Page 1 of 2

Micamo's Magnificent Mithara Blog!

Posted: 17 Dec 2014 09:56
by Micamo
Well, you folks asked for it, so here it is. Mithara changes according to my whims quite often, so I'll try to only talk about things that are more or less set in stone. Affixes and vocabulary will probably change, as are things I haven't specifically talked about yet (and sometimes things I have). I'll probably contradict myself; If I do, and you're confused, ask and I'll clarify.

First up: Noun Incorporation!

A transitive verb can always incorporate its object:

šiƛašk̓iš
sh-i-dush-k'ish
1s.A-REAL-cat-pet
I pet/am petting the cat.

Some intransitive verbs can incorporate their subjects. All so-called "descriptive" verbs fall into this category; In Mithara, any word can be used in the place of a verb stem to create a predicate meaning "to be X"; I call such predicates descriptive verbs.

ʔiceckʷil̕
i-chi-ckwil'
REAL-house-white
The house is white.

ʔiƛuskayus
i-dush-kayus
REAL-cat-domestic_animal
The cat is a domestic animal.

Intransitive verbs denoting an activity are among the class that usually can't incorporate their subjects:

*niƛašm̓ał
ni-i-dush-m'az
PERF-REAL-cat-work
The cat worked.

"Possessor raising", where an external object NP is interpreted as the possessor of the incorporated object, is generally impossible in Mithara, with three exceptions. First, descriptive verbs allow possessor raising:

Skalir ʔipa:sks̓x
skalir i-paash-ks'x
NAME REAL-dog-be.sick
Skalir's dog is sick.

Second, there is a small class of special nouns (mostly body part words) that, when incorporated, always allow possessor raising, even in a transitive verb:

Skalir šiqʷnx̓i:
skalir sh-i-qwn-x'ii
NAME 1s.A-REAL-hand-cut
I cut Skalir's hand.

Note, however, that "double incorporation" is utterly impossible in this case:

*šipa:šqʷnx̓i:
sh-i-paash-qwn-x'ii
1s.A-REAL-dog-hand-cut
I cut the dog's paw.

Third, either of these can also allow possessor raising when they take part in verb serialization (more on this later):

x̱ʷiša:n skalir ʔanipa:sks̓xt
xhwishaan skalir a-ni-i-paash-ks'x-t
witch NAME 3>3-PERF-REAL-dog-be.sick-CAUS
The witch made Skalir's dog sick.

skalir šineyiqʷnx̓i:t
skalir shi-ni-y-i-qwn-x'ii-t
NAME 1s.P-PERF-2s.A-REAL-hand-cut-CAUS
You made me cut Skalir's hand.

Mithara is a "Type IV" incorporating language, under Mithun's classification. This means that modifiers can be external, and external NPs can be coreferent with the incorporated object (with the external NP usually more specific)

nišep̓ikšt sel̕es
ni-sh-i-p'iksh-t sel'-es
PERF-1s.A-REAL-bread-make two-CL:ROUND
I made two loaves of bread.

akʷeʔ ƛ̕xay̓ nišełukyan
akwe' d'xay' ni-sh-i-zuk-yan
all red_salmon PERF-1s.A-REAL-fish-eat
I ate all the red salmon.

Note also that Mithara inflectionally treats transitive verbs that incorporate an object as grammatically transitive:

ʔaniłakyan
a-ni-i-zuk-yan
3>3-PERF-REAL-fish-eat
He ate the fish.

(a- is an allomorph of the usually null 3.P prefix, which only occurs when the agent is also third person. A minor detail, but a complication to building inflectionally correct verbs nonetheless.)

The final thing I'm gonna talk about here today is verb serialization; Mithara handles complex, multi-predication clauses by incorporating a lower predicate into the higher, predicate-selecting verb.

yinešiyaqriwax̱̓p
yi-ni-sh-i-yaq-riwaxh'p
2s.P-PERF-1s.A-REAL-kill-decide
I've decided to kill you.

As we've seen above incorporated noun can be "dragged along" into the higher predicate. However, if the lower predicate is transitive, the agent of the lower predicate raises to object of the higher predicate (and can thus be incorporated), and then the lower predicate (along with its own IN) incorporates. This is the only situation in which transitive agents can ever be incorporated:

nišešw̓ilčxʷč̕aθ
ni-sh-i-sw'il-chxw-t-y'ath
PERF-1s.A-REAL-boy-arrow-make-teach
I taught the boy to make arrows.

nišemiθp̓ekš̕kʷskit
ni-sh-i-mith-p'iksh'-kwski-t
PERF-1s.A-REAL-man-bread-sell-CAUS
I made the man sell the bread.

Alright, so any ideas on what you guys want to hear about next? Feel free to ask any questions you might have about any of this stuff, by the way.

Re: Micamo's Magnificent Mithara Blog!

Posted: 17 Dec 2014 16:59
by shimobaatar
Micamo wrote:šiƛašk̓iš
sh-i-dush-k'ish
1s.A-REAL-cat-pet
I pet/am petting the cat.
This is an example of the 3rd person patient not being marked (at least not with an affix) if the agent isn't also 3rd person, right?
Micamo wrote:Intransitive verbs denoting an activity are among the class that usually can't incorporate their subjects:

*niƛašm̓ał
ni-i-dush-m'az
PERF-REAL-cat-work
The cat worked.
Would the correct version of that sentence be

dush ni-i-m'az

or would "cat" go after the verb?

