Have You Made a Polysynthetic Conlang?

A forum for all topics related to constructed languages
User avatar
Arayaz
roman
roman
Posts: 1225
Joined: 07 Sep 2022 00:24
Location: Just south of the pin-pen merger
Contact:

Have You Made a Polysynthetic Conlang?

Post by Arayaz »

It has long been one of my goals to create a polysynthetic conlang. Still, despite my best efforts, I have never been able to pass five or six morphemes per word in a naturalistic language.

My first attempt at a polysynthetic language was Uniita. It was from back when I was much less talented than I am now - I don't think I'd even heard of sound changes at the time. I had just learned about Inuktitut, and was hoping to try it out. What I ended up doing was deleting all the spaces between words and calling it polysynthetic.

Since then, I've made two or three (depending on how you count it) languages that I'm comfortable calling at least kind of synthetic: Ṛikuḍa H'Aŋos, its descendant Aodyəi, and the yet-to-be-named isolate language I began to make for my D&D campaign that never ended up happening.

Ṛikuḍa H'Aŋos had about five maximum morphemes per verb. Nouns could in theory have infinite (all adjectives had prefixed forms that could attach to nouns), but in practice this would be pretty hard. Aodyəi was slightly more synthetic, since mandatory evidentials were introduced, but it's not much. The isolate was a lot better, even though it was the first of the three: I used a very extensive noun class system and inflected the verbs for it. Each verb had hundreds of thousands of possible forms, compared to the wimpy 3,888 forms of each verb in Aodyəi. I also played with the syntax to try to use as few words as possible (e.g. if there were two subjects connected by "and," each individual one could be negated on its own), and I had a lot more inflections than some languages, but I don't think I ever reached twelve morphemes per word.

I created this thread to hear about other polylangs (what I call polysynthetic languages, since "polysynthetic" is a pain to type, at least on my keyboard) and to try to learn from them.

Also, this is the most morphemes I could get on a verb in Aodyəi, my favorite of the three:

Gwaomya fazdiŋesubawɨ.
short.wide-iness-river-iness 1pla.2sge-inc-fut-flee-ded
I deduce you will begin to flee us in the short, wide river.
Proud member of the myopic-trans-southerner-Viossa-girl-with-two-cats-who-joined-on-September-6th-2022 gang

:con: Ruykkarraber languages, Izre, Ngama, Areyaxi languages, ???, 2c2ef0
my garbage

she/her
User avatar
Man in Space
roman
roman
Posts: 1304
Joined: 03 Aug 2012 08:07
Location: Ohio

Re: Have You Made a Polysynthetic Conlang?

Post by Man in Space »

Yes, Kgáweq'.
Twin Aster megathread

AVDIO · VIDEO · DISCO

CC = Common Caber
CK = Classical Khaya
CT = Classical Ĝare n Tim Ar
Kg = Kgáweq'
PB = Proto-Beheic
PO = Proto-O
PTa = Proto-Taltic
STK = Sisỏk Tlar Kyanà
Tm = Təmattwəspwaypksma
Knox Adjacent
cuneiform
cuneiform
Posts: 172
Joined: 24 Oct 2022 04:34

Re: Have You Made a Polysynthetic Conlang?

Post by Knox Adjacent »

Have, no. But I'm working on one.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Have You Made a Polysynthetic Conlang?

Post by Creyeditor »

Polysynthetic as in having noun incorporation and affixes? Or having a lot of morphosyntactic categories expressed per "word"? Or as in having a lot of agglutinating affixes on a word? In either case yes, some sketches. In the last two cases: yes, and I might be working on one (polypersonal agreement, voice, aspect marked on verbs and definiteness, number, possessor, size marked on nouns). Might be below your cut-off point, idk.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
Arayaz
roman
roman
Posts: 1225
Joined: 07 Sep 2022 00:24
Location: Just south of the pin-pen merger
Contact:

Re: Have You Made a Polysynthetic Conlang?

Post by Arayaz »

My understanding of a "polysynthetic" language is that it has 12 or more morphemes per word.

Edit from far-future me: this is dumb, and I don't know where I got the idea.
Last edited by Arayaz on 26 Sep 2023 23:45, edited 1 time in total.
Proud member of the myopic-trans-southerner-Viossa-girl-with-two-cats-who-joined-on-September-6th-2022 gang

:con: Ruykkarraber languages, Izre, Ngama, Areyaxi languages, ???, 2c2ef0
my garbage

she/her
User avatar
Sequor
sinic
sinic
Posts: 438
Joined: 30 Jun 2012 06:13

Re: Have You Made a Polysynthetic Conlang?

