I think those are two words in the same language, but other than that, I'm not sure I understand the question either.Dormouse559 wrote: ↑16 Oct 2018 18:14I don't understand the question. Are jarxvalma and Szor languages? If so, I don't immediately see how sound changes in separate languages involving very different phones, and very different changes to those phones, can be analogous. The sound shifts might both be assimilatory, but the former involves a change in place and manner of articulation with no change in voicing; the latter changes neither POA nor MOA but does change voicing.yangfiretiger121 wrote: ↑16 Oct 2018 17:20 Is [l̥ː] or [ʎ̥] the more natural outcome of jarxvalma's [m̥>l̥] assimilation when it should be analogous to the [s>z] assimilation in Szor [zːœ̠r]?
(Conlangs) Q&A Thread - Quick questions go here [2010-2020]
-
- korean
- Posts: 10373
- Joined: 12 Jul 2013 23:09
- Location: UTC-04:00
Re: (Conlangs) Q&A Thread - Quick questions go here
Re: (Conlangs) Q&A Thread - Quick questions go here
As you can see from the other responses, there isn't a satisfactorily clear-cut definition for what polysynthesis should mean. The usage of the term circles around high morphological complexity and a frequent use of single words sentences. The latter pretty much implies verb-centric morphology which allows you to put enough grammatical information on the predicate verb so that other words may be omitted from the sentence. Polypersonal agreement and noun incorporation help here, but there's no particular reason why either of them should be obligatory for a language labeled as polysynthetic.CarsonDaConlanger wrote: ↑15 Oct 2018 16:08 What's the boundary between polysynthetic and highly agglutinative? Like, when does a language start being polysynthetic?
A while ago I did an informal experiment where I calculated morpheme-to-word ratios from a bunch of languages from around the world representing different grammatical types. At least as far as I got, I couldn't find any separate class of languages with very high morphological complexity that I could have identified as representing a "polysynthetic" type. Rather, the experiment uncovered a full spectrum of languages ranging from those with next to no morphology to the extremely morphologically complex ones. Even though the analysis wasn't that rigorous (I didn't pay too much attention if some sources used very different definitions for word borders from the others, and my sample consisted basically from whichever language I ran across with reasonably long glossed example texts), it pointed out that at least in the sense of morphological complexity polysynthetic languages don't form a natural class distinct from other languages. Instead, it comes out as a convenient label given to a lot of languages that fall close to the upper end of the morphological complexity spectrum, very much like what Frislander pointed out,
It could be interesting to repeat the experiment but now counting the frequency of the single word sentences instead of the morpheme-to-word ratio. It's unlikely, though, that that would uncover any distinct classes of languages either.Frislander wrote: ↑16 Oct 2018 16:06 ... the definition of what counts as polysynthetic can be best described as "how similar is this language to these other languages we've called 'polysynthetic'".
Last edited by gach on 16 Oct 2018 19:25, edited 1 time in total.
- Dormouse559
- moderator
- Posts: 2945
- Joined: 10 Nov 2012 20:52
- Location: California
Re: (Conlangs) Q&A Thread - Quick questions go here
Hmm, good point. I also am still somewhat stumped. With only two changes to compare, and very different ones at that, there aren't a lot of parallels to draw. If these two changes were part of a broader simplification of consonant clusters, you would probably want to develop some underlying principles to determine outcomes, but that's hard to do with just two dissimilar sound shifts.shimobaatar wrote: ↑16 Oct 2018 18:29I think those are two words in the same language, but other than that, I'm not sure I understand the question either.Dormouse559 wrote: ↑16 Oct 2018 18:14I don't understand the question. Are jarxvalma and Szor languages? If so, I don't immediately see how sound changes in separate languages involving very different phones, and very different changes to those phones, can be analogous. The sound shifts might both be assimilatory, but the former involves a change in place and manner of articulation with no change in voicing; the latter changes neither POA nor MOA but does change voicing.yangfiretiger121 wrote: ↑16 Oct 2018 17:20 Is [l̥ː] or [ʎ̥] the more natural outcome of jarxvalma's [m̥>l̥] assimilation when it should be analogous to the [s>z] assimilation in Szor [zːœ̠r]?
