Polysynthetic Conlang

jburke · Post by **jburke** » Tue Sep 09, 2003 9:39 pm

Eddy the Great wrote:I was surprised that this existed in Cheyenne considering how you said that many things English expresses aren't in other langs and vice versa.

Well, you're not using a that-phrase in the Cheyenne; you're essentially saying two separate things: 'He's so bright, he X...' It's akin to a topic-comment structure, but it's called something else (don't recall the name).

But those triple consonant clusters still throw me for a loop; how about using a rule of vowel epenthesis or deleting one consonant for every three?

Aurora Rossa · Post by **Aurora Rossa** » Tue Sep 09, 2003 9:57 pm

But those triple consonant clusters still throw me for a loop; how about using a rule of vowel epenthesis or deleting one consonant for every three?

Compared to Georgian, it's actually quite reasonable, although it's still tough. Triple clusters are relatively rare and normally, the worst you'll see is tsfa, which is an africate + a fricative. I'm working on a conlang that is more vowel-heavy.

jburke · Post by **jburke** » Wed Sep 10, 2003 6:15 am

Compared to Georgian, it's actually quite reasonable,

Well, Georgian is a natlang with an historical phonology that led to its impressive clusters. The same is true of the Salishan languages, which (according to some accounts) split off from Proto Algonquian's ancestor millennia ago. The Salishan languages developed their tolerance for crazy consonant clusters because of vowel reduction and the dropping-out of schwas (which I understand is also true of Georgian and other Caucasian languages). If T?l@uigo is a conlang in the context of your world, I'd have to wonder why the creator of it would dream of making the phonology so hard to pronounce.

Aurora Rossa · Post by **Aurora Rossa** » Wed Sep 10, 2003 4:02 pm

If T?l@uigo is a conlang in the context of your world, I'd have to wonder why the creator of it would dream of making the phonology so hard to pronounce.

For one thing, initial l's are syllabic, so the lka clusters you may see aren't that hard, and since most roots and affixes end in a vowel, they bare the dificulty of many clusters. Also, it's a relict from a time when I had only 17 phonemes so I had very permissive phonotactics to make the best use of them. Now that isn't so necessary. Now I don't need that many clusters.

jburke · Post by **jburke** » Wed Sep 10, 2003 4:16 pm

Eddy the Great wrote:
If T?l@uigo is a conlang in the context of your world, I'd have to wonder why the creator of it would dream of making the phonology so hard to pronounce.
For one thing, initial l's are syllabic, so the lka clusters you may see aren't that hard, and since most roots and affixes end in a vowel, they bare the dificulty of many clusters. Also, it's a relict from a time when I had only 17 phonemes so I had very permissive phonotactics to make the best use of them. Now that isn't so necessary. Now I don't need that many clusters.

The best way to create a phonetically distinct language is to choose a small phoneme inventory and really milk it. Cheyenne does this; it has only 14 phonemes, and there's a sound and feel to its words that's unmistakable (e.g., you knew that eve's'he was Cheyenne without my ever telling you). The same is true of Mohawk. I choose a very strict set of phonological rules for Noyatukah because I wanted that "phonetically claustrophobic" feel; I wanted a language whose words would be unmistakable.

Aurora Rossa · Post by **Aurora Rossa** » Wed Sep 10, 2003 6:06 pm

K'?latğsuf/imi.
K'?la-tğsu-f/i-mi.
building-around(right)-walk-3SSA
He is walking to the other side of the building by going around the right side.

I have a problem. I need to add at least 2 new pseudocases(They act as cases on noun roots, but can be used on verbs) but I already have 35 and thus it is already pushing for kitchen sinker status. How do I get around this?

Dudicon · Post by **Dudicon** » Thu Sep 11, 2003 2:10 pm

Eddy the Great wrote:I have a problem. I need to add at least 2 new pseudocases(They act as cases on noun roots, but can be used on verbs) but I already have 35 and thus it is already pushing for kitchen sinker status. How do I get around this?

I know a way--get rid of your cases. You seem to frequently (and sometimes viciously) criticize IE and all it stands for, promoting clicks and polysynthesis and everything unusual, so why do you still use cases? In the big scheme of things, they're one of the most IE-like things you could use, especially in the context of the rest of your lang.

