Language of three

A forum for all topics related to constructed languages
Post Reply
Oneiros K
rupestrian
rupestrian
Posts: 19
Joined: 04 Aug 2017 08:59

Language of three

Post by Oneiros K »

Language of three

I’m developing a language (family) for my scifi/fantasy story.

Ancestral language

The original language from which all spoken languages in that story descend was a conlang based around symbolic use of the number three.

It is a conlang even within the story. It is not intended to be natural or to have naturally evolved.
Its consonants use three places of articulation (labial, dental, velar),
three manners (plosive, fricative, nasal),
and each consonants comes in three versions (plain, palatalized, labialized).

Consonants

Plosive b bj bw d dj dw g gj gw
Fricative v vj vw z zj zw ɣ ɣj ɣw
Nasal m mj mw n nj nw ŋ ŋj ŋw

It has three phonemic vowels which come in three versions, depending on the version of the preceding consonant (plain, followed by j-glide, followed by w-glide).
When preceded by a palatalized consonant, a j-glide is inserted after the vowel.
A w-glide is added when the preceding consonant is labialized.

Vowels

a aj aw
e ej ew
o oj ow

Each basic word consists of three consonants and three vowels with the syllable structure
C-V-C-V-C-V

Its a register tone language which marks grammatical case by low, mid or high tone.
Each syllable in a word receives one tone, which together mark the grammatical case.

Language of story

At the time of the story the original language is practically extinct and instead one of its descendents is used.

The following is a description of its phonological changes and new phoneme inventory.

It would be really helpful for me to receive a feedback on the (im)plausibility of the phonetic changes. I didn’t study linguistics and just started to learn about sound changes.

I created the daughter language with my own sense of euphony in mind. I know it’s highly subjective but it’s interesting for me if you personally like the sounds of the daughter language and/or would suggest changes.
I designed the ancestral with symmetry in mind, not necessarily with phonological beauty.

Consonant changes

Change of fricatives (plain, labialized and palatalized - except of ɣw)
v z ɣ -> ð ʒ ʁ

ɣw -> xw -> ʍ
plain ʁ -> r
ʁj -> ʎ -> l

Devoicing of plain and palatalized fricatives
ð ðj ʒ ʒj -> θ θj ʃ ʃj

Loss of labialization of fricatives
ðw ʒw -> ð ʒ

Devoicing of palatalized plosives
bj dj gj -> pj tj kj

The labialization of the plosives became replaced by a consonant cluster of the plosive + v, which led to the assimilation of the plosives, turning them into fricatives. The v got lost in that process, leaving only plain fricatives.
bw dw gw -> v z x

nj -> ɲ

Loss of ŋ
ŋ ŋj ŋw-> ʔ j w

Vowel changes

aj ej oj -> ai ɛi y
aw -> ʌw -> ɔ
ew -> ɛw -> ɛ
ow -> u

Which leads to:

Consonants
b pj v d tj z g kj x
θ θj ð ʃ ʃj ʒ r l ʍ
m mj mw n ɲ nw ʔ j w

Vowels
a ai ɔ
e ɛi ɛ
o y u

Labialization of the nasals is lost before u.
mwu, nwu -> mu, nu
Last edited by Oneiros K on 05 Aug 2017 20:48, edited 4 times in total.
User avatar
Lambuzhao
korean
korean
Posts: 5405
Joined: 13 May 2012 02:57

Re: Language of three

Post by Lambuzhao »

3+1 things:

1) Looks like you have 9 vowels, and 9 consonants in a few places (3x3).
Language of 9?
{Three's a crowd; Nine's more of a party}

2) "the following universal rule: no language has voiced stops without voiceless stops."
Finegan (2014), p. 242 ; among other places.

:?:
3) Maybe the Speakers of 3 may simply not distinguish between voiceless and voiced stops, which occurs in some natlangs.

4) Regarding Ling 'o' 3's inventory, it's six in one, half a dozen the other 2 me. Conlang away, I say!
User avatar
Vlürch
greek
greek
Posts: 452
Joined: 09 Mar 2016 21:19
Location: Finland
Contact:

Re: Language of three

Post by Vlürch »

Oneiros K wrote:Sadly the superscript j and w get copied as w and j, so I use w and j instead of the superscripts to denote palatalization and labialization.
You can use unicode input tools to get the superscript characters used for IPA, like this one.
Oneiros K
rupestrian
rupestrian
Posts: 19
Joined: 04 Aug 2017 08:59

Re: Language of three

Post by Oneiros K »

Lambuzhao wrote:3+1 things:

1) Looks like you have 9 vowels, and 9 consonants in a few places (3x3).
Language of 9?
{Three's a crowd; Nine's more of a party}

2) "the following universal rule: no language has voiced stops without voiceless stops."
Finegan (2014), p. 242 ; among other places.