By the way, what exactly determines how the morphemes surface in the bolded line/actual words? It's fine if you haven't set it in stone yet, but I'd be curious to know if you have. It seems like some interesting morphophonology and/or harmony.

Micamo wrote:Note, however, that "double incorporation" is utterly impossible in this case:

*šipa:šqʷnx̓i:
sh-i-paash-qwn-x'ii
1s.A-REAL-dog-hand-cut
I cut the dog's paw.
Would I be correct in assuming that:

paash shi-i-qwn-x'ii

would be the correct way to phrase that?
Micamo wrote:Alright, so any ideas on what you guys want to hear about next? Feel free to ask any questions you might have about any of this stuff, by the way.
You've explained and presented everything quite well, so I can't think of many specific questions other than the ones I've asked above. [:)]

One thing in particular that comes to mind in terms of what I'd be interested in hearing about is the verb template (if that's the correct term), if you have it at a point where you're comfortable with it.

Re: Micamo's Magnificent Mithara Blog!

Posted: 18 Dec 2014 00:09
by Micamo
shimobaatar wrote:This is an example of the 3rd person patient not being marked (at least not with an affix) if the agent isn't also 3rd person, right?
Yes.
Would the correct version of that sentence be

dush ni-i-m'az

or would "cat" go after the verb?
It can go in either position, but the external form of "cat" is du:she; Many nouns undergo phonetic reduction when they're incorporated, the most common such alteration being final vowel loss.
By the way, what exactly determines how the morphemes surface in the bolded line/actual words? It's fine if you haven't set it in stone yet, but I'd be curious to know if you have. It seems like some interesting morphophonology and/or harmony.
How consonants interact is not quite set in stone yet, but here's how vowel harmony works: There are 4 vowel phonemes, /a e i u/ (and the epenthetic schwa which does not count as a vowel phoneme since it's always optionally pronounced). There's regressive harmony between /a e/ that's blocked by the presence of a high vowel (or a syllabic resonant). Finally, there's a dissimilation rule; If two high vowels are in adjacent syllables, take the one on the right and lower it (i -> e, u -> a). All of these vowels operate the same whether the vowel is long or short. Some examples:

a-e-i -> e-e-i
i-i-u -> i-e-u
e-i-u -> e-i-a
u-i-a -> u-a-a

(As you can see, vowel phonemes in Mithara have a very small functional load.)
Micamo wrote:Would I be correct in assuming that:

paash shi-i-qwn-x'ii

would be the correct way to phrase that?
Right again, except the independent form of "dog" is paashi.
One thing in particular that comes to mind in terms of what I'd be interested in hearing about is the verb template (if that's the correct term), if you have it at a point where you're comfortable with it.
The exact slot-by-slot breakdown is up in the air right now, but the basic template is as follows:

MOOD1-ASPECT-PATIENT-INCEP-IRRL-PERF-AGENT-REAL-(APPL)-IN-(APPL)-VERB-EVIDENTIAL-MOOD2

(This template ignores the possibility of predicate incorporation which complicates matters a bit)

The part that's up in the air is the way the "mood" and aspectual morphemes interact; Lots of them can co-occur with each other and I've yet to completely decide exactly what combinations are possible and what order they occur in. I may or may not get rid of the suffixial moods and make them all prefixes; Evidentials will probably stay suffixes.

Re: Micamo's Magnificent Mithara Blog!

Posted: 18 Dec 2014 02:20
by shimobaatar
Micamo wrote:u-i-a -> u-a-a
The vowel harmony rules in general are interesting, but I especially like how the two come together here.
Micamo wrote:MOOD1-ASPECT-PATIENT-INCEP-IRRL-PERF-AGENT-REAL-(APPL)-IN-(APPL)-VERB-EVIDENTIAL-MOOD2
For future posts, could you go into more detail about some of the things bolded in the quote above, particularly evidentiality?

Re: Micamo's Magnificent Mithara Blog!

Posted: 18 Dec 2014 06:01
by Micamo
Well, I can talk about applicatives since that ties into noun incorporation: I've not yet decided on exactly what the full inventory of applicative morphemes is, but I know it'll be large (possibly unnaturally so), because Mithara lacks morphological case (with the exception of an optional ergative case) and adpositions of any kind, so applicatives need to fill some of the gap (the locative "moods" fill the rest, but I'll discuss them later).

Applicatives in Mithara can be applied to both transitives and intransitives. When applied to intransitives they transitivize the verb:

tishax'max'
t-i-sh-i-x'-max'
TRANSLOC-2s.P-1s.A-REAL-APPL:LOC-walk
I walked/am walking toward you.

When applied to a transitive verb, however, there are two possibilities. The first possibility is that the order of morphemes is APPL-IN-VERB, with the IN interpreted as the theme (and incorporation of the theme here is mandatory). The patient agreement agrees with the promoted "oblique" argument (except that independent obliques can't exist in Mithara).

k'siyex'kwe'emcn
k'-shi-yi-i-x'-kwe'em-cn
CISLOC-1s.P-2s.A-REAL-APPL:LOC-basket-carry
Bring that basket over to me.

