The CBB

Posted: **20 Oct 2018 05:21**

Here's Doan's revised phoneme inventory after leniting all plosives into fricatives, merging [ s] and [ç] into [ɬ], merging [z] and [ʑ] into [ɮ], and leniting [r] into [[ɮ˔]:

/m n/
/ɸ β f v θ ð x ɣ h/
/ɬ ɮ l ʎ ɮ˔/

How strange is this, especially with the unpaired glottal phoneme?

Posted: **20 Oct 2018 05:35**

Pretty strange to not have plosives. Perhaps you could have something like /m n/ > /b d/ to correct part of that, but then you wouldn't have nasal consonants.

Posted: **20 Oct 2018 05:49**

yangfiretiger121 wrote: ↑20 Oct 2018 05:21 How strange is this, especially with the unpaired glottal phoneme?

The fact that there's no phonemic voiced counterpart to /h/ is the exact opposite of strange.

Posted: **20 Oct 2018 13:16**

shimobaatar wrote: ↑20 Oct 2018 05:49
yangfiretiger121 wrote: ↑20 Oct 2018 05:21 How strange is this, especially with the unpaired glottal phoneme?
The fact that there's no phonemic voiced counterpart to /h/ is the exact opposite of strange.

In fact it's the least weird thing about the whole inventory.

I don't think it's unreasonable to have through-going lention, however I would not consider it natural for it to not have any environments where it is blocked, because even languages which do do things like this (Kabyle for example) they always have some environments which block the lenition.

Posted: **20 Oct 2018 14:32**

Frislander wrote: ↑20 Oct 2018 13:16
shimobaatar wrote: ↑20 Oct 2018 05:49
yangfiretiger121 wrote: ↑20 Oct 2018 05:21 How strange is this, especially with the unpaired glottal phoneme?
The fact that there's no phonemic voiced counterpart to /h/ is the exact opposite of strange.
In fact it's the least weird thing about the whole inventory.

I don't think it's unreasonable to have through-going lention, however I would not consider it natural for it to not have any environments where it is blocked, because even languages which do do things like this (Kabyle for example) they always have some environments which block the lenition.

Okay. Thanks. I, probably, should have mentioned that the lenition isn't blocked anywhere because the language is now mora-based, like Japanese, with a moraic nasal, a moraic fortis/lenis (lateral) fricative pair, a moraic alveolar lateral approximant, and a moraic palatal lateral approximant. For some reason, I forgot to specify it as the language's consonant inventory or add vowels as well. Thus, the vowels are as follows: /ɪ ʊ ɛ̠ œ̠ ɑ̟/.

Posted: **20 Oct 2018 15:42**

yangfiretiger121 wrote: ↑20 Oct 2018 14:32
Frislander wrote: ↑20 Oct 2018 13:16
shimobaatar wrote: ↑20 Oct 2018 05:49
yangfiretiger121 wrote: ↑20 Oct 2018 05:21 How strange is this, especially with the unpaired glottal phoneme?
The fact that there's no phonemic voiced counterpart to /h/ is the exact opposite of strange.
In fact it's the least weird thing about the whole inventory.

I don't think it's unreasonable to have through-going lention, however I would not consider it natural for it to not have any environments where it is blocked, because even languages which do do things like this (Kabyle for example) they always have some environments which block the lenition.
Okay. Thanks. I, probably, should have mentioned that the lenition isn't blocked anywhere because the language is now mora-based, like Japanese, with a moraic nasal, a moraic fortis/lenis (lateral) fricative pair, a moraic alveolar lateral approximant, and a moraic palatal lateral approximant. For some reason, I forgot to specify it as the language's consonant inventory or add vowels as well. Thus, the vowels are as follows: /ɪ ʊ ɛ̠ œ̠ ɑ̟/.

As far as I can tell, being mora-based rather than syllable-based still shouldn't have this sort of effect. Japanese is mora-based, but it does have environments in which its own process of morphophonological lenition, rendaku, is blocked.