Post by Sequor »

It's a vague but useful term. There's no hard cut-off like (the possibility of) having 12 morphemes (in the verb).
hīc sunt linguificēs. hēr bēoþ tungemakeras.
Ælfwine
roman
roman
Posts: 940
Joined: 21 Sep 2015 01:28
Location: New Jersey

Re: Have You Made a Polysynthetic Conlang?

Post by Ælfwine »

For the "lols" I once tried to make a polysynthetic romlang. Didn't go too far with it, though.
My Blog

A-posteriori, alternative history nerd
User avatar
Arayaz
roman
roman
Posts: 1225
Joined: 07 Sep 2022 00:24
Location: Just south of the pin-pen merger
Contact:

Re: Have You Made a Polysynthetic Conlang?

Post by Arayaz »

That's not that far-fetched. Arguments have been made that spoken French is polysynthetic.
Proud member of the myopic-trans-southerner-Viossa-girl-with-two-cats-who-joined-on-September-6th-2022 gang

:con: Ruykkarraber languages, Izre, Ngama, Areyaxi languages, ???, 2c2ef0
my garbage

she/her
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: Have You Made a Polysynthetic Conlang?

Post by eldin raigmore »

No, but I want to!
User avatar
kiwikami
roman
roman
Posts: 1203
Joined: 26 May 2012 17:24
Location: Oh, I don't know, I'm probably around here somewhere.

Re: Have You Made a Polysynthetic Conlang?

Post by kiwikami »

Alál is technically polysynthetic given the number of possible morphemes per word and the way in which its verbs are constructed (including polypersonal agreement), but most verbs tend to have only about 4-6 morphemes (root, agreement, aspect, voice/volition, maybe a derivational morpheme or two).

A more complex but realistically usable verb would look something like... mm... let's see if I can make this look vaguely pretty.

Ṭaıtḳâkattúaṭákamàḷaıxxìṭ·kmaakakkuûkurḳama.
ṭa<vı-tḳa-ŕkvr̀-ttúv-1vŕkv-mà<ḷaı>x>x-ìṭ-k-ma<aka-k<~ku>ú~kù>r̀-ḳ-ḳ-a
remember<3-s2-sAll-o6-oAll-egg<scrambled>>-NEG-purpose-eat<want-PST.EVID<~EXC>~EXT>-ITER-DSTR-V.A
Long ago, both of them apparently used to go about wanting to eat a whole half-dozen scrambled eggs one after another, in order to avoid remembering.

But that's not going to show up. Much.
Edit: Substituted a string instrument for a French interjection.

:eng: :mrgreen: | :fra: [:)] | ASL [:S] | :deu: [:|] | :tan: [:(] | :nav: [:'(]
Ælfwine
roman
roman
Posts: 940
Joined: 21 Sep 2015 01:28
Location: New Jersey

Re: Have You Made a Polysynthetic Conlang?

Post by Ælfwine »

Üdj wrote: 29 Oct 2022 04:14 That's not that far-fetched. Arguments have been made that spoken French is polysynthetic.
Yeah that was my motivation, a future French dialect that became completely agglutinating and polysynthetic.

Feel free to take the idea
My Blog

A-posteriori, alternative history nerd
Nortaneous
greek
greek
Posts: 674
Joined: 14 Aug 2010 13:28

Re: Have You Made a Polysynthetic Conlang?

Post by Nortaneous »

'Polysynthesis' is a vague typological term. The languages of the Swamp Continent generally come close in various ways. Hlu has extensive verbal morphology, but doesn't really have polypersonal agreement, obligatorily marking only subject and benefactor:

noechuehmaehlaeboqwiyayalekhi
[nɤȶʰɯm̥æɬæɓoʔwijɔjɔlekʰi]
2-PST-NEG-1.BEN-ABIL-CAUS-scatter-AND-COND
Had you not dispersed [them] for our benefit...

The one language known to have marking of both subject and object on the verb, Zot, is one of the least typically polysynthetic: person marking is essentially derivational, with e.g. direct nominalization of person-marked forms being permitted, and frequently suppletive, and there are about thirty verbs, generally with vague and abstract meanings. For example:

nyuj
[nju˧˩]
you hit me

zos
[ʐo˧˥]
I hit you

yvpxp
[jɵ˩pæ˩]
we hit her

zomqf
[ʐo˥mɒ˧˥]
I who hit you

wxn'yvt
[wæ̃˥jɵ˥]
caning

wxn'yvzof
[wæ̃˥jɵ˧ʐo˦˥]
I cane you

wxn'yvyisinmqf
[wæ̃˥jɵ˧ji˥ɕĩ˥mɒ˥]
he who canes them
Üdj wrote: 27 Oct 2022 23:57 What I ended up doing was deleting all the spaces between words and calling it polysynthetic.
not entirely invalid - cf. Coptic and Kiranti
Trailsend
moderator
moderator
Posts: 1451
Joined: 18 Aug 2010 05:22

Re: Have You Made a Polysynthetic Conlang?