The second possibility is that the promoted oblique is incorporated, and the order is IN-APPL-VERB, and the patient agreement agrees with the theme.

tsiwapstax'cn ne?
t-shi-u-i-psta-x'-cn ne
TRANSLOC-1s.P-IRRL-2s.A-riverbank-APPL:LOC-carry Q
Would you carry me to the riverbank?

(Mithara distinguishes stop+fricative clusters from affricates.)

Note that it's impossible to do both at once, incorporating two nouns:

*tuypstex'kwe'emcn ne?
t-u-i-psta-x'-kwe'em-cn ne
TRANSLOC-IRRL-2s.A-riverbank-APPL:LOC-basket-carry Q
Would you carry the basket to the riverbank? (OK: kwe'em ne tuypstax'cn? OR pstaani ne tuyx'kwe'emcn?)


EDIT: Oh yeah, almost forgot. You can incorporate the oblique in an applicativized intransitive as well: If the intransitive could incorporate, then you have both choices available that you have for transitive verbs (with required incorporation of the original subject if you don't incorporate the oblique). If it couldn't, then you only have the second option available. Descriptive verbs lose their ability to possessor raise if you take the second option with them, but body part nouns don't so long as the verb was intransitive.

Re: Micamo's Magnificent Mithara Blog!

Posted: 18 Dec 2014 12:43
by DesEsseintes
I sketched a phoneme inventory in IPA from what I can gather from your examples. I hope you don't mind. [:)]

Code: Select all

p   t   t͡s  t͡ɬ  t͡ʃ  k   kʷ  q   qʷ  ʔ
pʼ  tʼ  t͡sʼ t͡ɬʼ t͡ʃʼ kʼ  kʷʼ qʼ  qʷʼ
    θ   s   ɬ   ʃ   x   xʷ  X   Xʷ  h
    θʼ  sʼ  ɬʼ  ʃʼ  xʼ  xʷʼ Xʼ  Xʷʼ
m   n       l   j       w
mʼ  nʼ      lʼ  jʼ      wʼ
            ɾ
            ɾʼ

a   e   i   u
I used X for the uvular fricatives, as I can't get the proper symbol to work.

Questions:
- I'm assuming standard usage of Americanist notation, especially for c č š ł ƛ λ. Was I right in doing so?
- I saw no evidence of /h ɸ ɸʼ/ so I didn't include them. Do they occur?
- What is the nature of the rhotic?
- What did I miss?

Re: Micamo's Magnificent Mithara Blog!

Posted: 18 Dec 2014 12:49
by Micamo
DesEsseintes wrote:- I'm assuming standard usage of Americanist notation, especially for c č š ł ƛ λ. Was I right in doing so?
Mhmm!
- I saw no evidence of /h ɸ ɸʼ/ so I didn't include them. Do they occur?
/h/ does but the others don't.
- What is the nature of the rhotic?
I go back and forth between it being an alveolar tap and an alveolar approximant. It's definitely not a trill or a uvular rhotic.
- What did I miss?
Just /h rʼ/. There used to be voiceless versions of all the sonorants too but I got rid of them.

Re: Micamo's Magnificent Mithara Blog!

Posted: 18 Dec 2014 13:30
by DesEsseintes
Micamo wrote:
- What did I miss?
Just /h rʼ/. There used to be voiceless versions of all the sonorants too but I got rid of them.
I've filled them in. [:)]

Re: Micamo's Magnificent Mithara Blog!

Posted: 19 Dec 2014 02:16
by shimobaatar
I'm not the most knowledgeable when it comes to some linguistic terminology/polysynthetic languages in general, but the examples are definitely helping me work things out. I've learned quite a few things just by Googling certain glossing terms. Still, I apologize if some of my questions are stupidly simple. Thanks for your patience.
Micamo wrote:(Mithara distinguishes stop+fricative clusters from affricates.)
Cool.
Micamo wrote:EDIT: Oh yeah, almost forgot. You can incorporate the oblique in an applicativized intransitive as well: If the intransitive could incorporate, then you have both choices available that you have for transitive verbs (with required incorporation of the original subject if you don't incorporate the oblique). If it couldn't, then you only have the second option available. Descriptive verbs lose their ability to possessor raise if you take the second option with them, but body part nouns don't so long as the verb was intransitive.
Could we see examples of some of the situations mentioned here, please?
DesEsseintes wrote:I sketched a phoneme inventory in IPA from what I can gather from your examples.
You read my mind! [:)] I was going to ask if the inventory had been fully decided upon yet.

I like it, especially the glottalized sonorants, the "missing" labial fricatives, and the ejective labialized dorsal obstruents.

Re: Micamo's Magnificent Mithara Blog!

