Hoskh

A forum for all topics related to constructed languages
Post Reply
HoskhMatriarch
mayan
mayan
Posts: 1779
Joined: 16 May 2015 17:48

Hoskh

Post by HoskhMatriarch » 12 Dec 2015 05:35

I'm making a new thread since I've redone everything to try to make it naturalistic.

Here's the sort of ridiculous phoneme inventory that supposedly looks like German and Arabic but with aspiration (I'm still tweaking this some, mostly with the diphthongs):

Consonants:

/m n ŋ/ <m n ŋ>
/p pʰ b t tʰ d k k͡x g q q͡χ ʡ/ <p ph b t th d k kh g ⱪ ⱪh ḳ>
/p͡f t͡s t͡ʃ d͡ʒ/ <pf z zj dj>
/f v s z ʃ ʒ x ɣ χ ʜ h/ <f w/v s ż sj żj ch gh ⱨ ħ h>
/j l ʀ/ <j l r>


Vowels:

/a aː æ æː ɛ eː ɶ øː ɪ iː ʏ yː ɔ oː ʊ uː/ <a a ä ä e e ö ö i i ü ü o/å o u u>
/aɪ æɪ aʊ ɔʏ ɔʊ ɔa/ <ai äi au äu åu å>

So there's the obscene phoneme inventory. But, you may be wondering, why are different sounds represented by the same letters? Is this English? It's a romanization, it's supposed to be logical! And, it is. Vowels before consonant clusters or consonants that are written doubled are short, and other vowels are long. Vowels before doubled consonants can still be long if they're written doubled, which is common with affixes. The reason for this is that originally vowel length was only allophonic, and vowels before clusters, geminates, and voiceless consonants were allophonically short, and the others were allophonically long. I tried to make my language look like it's natively written in Latin by introducing etymological quirks like that. I probably could have done better than making the geminate <k> written as <ck> though, even though I really like how it looks.

Phonology:

Obstruents are devoiced in the coda position (not just word-finally, but any syllable coda).
A syllable is something like (C)(C)(C)(C)(V)(C)(C)(C)(C). The vowel is optional, but no syllable may begin with one, so in a sense the consonants are not. Consonants may be the nucleus, even in complex syllables (but generally not quite as complex as with a vowel nucleus), and syllabic consonants are extremely common.
When a syllable or word is written as starting with a vowel, there's actually a glottal stop. Thus, no word begins with a vowel, and no vowels may be in hiatus.
There is no nasal assimilation. /ŋ/ is [ɴ] before uvular stops, however.
There is no voicing assimilation either, but considering it only makes a difference in the syllable onset, that's not such a big deal.
/v/ is an approximant somewhat more than half the time, as in, it appears in places like /w/ would. A smaller, but significant, portion of the time it actually patterns as a fricative. This is because originally there was no /v/ (or any other voiced fricative), only /w/, then /w/ became /v/, then there was more /v/ because b > v V_V.

Sound changes:

Here is a giant page of sound changes: http://conlang.wikia.com/wiki/Hoskh/Sandbox
I'm not sure how many centuries or millennia that is, but it starts out with five vowels and ends up with 22 so probably a lot. It probably doesn't have enough umlaut and lenition.

I'm not really sure where to start with the grammar so here's a table of pronouns:

1S 1P 2S 2P 3S F 3S M 3S N 3P F 3P M 3P N 5S
N akh jon sa wan thu thi thoss chu chi chon rin
A khen jonṇ sach wanṇ thuch thin thoss chun chin chon rich
D khor jor sern worn thurn thel thor chul chir chor rurn
G khes jom sas waans thos this thosas chus chis chom ros
L khom jom sas wam thos thes thom chum chim chom rom

I'll also post the numbers as soon as I come up with some, which I haven't yet. I kind of want to model them on Hindi because Hindi numbers are so wonderfully weird, even better than German, Old English, or Danish, but that means I'll be coming up with basically seperate roots for 1 to 100, which will probably be half as hard as coming up with logograms.

Script:

It's going to be a logography, but there's not much progress so far, except it'll have lots of ligatures and use bold lines like blackletter writing and it's for a "polysynthetic" language (about as synthetic as Georgian from what I've seen, but not quite as synthetic as Mohawk) that's nothing like what people normally associate with logographies (aka Chinese languages that are stereotyped as having monosyllabic words and don't have verbs like rezjmäcktŋüültzjmackrarŋulŋochwṇchin and long Germanic and Sanskrit style compound nouns left and right).