Post by Trailsend »

Üdj wrote: 28 Oct 2022 13:59 My understanding of a "polysynthetic" language is that it has 12 or more morphemes per word.
Where did you find this definition? It's not one I'm aware of.

As far as I know there isn't a consensus on a definition for polysynthesis. There's Baker's idea that a polysynthetic language is one that fits a very particular criterion involving agreement marking between phrasal heads and arguments, but that one's pretty tailor-made for generative grammar theory. There are folks that say there's no good definition for "polysynthetic" because the term is too vague to really be useful. Some folks connect polysynthesis to particular operations like productive noun-incorporation, and some do make vague handwaves about polysynthesis meaning, like, super synthetic. But I haven't heard the number 12 given as like...a threshold for polysynthesis.


My main project Hiding Waters is polysynthetic. Incorporation is the primary driver for derivation, and content words have quite a lot of morphemes on them. Not arbitrarily many, though; they have a templatic structure, you can't do the kind of arbitrary post-base stacking you can do in Alutiiq or Inuktitut and company.

Sọ̀guóḳȯdọqjkukslọtạnotlhqhnạwhọ́ṇg.
s<ọ̀>-g<uó-ḳ<ȯ>dọq-jk-u-k-sl-ọ-tạn<o>tlh>q-hnạwh<ọ́>ṇg
AUX<ABL.INAN>-eat<LEAD.VOL-last_night<VIA.INAN>-1.PL-LEAD-PFV-CLF(fish)-INAN-house<LOC.INAN>>-spent<LAT.INAN>
Because of that, we wanted to spend last night in the house eating the fish until it was all gone.

The root of the word is g*q, "eat", and it has four other roots incorporated onto it:

- sọ̀-, a semantically blank root that in this position means that something that was already stated caused this to happen
- -ḳȯdọq-, "last night", which in this position means the action happened throughout the full duration of last night
- -tạnotlh-, "house", which in this position means the house was the location of the action
- -hnạwhọ́ṇg, "spent/gone/exhausted/used up", which in this position means everything being used up was the end point of the action

-jku- marks the agent of the action: first person plural, "we"

-slọ- marks the patient of the action, something carrying the "fish" classifier.
任何事物的发展都是物极必反,否极泰来。
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Have You Made a Polysynthetic Conlang?

Post by Creyeditor »

I just noticed that my polysynthetic lang does not have 12 morphemes, at least in the sentences that I have so far. It has noun incorporation and agreement though.

:con: K'úk'uts'

Wátsmuk'útásk'op'táptsatku__s'énits's'ats'íf'ets'iki'áptsatki.
[wa˦ts.mu˧.kʼu˦.ta˦s.kʼo˧pʼ.ta˦ˈ.tsa˧t.ku˧__sʼe˦.ni˧tsʼ.sʼa˧.tsʼi˦.fʼe˧.tsʼi˧.ki.ʔa˦ˈ.tsa˧t.ki˧]
wáts-muk'útás-k'ó-p't-á-p-tsa-tku_s'éníts'-s'ats'íf'ets'-iku-á-p-tsa-tku
tree-go-FAST-AROUND-FEM-HUM-CHILD-PST__ball-kick-AGAINST-FEM-HUM-CHILD-PST
The girl ran around the tree and kicked the ball.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
Omzinesý
mongolian
mongolian
Posts: 4081
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Have You Made a Polysynthetic Conlang?

Post by Omzinesý »

Nortaneous wrote: 07 Nov 2022 00:14
The one language known to have marking of both subject and object on the verb, Zot, is one of the least typically polysynthetic: person marking is essentially derivational, with e.g. direct nominalization of person-marked forms being permitted, and frequently suppletive, and there are about thirty verbs, generally with vague and abstract meanings. For example:

nyuj
[nju˧˩]
you hit me

zos
[ʐo˧˥]
I hit you

yvpxp
[jɵ˩pæ˩]
we hit her

zomqf
[ʐo˥mɒ˧˥]
I who hit you

wxn'yvt
[wæ̃˥jɵ˥]
caning

wxn'yvzof
[wæ̃˥jɵ˧ʐo˦˥]
I cane you

wxn'yvyisinmqf
[wæ̃˥jɵ˧ji˥ɕĩ˥mɒ˥]
he who canes them
Zot sounds interesting. I didn't find it with quick googling. Where is it spoken. Or is it a conlang of yours? In any case I'm interested to see more.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
LinguoFranco
greek
greek
Posts: 613
Joined: 20 Jul 2016 17:49
Location: U.S.