Posted: 19 Dec 2014 04:00
by Micamo
shimobaatar wrote:I'm not the most knowledgeable when it comes to some linguistic terminology/polysynthetic languages in general, but the examples are definitely helping me work things out. I've learned quite a few things just by Googling certain glossing terms. Still, I apologize if some of my questions are stupidly simple. Thanks for your patience.
No problem!
Micamo wrote:EDIT: Oh yeah, almost forgot. You can incorporate the oblique in an applicativized intransitive as well: If the intransitive could incorporate, then you have both choices available that you have for transitive verbs (with required incorporation of the original subject if you don't incorporate the oblique). If it couldn't, then you only have the second option available. Descriptive verbs lose their ability to possessor raise if you take the second option with them, but body part nouns don't so long as the verb was intransitive.
Could we see examples of some of the situations mentioned here, please?
Sure thing! Applicativized intransitive verbs incorporating an oblique:

sicax̓m̓ał
sh-i-ci-x'-m'az
1s.A-REAL-house-APPL:LOC-work
I worked in the house.

šmiʔx tiłtawsw̓ilsƛk̓ʷ ʔik̓ʷaʔ
sh-mi'x tiz=tu-u-sw'il-s-dkw' i-kw'a'
1s.POSS-liver NEG=3INV-IRRL-boy-APPL:BEN-tasty REAL-die
I am sad because of the boy who died. (lit. "My liver is not tasty to the boy who died.")

Applicativized unergatives (intransitives that can't incorporate) still can't incorporate their subjects when applicativized:

*ci:qʷaʔ aix̓sw̓ilm̓ał
ciiqwa' a-i-x'-sw'il-m'az
house 3>3-REAL-APPL:LOC-boy-work
The boy worked in the house. (OK: sw̓il ʔicax̓m̓ał)

Applicativized unaccusatives (intransitives that can incorporate) can:

sw̓il ʔik̓ʷaʔ tiłasəsmiʔxƛk̓ʷ
sw'il i-kw'a' tiz=u-sh-s-mi'x-dkw'
boy REAL-die NEG=IRRL-1s.A-APPL:BEN-liver-tasty
I am sad because of the boy who died.

But an applicativized unaccusative must incorporate one argument or the other, it can't leave both external:

*šmiʔx tiłtawsƛk̓ʷ sw̓il ʔik̓ʷaʔ
sh-mi'x tiz=tu-u-s-dkw' sw'il i-kw'a'
1s.POSS-liver NEG=3INV-IRRL-APPL:BEN-tasty boy REAL-die
I am sad because of the boy who died.

An applicativized unergative however can leave both arguments external without trouble:

sw̓il ʔaix̓m̓ał ci:qʷaʔ
sw'il a-i-x'-m'az ciiqwa'
boy 3>3-REAL-APPL:LOC-work house
The boy worked in the house.

Applicativized descriptive verbs can possessor raise if they incorporate their subject, but not if they incorporate the oblique:

sw̓il ʔaix̓pa:sci:qʷaʔkaʔ č̕ki:t̕
sw'il a-i-x'-paash-ciiqwa'-ka' ch'kiit'
boy 3>3-REAL-APPL:LOC-dog-house-CHAR winter
The boy's dog lives in the house during winter. (But stays outside the rest of the year.)

*šmiʔx sw̓il tiłtawk̓ʷaʔcəsƛk̓ʷ
sh-mi'x sw'il tiz=tu-u-kw'a'ch-s-dkw'
1s.POSS-liver boy NEG=3INV-IRRL-death-APPL:BEN-tasty
I am sad because of the boy's death

A body part noun (including silc "tent") can still possessor raise in an applicativized intransitive verb, whether they're incorporated as the subject or as the oblique:

ʔisesilcx̓max̓
i-sh-i-silc-x'-max'
2s.P-1s.A-REAL-tent-APPL:LOC-walk
I walked to your tent.

ʔišex̓qʷnłeq̓
i-sh-i-x'-qwn-zeq'
2s.P-1s.A-REAL-APPL:LOC-hand-lie
My hand is laying on you.

Notice that subject incorporation possessor raising identifies with the agent prefix, and oblique raising with the patient prefix. Such raising becomes impossible when the verb is a transitive applicativized verb:

#ʔisesilcx̓cn
i-sh-i-silc-x'-cn
2s.P-1s.A-REAL-tent-APPL:LOC-carry
I carried it to your tent. (OK as: I carried you to the tent.)

#ʔisax̓at̕cn
i-sh-i-x'-at'-cn
2s.P-1s.A-REAL-APPL:LOC-body-carry
I carried your body to it. (OK as: I carried the body to you.)

Any questions?
You read my mind! [:)] I was going to ask if the inventory had been fully decided upon yet.

I like it, especially the glottalized sonorants, the "missing" labial fricatives, and the ejective labialized dorsal obstruents.
For the longest time it was missing /p p'/ too but then I put them in. Also, it had aspirated versions of all the fricatives (but not the stops).


EDIT: I should also mention something interesting happens when you combine applicatives and predicate raising.

If you use a verb that raises to object, for example -t "make", with an intransitive verb, you can incorporate the subject (even if the lower verb is unergative) and use an applicative on the lower stem, incorporating the oblique. For example:

šiƛascix̓ne:t
sh-i-dush-ci-x'-nii-t
1s.A-REAL-cat-house-APPL:LOC-sleep-CAUS
I made the cat sleep in the house.