:?:
3) Maybe the Speakers of 3 may simply not distinguish between voiceless and voiced stops, which occurs in some natlangs.

4) Regarding Ling 'o' 3's inventory, it's six in one, half a dozen the other 2 me. Conlang away, I say!

Thanks for the response. [:)]

1) I think of it as 3x3x3 = 33 consonants. 3 places of articulation(labial, dental, velar) x 3 manners(plosive, fricative, nasal) x 3 types of secondary articulation*(plain, palatalized, labialized)
*or the lack thereof

2) I should have made this clearer. The original/ancient language of three is a conlang even within it's own world. It is not intended to be natural.
It's not designend to be linguistically plausible but rather to be systematic and leading through plausible sound changes to a daughter language which phoneme inventory I like. I thought of phonemically distinguishing voicing in the ancestral language. First sketches of the language had a voiced, unvoiced, unvoiced-aspirated distinction among consonants. Allophonic voiceless/voiced distinction could also be possible. However I somehow would feel forced to to distinguish 3 types of voicing; e.g. voiced, unvoiced, prevoiced. I tried to avoid rare sounds as prevoiced consonants for simplicity's sake but this is still an option. As is the allophonic distinction of voiced/voiceless/voiceless-aspirated consonants.
I could use only voiceless consonants instead of only voiced ones(and personally would prefer this for euphony reasons) but than I would use the rare voiceless nasals m̥ n̥ ŋ̥ to keep voicing constant in the language (for symmetry reasons). Which I'm not sure about, again for simplicity's sake.

4) ? I'm sorry I don't get that comment. :wat:
Why six in one and not 3 in one?
It's
o oj ow
a aj aw
e ej ew

The vowels with glides can be analyzed more as allophones. (or as a result of infixes j, w)
with a, o, e beeing the underlying phonemes. aj, aw; oj, ow; ej, ew would be allophones of the plain vowels after palatalized and labialized consonants respectively.
I personally liked this, leading to syllables like:
vjaj, mjaj, njej; bwaw, ...
or (in the daughter language)
θjai, mjai, ɲɛi

However, the phonology of the original (unnatural-even-within-the-story) language could be changed rather arbitrarily. It just has to follow the rule of 3x3x3 for the consonants; 3x3 for the vowels. It's pretty open to advice and changes.

What would be really helpful for me is a feedback on whether the historical sound changes from that (unnatural) ancestral language to the (natural-within-the-story) daughter language are plausibel. And it's resulting phoneme inventory, given it's unnatural origion.

I'm pretty sure that it isn't in all details. Just my best attempt given my lack of formal linguistical background [;)]

I avoided rounding e to ø, a to ɶ after labialized consonants/before w glides as I just don't like these sounds...
User avatar
Lao Kou
mongolian
mongolian
Posts: 5089
Joined: 25 Nov 2012 10:39
Location: 蘇州/苏州

Re: Language of three

Post by Lao Kou »

Oneiros K wrote:I avoided rounding e to ø, a to ɶ after labialized consonants/before w glides as I just don't like these sounds...
Then you shall have no pie. [B)]
道可道,非常道
名可名,非常名
User avatar
Lambuzhao
korean
korean
Posts: 5405
Joined: 13 May 2012 02:57

Re: Language of three

Post by Lambuzhao »

Oneiros K wrote:
Thanks for the response. [:)]
You're welcome!


4) ? I'm sorry I don't get that comment. :wat:
Why six in one and not 3 in one?
That's because "Six in one, 1/2 dozen the other" is an idiom. It means something like "either way, it's okay" or "either way, it makes that same difference". Whatever way you work out your conlang, I'll be cheering from the sidelines and waiting to fling some of it back at you on the Conlang Conversation Thread (you have been officially pre-Lamboozled).
[;)]

It's
o oj ow
a aj aw
e ej ew

The vowels with glides can be analyzed more as allophones. (or as a result of infixes j, w)
with a, o, e beeing the underlying phonemes. aj, aw; oj, ow; ej, ew would be allophones of the plain vowels after palatalized and labialized consonants respectively.
I personally liked this, leading to syllables like:
vjaj, mjaj, njej; bwaw, ...
or (in the daughter language)
θjai, mjai, ɲɛi

However, the phonology of the original (unnatural-even-within-the-story) language could be changed rather arbitrarily. It just has to follow the rule of 3x3x3 for the consonants; 3x3 for the vowels. It's pretty open to advice and changes.