-
- sinic
- Posts: 337
- Joined: 17 Jun 2018 03:04
Re: (Conlangs) Q&A Thread - Quick questions go here
Yes. Both words are in Dhoan, my conlang. Was "assimilation" the correct terminology? If not, what's the correct terminology?Dormouse559 wrote: ↑16 Oct 2018 19:24Hmm, good point. I also am still somewhat stumped. With only two changes to compare, and very different ones at that, there aren't a lot of parallels to draw. If these two changes were part of a broader simplification of consonant clusters, you would probably want to develop some underlying principles to determine outcomes, but that's hard to do with just two dissimilar sound shifts.shimobaatar wrote: ↑16 Oct 2018 18:29I think those are two words in the same language, but other than that, I'm not sure I understand the question either.Dormouse559 wrote: ↑16 Oct 2018 18:14I don't understand the question. Are jarxvalma and Szor languages? If so, I don't immediately see how sound changes in separate languages involving very different phones, and very different changes to those phones, can be analogous. The sound shifts might both be assimilatory, but the former involves a change in place and manner of articulation with no change in voicing; the latter changes neither POA nor MOA but does change voicing.yangfiretiger121 wrote: ↑16 Oct 2018 17:20 Is [l̥ː] or [ʎ̥] the more natural outcome of jarxvalma's [m̥>l̥] assimilation when it should be analogous to the [s>z] assimilation in Szor [zːœ̠r]?
It's my fault for not thinking to provide the relevant rules for the shifts, which are below.
Fortis-lenis pairs, such as [sz], condense into the trailing sound, yielding sounds such as [zː].
Sequences of two voiceless allophones, such as [l̥m̥], condense into the leading sound, yielding sounds such as [].
Post-fricativization(sp?) [ɸf], [fɸ], [βv], and [vβ] are meant to condense into the latter sound, similar to [sz] but based on sound similarity.
Slightly off topic, but someone posted something about their conlang while I was typing. Does anyone else not see that post?
Alien conlangs (Font may be needed for Vai symbols)
Re: (Conlangs) Q&A Thread - Quick questions go here
I'm working on an Old Norse-Kalaallisut creole, and I am trying to create sound correspondences for each language.
Could it be possible for one language to take in another languages /q/ sound as something like /kʰ/? I've noticed the loanword for "kayak" in Faroese is taken as "kajakk" [kʰajaʰkː] from the original "qajaq" [qajaq]. However, this does not seem to happen in the reverse: Kalaallisut "kona" [kona] is a loan from Old Norse kona [kʰona] (as opposed to "qona").
Or, should I expect aspiration to drop entirely meaning a merger between /kʰ/, /k/ and /q/? This seems the most likely scenario for a creole, no?
EDIT: Yes that was me. I took it down, but then put it back up again. Sorry!
Could it be possible for one language to take in another languages /q/ sound as something like /kʰ/? I've noticed the loanword for "kayak" in Faroese is taken as "kajakk" [kʰajaʰkː] from the original "qajaq" [qajaq]. However, this does not seem to happen in the reverse: Kalaallisut "kona" [kona] is a loan from Old Norse kona [kʰona] (as opposed to "qona").
Or, should I expect aspiration to drop entirely meaning a merger between /kʰ/, /k/ and /q/? This seems the most likely scenario for a creole, no?
EDIT: Yes that was me. I took it down, but then put it back up again. Sorry!
Re: (Conlangs) Q&A Thread - Quick questions go here
This question is a bit beyond my knowledge base but my understanding is that that the /f/ /v/ distinction occurred after French loan words came into English. It seems plausible enough loan words could have a similar impact on Norse settlers in Greenland. I don't know how likely it would be to recognise an aspiration distinction on just velar consonants without also being sensitive to aspiration on other stops.Ælfwine wrote: ↑16 Oct 2018 22:34 I'm working on an Old Norse-Kalaallisut creole, and I am trying to create sound correspondences for each language.
Could it be possible for one language to take in another languages /q/ sound as something like /kʰ/? I've noticed the loanword for "kayak" in Faroese is taken as "kajakk" [kʰajaʰkː] from the original "qajaq" [qajaq]. However, this does not seem to happen in the reverse: Kalaallisut "kona" [kona] is a loan from Old Norse kona [kʰona] (as opposed to "qona").
Or, should I expect aspiration to drop entirely meaning a merger between /kʰ/, /k/ and /q/? This seems the most likely scenario for a creole, no?
EDIT: Yes that was me. I took it down, but then put it back up again. Sorry!