CALS: http://cals.conlang.org/language/hoskh/feature/

As you can see from the CALS page, verbs conjugate for 11 things. Don't even ask me for a verb conjugation table, since any verb can have thousands of forms. And those forms are pretty fusional too, they're just fusional in really predictable ways, for all the umlaut and ablaut and lenition.

If you have any questions, feel free to ask them, even if it's just "why do you need so many vowels and consonants anyways" or "how many words for the are there".

Ütrṇnönn, nönn otżorriinnönn? Ⱨoss sais ⱨi ritt otlöszoss.
[ˈʔʏt.ʀn̩ˌnɶn nɶn ˈʔɔt.zɔːˤˌʀiːn.nɶn χɔ saɪ̯s χiː ʀɪt ˈʔɔt.lœsˌt͡sɔs]
üt-rṇ-nönn nönn ot-żor-riin-nönn sj-oss ⱨ-i ritt ot-lös-zoss
be-INDEF.AGT-Q or be-NEG-INDEF.AGT-Q that-NEUT there-ADV the-MASC ask/NOM be-INFER-3S.NEUT
"To be, or not to be? That is the question."

(That's what happens when you don't have infinitives.)
Last edited by HoskhMatriarch on 12 Dec 2015 19:35, edited 1 time in total.
No darkness can harm you if you are guided by your own inner light

User avatar
druneragarsh
greek
greek
Posts: 619
Joined: 01 Sep 2015 14:56
Location: Finland

Re: Hoskh

Post by druneragarsh » 12 Dec 2015 12:13

/p pʰ b t tʰ d k k͡x g q q͡χ ʡ/ <p ph b t th d k kh g ⱪ ⱪh>
I think you're missing a romanization for ʡ.

Even if your verbs inflect for everything under the sun, could you give us some simple conjugational tables, eg Indicative Present, Indicative Past, Imperative, etc, maybe with some description on how the fusional ablauts etc work?

I'd be interested in hearing some spoken Hoskh. Do you have a sample you could upload?
drúne, rá gárš
drun-VOC I.ERG read

List of conlangs with links!
Refer to me with any sex-neutral (or feminine) 3s pronoun, either from English (no singular they please, zie etc are okay) or from one of your conlangs!
CWS

cromulant
cuneiform
cuneiform
Posts: 157
Joined: 13 Jan 2012 00:04

Re: Hoskh

Post by cromulant » 12 Dec 2015 16:32

HoskhMatriarch wrote:A syllable is something like (C)(C)(C)(C)(V)(C)(C)(C)(C).
So, everything is optional?

If CCCCCCCC (8 consonants) is a valid syllable, why stop there, since it can be followed by another "syllable" consisting only of consonants?

If the nucleus is a syllabic consonant, where in the (C)(C)(C)(C)(V)(C)(C)(C)(C) can be the nucleus be?

Word-medially, how are syllable boundaries determined, given a string of consonants which could be part of either a coda or an onset?

What are the phonotactics beyond this, i.e. the rules for which consonants can go where and adjacent to which other consonants?

User avatar
Micamo
MVP
MVP
Posts: 7201
Joined: 05 Sep 2010 18:48
Contact:

Re: Hoskh

Post by Micamo » 12 Dec 2015 17:50

cromulant wrote:If CCCCCCCC (8 consonants) is a valid syllable, why stop there, since it can be followed by another "syllable" consisting only of consonants?
I don't know the answer for Hoskh, but I analyze Mithara syllables as (C)(V)(C) because of distributives: This is the maximum length of material that can be reduplicated to form distributives.

chkay (berry) -> kaychkay (mixed berries)
lhuk (fish) -> lhuxlhak (fish of different species)
sxs (reindeer fat) -> sixsxs (reindeer fat spread out over a surface)
My pronouns are <xe> [ziː] / <xym> [zɪm] / <xys> [zɪz]

My shitty twitter

HoskhMatriarch
mayan
mayan
Posts: 1779
Joined: 16 May 2015 17:48

Re: Hoskh

Post by HoskhMatriarch » 12 Dec 2015 20:07

druneragarsh wrote:
/p pʰ b t tʰ d k k͡x g q q͡χ ʡ/ <p ph b t th d k kh g ⱪ ⱪh>
I think you're missing a romanization for ʡ.

Even if your verbs inflect for everything under the sun, could you give us some simple conjugational tables, eg Indicative Present, Indicative Past, Imperative, etc, maybe with some description on how the fusional ablauts etc work?

I'd be interested in hearing some spoken Hoskh. Do you have a sample you could upload?
OK, I fixed that.