However, you utterly lose the ability to applicativize at all if you raise a transitive verb:

*tisesw̓ilx̓ƛascnt
t-i-sh-i-sw'il-x'-dush-cn-t
TRANSLOC-2s.P-1s.A-REAL-boy-APPL:LOC-cat-carry-CAUS
I made the boy carry the cat to you. (Also Bad: *tisesw̓ilƛasx̓cnt "I made the boy carry you to the cat.")

Re: Micamo's Magnificent Mithara Blog!

Posted: 20 Dec 2014 20:44
by shimobaatar
Micamo wrote:Any questions?
So applicativized unergative intransitive verbs can either incorporate their oblique arguments or neither argument, but not their subjects? And applicativized unaccusative intransitive verbs can incorporate their subjects or oblique arguments, but not both, and not neither?

But any verb like "make" can result in an intransitive verb of either kind being able to incorporate both arguments?
Micamo wrote:I am sad because of the boy who died. (lit. "My liver is not tasty to the boy who died.")
Interesting expression!
Micamo wrote:1s.POSS-liver NEG=3INV-IRRL-boy-APPL:BEN-tasty REAL-die
Why are all of the examples concerning the boy's death marked as irrealis? I'd assume it has something to do with the fact that the sentences are describing "being sad because of the boy's death", but I'm not sure exactly what subcategory of irrealis that would fall under.

Also, what does 3INV mean?
Micamo wrote:boy 3>3-REAL-APPL:LOC-dog-house-CHAR winter
What does the abbreviation CHAR mean? None of my searches are turning up anything, and there are several things it could mean based on the context of that sentence.
Micamo wrote:A body part noun (including silc "tent")
That's cool! What made you decide to include "tent" within that category?
Micamo wrote:Also, it had aspirated versions of all the fricatives (but not the stops).
I like the idea of aspirated fricatives in general, but I never know what to do with them diachronically.

Re: Micamo's Magnificent Mithara Blog!

Posted: 21 Dec 2014 01:26
by Micamo
shimobaatar wrote:So applicativized unergative intransitive verbs can either incorporate their oblique arguments or neither argument, but not their subjects? And applicativized unaccusative intransitive verbs can incorporate their subjects or oblique arguments, but not both, and not neither?

But any verb like "make" can result in an intransitive verb of either kind being able to incorporate both arguments?
Yes.
Why are all of the examples concerning the boy's death marked as irrealis? I'd assume it has something to do with the fact that the sentences are describing "being sad because of the boy's death", but I'm not sure exactly what subcategory of irrealis that would fall under.
It's because the sentence is grammatically negative; All negated sentences take the irrealis forms.
Also, what does 3INV mean?
It's the transitive agreement form used when the patient outranks the agent in animacy or obviateness (more on both of these later), but only when both are third person.
What does the abbreviation CHAR mean? None of my searches are turning up anything, and there are several things it could mean based on the context of that sentence.
It's a derivational affix I call the "characterizer". It has several meanings and translations depending on the context, but it works sorta like this:

dog-house "The dog is a house", dog-house-CHAR "the dog is a house-dog" (it lives in a house, as opposed to an outdoor dog)
fish-river "The fish is a river", fish-river-CHAR "the fish is a river-fish" (it's of a species that lives in a river as opposed to a pond, lake, or ocean)
man-hammer "The man is a hammer", man-river-CHAR "the man is a hammer-man" (he has an affinity toward hammers)
man-dog "The man is a dog", man-dog-CHAR "the man is a dog person"
That's cool! What made you decide to include "tent" within that category?
*shrug* I just thought it would be cool.
I like the idea of aspirated fricatives in general, but I never know what to do with them diachronically.
One of the reasons I got rid of them.

Re: Micamo's Magnificent Mithara Blog!

Posted: 21 Dec 2014 01:56
by shimobaatar
Micamo wrote:It's because the sentence is grammatically negative; All negated sentences take the irrealis forms.
Ah, OK, that makes sense.
Micamo wrote:It's the transitive agreement form used when the patient outranks the agent in animacy or obviateness (more on both of these later), but only when both are third person.
Cool! And I'm looking forward to hearing about Mithara animacy and obviateness in the future.
Micamo wrote:It's a derivational affix I call the "characterizer". It has several meanings and translations depending on the context, but it works sorta like this:

dog-house "The dog is a house", dog-house-CHAR "the dog is a house-dog" (it lives in a house, as opposed to an outdoor dog)
fish-river "The fish is a river", fish-river-CHAR "the fish is a river-fish" (it's of a species that lives in a river as opposed to a pond, lake, or ocean)
man-hammer "The man is a hammer", man-river-CHAR "the man is a hammer-man" (he has an affinity toward hammers)
man-dog "The man is a dog", man-dog-CHAR "the man is a dog person"
I really like this; it's such a cool idea!
Micamo wrote:*shrug* I just thought it would be cool.
It is! Is "tent" the only outlier in the body parts category, or are there other similar nouns (such as other buildings or residential structures) in there as well?

Re: Micamo's Magnificent Mithara Blog!