Re: Have You Made a Polysynthetic Conlang?

Post by LinguoFranco »

I've made a few sketches of polysynthetic conlangs, but scrapped them pretty early due to boredom. I don't think it had anything to do with polysynthesis itself, I think it's just my ADHD that leads me to abandoning most of my projects before making any real progress.

The closest I've come is with my personal conlang, as it has rich verb conjugation and polypersonal agreement, but even then, I think it's more agglutinative than polysynthetic.
User avatar
Mándinrùh
cuneiform
cuneiform
Posts: 161
Joined: 21 Aug 2016 20:37
Location: New England

Re: Have You Made a Polysynthetic Conlang?

Post by Mándinrùh »

Nortaneous wrote: 07 Nov 2022 00:14 nyuj
[nju˧˩]
you hit me

zos
[ʐo˧˥]
I hit you

yvpxp
[jɵ˩pæ˩]
we hit her
You're doing that Hmong thing, aren't you, with the illegal coda consonants to mark tone? And also that Thai Romanization thing where they use <v> and <x> as vowels?
Creator of Image Atili
My website | My blog
User avatar
Man in Space
roman
roman
Posts: 1304
Joined: 03 Aug 2012 08:07
Location: Ohio

Re: Have You Made a Polysynthetic Conlang?

Post by Man in Space »

I see you've become familiar with Nort's phonological and orthographic flair. He has insane romanizations that I aspire to and fail to reach—I've always liked how he writes his languages.
Twin Aster megathread

AVDIO · VIDEO · DISCO

CC = Common Caber
CK = Classical Khaya
CT = Classical Ĝare n Tim Ar
Kg = Kgáweq'
PB = Proto-Beheic
PO = Proto-O
PTa = Proto-Taltic
STK = Sisỏk Tlar Kyanà
Tm = Təmattwəspwaypksma
chris_notts
hieroglyphic
hieroglyphic
Posts: 46
Joined: 14 Feb 2012 00:49

Re: Have You Made a Polysynthetic Conlang?

Post by chris_notts »

Kind of? I think my most recent project, which started off as Mayan inspired with bits of Navajo, Algonquian etc., is closest. Verbs are inflected for agreement with two arguments, TAM, polarity, motion, for various subordinate/adverbial functions, plus it has a preference for complex predicates (coverb+verb, or medial+final in Algonquian terminology) and it has noun incorporation. An example:

nurinkóhsatzól in tacha' o'
nu-r-ink-óh-satz-ól in tacha' o'
SUB-3-PROX-NEG-medicine-trans.liquid DET man DEIC
"Since the man didn't bring medicine"
"The man not having brought medicine, ..."

Verbs don't tend to get much bigger than that though (6-9 syllables), and the inflectional potential of other word classes is pretty limited by comparison.
Nortaneous
greek
greek
Posts: 674
Joined: 14 Aug 2010 13:28

Re: Have You Made a Polysynthetic Conlang?

Post by Nortaneous »

Mándinrùh wrote: 09 Dec 2022 01:29 You're doing that Hmong thing, aren't you, with the illegal coda consonants to mark tone? And also that Thai Romanization thing where they use <v> and <x> as vowels?
I haven't seen those Thai romanizations, although I vaguely remember Google doing something odd with <x>.

I had a conlang with a romanization based on Nuosu (where <y v> are vowels and tones are marked <-p -t -x>), then H13 saw that and made one with <v w> as vowels (I assume based on Zhuang), then someone on IRC said why don't you use <q> for /Q/, and after that we found out about Natqgu and I added a schwa. So it's /a ɒ e ə o i u ɿ v̩/ <a q e z o i u y v>.

Zot, having long been a neighboring language to the more esteemed and widely spoken Zzyxwqnp, naturally has a lot of influence, so the orthographic norms should have something in common - hence tone letters (except only five word-level tone contours are possible, so you only need tone markers at the end of the word, as in wxn'yvyisinmqf) and <q>. <v> for /ɵ/ is because that's how Zzyxwqnp /v̩/ is borrowed.
Post Reply