Re: (Conlangs) Q&A Thread - Quick questions go here
Okay, let me explain.Nachtuil wrote: ↑16 Oct 2018 22:48This question is a bit beyond my knowledge base but my understanding is that that the /f/ /v/ distinction occurred after French loan words came into English. It seems plausible enough loan words could have a similar impact on Norse settlers in Greenland. I don't know how likely it would be to recognise an aspiration distinction on just velar consonants without also being sensitive to aspiration on other stops.Ælfwine wrote: ↑16 Oct 2018 22:34 I'm working on an Old Norse-Kalaallisut creole, and I am trying to create sound correspondences for each language.
Could it be possible for one language to take in another languages /q/ sound as something like /kʰ/? I've noticed the loanword for "kayak" in Faroese is taken as "kajakk" [kʰajaʰkː] from the original "qajaq" [qajaq]. However, this does not seem to happen in the reverse: Kalaallisut "kona" [kona] is a loan from Old Norse kona [kʰona] (as opposed to "qona").
Or, should I expect aspiration to drop entirely meaning a merger between /kʰ/, /k/ and /q/? This seems the most likely scenario for a creole, no?
EDIT: Yes that was me. I took it down, but then put it back up again. Sorry!
The Kalaallisut have only /p t k q/ without voiced or aspirated pairs. Old Norse likely had /pʰ p tʰ t kʰ k/.
Of course I could have /q/ in the conlang, but it is far more likely to have /q/ -> /k/ on the Kalaallisut side and /pʰ tʰ kʰ/ -> /p t k/ on the Norse side. I agree /p t k kʰ/ doesn't seem any more likely than /p t k q/. So I guess the likely inventory would be merely /p t k/.
Since Old Norse geminate plosives were pre-aspirated, I'll probably have qajaq -> kajakk like in Faroese, but without the aspiration, the geminate /k/ being a remnant of that.
-
- sinic
- Posts: 337
- Joined: 17 Jun 2018 03:04
Re: (Conlangs) Q&A Thread - Quick questions go here
'Tis perfectly fine. Just got confused when the second preview screen appeared without the post on it.Ælfwine wrote: ↑16 Oct 2018 22:34 I'm working on an Old Norse-Kalaallisut creole, and I am trying to create sound correspondences for each language.
Could it be possible for one language to take in another languages /q/ sound as something like /kʰ/? I've noticed the loanword for "kayak" in Faroese is taken as "kajakk" [kʰajaʰkː] from the original "qajaq" [qajaq]. However, this does not seem to happen in the reverse: Kalaallisut "kona" [kona] is a loan from Old Norse kona [kʰona] (as opposed to "qona").
Or, should I expect aspiration to drop entirely meaning a merger between /kʰ/, /k/ and /q/? This seems the most likely scenario for a creole, no?
EDIT: Yes that was me. I took it down, but then put it back up again. Sorry!
Alien conlangs (Font may be needed for Vai symbols)
- eldin raigmore
- korean
- Posts: 6352
- Joined: 14 Aug 2010 19:38
- Location: SouthEast Michigan
Re: (Conlangs) Q&A Thread - Quick questions go here
I would just call it “adding a patient”.TwistedOne151 wrote: ↑15 Oct 2018 00:49 So, I'm looking at valency-increasing operations, and the general example is the causative (which seems to cover well most intransitive-transitive pairs). Applied to an intransitive verb, it adds an agent and turns the former subject into the patient: "the chair broke" > "I broke the chair," "the ball dropped" > "I dropped the ball," "the book lay on the shelf" > "I laid the book on the shelf," and so on.
The question is, what would be the term for the operation that turns an intransitive verb transitive by adding a patient, such as turning intransitive "I ran" into transitive "I ran a marathon," or "the team won" into "the team won the tournament," and so on?
P.S. Also what Ahzoh says.
Edit: Also, what Omzinesey says in the next post.
Last edited by eldin raigmore on 17 Oct 2018 16:42, edited 2 times in total.
My minicity is http://gonabebig1day.myminicity.com/xml
Re: (Conlangs) Q&A Thread - Quick questions go here
Aikhenvald&Dixon use term E [xtension] for those trivial objects.TwistedOne151 wrote: ↑15 Oct 2018 00:49 So, I'm looking at valency-increasing operations, and the general example is the causative (which seems to cover well most intransitive-transitive pairs). Applied to an intransitive verb, it adds an agent and turns the former subject into the patient: "the chair broke" > "I broke the chair," "the ball dropped" > "I dropped the ball," "the book lay on the shelf" > "I laid the book on the shelf," and so on.