Posted: **20 Oct 2018 19:29**

sangi39 wrote: ↑20 Oct 2018 15:42
yangfiretiger121 wrote: ↑20 Oct 2018 14:32
Frislander wrote: ↑20 Oct 2018 13:16
shimobaatar wrote: ↑20 Oct 2018 05:49
yangfiretiger121 wrote: ↑20 Oct 2018 05:21 How strange is this, especially with the unpaired glottal phoneme?
The fact that there's no phonemic voiced counterpart to /h/ is the exact opposite of strange.
In fact it's the least weird thing about the whole inventory.

I don't think it's unreasonable to have through-going lention, however I would not consider it natural for it to not have any environments where it is blocked, because even languages which do do things like this (Kabyle for example) they always have some environments which block the lenition.
Okay. Thanks. I, probably, should have mentioned that the lenition isn't blocked anywhere because the language is now mora-based, like Japanese, with a moraic nasal, a moraic fortis/lenis (lateral) fricative pair, a moraic alveolar lateral approximant, and a moraic palatal lateral approximant. For some reason, I forgot to specify it as the language's consonant inventory or add vowels as well. Thus, the vowels are as follows: /ɪ ʊ ɛ̠ œ̠ ɑ̟/.
As far as I can tell, being mora-based rather than syllable-based still shouldn't have this sort of effect. Japanese is mora-based, but it does have environments in which its own process of morphophonological lenition, rendaku, is blocked.

Touché.

I know an /ɑ̟/ will block it (mora-based, tentatively) globally and a preceding nasal additionally blocks in for velar stops, which assimilate to /ŋ/. How's that for, at least, a start?

Posted: **21 Oct 2018 21:17**

Basically what people are saying is that it's implausible for any sound change or series of sound changes to cause a language to not have any stop consonants. Languages have stop consonants, period.

Or well, real-life languages do, anyways. With a conlang you're free to do what you want - but, when you ask for people's opinion on the phoneme inventory of a conlang, the presumption is that you're asking people to evaluate it by the standard "Is this plausible for a natural (real-life) language?" - Since if you weren't trying to create a naturalistic language, then you could do whatever you wanted and you wouldn't need people to evaluate the consonant inventory for realism or "strangeness".

Could you make an otherwise naturalistic language with no stop consonants? Yes, though the lack of stop consonants would still place it in the realm of fantasy, since no such language exists and presumably has never existed.

So, long story short, if you don't want the language to be extremely strange and implausible, you must give it at least two stop consonants, and you cannot have a sound change that removes all stops from the language. With just two stops the likely pairings would be /p t/ (or /b d/ or /b t/), /t k/, or /t ʔ/. Although a language with your stop system would almost certainly have more - likely but not necessarily one stop at each place of articulation, and likely but not necessarily a voicing contrast in the stops.

Now, you could have some major allophony, possibly representing a stage of phonological transition, where all stops are lenited to fricatives in a large number of cases, while preserved in some others - e.g. Kabyle, where iirc it's just the non-emphatic stops that are lenited, or Hebrew, where all stops became fricatives after a vowel, or some Oceanic languages, where stops lenited in all positions except after nasals (though the nasals did eventually drop from those clusters, resulting in a new series of plain stops).

Edit: Actually, based on your initial comment, I think a Hebrew-type lenition would be great, since it would give you a ton of fricatives (while maintaining stops) and would fit in with the sound changes you described.

So, you could have an earlier stage of Doan that looks like this:

/m n/
/p b t d k g/
/f v s z ɬ ɮ ç ʑ h/
/l ʎ r/

Then the non-lateral fricatives and the rhotic merge with the laterals, creating a "gap", i.e. a paucity of non-lateral fricatives articulated in the middle of the mouth - which wouldn't necessarily cause anything to happen, but in this case it happens to be "fixed" by stops leniting to fricatives after a vowel, creating a series of major fricative allophones:

/m n/
/p b t d k g/
[ɸ β θ ð x ɣ]
/f v ɬ ɮ h/
/l ʎ ɮ˔/

Posted: **21 Oct 2018 22:12**

I'm working on a descendant of K'otê, and I've run into a problem with vowel harmony. K'otê had vowel harmony between /i(ː) ɨ(ː) u(ː) ə(ː)/ and /e(ː) ə(ː) o(ː) a(ː)/ respectively. The descendant went through the following vowel related sound changes:

