zompist bboard

Posted: **Sat May 12, 2012 4:34 pm**

treegod wrote:I've now read through Whimemsz work; well done, that's given me something to read again and again. :)

I didn't understand all of it, but what I did understand fascinated me. Now I'll go through this thread with some (little) idea about what is being said. :-D

Thanks, glad it's been of help. Incidentally I'd like to take this opportunity to apologize for being such a dick to you a few weeks ago -- it was far out of proportion to what you deserved, and I regret it. I still think my basic point of encouraging newbs to do their own outside research was an important one, but yours was a fairly minor transgression :). Part of it is just that I remember [oldmanvoice] Back In the Day, when the highest rank was after 200 posts [/oldmanvoice], so it's always a little disconcerting to see people with hundreds of posts who still lack a lot of fairly basic linguistics knowledge. But anyway, my point is you weren't being a jerk, you just weren't understanding something, and my rudeness was far out of proportion to that, and I'm sorry.

Serafín wrote:I'd suggest that this section could be better improved with examples where not only these verbs in Spanish and French have pronoun clitics (the verb and the pronouns form really one word), but moreover, these clitics in both languages are developing into affixes where the verb agrees with its arguments in polypersonal agreement.

Hm, that's an interesting point. I was aware of this phenomenon in Spanish, but it just.. didn't occur to me when I was writing the section. I'll try to incorporate some of that info into the post.

Posted: **Tue May 15, 2012 5:09 pm**

(Okay, I added a brief mention of Spanish clitic doubling to section 8.2, as well as another example of non-configuration from Warlpiri to...whichever post talks about non-configurationality)

Posted: **Wed May 16, 2012 4:08 pm**

Whimemsz wrote:Thanks, glad it's been of help. Incidentally I'd like to take this opportunity to apologize for being such a dick to you a few weeks ago -- it was far out of proportion to what you deserved, and I regret it. I still think my basic point of encouraging newbs to do their own outside research was an important one, but yours was a fairly minor transgression . Part of it is just that I remember [oldmanvoice] Back In the Day, when the highest rank was after 200 posts [/oldmanvoice], so it's always a little disconcerting to see people with hundreds of posts who still lack a lot of fairly basic linguistics knowledge. But anyway, my point is you weren't being a jerk, you just weren't understanding something, and my rudeness was far out of proportion to that, and I'm sorry.

I've just seen this. Thanks for apologising. I owe my apologies too, I did write an apology on another thread (probably deleted), I did a fair bit of provoking after asking all the newb questions, and I apologise for that.

Posted: **Thu May 17, 2012 9:51 am**

How does NI affect object agreement? WOuld the agreement affixes be amputated, or no?

Posted: **Thu May 17, 2012 10:01 am**

Solarius wrote:How does NI affect object agreement? WOuld the agreement affixes be amputated, or no?

I have generally heard that transitive verbs with incorporated objects lose pronominal agreement for the object though not for the subject. If you have the phrase "house she-sees-it" then it becomes "she-house-sees" with incorporation. Some languages can incorporate arguments besides the object, though, and those behave in various other ways. One can apparently incorporate the subject of some intransitive verbs in Mohawk, in which case the subject agreement prefix remains and incorporation of oblique arguments does not typically affect agreement.

Posted: **Thu May 17, 2012 10:36 am**

Jabechasqvi wrote:
Solarius wrote:How does NI affect object agreement? WOuld the agreement affixes be amputated, or no?
I have generally heard that transitive verbs with incorporated objects lose pronominal agreement for the object though not for the subject. If you have the phrase "house she-sees-it" then it becomes "she-house-sees" with incorporation. Some languages can incorporate arguments besides the object, though, and those behave in various other ways. One can apparently incorporate the subject of some intransitive verbs in Mohawk, in which case the subject agreement prefix remains and incorporation of oblique arguments does not typically affect agreement.