The question is, what would be the term for the operation that turns an intransitive verb transitive by adding a patient, such as turning intransitive "I ran" into transitive "I ran a marathon," or "the team won" into "the team won the tournament," and so on?
Agent and patient are semantic terms. Subjects of intransitive verbs are agents or patients too, so the former subject doesn't turn to the patient but an object.
Terms unergative verb (i.e. intransitive having an agent) and unaccusative (i.e. intraintransitive having a patient) are handy for speaking.
My meta-thread: viewtopic.php?f=6&t=5760
Re: (Conlangs) Q&A Thread - Quick questions go here
I'm working on my first real conlang, and I was wondering if my consonant inventory seemed reasonable. It's a language for my own personal brand of orcs.
I know labials are extremely common in natlangs, but I've decided not to use any both for a distinct style and because they're harder to pronounce with giant tusks.
Here they are:
ʋ, t, d, (θ), ð, z, tʃ, dʒ, n, r, l, ɖ, ʐ, (ʃ), ʒ, j, k, g, x, ʔ, h
Parentheses indicate letter that will be phones but not seperare phonemes. (Probably.) I feel like it's funky but honestly I have no idea. Any suggestions/thoughts appreciated!
I know labials are extremely common in natlangs, but I've decided not to use any both for a distinct style and because they're harder to pronounce with giant tusks.
Here they are:
ʋ, t, d, (θ), ð, z, tʃ, dʒ, n, r, l, ɖ, ʐ, (ʃ), ʒ, j, k, g, x, ʔ, h
Parentheses indicate letter that will be phones but not seperare phonemes. (Probably.) I feel like it's funky but honestly I have no idea. Any suggestions/thoughts appreciated!
- LinguistCat
- sinic
- Posts: 325
- Joined: 06 May 2017 07:48
Re: (Conlangs) Q&A Thread - Quick questions go here
One thing is that it's a bit odd to have [z] and [ʃ] (among others) but not [s*] even as a major allophone, especially since you also have θ/ð. Is there any specific reason for that choice outside of aesthetics? (Which is fine if that's why you chose to leave it out, but it's good to know the rules before breaking them consciously.)
----
On my own note, what are good ways to derive /l/ (or other laterals) in a phonology that doesn't have it. I figure coronals would be a good place to start, but am not sure in what environments this would be likely to happen.
* put this in there to stop bbcode hopefully it works
- DesEsseintes
- mongolian
- Posts: 4331
- Joined: 31 Mar 2013 13:16
Re: (Conlangs) Q&A Thread - Quick questions go here
In Hong Kong Cantonese, n → l occurs near unconditionally in initial position.LinguistCat wrote: ↑18 Oct 2018 07:00 On my own note, what are good ways to derive /l/ (or other laterals) in a phonology that doesn't have it. I figure coronals would be a good place to start, but am not sure in what environments this would be likely to happen.
* put this in there to stop bbcode hopefully it works
Any coronal rhotic could become /l/, either unconditionally or in coda position.
Another candidate would be lenition of/d/.
Re: (Conlangs) Q&A Thread - Quick questions go here
I dropped [s*] because I determined the [s*]/[z] distinction would be harder to make with a pronounced underbite & tusks. (I stuck some chopsticks in my mouth and tried to pronounce sounds. Very. Scientific.)LinguistCat wrote: ↑18 Oct 2018 07:00
One thing is that it's a bit odd to have [z] and [ʃ] (among others) but not [s*] even as a major allophone, especially since you also have θ/ð. Is there any specific reason for that choice outside of aesthetics? (Which is fine if that's why you chose to leave it out, but it's good to know the rules before breaking them consciously.)
(I understand the * after the s, now, lol.)
- Creyeditor
- MVP
- Posts: 5091
- Joined: 14 Aug 2012 19:32
Re: (Conlangs) Q&A Thread - Quick questions go here
*r > l is also a good way. In extreme cases of lenition you could end up with *t > l maybe via r.DesEsseintes wrote: ↑18 Oct 2018 07:44In Hong Kong Cantonese, n → l occurs near unconditionally in initial position.LinguistCat wrote: ↑18 Oct 2018 07:00 On my own note, what are good ways to derive /l/ (or other laterals) in a phonology that doesn't have it. I figure coronals would be a good place to start, but am not sure in what environments this would be likely to happen.
