Sound Change Quickie Thread

[Nortaneous]

I'm trying to get rid of /T/ without going to /s/ in most of the cases. What about these changes?

I doubt it, but you could shift it to any of /f t x/ or maybe even /K/.

[Thry]

I'm trying to get rid of /T/ without going to /s/ in most of the cases. What about these changes?

I doubt it, but you could shift it to any of /f t x/ or maybe even /K/.

I know, I'm already shifting it to [f] word-initially.

Then I'll do [faTas] > [faas] > [fajas], which is not implausible I hope...
and then [paT] > [pa] or [paj] by emphasis or analogy with something.

[Pole, the]

But the language has also /r d/, which remained unchanged. So, maybe the *z > ð > l would be more plausible?

(It's something like a peripheral Slavic conlang, example changes: *voz- > wolo, *zemja > lemi, *polz- (?) > plalo, *vezǫ > welo [velo].)

[Tropylium]

Is /T/ and/or /D/ becoming /l/ attested? I'm using it for English > Mekoshan.

*ð > l is well attested in Tavastian Finnish. Its voiceless pair goes *θ > t̪, however. (These are independant developments from the possible *ð > l reconstructed for older stages of Uralic elsewhere.)

[Qwynegold]

Thanks for all the answers to my question. I'm doing something similar to a few of the suggestions.

Is /T/ and/or /D/ becoming /l/ attested? I'm using it for English > Mekoshan.

Not sure if it's attested, but D > l sounds fine to me.

It seems to have happened in some Uralic languages which reflect Proto-Uralic *ð as /l/. It is uncertain whether *ð really was /ð/ or something else, though. Some Uralicists assume that PU *ð really was a lateral obstruent instead.

Some dialect of Finnish did that, I think. EDIT: Never mind.

I'm trying to get rid of /T/ without going to /s/ in most of the cases. What about these changes?

[paT] > [paj]
[faTas] > [fajas]

T > j, V#

If it's too implausible, I could have it elide and explain the [j] some different way.

Another dialect of Finnish did ð > j , maybe the same could happen to θ?

[finlay]

Any fricative can debuccalise to h. θ already does in many scottish varieties.

[Qwynegold]

I have a question! I have a conlang where vowels in unstressed initial syllables become the same as the following syllable's vowel, eg. V₁[-stressed] > V₂ / _CV₂. But then a lot of sound changes happen and, I don't know, maybe in half of the words the vowel in the initial syllable is no longer same as the following syllable's vowel. Would it be plausible then to have a sound change where this initial vowel changes to ə, but only if the following vowel is different, eg. V₁ > ə / #(C)_CV !_CV₁?

[----]

Soo, I'm trying to decide some sound changes for Kàmo, and I can't figure out what to do with /ɾ/. What can this phoneme turn into? For the purpose of convenience, assume that all syllables in the language are (C)V (this is pretty close to being the case anyway).

[Thry]

Soo, I'm trying to decide some sound changes for Kàmo, and I can't figure out what to do with /ɾ/. What can this phoneme turn into? For the purpose of convenience, assume that all syllables in the language are (C)V (this is pretty close to being the case anyway).

Have you already decided not to turn it into [l], [n] (through [l] maybe), [x], [X], [h]?

[Click]

Would it be plausible to derive clicks from clusters of affricates?

[Nortaneous]

Soo, I'm trying to decide some sound changes for Kàmo, and I can't figure out what to do with /ɾ/. What can this phoneme turn into? For the purpose of convenience, assume that all syllables in the language are (C)V (this is pretty close to being the case anyway).

g ɣ gʟ

[----]

Soo, I'm trying to decide some sound changes for Kàmo, and I can't figure out what to do with /ɾ/. What can this phoneme turn into? For the purpose of convenience, assume that all syllables in the language are (C)V (this is pretty close to being the case anyway).

g ɣ gʟ

I could conceivably see it turning into the first two, but I can't think of a relatively short line of sound changes that would give me that third one. How is it going to get to that?

[*Ceresz]

Soo, I'm trying to decide some sound changes for Kàmo, and I can't figure out what to do with /ɾ/. What can this phoneme turn into? For the purpose of convenience, assume that all syllables in the language are (C)V (this is pretty close to being the case anyway).

g ɣ gʟ

I believe /ɾ/ → [ʃ] and /ɾʲ/ → [ð] are attested in a few Scottish Gaelic dialects, just to give you a few other options. I could see /ɾ/ becoming a voiced dental fricative without the palatalization too.