Also, in classificatory incorporation (i.e. Type 4 according to Mithun's classification), agreement affixes are usually not dropped (or the incorporated classifying noun stem is used as an object agreement affix), and overt NPs associated with the same referent may still be present. "She sees a dog" might thus become something like "she-<animal>-sees-it dog" under classificatory incorporation.

Posted: **Sun May 27, 2012 5:52 pm**

Hey Whimemsz, just popping in to say thanks for this thread. Interesting discussions.

Posted: **Mon May 28, 2012 3:27 am**

This is a fascinating thread. I have another question. One of the conlangs I am playing with will have a large number of noun classes. Do you know if natlangs with more than four noun classes tend towards head-marking (as in Swahili verb affixes) or to polysynthesis?

Posted: **Wed May 30, 2012 3:43 pm**

Australian languages with noun classes such as apparently Dyirbal tend to be dependent marking and polysynthetic!

Posted: **Wed May 30, 2012 4:14 pm**

Radagast revived wrote:Australian languages with noun classes such as apparently Dyirbal tend to be dependent marking and polysynthetic!

Polysynthetic in what sense?

Posted: **Thu May 31, 2012 5:46 am**

Jabechasqvi wrote:
Radagast revived wrote:Australian languages with noun classes such as apparently Dyirbal tend to be dependent marking and polysynthetic!
Polysynthetic in what sense?

in the sense that you'd exclude them on some bullshit criterion like you do eskimo-aleut

Posted: **Thu May 31, 2012 9:30 am**

Drydic Guy wrote:in the sense that you'd exclude them on some bullshit criterion like you do eskimo-aleut

Even though some professional linguists have also argued based on those same criteria that Eskimo-Aleut does not qualify as polysynthetic? What makes them reasonable but me full of shit when I am merely suggesting the same thing they've been arguing?

Posted: **Thu May 31, 2012 9:35 am**

presumably they are reasoning

Posted: **Thu May 31, 2012 11:25 am**

Pthug wrote:presumably they are reasoning

But we are putting forth precisely the same conclusions. You could argue that I reached the conclusion for the wrong reasons, but not that an otherwise sound conclusion becomes wrong when I say it.

But putting aside the Baker issue, what definition is Radagast using when he describes Dyirbal as polysynthetic?

Posted: **Thu May 31, 2012 2:02 pm**

Back to my question - out of the polysynthetic languages known to ZBBers, how many of them have multiple noun classes (i.e. >4)? and out of the multiple NC languages, what proportion of them are a) head-marking b) agglutinative and c) polysynthetic? Because my (non-specialist) intuition suggests gender and especially multiple NC may be a factor in making languages evolve towards synthesis and I wondered, if that was the case, how far might the process go?

Posted: **Thu May 31, 2012 2:19 pm**

Drydic Guy wrote:
Jabechasqvi wrote:
Radagast revived wrote:Australian languages with noun classes such as apparently Dyirbal tend to be dependent marking and polysynthetic!
Polysynthetic in what sense?
in the sense that you'd exclude them on some bullshit criterion like you do eskimo-aleut

I wouldn't call Dyirbal polysynthetic at all. Like most Pama-Nyungan languages it's very noun-heavy; there is not a tremendous amount of verbal morphology; and the morpheme-to-word ratio isn't particularly astronomical. If you're going to call it "polysynthetic", then it's certainly a different kind of polysynthesis than the kind I've described in this thread. You're not required to knee-jerk dismiss everything Eddy says ever, guys. Like, at least look at his post before deciding whether to mock it -- this was a completely valid question on his part.

Here's an example Dyirbal sentence:

balan dyugumbil banggul yarangu banggu yugunggu balgan
"The man hit the woman with a stick"
(the.CL:II.NOM woman(II).NOM the.CL:I.ERG man(I).ERG the.CL:IV.INSTR stick(IV).INSTR hit)

EDIT: That being said -- Eddy, Eskimo-Aleut languages are totally polysynthetic, basically everyone but Baker agrees on that, you don't even know what his arguments are against it ... so, stop making that argument. Appeal to authority is a logical fallacy; if you're going to argue that Eskimo-Aleut languages aren't polysynthetic, then do it using evidence, godddd.