ɨ(ː) ə(ː) → i(ː) e(ː)
i u → e o
eː oː → i u
iː uː aː → i u a

Basically, most corresponding vowel pairs merge with each other with the exception of /ə(ː)/ and /a(ː)/. /ə/ becomes /e/, /əː/ becomes /i/, and /a(ː)/ becomes /a/. This means the only remaining vowel harmony is between /a/ and /e/~/i/. Not only this but you wouldn't be able to tell by looking at a word whether it belongs in the /a/-group or the /e/~/i/-group. Here's an example of some hypothetical words that show what I'm talking about:

Words: /tun/, /ton/, /tuːn/, /toːn/
Affix: -/nəː/ or -/naː/
/tunnəː/ → /tonni/
/tonnaː/ → /tonna/
/tuːnnəː/ → /tunni/
/toːnna/ → /tunna/

If I wanted to just get rid of vowel harmony completely, how would I do that? What would the reflex of the hypothetical affix, "-/nəː/ or -/naː/?" If I wanted to keep the distinction, would it be naturalistic? I imagine it wouldn't be that unnaturalistic for this to become essentially a declension system.

Posted: **21 Oct 2018 22:33**

The vowel harmony system would probably just cease to be productive. Suffixes displaying alternations would likely just remain as they are, at least initially, but newly-derived words, or newly-adopted loanwords, would probably just take one of the variants by default, or would take one of them depending on some other criterion, like whether they resembled a word with the a-variant or the e/i-variant. It could definitely become, essentially, a declensional system, with two classes of words, one regularly taking the historical a-variant and the other the historical e/i-variant.

Think of Germanic i-umlaut, which was essentially backwards vowel harmony: After the conditioning environment was lost, initially all of the words affected by it kept their vowel assimilation, but over time some ceased to have this alternation, with either the non-umlauted or umlauted variant appearing in all forms of the word. There ended up being two "declensions" of nouns, one that had umlauted forms and one that didn't. Some words that historically never had umlaut were given umlauted plurals or whatever, as it simply came to be a purely morphological process.

With the loss of progressive vowel harmony this process is just happening largely with suffixes instead, as opposed to the root words.

Posted: **22 Oct 2018 00:18**

Starting with a mostly CV language and dropping some vowels in some places, could the consonant that had proceeded the vowel become one thing before another consonant and another at the end of words, but still change in both places?

Posted: **22 Oct 2018 01:13**

Porphyrogenitos wrote: ↑21 Oct 2018 21:17 Basically what people are saying is that it's implausible for any sound change or series of sound changes to cause a language to not have any stop consonants. Languages have stop consonants, period.

Or well, real-life languages do, anyways. With a conlang you're free to do what you want - but, when you ask for people's opinion on the phoneme inventory of a conlang, the presumption is that you're asking people to evaluate it by the standard "Is this plausible for a natural (real-life) language?" - Since if you weren't trying to create a naturalistic language, then you could do whatever you wanted and you wouldn't need people to evaluate the consonant inventory for realism or "strangeness".

Could you make an otherwise naturalistic language with no stop consonants? Yes, though the lack of stop consonants would still place it in the realm of fantasy, since no such language exists and presumably has never existed.

So, long story short, if you don't want the language to be extremely strange and implausible, you must give it at least two stop consonants, and you cannot have a sound change that removes all stops from the language. With just two stops the likely pairings would be /p t/ (or /b d/ or /b t/), /t k/, or /t ʔ/. Although a language with your stop system would almost certainly have more - likely but not necessarily one stop at each place of articulation, and likely but not necessarily a voicing contrast in the stops.

Now, you could have some major allophony, possibly representing a stage of phonological transition, where all stops are lenited to fricatives in a large number of cases, while preserved in some others - e.g. Kabyle, where iirc it's just the non-emphatic stops that are lenited, or Hebrew, where all stops became fricatives after a vowel, or some Oceanic languages, where stops lenited in all positions except after nasals (though the nasals did eventually drop from those clusters, resulting in a new series of plain stops).

Edit: Actually, based on your initial comment, I think a Hebrew-type lenition would be great, since it would give you a ton of fricatives (while maintaining stops) and would fit in with the sound changes you described.