What would be really helpful for me is a feedback on whether the historical sound changes from that (unnatural) ancestral language to the (natural-within-the-story) daughter language are plausibel. And it's resulting phoneme inventory, given it's unnatural origion.

I'm pretty sure that it isn't in all details. Just my best attempt given my lack of formal linguistical background [;)]

I avoided rounding e to ø, a to ɶ after labialized consonants/before w glides as I just don't like these sounds...[/quote]
User avatar
Frislander
mayan
mayan
Posts: 2088
Joined: 14 May 2016 18:47
Location: The North

Re: Language of three

Post by Frislander »

Oneiros K wrote:Language of three

I’m developing a language (family) for my scifi/fantasy story.

Ancestral language

The original language from which all spoken languages in that story descend was a conlang based around symbolic use of the number three.

It is a conlang even within the story. It is not intended to be natural or to have naturally evolved.
Its consonants use three places of articulation (labial, dental, velar),
three manners (plosive, fricative, nasal),
and each consonants comes in three versions (plain, palatalized, labialized).

Consonants

Plosive b bj bw d dj dw g gj gw
Fricative v vj vw z zj zw ɣ ɣj ɣw
Nasal m mj mw n nj nw ŋ ŋj ŋw

It has three phonemic vowels which come in three versions, depending on the version of the preceding consonant (plain, followed by j-glide, followed by w-glide).
When preceded by a palatalized consonant, a j-glide is inserted after the vowel.
A w-glide is added when the preceding consonant is labialized.

Vowels

a aj aw
e ej ew
o oj ow
Its a register tone language which marks grammatical case by low, mid or high tone.
Each syllable in a word receives one tone, which together mark the grammatical case.
I like this, though I'd be tempted to analyse the syllable structure as C(G)V(G)T, where C is one of /b d g v z ɣ m n ŋ/, G one of /j w/ and V one of /a e o/ and T is the tone.
Each basic word consists of three consonants and three vowels with the syllable structure
C-V-C-V-C-V
Hmm, with a phonology like that honestly I reckon CVCV would be perfectly fine, honestly, unless you have a really baroque case system which requires lots of tone contrasts (which I guess you could have given your stated unnaturalness of this).
I created the daughter language with my own sense of euphony in mind. I know it’s highly subjective but it’s interesting for me if you personally like the sounds of the daughter language and/or would suggest changes.
I designed the ancestral with symmetry in mind, not necessarily with phonological beauty.

Consonant changes

Change of fricatives (plain, labialized and palatalized - except of ɣw)
v z ɣ -> ð ʒ ʁ
Err... well I supposed it could happen like this, but it just seems a bit of an arbitrary set, and not having it interact with the a following /j w/ honestly stretches credulity. Something like /v vʲ vʷ z zʲ zʷ ɣ ɣʲ ɣʷ/ > /v vʲ w z ʒ ʒʷ ɣ j ɣʷ/ or /ð vʲ v z ʒ ʒʷ ɣ j w/.
ɣw -> xw -> ʍ
plain ʁ -> r
ʁj -> ʎ -> l
ɣʷ > ʍ seems a rather arbitrary devoicing, though ʁ > r is probably alright. ʁʲ > ʎ is particularly strange.
Devoicing of plain and palatalized fricatives
ð ðj ʒ ʒj -> θ θj ʃ ʃj
This is totally unnatural and there is no phonetic motivation for labialisation blocking devoicing.
Loss of labialization of fricatives
ðw ʒw -> ð ʒ
This is OK as is, though the context is strange.
Devoicing of palatalized plosives
bj dj gj -> pj tj kj
Again, this is unnatural and there is no phonetic motivation for it, especially given that the environment here is different from that given above.
The labialization of the plosives became replaced by a consonant cluster of the plosive + v, which led to the assimilation of the plosives, turning them into fricatives. The v got lost in that process, leaving only plain fricatives.
bw dw gw -> v z x

nj -> ɲ
Perfectly reasonable and common the world over.
Loss of ŋ
ŋ ŋj ŋw-> ʔ j w
Also perfectly fine as is.
Vowel changes

aj ej oj -> ai ɛi y
aw -> ʌw -> ɔ
ew -> ɛw -> ɛ
ow -> u
Mostly fine (though ew > ɛ seems a bit shaky to me and I'm no sure it's attested, unlike the others which definitely are), though the resulting system with /y/ but not /i/ is highly imbalanced to the point of imo being unnatural. might I suggest ej > i, aj > ɛ and ew > ø instead?
Which leads to:

Consonants
b pj v d tj z g kj x
θ θj ð ʃ ʃj ʒ r l ʍ
m mj mw n ɲ nw ʔ j w

Vowels
a ai ɔ
e ɛi ɛ
o y u

Labialization of the nasals is lost before u.
mwu, nwu -> mu, nu
Probably better to arrange the system in a matter which fits the phonology of the daughter, rather than mapping it onto the parent.

/pʲ tʲ kʲ ʔ/
/b d g/
/θ θʲ ʃ ʃʲ x/
/v ð z ʒ/
/m mʲ n ɲ nʷ/
/w l r j/

/y u/
/e o/
/ɛ a ɔ/
/aɪ̯ ɛɪ̯/

Hmm, as it stands this system is actually a good deal more unnatural than the parent language (notable points include: the only voiceless consonants being palatalisaed; having a /ʃ ʃʲ/ contrast but no /s/; having a separate /nʷ/ phoneme (seriously try pronouncing that at the start of a word; and as mentioned having /y/ but no /i/), which is in itself weird; if you start from a system like what you started with you would expect the result of natural phonological changes to be more natural, not less. And don;t get me wrong you do have a lot you can work with here, and there's a lot of potential in the system; it's just that you've foisted some sound changes which are perfectly reasonable in and of themselves into environments which make no sense phonetically.

Your best help in these circumstances is to look at natural languages which have similar systems to your protolang and see what changes they go through.
User avatar
MrKrov
banned
Posts: 1929
Joined: 12 Aug 2010 02:47
Location: /ai/ > /a:/
Contact:

Re: Language of three

Post by MrKrov »

Frislander wrote:Hmm, as it stands this system is actually a good deal more unnatural than the parent language (notable points include: the only voiceless consonants being palatalisaed; having a /ʃ ʃʲ/ contrast but no /s/;
Basically everything you said was otherwise fine, but this bit was not so egregious. Having /ʃ/ as the only sibilant is okay. And having /θ ʃ/ minus /s/ is fine. And if one pair of the set can have a palatalization contrast, it's fair if the other can too.
If it helps, think of it like one of those pairs patterns like it was really /s sʲ/ but the actual form is to enhance perceptual salience to maintain contrast.
Last edited by MrKrov on 16 Aug 2017 02:09, edited 1 time in total.
cromulant
cuneiform
cuneiform
Posts: 143
Joined: 13 Jan 2012 00:04

Re: Language of three

Post by cromulant »

MrKrov wrote:
Frislander wrote:Hmm, as it stands this system is actually a good deal more unnatural than the parent language (notable points include: the only voiceless consonants being palatalisaed; having a /ʃ ʃʲ/ contrast but no /s/;
Basically everything you said was otherwise fine, but this bit was not so egregious. Having /ʃ/ as the only sibilant is okay. And having /θ ʃ/ minus /s/ is fine.
Indeed: attested in Karajá.
User avatar
Frislander
mayan
mayan
Posts: 2088
Joined: 14 May 2016 18:47
Location: The North

Re: Language of three

Post by Frislander »

cromulant wrote:
MrKrov wrote:
Frislander wrote:Hmm, as it stands this system is actually a good deal more unnatural than the parent language (notable points include: the only voiceless consonants being palatalisaed; having a /ʃ ʃʲ/ contrast but no /s/;
Basically everything you said was otherwise fine, but this bit was not so egregious. Having /ʃ/ as the only sibilant is okay. And having /θ ʃ/ minus /s/ is fine.
Indeed: attested in Karajá.
And Shawnee, and more solidly too (SAPHON doesn't give /ʃ/ as a phoneme in Karajá).

It's not the lack of /s/ that I object to per se, it's just that I think the added palatalisation dimension doesn't sit well with me, though I guess it could happen and I guess I'd be OK with it if it was part of a more through-going palatalisation distinction.
Post Reply