Posted: **Thu May 31, 2012 4:18 pm**

Mornche Geddick wrote:Back to my question - out of the polysynthetic languages known to ZBBers, how many of them have multiple noun classes (i.e. >4)? and out of the multiple NC languages, what proportion of them are a) head-marking b) agglutinative and c) polysynthetic? Because my (non-specialist) intuition suggests gender and especially multiple NC may be a factor in making languages evolve towards synthesis and I wondered, if that was the case, how far might the process go?

I don't think polysynthesis and gender correlate very well (but I may be wrong; this isn't a systematic survey). Algonquian languages have just two genders; Eskimo has none IIRC. Ainu and Quechua don't have gender. Abkhaz and Abaza just have three.

The Bantu languages have high number of genders (the classic number I think is 10), but I wouldn't consider them polysynthetic. Tsova-Tush (Bats) has 8 (or less, some of the 'genders' are better seen as irregularities), and its verbs are highly agglutinative but I don't know if it's considered polysynthetic.

Think of it this way— a gender system generally can add only three morphemes max to the verb (for subject and two object agreement markers). That's enough to add some complexity but not the out-of-control affixation we call polysynthetic.

Posted: **Thu May 31, 2012 4:22 pm**

This may be a bit of a parentheses, but... how's about making a purely polysynthetic lang with just ablaut and other non-con morphology? wouldn't that be neat ? can it be done?

Posted: **Thu May 31, 2012 4:31 pm**

Jabechasqvi wrote:But we are putting forth precisely the same conclusions. You could argue that I reached the conclusion for the wrong reasons, but not that an otherwise sound conclusion becomes wrong when I say it.

Just a philosophicolinguistical remark: it becomes unsound, by definition, if you have concluded it for the wrong reasons. In order for something to be sound it has to be valid too.

Posted: **Fri Jun 01, 2012 8:25 am**

Torco wrote:This may be a bit of a parentheses, but... how's about making a purely polysynthetic lang with just ablaut and other non-con morphology? wouldn't that be neat ? can it be done?

That would probably be quite a messy affair, if possible at all. There are only so many phonemes that can be changed in so many ways, so you'd probably run out of forms if you try to do this.

Posted: **Fri Jun 01, 2012 8:34 am**

WeepingElf wrote:
Torco wrote:This may be a bit of a parentheses, but... how's about making a purely polysynthetic lang with just ablaut and other non-con morphology? wouldn't that be neat ? can it be done?
That would probably be quite a messy affair, if possible at all. There are only so many phonemes that can be changed in so many ways, so you'd probably run out of forms if you try to do this.

well, yes, but there's many ways in which a root, especially a longish root, might change in order to reflect things. imagine. plus there's ablaut, and irregular forms, and consonant gradation. maybe not exclusively non-con, but mostly ?

after all, what's the average amount of morphemes a polysynthetic lang has to have in order to be poly ?

Posted: **Fri Jun 01, 2012 9:32 am**

Torco wrote:This may be a bit of a parentheses, but... how's about making a purely polysynthetic lang with just ablaut and other non-con morphology? wouldn't that be neat ? can it be done?

I actually toyed with something like that once, after having a ZBB-inspired dream about the concept. I didn't get very far, though, for the reasons that WeepingElf suggests.

Posted: **Mon Oct 13, 2014 9:10 pm**

Great work, Whimemsz. You've put together well-researched, intriguing material. Thank you for all the time and effort you put into the project.

A few topics I'd very much like to find out more about:

- In head-marking polysynthetic languages without oblique cases, how are subtle shades of meaning in the domains of relationship, location, and time expressed? Are they more creative than using prepositions, postpositions, or set phrases like "in the centre of town"? If these items are marked on the verb, how complex are the systems in use, and how are multiple oblique phrases in one sentence handled?