Posted: 21 Dec 2014 04:09
by Micamo
shimobaatar wrote:It is! Is "tent" the only outlier in the body parts category, or are there other similar nouns (such as other buildings or residential structures) in there as well?
Things produced by the body (such as eggs, milk, or waste) also count as "body parts" for the purposes of possessor raising and other possessive phenomena (more on these later). As do certain common cultural items everybody has such as tents, blankets, clothes, fire pits, and canoes.

Re: Micamo's Magnificent Mithara Blog!

Posted: 23 Dec 2014 06:46
by DesEsseintes
What's nominal morphology like?

I can see that possession is marked on the noun. Are the possessive prefixes the same as the person affixes on verbs? If so, is the 3rd person possessive also zero?

Is number marked on the noun?

Obviateness?

I'm particularly interested in this at the moment, as I'm trying to get some nominal morphology down for Hííenununóóoþa (so far success is zero), and I'm trying not to just copy Blackfoot.

Re: Micamo's Magnificent Mithara Blog!

Posted: 23 Dec 2014 07:55
by Micamo
DesEsseintes wrote:What's nominal morphology like?

I can see that possession is marked on the noun. Are the possessive prefixes the same as the person affixes on verbs? If so, is the 3rd person possessive also zero?

Is number marked on the noun?

Obviateness?

I'm particularly interested in this at the moment, as I'm trying to get some nominal morphology down for Hííenununóóoþa (so far success is zero), and I'm trying not to just copy Blackfoot.
Some of the inflectional categories a noun can be inflected for include:

Possessive affixes (incl. depossessive)
Possessor classification
Distributive
Diminutive
Agentive(?)
Obviative

As well as various derivational processes. (Diminutives are classified as a type of inflection because they interact with verb agreement but I'll talk about that later.)

The first person singular (sh-~s-~si-~shi-) and second person singular (i-~y-~yi-) are, but I'm unsure about the rest. I'm actually a bit embarassed to admit this, but I don't know what all the various phonological forms for the pronominal prefixes are. There's singular/dual/plural for first and second-person pronominals (but no number distinction for third or indefinite persons), and an inclusive/exclusive distinction. First person exclusive is probably going to be haa-, but I'm not completely sure about it and the forms of the others change every other day. That's pretty much the state of the project and why I'm so hesitant to try to write a grammar or something; Certain areas of the language are defined in extreme detail while other, basic things languish in limbo.

Number as such is not marked on nouns in Mithara; There is, however, an associative suffix (phonological form unknown) for proper names only, and distributive forms which are often translatable into english as plurals. Some examples of distributive forms given below:

huc "cloud", hu'c "clouds of different shapes"
chxw "arrow", chichxw "arrows of different tip or feather designs"
sxs "animal fat", sixsxs "fat spread around onto a surface"
kwe'em "basket", kwekwe'em "baskets of different sizes"
chkay "blueberry", kaychkay "blueberries from different bushes"

But note that the singular forms can also be interpreted as collective plurals (a bunch of blueberries from the same bush, or identical baskets), and there's no inflectional difference between those. (If you want to specify there's only one of something, you have to use a numeral nec'- "one" with an appropriate numeral classifier suffix).

Obviation (in the basic case) is straightforward; I'll have a fuller post discussing the various complications later this week.

duushe 'isxhxhxh duushexhw, tup'qxe'
duushe i-s-xhxhxh duushe-xhw | tu-i-p'qxi'
cat REAL-APPL:BEN-hiss cat-OBV | 3INV-REAL-smack
A cat hissed at another cat, then the cat that was hissed at smacked the cat who hissed. ('ip'qxe' the cat that did the hissing did the smacking)

Re: Micamo's Magnificent Mithara Blog!

Posted: 23 Dec 2014 08:14
by DesEsseintes
Micamo wrote:I'm actually a bit embarassed to admit this, but I don't know what all the various phonological forms for the pronominal prefixes are

I know exactly what you're talking about. Rant in spoiler:
Spoiler:
I've never understood how people can post a lang with personal endings in the first post. I can never decide on these affixes because I think they're so important to the look of the language. This is why I always lie to myself by saying "oh this one is going to be a speedlang, so I don't really care about getting the affixes absolutely right." Then I grow fond of the speedlang, want to get it absolutely right, start hating the person affixes, and end up crying myself to sleep. (← slight exaggeration)
huc "cloud", hu'c "clouds of different shapes"
chxw "arrow", chichxw "arrows of different tip or feather designs"
sxs "animal fat", sixsxs "fat spread around onto a surface"
kwe'em "basket", kwekwe'em "baskets of different sizes"
chkay "blueberry", kaychkay "blueberries from different bushes"
Ooh, these look nice! I can see that there are many different and unpredictable processes at work here. If I'm not mistaken, you described some of these elsewhere on the forum at some point.

I look forward to the post on obviation. In the meantime, I'm going to have speaking to my cat in Mithara with the verb xhxhxh. Can I say this?

ʔisesxhxhxh. - I'm hissing at you

Siyesxhxhxh ne? - Are you hissing at me?

I must be doing sth wrong. I have no idea where the realis affix went...
Edit: I got the person prefixes back to front! And forgot a glottal stop!! Disaster... [:O]

Re: Micamo's Magnificent Mithara Blog!