[----]

I'd rather not have /ʃ/ since the sound changes I already have planned result in quite a lot of /ɕ/ and I honestly can't tell the difference between them.
Those other suggestions are cool though. Kind of a tangent, but are there any natlangs that have /g/ as a reflex of earlier /r/?

[Nortaneous]

Soo, I'm trying to decide some sound changes for Kàmo, and I can't figure out what to do with /ɾ/. What can this phoneme turn into? For the purpose of convenience, assume that all syllables in the language are (C)V (this is pretty close to being the case anyway).

g ɣ gʟ

I could conceivably see it turning into the first two, but I can't think of a relatively short line of sound changes that would give me that third one. How is it going to get to that?

I have no idea, but Hiw has gʟ for Proto-Austronesian *r.

Kind of a tangent, but are there any natlangs that have /g/ as a reflex of earlier /r/?

I think there are some Austronesian langs that do. Atayal reflects Proto-Austronesian *R (probably an alveolar trill) as /g/ everywhere but before /i/, where it has /r/. Other options attested in Formosan langs: x l ɬ ɽ ʀ j Ø. Also, Marquesan merges Proto-Polynesian *l *r as a glottal stop.

[Qwynegold]

I have a question! I have a conlang where vowels in unstressed initial syllables become the same as the following syllable's vowel, eg. V₁[-stressed] > V₂ / _CV₂. But then a lot of sound changes happen and, I don't know, maybe in half of the words the vowel in the initial syllable is no longer same as the following syllable's vowel. Would it be plausible then to have a sound change where this initial vowel changes to ə, but only if the following vowel is different, eg. V₁ > ə / #(C)_CV !_CV₁?

Or would it just be better to just have V[-stressed] > ə / #(C)_, even though this would be a reversal of an earlier sound change? Because it went like this:
Rule 1: V[-stressed] > ə
Rule 2: ə[-stressed] > V₂ / #(C)_CV₂
So if I have Rule 3: V[-stressed] > ə / #(C)_, it would undo that Rule 2, which apparently does not happen. Though Rule 3 happens a lot later than Rule 2, with a lot of other sound changes between.

Then I have another question. Which one of these is most plausible:
1. ʔC > C[+glottalized]
2. Cʔ > C[+glottalized]
3. Both

Or can I just pick any of these three options?

[Thry]

I'm trying to get rid of /T/ without going to /s/ in most of the cases. What about these changes?

[paT] > [paj]
[faTas] > [fajas]

T > j, V#

If it's too implausible, I could have it elide and explain the [j] some different way.

Another dialect of Finnish did ð > j , maybe the same could happen to θ?

Oh, that's great (hadn't seen it before), because I in fact have had /θ/ voice to [ð] between vowels so we don't even have to wonder. I'll have it turn to [f] word-initially, [s] in clusters, [ð] then [j] between vowels and disappear word-finally, then having [j] be formed by analogy. For example:

Latin faciat > ['fa.ki.at] > ['fa.ki.a] > ['fa.kja] > ['fa.cʲa] > ['fa.tsa] > *['fa.sɐ] > ['fa.θɐ] > ['fa.ðɐ] > ['fa.jɐ] > ['fɛ.jɐ] spelt faça.

Then facit > ['fa.kit] > ['fa.ke] > ... > *['fa.sə] > ['fa.θə] > [faθ] > [fa] then variant [faj] > [fɛj], spelt faç or fa (fai would not be normative spelling).

*Laminal [s]

[Melteor]

I had an idea for a condialect for interlingua. Basically, what interlingua does is regularize words and their derivatives, so you get 'nocte' for 'night' because of 'nocturnal', even though all the modern languages have lost the /k/. Im leaving the vowels alone but the consonants I feel are up to be played with. Are there regular sound mergers which would be legal like simplifying clusters ie /kt/>/t/? I was thinking of turning /p t d k/ into fricatives [ɸ θ D x] with the resulting language sounding kind of like parseltongue. How valid is this radical sound change and how understandable is it to a native speaker of say Spanish?