(1) a. The boss bought a cup of coffee for all his employees.
b. The teacher made a request to the principal on behalf of her students.

(2) a. His friends arrived by train around ten o'clock in the morning.
b. The soldier killed the lion without using a weapon of any kind.

(3) a. The prime minister fired the minister of defense for his incompetence.
b. Schools were closed because of the snowstorm.
c. The government worked against its own interests by supporting the foreign rebellion.

(4) a. Most of the hikers came to the campsite with all their gear.
b. A few of the hikers came to the campsite without all their gear.

(5) a. The family sat in the waiting room for a long time.
b. I walked toward the house's front door but then realized it wasn't the right address.
c. We wandered into the living room when we got home early in the evening.
d. We walked up to the end of the train tracks and then turned around.

(6) a. The student and his friends worked during the summer.
b. The student and his friends vacationed from late July into early August.
c. The student and his friends practiced their instruments from the start of May until the end of August.

(7) a. Everyone trekked across the forest and swam across the river in five hours.
b. News of the king's assassination quickly traveled across the country.
c. The BBC can be heard around the world.
d. Houses and farms are scattered throughout the area.

(8) a. We ran past the shopping mall on our way home.
b. The student and his friends stayed up late beyond the start of the new school year.

(9) a. A strange man came out of the building.
b. A strange man walked away from the building, after standing near the entrance for a while.
c. We leaned the suitcases up against the wall, far out of reach of the rain.
d. The doctor has been up since eight o'clock in the morning.

(10) a. He stood on the roof of his house and looked out at the scenes of destruction in the city around him.
b. The helicopter hovered above the enemy air base.
c. The clouds floated in over the lake as we were swimming in to the beach.

(11) a. He hid under the bed.
b. A tunnel runs underneath the river.
c. A racoon came out from under the park bench as I sat down.

(12) He moved the bricks out of the warehouse and up the hill.

(13) We carried the piano down the stairs and into the basement.

- What kinds of morphophonology are common in polysynthetic languages? How do roots, affixes, or both change?

- What polysynthetic languages are more fusional and less agglutinative? In what areas of the nominal or verbal morphology are more or less likely to be fusional or agglutinative?

Posted: **Mon Oct 13, 2014 9:14 pm**

Radagast revived wrote:Non-true quantifiers are quantifiers that are not heads of a noun phrase as quantifiers usually are supposed to be - that is they do not form an actual phrase with the noun they modify and the noun does not express any agreement with the quantifier. There is also some semantic tricks about the scopes of quantifiers such as "some", "every", "each" that become complicated by the fact that they don't form phrase with the nouns but are simply co-adjuncts to a main phrasal head.

So with a true quantifyier you'd get a tree structure like this
Code: Select all
[[i] broke [two [cups]]
But in Nahuatl you'd get something like this [(it was) two][(it was) cups]

Can you recommend any links/papers on this subject or provide some illustrative natlang examples? I'm still not grasping this concept.

If Whimemsz or others know something about the topic or can point me to one or more resources on it, please feel free to chime in.

Posted: **Wed Oct 22, 2014 7:41 am**

Lakota has dative and benefactive affixes that add an argument to the verb for (1). The rest of your examples are just adverbial or postpositional phrases...

a. opȟétȟuŋ X buys Y : opȟékičatȟuŋ X buys Y for Z (dative)
Itȟáŋčhaŋ kiŋ | tȟawówaši kiŋ oyás'iŋ | wakȟályapi waŋžígžila | opȟéwičhakičatȟuŋ.
"The boss | all his employees | one coffee each | he bought it for them."

b. walá X makes a request : wakíčila X makes a request on behalf of Z (benefactive)
Waúŋspewičhakhiye kiŋ | tȟawáyawa kiŋ hená | itȟáŋčhaŋ kiŋ etáŋ | wawíčhakičila.
"The teacher | her students | from the principal | she makes a request on their behalf."

zompist bboard

Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices

Re: Polysynthesis for Novices