Aurora Rossa · Post by **Aurora Rossa** » Thu Sep 11, 2003 4:07 pm

I know a way--get rid of your cases. You seem to frequently (and sometimes viciously) criticize IE and all it stands for, promoting clicks and polysynthesis and everything unusual, so why do you still use cases? In the big scheme of things, they're one of the most IE-like things you could use, especially in the context of the rest of your lang.

I don't viciously criticize IE, only the grammatical gender. If I got rid of cases, how would I express relationships and still keep the lang poly?

Dudicon · Post by **Dudicon** » Thu Sep 11, 2003 4:29 pm

Eddy the Great wrote:
I know a way--get rid of your cases. You seem to frequently (and sometimes viciously) criticize IE and all it stands for, promoting clicks and polysynthesis and everything unusual, so why do you still use cases? In the big scheme of things, they're one of the most IE-like things you could use, especially in the context of the rest of your lang.
I don't viciously criticize IE, only the grammatical gender. If I got rid of cases, how would I express relationships and still keep the lang poly?

Well, for instance, Jeff's Noyatukah gets along just fine without any case, as do plenty of other languages. I don't see how polysynthesis necessitates having a case system.

Aurora Rossa · Post by **Aurora Rossa** » Thu Sep 11, 2003 4:31 pm

Well, for instance, Jeff's Noyatukah gets along just fine without any case, as do plenty of other languages. I don't see how polysynthesis necessitates having a case system.

He must use prepositions or something.

Dudicon · Post by **Dudicon** » Thu Sep 11, 2003 4:43 pm

Eddy the Great wrote:He must use prepositions or something.

Not exactly--as usual, polysynthetic languages split things up differently than IE languages, not just rearrange them. For example, things indicating location and movement, such as IE prepositions like in and onto are represented in Noyatukah with locatives and directionals, morphemes within the verb phrase. Possession is handled with possessive affixes, as if in English we said things like book-my for "my book." And the ideas corresponding to the "core cases" of IE, ie. agent, patient, and the like are handled by the agreement-marking finals, and specifically on the complex interplay between the proximate and obviate persons in many instances, when more than one verb is involved in a phrase.

jburke · Post by **jburke** » Thu Sep 11, 2003 5:23 pm

Dudicon wrote:
Eddy the Great wrote:He must use prepositions or something.
Not exactly--as usual, polysynthetic languages split things up differently than IE languages, not just rearrange them. For example, things indicating location and movement, such as IE prepositions like in and onto are represented in Noyatukah with locatives and directionals, morphemes within the verb phrase. Possession is handled with possessive affixes, as if in English we said things like book-my for "my book." And the ideas corresponding to the "core cases" of IE, ie. agent, patient, and the like are handled by the agreement-marking finals, and specifically on the complex interplay between the proximate and obviate persons in many instances, when more than one verb is involved in a phrase.

Thanks, Dudicon. You're essentially right: polysynthetic langauges cannot be described as having cases in the most rigorous sense (though there is a larger, vaguer sense in which all languages, even isolating ones, have "case"). As for Noyatukah locatives, yes, these attach to initials (incorporated forms of nominals) within the verb; a notion like "upon the hill" would only ever appear as a locative inside the verb, never as a free-standing element (only subjects/agents and objects/patients ever appear as free-standing elements outside a verb).

Possessives are a little trickier. I just formalized the lists of them yesterday, after much internal debate over their derivation. They also can attach to initials, but also can appear infixed to a final in a free-standing nominal. Noyatukah recognizes two kinds of possession: alienable and inalienable; the lists go:

INALIENABLE

1st singular: -n?-
1st pl. inc: -w?-
1st pl. exc. -ni-
2nd formal: -w?-
2nd informal: -m?-
3rd animate: -s?-
3rd inanimate: -m?-
4th: -s?-
Indefinite: -n?-

ALIENABLE

1st singular: -?no-
1st pl. inc: -?ya-
1st pl. exc. -?ni-
2nd formal: -?ye-
2nd informal: -?mo-
3rd animate: -?se-
3rd inanimate: -?me-
4th: -?so-
Indefinite: -?ne-

As you can see, the number distinction collapses for possessives in the 3rd and 4th persons (2nd does not recognize number at all, and 4th does not recognize animacy at all); and further that the alienable possessives are related to their inalienable counterparts.