Posted: 23 Dec 2014 09:57
by Micamo
DesEsseintes wrote:Ooh, these look nice! I can see that there are many different and unpredictable processes at work here. If I'm not mistaken, you described some of these elsewhere on the forum at some point.
Yes:
Spoiler:
The general procedure of distributive formation: Copy the first syllable in the noun, then append it to the left of the noun, ignoring vowel length. Where "syllable" is one of the following shapes:

CV
CN
CVR
CVF

Where C = any consonant, V = any vowel, R = {m mh m' n nh n' l lh l' w wh w' y yh y'}, N = {m mh m' n nh n' l lh l'}, F = {c c' ch ch' d d' th th' z z' s s' sh sh' x x' xw xw' xh xh' xhw xhw'}

Mithara prosodic phonology tolerates unsyllabified obstruents that are thus ignored by the reduplication process:

q̓ʷsa "head" -> saq̓ʷsa

It does not, however, tolerate unsyllabified sonorants.

In addition, two further mutations generally apply:

If the reduplicated syllable attaches directly to a sonorant (or h) that is not already glottalized, it becomes glottalized (or ʔ in the case of /h/)

yat "child" -> yay̓at

If the reduplicated vowel is /i/ or /u/, reduplicate the vowel as normal but change stem i -> e and u -> a:

miq "creek" -> mim̓eq
huc "cloud" -> huʔac

For obstruent-only stems, generally, one consonant is selected as an onset for a syllable with /i/ as its nucleus: It is generally not predictable which consonant will be chosen for the onset, but there is a clear preference for plosives over affricates, and affricates over fricatives:

łk̓ "trail" -> k̓iłk̓

Exceptions: Occasionally you see a stem which uses a "wrong" reduplication pattern, not predictable from their shape: These are referred to as so-called irregular distributives.

Some stems which should use the CN type instead use a CiN type:

c̓m "tree" -> c̓imc̓m

Some stems which should use a CVR or CVF type inexplicably lose their coda consonant during reduplication:

čkay "blueberry" -> kačkay

Some stems with a vowel nonetheless use a Ci reduplication pattern:

x̱me:p "teardrop" -> mix̱me:p

Some stems with an unsyllabified P after their first vowel (where P is any plosive) include the plosive in the reduplicated syllable, but spirantize it (no form CVp or CVp' ever does this):

ƛ̕mut "cedar branch" -> muθƛ̕mut
łxʷaq̓ʷ "arctic fox" -> xʷax̱̓ʷłxʷaq̓ʷ

Finally, a some obstruent-only stems include a fricative coda in the reduplicated syllable, always one immediately to the right of the chosen onset.

sxs "fat" -> sixsxs
This is all mostly still valid; There are two changes. First is that RV type reduplication deletes the stem vowel. Hence hu'ac -> hu'c and yay'at -> yay't. Second is that RV reduplication also causes any other resonant consonants in the word to become glottalized:

yak'am "boots", yay'k'am' "mismatched boots"
wer "wind", wew'r' "winds blowing from different directions"
I look forward to the post on obviation. In the meantime, I'm going to have speaking to my cat in Mithara with the verb xhxhxh. Can I say this?

ʔisesxhxhxh. - I'm hissing at you

Siyesxhxhxh ne? - Are you hissing at me?

I must be doing sth wrong. I have no idea where the realis affix went...
These are both correct, actually:

'isesxhxhxh
i-sh-i-s-xhxhxh
2s.P-1s.A-REAL-APPL:BEN-hiss

siyesxhxhxh ne?
shi-i-i-s-xhxhxh
1s.P-2s.A-REAL-APPL:BEN-hiss Q

Re: Micamo's Magnificent Mithara Blog!

Posted: 07 Jan 2015 04:12
by Micamo
Well so much for getting one of these a week! Sorry for the delay but I've been pouring my energy into writing Tazaric stuff (and playing Dragon Age <.<) instead. But without further ado:

Obviateness and Animacy

Obviateness (of which there are two levels, proximate and obviate) and Animacy (of which there are three levels; Human, Non-Human Animate, and Inanimate) are inherent features (phi-features) of every Mithara noun.

Animacy is (generally) unmarked, but Obviateness (generally) is. The obviative marker for non-human nouns is -xh or -xhw, depending on whether it follows a plain or labial consonant/vowel. Examples:

ƛu:šex̱ "the other cat"
łukx̱ "the other fish"
heʔw̓x̱ʷ "all of those other reindeer"

Note that for the purposes of allomorphy, long a: counts as a rounded vowel but short a is a plain vowel.

For human nouns the suffix is -xek~-xwekw~-xak~-xwakw. Both consonants harmonize to the end of the stem, and the vowel changes depending on whether the final vowel (even if the nominal isn't vowel-final) is front or back. This is a rare instance of vowel harmony that's progressive rather than regressive. Note further that the obviative suffixes can even be applied to pseudonominals:

ka:yc̓mx̓i:xek
kaa-i-c'm-x'ii-xek
INDEF.H-REAL-tree-cut-OBV
The other lumberjack (one who cuts trees).

ke:hew̓č̕u:xʷakʷ
kaa-i-hew'-ch'uu-xwakw
INDEF.H-REAL-reindeer-collect-OBV
The other reindeer herder.