* put this in there to stop bbcode hopefully it works
Any coronal rhotic could become /l/, either unconditionally or in coda position.
Another candidate would be lenition of/d/.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 2 3 4 4
Ook & Omlűt & Nautli languages & Sperenjas
Papuan languages, Morphophonology, Lexical Semantics
"Thoughts are free."
Produce, Analyze, Manipulate
1 2 3 4 4
Ook & Omlűt & Nautli languages & Sperenjas
Papuan languages, Morphophonology, Lexical Semantics
Re: (Conlangs) Q&A Thread - Quick questions go here
Several Slavic languages have laterals that are a by-product of palatalization; specifically, *pʲ *bʲ *mʲ > pʎ bʎ mʎ (> pl bl ml)Creyeditor wrote: ↑18 Oct 2018 09:11*r > l is also a good way. In extreme cases of lenition you could end up with *t > l maybe via r.DesEsseintes wrote: ↑18 Oct 2018 07:44In Hong Kong Cantonese, n → l occurs near unconditionally in initial position.LinguistCat wrote: ↑18 Oct 2018 07:00 On my own note, what are good ways to derive /l/ (or other laterals) in a phonology that doesn't have it. I figure coronals would be a good place to start, but am not sure in what environments this would be likely to happen.
* put this in there to stop bbcode hopefully it works
Any coronal rhotic could become /l/, either unconditionally or in coda position.
Another candidate would be lenition of/d/.
Blog: audmanh.wordpress.com
Conlangs: Ronc Tyu | Buruya Nzaysa | Doayâu | Tmaśareʔ
Conlangs: Ronc Tyu | Buruya Nzaysa | Doayâu | Tmaśareʔ
-
- korean
- Posts: 10373
- Joined: 12 Jul 2013 23:09
- Location: UTC-04:00
Re: (Conlangs) Q&A Thread - Quick questions go here
Welcome to the board!cucy wrote: ↑18 Oct 2018 06:24 I'm working on my first real conlang, and I was wondering if my consonant inventory seemed reasonable. It's a language for my own personal brand of orcs.
I know labials are extremely common in natlangs, but I've decided not to use any both for a distinct style and because they're harder to pronounce with giant tusks.
Here they are:
ʋ, t, d, (θ), ð, z, tʃ, dʒ, n, r, l, ɖ, ʐ, (ʃ), ʒ, j, k, g, x, ʔ, h
Parentheses indicate letter that will be phones but not seperare phonemes. (Probably.) I feel like it's funky but honestly I have no idea. Any suggestions/thoughts appreciated!
There are a few natural languages and language families which largely lack labials, but I should point out that the labiodental approximant [ʋ] technically falls under the umbrella of "labial". Are you just trying to avoid bilabial consonants?
As LinguistCat said, I'd expect [s] to at least appear allophonically, especially since you have [θ ʃ]. Some other things that stand out to me are the fact that /t͡ʃ/ is a phoneme, but [ʃ] is not, the fact that /ɖ/ doesn't have a voiceless counterpart like /d g/ do, and the fact that (not counting /h/) /x/ is the only fricative phoneme represented as voiceless.
These aren't huge, unreasonable issues, at least not in my opinion, and this is your language, so you can do whatever you'd like, especially since it sounds like you're making it for a fantasy setting.
Have fun!
- LinguistCat
- sinic
- Posts: 325
- Joined: 06 May 2017 07:48
Re: (Conlangs) Q&A Thread - Quick questions go here
These are all pretty helpful. I might avoid r > l just because I'm already turning a lot of /r/'s or consonant clusters with /r/ in them into /r:/. So I don't want to lose all my /r/'s or have to pull too many things over to fill in. The palatalization route might be my best choice and I also thought of having some /s/'s become /l/'s via /ɬ/ somehow.cedh wrote: ↑18 Oct 2018 11:32Several Slavic languages have laterals that are a by-product of palatalization; specifically, *pʲ *bʲ *mʲ > pʎ bʎ mʎ (> pl bl ml)Creyeditor wrote: ↑18 Oct 2018 09:11*r > l is also a good way. In extreme cases of lenition you could end up with *t > l maybe via r.DesEsseintes wrote: ↑18 Oct 2018 07:44In Hong Kong Cantonese, n → l occurs near unconditionally in initial position.LinguistCat wrote: ↑18 Oct 2018 07:00 On my own note, what are good ways to derive /l/ (or other laterals) in a phonology that doesn't have it. I figure coronals would be a good place to start, but am not sure in what environments this would be likely to happen.