The inalienable possessives are derived from the subjective finals:

1st singular: -no
1st pl. inc: -wam
1st pl. exc.: -nim
2nd formal: -?we
2nd informal: -?mo
3rd animate sing: -nan
3rd inanimate sing: -mex
3rd animate pl: -nasho
3rd inanimate pl:-mesho
4th singular: -so
4th plural: -sotlo
Indefinite: -newe

Aurora Rossa · Post by **Aurora Rossa** » Thu Sep 11, 2003 6:33 pm

If I have 36 case-like endings, my lang won't be a kitchen sinker, will it?

Drydic · Post by **Drydic** » Fri Sep 12, 2003 1:52 am

Eddy the Great wrote:If I have 36 case-like endings, my lang won't be a kitchen sinker, will it?

Not if you do a good job on it. 36 is probably high on the scale of acceptable conlangs rather than Kitchen Sinkers, but if they look right, feel right, and interact well (with each other and the verb system) then it can be done.

jburke · Post by **jburke** » Fri Sep 12, 2003 2:48 am

Drydic_guy wrote:
Eddy the Great wrote:If I have 36 case-like endings, my lang won't be a kitchen sinker, will it?
Not if you do a good job on it. 36 is probably high on the scale of acceptable conlangs rather than Kitchen Sinkers, but if they look right, feel right, and interact well (with each other and the verb system) then it can be done.

Ran, I think, best expressed the definition of a Kitchen Sinker as a language with much bulk and little depth. Thirty-six case-like endings pales in comparison to Mohawk's (or Noyatukah's) arsenal of locatives, postverbs and misc. inflections. Remember that a poly language is one where almost all grammatical relations are expressed by inflections on the verb--so naturally, you're going to have a truckload of inflections; this is no different from English's assortment of prepositions, adverbs, etc.

What makes a conlang a KSC is when it tries to cram too many unnecessary features in, such as making all of the pronominal distinctions that Mark lists in the LCK; or having umpteen different tenses; etc. Use the features that are required for the language to make sense to its speakers and no more; this is more than a little artistic, so there are no quick answers, but I suggest that if you know your speakers well, you'll know when you're going too far.

Aurora Rossa · Post by **Aurora Rossa** » Fri Sep 12, 2003 5:32 pm

Ran, I think, best expressed the definition of a Kitchen Sinker as a language with much bulk and little depth. Thirty-six case-like endings pales in comparison to Mohawk's (or Noyatukah's) arsenal of locatives, postverbs and misc. inflections. Remember that a poly language is one where almost all grammatical relations are expressed by inflections on the verb--so naturally, you're going to have a truckload of inflections; this is no different from English's assortment of prepositions, adverbs, etc.

How do locatives work?

jburke · Post by **jburke** » Fri Sep 12, 2003 9:01 pm

Eddy the Great wrote:How do locatives work?

They are nominals + markers that tell where the action of the verb happens. In every language I know, they're incorporated in a different way from objects or subjects; Noyatukah incorporates them between the postverb and final, and they can be freely ordered with medials and directionals.

Aurora Rossa · Post by **Aurora Rossa** » Fri Sep 12, 2003 9:44 pm

They are nominals + markers that tell where the action of the verb happens. In every language I know, they're incorporated in a different way from objects or subjects; Noyatukah incorporates them between the postverb and final, and they can be freely ordered with medials and directionals.

So the k'?lats@i- id k'?lats@iksami(He is inside the building) would be a locational as -ts@i indicates being inside?

jburke · Post by **jburke** » Fri Sep 12, 2003 9:47 pm

Eddy the Great wrote:
They are nominals + markers that tell where the action of the verb happens. In every language I know, they're incorporated in a different way from objects or subjects; Noyatukah incorporates them between the postverb and final, and they can be freely ordered with medials and directionals.
So the k'?lats@i- id k'?lats@iksami(He is inside the building) would be a locational as -ts@i indicates being inside?