The Diminutive suffix has an interesting effect; Diminutizing a noun causes that noun to become grammatically inanimate. Diminutive forms are achieved by a combination of metathesis and affixation. All diminutive nouns take the suffix -'is, and stems of the form CCV or CCVC change into CVC or CVCC instead:

ƛ̕mut "cedar branch", ƛ̕umt̕es "little cedar branch"
hew̓ "reindeer", hew̓is "little reindeer"
q̓ʷsa "head", q̓ʷas̓is "little head"

Diminutive formation applies before distributive formation:

muƛ̕mat "cedar branches from different trees", ƛ̕umƛ̕amt̕is "little cedar twigs from different trees"

Animacy and Agreement

Animacy is only relevant to agreement on transitive verbs in which both of the arguments which carry agreement factors are third person. In all other situations, third person marking is always null and the direct prefix a- and the inverse prefix tu- do not apply.

If the agent outranks the patient in animacy (Human > Non-Human > Inanimate), or they're on the same animacy level but the patient outranks the agent in obviateness (Proximate > Obviate), or both arguments the direct form a- is used. Otherwise, the inverse form tu- is used.

Note that this only applies to the relationship between the nominals that are co-indexed on the agreement factors, and only indirectly to the relationship between the agent and patient of the clause. In prototypical cases these two line up, but in situations where they don't, this distinction is very important. Note also that incorporated nouns in Mithara leave behind a trace that retains its phi-features and is thus a valid target for an agreement probe.

Let's look at prototypical examples first:

miθe ʔaipa:scn
mithe a-i-paash-cn
man DIR-REAL-dog-carry
The man carried the dog.

pa:še tumaθk̓xat̕
paashe tu-i-mith-k'xat'
dog INV-REAL-man-bite
The dog bit the man.

When a transitive verb is applicativized and incorporates its theme (P-A-APPL-N-V), the P agreement factor coindexes with the applicativized object, and not with the true patient of the sentence:

łxʷaq̓ʷč tux̓ʷercn miθe
zxwaqw'-ch tu-i-x'-wir-cn mithe
fox-AGT INV-REAL-APPL:LOC-baby-carry man
The fox carried the baby (inanimate) to the man.

The inverse prefix is used even though the agent outranks the patient, because the P agreement factor is actually coindexed with the location (man) and not with the baby. If you incorporate the oblique argument instead (P-A-N-APPL-V), the relationship once again reflects the relationship between semantic A and P:

łxʷaq̓ʷ ʔaymiθx̓cn wiʔr
zxwaqw' a-i-mith-x'-cn wi'r
fox INV-REAL-man-APPL:LOC-carry
The fox carried the baby (inanimate) to the man.

Inverse marking is also affected by possessor raising; In either case (from a body part noun or from a descriptive verb root), possessor raising causes an agreement factor to coindex with a possessor rather than with its ordinary target. Again the relationship is between the agreement factors and not between the agents and patients:

pa:šeč miθe tuqʷnk̓xat̕
paashe-ch mithe tu-i-qwn-k'xat'
dog-AGT man INV-REAL-hand-bite
The dog bit the man's hand.

Once again, we have inverse marking even though "dog" outranks "hand", because the actual relationship is between the agreement factors; "man" outranks "dog" so the inverse marking is used.

The last case I want to discuss is predicate raising; In a predicate raising verb that raises to object, the higher incorporated object is the target of the P agreement factor, not the lower object. This again affects inverse marking:

pa:še tumeθciłtne:t
paashe tu-i-mith-ci-zt-nii-t
dog INV-REAL-man-house-APPL:OUTSIDE-sleep-CAUS
The dog made the man sleep out of the house.

In a raising verb that raises to subject, though, the P agreement factor does target the lower object:

pa:še tumaθk̓xat̕riwax̱̓p
paashe tu-i-mith-k'xat'-riwaxh'p
dog INV-REAL-man-bite-decide
The dog decided to bite the man.

The Agentive suffix

The agentive marker -ch is an optional "case" suffix that's used mostly for disambiguation. It's used only when there are two or more external NPs; The agentive is never used when a clause only has one unincorporated argument. It's used in two situations:

- The agent is outranked by the patient and inverse marking is used. This usage is conservative and is not usually used by younger speakers.

- The agent and patient are equal in animacy and in obviateness, so the sentence is ambiguous without the agentive marker (since word order doesn't help). This usage is still common even in younger speakers. It's less commonly used when the agent is clause-initial, but almost always used when the agent is in some other position.


Any questions?

Re: Micamo's Magnificent Mithara Blog!

Posted: 07 Jan 2015 08:36
by DesEsseintes
Micamo wrote:łxʷaq̓ʷč tux̓ʷercn miθe
zxwaqw'-ch tu-i-x'-wir-cn mithe
fox-AGT INV-REAL-APPL:LOC-baby-carry man
The fox carried the baby (inanimate) to the man.

Here we have a sentence with two explicit NPs and neither is marked for obviateness. Does that mean obviation marking is optional? If so, is it really obviation and not just a suffix meaning "other"? I'm asking in earnest, as I'm still learning how obviation works.