Any coronal rhotic could become /l/, either unconditionally or in coda position.
Another candidate would be lenition of/d/.
Re: (Conlangs) Q&A Thread - Quick questions go here
cucy wrote: ↑18 Oct 2018 06:24 I'm working on my first real conlang, and I was wondering if my consonant inventory seemed reasonable. It's a language for my own personal brand of orcs.
I know labials are extremely common in natlangs, but I've decided not to use any both for a distinct style and because they're harder to pronounce with giant tusks.
Here they are:
ʋ, t, d, (θ), ð, z, tʃ, dʒ, n, r, l, ɖ, ʐ, (ʃ), ʒ, j, k, g, x, ʔ, h
Parentheses indicate letter that will be phones but not seperare phonemes. (Probably.) I feel like it's funky but honestly I have no idea. Any suggestions/thoughts appreciated!
Code: Select all
n
t tʃ k ʔ
d dʒ ɖ g
(θ) (ʃ) x h
ð z ʒ ʐ
r
ʋ l j
The lack of a voiceless counterpart to /ɖ/ is attested in Somali (which also lacks a voiceless counterpart to /dʒ/), although if I remember correctly /ɖ/ derives from an older implosive.
As others have said, the only thing that really strikes me as odd is the fact that all of the fricatives barring /x/ and /h/ (and those marked in rounded brackets) are voiced. Personally, what I'd do, is to say you have /θ s ʃ ʂ/ and have these become voiced allophonically, say, between vowels and adjacent to other voiced consonants (/d dʒ ɖ g r ʋ l j/). That way you keep the phoneme inventory the same size, and keep those voiced fricatives in a number of environments. And I don't see a reason why not having /x/ and /h/ undergo those changes would be implausible either (say they just don't become voiced between vowels and never occur adjacent to other consonants).
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Re: (Conlangs) Q&A Thread - Quick questions go here
Whoops, yes, ʋ is a labial, I have that as my only labial sound! I misspoke, I think I did mean bilabial.shimobaatar wrote: ↑18 Oct 2018 14:12 There are a few natural languages and language families which largely lack labials, but I should point out that the labiodental approximant [ʋ] technically falls under the umbrella of "labial". Are you just trying to avoid bilabial consonants?
As LinguistCat said, I'd expect [s] to at least appear allophonically, especially since you have [θ ʃ]. Some other things that stand out to me are the fact that /t͡ʃ/ is a phoneme, but [ʃ] is not, the fact that /ɖ/ doesn't have a voiceless counterpart like /d g/ do, and the fact that (not counting /h/) /x/ is the only fricative phoneme represented as voiceless.
Okay, what I think I'm doing is adding /s, ʂ/, and along with /θ, tʃ, ʃ/, will be allophones to their voiced counterparts. I'm not sure under which circumstances they'd change yet. (I see your (sangi39) suggestion but I don't know why you made it, so I want to do more research rather than just copying what you said.)sangi39 wrote: ↑18 Oct 2018 23:14 The lack of a voiceless counterpart to /ɖ/ is attested in Somali (which also lacks a voiceless counterpart to /dʒ/), although if I remember correctly /ɖ/ derives from an older implosive.
As others have said, the only thing that really strikes me as odd is the fact that all of the fricatives barring /x/ and /h/ (and those marked in rounded brackets) are voiced. Personally, what I'd do, is to say you have /θ s ʃ ʂ/ and have these become voiced allophonically, say, between vowels and adjacent to other voiced consonants (/d dʒ ɖ g r ʋ l j/). That way you keep the phoneme inventory the same size, and keep those voiced fricatives in a number of environments. And I don't see a reason why not having /x/ and /h/ undergo those changes would be implausible either (say they just don't become voiced between vowels and never occur adjacent to other consonants).
/ʈ/ will be a seperate phoneme, since the other plosives have voice and unvoiced as separate phonemes. /x/ will be /ɣ/ and /h/ will be /ɦ/, and I'll figure out what to do with /x/ and /h/, I suppose.
Thanks so much!