Yeah, that's a locative. (Though I don't know why you insist on having a copula in your language; why not use real verbs instead?)

Aurora Rossa · Post by **Aurora Rossa** » Fri Sep 12, 2003 10:09 pm

Yeah, that's a locative. (Though I don't know why you insist on having a copula in your language; why not use real verbs instead?)

I'm not sure how this would be expressed without a copula. I'm renaming my conlang to K?t?l@u?kn?ukus?.

K?t?l@u?kn?ukus?
K?-t?l@u-?-kn?-u-ku-s?
habitual-person-plural-speak-passive-3POA-3SSI
It is spoken by the Terps.

Notes: kn? means to say soemthing, or to speak using something. The Terps, members of Terra Pvlchra, call themselves the T?l@u? or the Tu'l?m?. Tu'li is a verb, to be a Terp or do what a Terp does and basically like the lang Mark mentions when suggesting how to get rid of nouns, verbs are used for what country you are from and what religion you practice, as in: Klistimi-He is Christian. /u'ami-He is Jewish.

jburke · Post by **jburke** » Fri Sep 12, 2003 10:13 pm

I'm not sure how this would be expressed without a copula.

'He-insides-the-building'. Cheyenne has a root, -hmeov-, that means 'inside' and is used just like that.

I'm renaming my conlang to K?t?l@u?kn?ukus?.

K?t?l@u?kn?ukus?
K?-t?l@u-?-kn?-u-ku-s?
habitual-person-plural-speak-passive-3POA-3SSI
It is spoken by the Terps.

How about just 'The-Terps-speak'? More succinct, less reliant on English-style expression.

Aurora Rossa · Post by **Aurora Rossa** » Fri Sep 12, 2003 10:28 pm

'He-insides-the-building'. Cheyenne has a root, -hmeov-, that means 'inside' and is used just like that.

I suppose that could work.

How about just 'The-Terps-speak'? More succinct, less reliant on English-style expression.

That could also work, although the attention is drawn to the Terps rather than what they speak. The passive really doesn't have much of a purpose when you look at it your way.

jburke · Post by **jburke** » Sat Sep 13, 2003 2:26 am

How about just 'The-Terps-speak'? More succinct, less reliant on English-style expression.

That could also work, although the attention is drawn to the Terps rather than what they speak. The passive really doesn't have much of a purpose when you look at it your way.[/quote]

The problem is that you're projecting your English thinking habits onto your very different language; and even trying to mimick English syntax with it. If 'speak' is the root of the word, speakers will identify _that_ as the main idea, as the emphasis, esp. if you incorporate 'Terps' (effectively backgrounding it). Now, if you were to use a free-standing noun for Terps, and place it before the verb, suddenly you draw more attention to that noun than to the verb (and its root 'speak').

Mecislau · Post by **Mecislau** » Sat Sep 13, 2003 7:27 am

How long can the words for a single noun get in polysynthetic languages? I'm just translating a few sentences, and I found the word for "worm" is weno-ranitléwatlarasivíto?axo, meaning "it crawls through the earth" when English-ized (The dash should be a macron over the vowel before it). The length comes for the word for "earth" itself, watlarasivíto?axo. Is this unrealistic for a single noun?

jburke · Post by **jburke** » Sat Sep 13, 2003 7:40 am

Maknas wrote:How long can the words for a single noun get in polysynthetic languages? I'm just translating a few sentences, and I found the word for "worm" is weno-ranitl?watlarasiv?to?axo, meaning "it crawls through the earth" when English-ized (The dash should be a macron over the vowel before it). The length comes for the word for "earth" itself, watlarasiv?to?axo. Is this unrealistic for a single noun?

In the Algonquian languages, when a noun is incorporated into a larger word, it is typically "abbreviated"; Cheyenne's medial forms are an example of this phenomenon (each noun has a free-standing form and one or more medial forms--sometimes the medials distinguish between singular and plural, sometimes not). The "abbreviations" are linked to a longer, usually verbal, word, and derive phonologically from that word.

In Noyatukah, the incorporated forms are called initials. E.g., in manwaesheamewaanasho, the manwa- element is the plural proximate initial of the word for 'ribbon snakes', emanosawasanasho.