qiihoskeh: myo: kanji (Default)
Here's another pending issue for project rho.

For each grammatical voice, there's an implied subject case and an implied object case (this grammatical voice is marked only on divalent and trivalent verbstems). The subject also may agree in gender. For trivalent verbstems, this is reasonably straightforward:
Act- subject is -Agt, object is -Pat if it's animate and -Thm otherwise
PsA- subject is -Pat, object is -Agt if it's animate and -Thm otherwise
PsI- subject is -Thm, object is -Agt or -Pat (this is ambiguous!)

For divalent verbstems denoting actions, it's also reasonably straightforward:
Act- subject is -Agt, object is -Pat
PsA- subject is -Pat used if it's animate, object is -Agt
PsI- subject is -Pat used if it's inanimate, object is -Agt

But for verbstems denoting perceptions, I'm undecided. What I have currently is:
Act- subject is -Pat, object is -Thm
PsA- subject is -Thm used if it's animate, object is -Pat
PsI- subject is -Thm used if it's inanimate, object is -Pat

For verbforms, the implied subject is the current topic if no explicit subject taking the implied subject case appears. Example:

"As for John, he gave the dog water."
John-Top, Act-give-Pst-Ind Def-dog-Pat water-Thm

This evaluates (ignoring topicality concerns) to:
"John gave the dog water."
John-Agt Act-give-Pst-Ind Def-dog-Pat water-Thm

(no subject)

2006-Dec-07, Thursday 17:59
qiihoskeh: myo: kanji (Default)
* I still haven't been getting much done.

* I reread Putting Up Roots by the late Charles Sheffield.

* Project Rho has a number of pending issues. Here's one:

I want to distinguish tangent vs. non-tangent spatial relations whenever the difference makes sense. I've come up with three possibilities.
(1) double the number of words for spatial relations, e.g. on_top_of vs. above
(2) double the genitive case; this would impact other uses of the genitive, for instance how would the cases apply to verbstem nouns? For divalent nounstems other than spatial relations the tangent case could be used for inalienable possession while the non-tangent case could be used for alienable possession.
(3) double the number of locational cases. This could be done easily if the dynamic cases were derived from the static case, because all I'd have to do is double the static (locative) case and add the same dynamic endings to both forms.

Here's another:

How do I make possessives of monovalent nouns? These are some possibilities, not necessarily mutually exclusive.
[1] use one of several specific relational words as adjective, taking the same case of the possessum (the possessor would take the genitive case of course)
[2] derive a possessive adjective from the possessor; this will work only for single words like pronouns, not multiword phrases
[3] derive a divalent wordstem from the possessum so that it can take a possessor
[4] define more nouns as being divalent; note that the monovalent word can be made by using an article proclitic instead of a possessor.
[5] see (2) above concerning inalienable vs. alienable possession.

While I'm at it, here's yet another issue:

I need to decide on the component order of phrases. Right now I have:

NOUN ADJECTIVES RELATIVE_CLAUSES

The NOUN either has a possessor preceding it (possibly a proclitic genitive pronoun) or a proclitic article. Requiring a noun in that position takes care of some ambiguities. The problem I have is with the placement of quantifiers and the construction of the partitive and the superlative. Some possibilities for the partitive:
(1) add a partitive case, to be used on the source phrase
(2) derive divalent quantifiers which take the source phrase as their possessors.

(no subject)

2006-Dec-05, Tuesday 17:32
qiihoskeh: myo: kanji (Default)
* For the past week I've been feeling stressed and very sleepy in the daytime, not at night. So I haven't gotten much done.

* I forgot to mention that I reread To Say Nothing of the Dog and Bellwether, both by Connie Willis (if you didn't know).

* I have started yet another conlang, project rho.

The ending determines the syntactical word class: mood endings if the word functions as a verb and case endings if the word functions as a noun. These can go on words of any lexical class except that particles are uninflected and pronouns can't take mood endings. The lexical classes are adjective, verb, noun, pronoun, and particle. Adjectives and verbs are marked for aspect, tense, voice, and gender, while nouns are marked for number and voice and definiteness and pronouns for number. Voice for nouns is limited to direct (unmarked), inverse, and reciprocal, and can appear only on divalent nouns, while voice for verbs includes active, passive, reflexive, and reciprocal.

Spatial relations are specified using divalent nouns rather than adpositions. The object takes the genitive case and the relationshship word takes one of the locational cases (at, to, from, or via).

I've layed out a bit of the grammar, although some things are undecided, and have the phonology partly done. What I haven't been able to start on is assigning morphemes.

Profile

qiihoskeh: myo: kanji (Default)
qiihoskeh

August 2017

M T W T F S S
 123456
78910111213
14151617181920
2122 2324252627
28293031   

Syndicate

RSS Atom

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated 2017-Sep-22, Friday 02:33
Powered by Dreamwidth Studios