Well, the verb inflection works by having semi-agglutinative affixes, then they fuse in different but really regular ways. I'm not sure what a "simple" conjugational table is, but here's some examples of how things fuse:

chaff - to run
choffuch - ran (the u rounds the a)
chaffchuch - was running (there used to be a reduplicated vowel blocking any umlaut to the stem)
choffth - she runs (the feminine nominative pronoun is thu, and the unstressed vowel is dropped)
chäffth - he runs (the masculine nominative pronoun is thi, and the unstressed vowel is dropped)
choffuchth - she ran
choffüchth - he ran

The neuter conjugations tend to give giant obstruent clusters, like:
choffochths - it ran


I'll upload some samples once I get better at speaking it. For now I make it sound too much like some combination of German and some Semitic language or something, unless I say it with a bad "English" accent. Of course, it probably will sound somewhat like that no matter what I do thanks to that sound inventory, but I can make it sound different ways by playing with fine details that don't show up in broad transcription.

cromulant wrote:
HoskhMatriarch wrote:A syllable is something like (C)(C)(C)(C)(V)(C)(C)(C)(C).
So, everything is optional?

If CCCCCCCC (8 consonants) is a valid syllable, why stop there, since it can be followed by another "syllable" consisting only of consonants?

If the nucleus is a syllabic consonant, where in the (C)(C)(C)(C)(V)(C)(C)(C)(C) can be the nucleus be?

Word-medially, how are syllable boundaries determined, given a string of consonants which could be part of either a coda or an onset?

What are the phonotactics beyond this, i.e. the rules for which consonants can go where and adjacent to which other consonants?
Not everything is optional. You have to have a sound to have a syllable. Also, giant strings of consonants don't really happen much. One, the maximal syllable is not that common (just like how the only maximal syllable in English is strengths and the average English word is more like sit or drag), and two, vowel deletion only happens in historically unstressed syllables, which happen in clitics and polysyllabic words only. The consonant nucleii are the C right after the optional V. A pretty average polysyllabic word is [ˈʀeːˌt͡ʃmækt.ŋyːltˌt͡ʃmak.ʀaːˤˌŋuːl.ŋɔxˌvn̩.xɪn]. As you can see, there's lots of complex codas and odd clusters, but there are no strings of 24 consonants, and while the syllabic consonants are fairly common, the syllables they occur in tend to be simpler than those with a vowel nucleus ([vn̩] vs. [t͡ʃmækt]). Shorter words tend to be more like [t͡ʃnɪt.ŋn̩] and there isn't even a phonetic four-consonant coda so far (although there are phonemic ones like /saʀntʰs/ [saːˤntʰs] "it is guilty". Yes, this is a non-rhotic language).

The phonotactic rules are extremely complicated. There definitely are rules, but it might be hard to see that from what kinds of clusters there are. The main defining phonotactic feature is the complete lack of intervocalic stops other than the glottal stop, which I can't really get to work with word generators, which will always spit out illegal words like *kete. *Kete would become kese, since stops between vowels are opened (this is the origin of all voiced fricatives other than /v/ and also adds many voiceless fricatives and affricates, like Begadkefat or the High German Consonant Shift). However, kente would be allowed, and keckte, and keffte, and others where the lenition of the stop is blocked by an intervening consonant. As far as which clusters are allowed... There are too many individual ones to list. As you can see, /ʀntʰs/ and /sk͡x/ and /ltt͡ʃm/ and /ktŋ/ are allowed, and I have other words with things like /k͡xfʀ/ and /xk͡xtʰ/. Pretty much anything can happen due to mass vowel deletions. It's nowhere near as bad as those Pacific Northwest languages, but there are still not many restrictions on phonotactics.
No darkness can harm you if you are guided by your own inner light

HoskhMatriarch
mayan
mayan
Posts: 1779
Joined: 16 May 2015 17:48

Re: Hoskh

Post by HoskhMatriarch » 15 Dec 2015 07:49

OK, I tried to make some recordings. Only this one turned out anything like OK (I don't think I got the lone pharyngealized vowel in though):

Hoskh translation of "To be or not to be...": http://vocaroo.com/i/s1qMkGaNrpJD

Mostly it just sounds like Germanic and Semitic languages had a baby with all those vowels and consonants (and also it has aspiration). One of the recordings I did I accidentally just went into a German accent, and then realized that that wasn't even an acceptable pronunciation for that since I did the stress all wrong. Does anyone have any advice for how I can do the vowel qualities and prosody without just copying German and/or English? It's not like I can just go to a Hoskh speaker and copy them. This language is really grammatically not European, but from how I end up pronouncing it you wouldn't be able to tell, even though European languages don't have /q/
No darkness can harm you if you are guided by your own inner light

Post Reply