVE PRONOMINAL CLITICS               Mulder 1988: 99
|
|     yagwa=*t*           dzap=dit             Rita waas.
|     PRESENT=*3ERGATIVE* make=ERGATIVE.PROPER      blanket
|     'Rita (*ERGATIVE*) is making a blanket.'
|
| Case on full noun phrases is specifed by clitics that precede the noun
| phrase but merge phonologically with the word to tehir left.  In (18)
| the clitic _=dit_ marks the following noun _Rita_ as ergative, but it is
| attached to the preceding word 'make'. (It is interesting that a
| similar pattern appears in the Northern Wakashan languages, including
| Kwakwala.  The languages are genetically unrelated, but they are
| geographicallly adjacent.)  The Sm'algyax clitics distinguish
| ergative, absolutive, genitive, and oblique (benefactive,
| instrumental, locative, temporal) categories.  The case enclitics,
| termet 'connectives' by Tsimshianists, also distinguish common nouns
| from proper nouns, which include independent pronouns. The enclitic
| _=dit_ in (18) identifies _Rita_ not only as ergative but also as a
| proper name.  Case enclitics for common noun phrases can be seen in
| (19): _=a_ absolutive common and _=da_ ergative common.
|
| (19)  SM'ALGYAX COMMON NOUN CLITICS	         Mulder 1988: 46, 58, 46
|
| a.  Nah  siipg=*a*                ƚguqoomƚk.
|     PAST sick=*ABSOLUTIVE.COMMON* child
|     'The child (*ABSOLUTIVE*) was sick.'
|
| b.  Yagwa   hadiks=*a*               sts'ool da      ts'm t'aaks.
|     PRESENT swim=*ABSOLUTIVE.COMMON* beaver  oblique in   pond
|     'A beaver (*ABSOLUTIVE*) is swimming in the pond'
|
| c.  Yagwa=t           huum=*da*               duus=*a*                hoon.
|     PRESENT=3ERGATIVE smell=*ERGATIVE.COMMON* cat=*ABSOLUTIVE.COMMON* fish
|     'The cat (*ERGATIVE*) is sniffing the fish.' (elicited)
|
| The language thus shows interesting distinctions among kinds of
| grammatical relations: ergative pronouns are clausal clitics, while
| absolutive pronouns are verbal suffixes;  core case markers are
| nominal clitics, while obliques are particles.