So, you could have an earlier stage of Doan that looks like this:

/m n/
/p b t d k g/
/f v s z ɬ ɮ ç ʑ h/
/l ʎ r/

Then the non-lateral fricatives and the rhotic merge with the laterals, creating a "gap", i.e. a paucity of non-lateral fricatives articulated in the middle of the mouth - which wouldn't necessarily cause anything to happen, but in this case it happens to be "fixed" by stops leniting to fricatives after a vowel, creating a series of major fricative allophones:

/m n/
/p b t d k g/
[ɸ β θ ð x ɣ]
/f v ɬ ɮ h/
/l ʎ ɮ˔/

The current working inventory is below noting that adjacency to /ɑ̟/ blocks lenition globally and a preceding nasal additionally blocks in for velar plosives, which assimilate to /ŋ/. Allophonic [w] occurs when [v] would follow a plosive allophone, [ɸ], [β] (possibly), [f], [θ], or [ð]. Alternatively, I could flip the plosive allophony, as you suggest. The lateral fricatives, lateral trills, and voiceless alveolar trill started out as phonetic mutations rather than distinct phonemes.

/m n ŋ/
/ɸ β v ɬ ɮ θ ð x ɣ/
[p b w ç ʑ t d k g]
/f h/
/l ʎ/
/ɬ̝ ɮ˔/
[r̥ r]
/ɪ ʊ ɛ œ ɑ/

Posted: **22 Oct 2018 01:14**

Do you ever post something in your conlang thread and then immediately change your mind about it? Is it possible to be a conlanger and be as indecisive as I am?

Posted: **22 Oct 2018 01:20**

KaiTheHomoSapien wrote: ↑22 Oct 2018 01:14 Do you ever post something in your conlang thread and then immediately change your mind about it? Is it possible to be a conlanger and be as indecisive as I am?

I may not be as indecisive as you, but I'm, probably, damn close to it. So, the answers are resounding yeses.

Posted: **22 Oct 2018 06:11**

KaiTheHomoSapien wrote: ↑22 Oct 2018 01:14 Do you ever post something in your conlang thread and then immediately change your mind about it? Is it possible to be a conlanger and be as indecisive as I am?

The very act of communicating ideas about my conlangs is exactly what makes me start to doubt and reconsider my decisions. This is a good thing. I think.

Posted: **22 Oct 2018 06:14**

KaiTheHomoSapien wrote: ↑22 Oct 2018 01:14 Do you ever post something in your conlang thread and then immediately change your mind about it? Is it possible to be a conlanger and be as indecisive as I am?

why hello there

Posted: **22 Oct 2018 06:47**

Well good to know; that does make me feel better

Posted: **22 Oct 2018 17:30**

Is "lateral mutation" an accurate way to describe changes like [lh → ɬ] and [lð → ɮ̪]? On a related note, is "rhotic mutation" an accurate way to describe changes like [kr → r̥]?

Posted: **22 Oct 2018 17:42**

KaiTheHomoSapien wrote: ↑22 Oct 2018 01:14 Do you ever post something in your conlang thread and then immediately change your mind about it? Is it possible to be a conlanger and be as indecisive as I am?

Do you expect one says you are impossible and therefore must stop conlanging immediately?

Will you listen anyway?

But if doing things randomly bothers you then you may think about reasons to pick this or that and make some principles up for yourself.
And there is conlang.wikia.com to change things when and where one wishes.

Posted: **22 Oct 2018 19:38**

yangfiretiger121 wrote: ↑22 Oct 2018 17:30 Is "lateral mutation" an accurate way to describe changes like [lh → ɬ] and [lð → ɮ̪]? On a related note, is "rhotic mutation" an accurate way to describe changes like [kr → r̥]?

I wouldn't call it a "mutation" unless is occurs over a word boundary and the conditioning environment is lost, as if the Spanish definite article lost the l but certain consonants mutated into lateral fricatives, e.g. el hijo > e ɬijo vs. un hijo (I know Spanish "h" is silent in the modern day but you get the point). Aside from that I'd be more likely to refer to it as a "lateral spirantisation" or something like that.

And by that logic the second change cannot be said to be "rhotic mutation", because there was a rhotic already there, and if it occurs over a morpheme boundary like above, then it is more a spirantisation mutation: ek ruba > e r̥uba.

The CBB

(Conlangs) Q&A Thread - Quick questions go here [2010-2020]

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here

Re: (Conlangs) Q&A Thread - Quick questions go here