Ex. 'honestitate' > [onesθi'θaθe]

[Thry]

I had an idea for a condialect for interlingua. Basically, what interlingua does is regularize words and their derivatives, so you get 'nocte' for 'night' because of 'nocturnal', even though all the modern languages have lost the /k/. Im leaving the vowels alone but the consonants I feel are up to be played with. Are there regular sound mergers which would be legal like simplifying clusters ie /kt/>/t/? I was thinking of turning /p t d k/ into fricatives [ɸ θ D x] with the resulting language sounding kind of like parseltongue. How valid is this radical sound change and how understandable is it to a native speaker of say Spanish?

Ex. 'honestitate' > [onesθi'θaθe]

What do you mean by "legal"? The clusters have been naturally simplified (save in my romlang, lol, it still has noct, but that doesn't count; however, Romanian still has a cluster, it has just changed the POA to labial, noapte). It could not be legal depending on what you mean. Interlingua aimed to convert them back to the Latin cluster. So what you're doing is going in the natlang direction again.

About the understandability, not much and not the same for everyone, consider that those <t>s are now regarded as /d/ by speakers of Spanish, Portuguese (honestidad(e))... only Italians and such would think of underlying /t/. And [θ] is harder to relate to /d/ than it is to /t/. Plus some languages, such as Spanish, already have /θ x/ so that plays a role in the interpretation.

Something as simple as "attack" atake > [aθaxe] has its understandability lost.

[Melteor]

I had an idea for a condialect for interlingua. Basically, what interlingua does is regularize words and their derivatives, so you get 'nocte' for 'night' because of 'nocturnal', even though all the modern languages have lost the /k/. Im leaving the vowels alone but the consonants I feel are up to be played with. Are there regular sound mergers which would be legal like simplifying clusters ie /kt/>/t/? I was thinking of turning /p t d k/ into fricatives [ɸ θ D x] with the resulting language sounding kind of like parseltongue. How valid is this radical sound change and how understandable is it to a native speaker of say Spanish?

Ex. 'honestitate' > [onesθi'θaθe]

What do you mean by "legal"? The clusters have been naturally simplified (save in my romlang, lol, it still has noct, but that doesn't count; however, Romanian still has a cluster, it has just changed the POA to labial, noapte). It could not be legal depending on what you mean. Interlingua aimed to convert them back to the Latin cluster. So what you're doing is going in the natlang direction again.

About the understandability, not much and not the same for everyone, consider that those <t>s are now regarded as /d/ by speakers of Spanish, Portuguese (honestidad(e))... only Italians and such would think of underlying /t/. And [θ] is harder to relate to /d/ than it is to /t/. Plus some languages, such as Spanish, already have /θ x/ so that plays a role in the interpretation.

Something as simple as "attack" atake > [aθaxe] has its understandability lost.

Well, what I meant was that the orthography would be etymological and would show things that would be underlying but wouldn't always show up a laFrench. Not sure how to get that to work.

I didn't know, these things are hard to judge. In Tuscany they would pronounce 'atake' like that. The fricative change only spplies intervocalically though, not to clusters. This is mostly an issue because interlingua didn't concern itself with the spoken languages, or so I read. It's not as big in scope a project as Slovianski for instance.
http://steen.free.fr/interslavic/nms.html

What romlang do you speak?

Another thing is, do speakers of Spanish appreciate nasal vowels as v+n or v+m? I would think so in Spain because of exposure to Portugese.

[Thry]

Yep, I guess that reading inelligibility is a primary goal is a reasonable idea.

What romlang do you speak?

My natlang is Spanish, and I'm Spanish. "Romlang" usually means "constructed romance language, a romance conlang".

Another thing is, do speakers of Spanish appreciate nasal vowels as v+n or v+m? I would think so in Spain because of exposure to Portugese.

Depends - Portuguese is more exposed to Spanish than Spanish to Portuguese, and they're so related that many times a person already knows the word and thinks of the Spanish one, so it's hard to tell. I may be a bit polluted because I already speak Portuguese. People here normally realize São paulo as /sao paulo/, but I'd be inclined to say that has influence from orthography (c.f. also Instituto camões as /instituto kamoes/. But I guess people hearing Portuguese sim would hear sin rather than si, so yes (more n than m).

[Melteor]

Yep, I guess that reading inelligibility is a primary goal is a reasonable idea.

Little known fact: Interlingua actually has two orthographies, but one is more popular than the other. The other one looks less like French and more like Spanish, and leaves the 'e' off the end of most words. Supposedly, it's a little harder to read without the 'e's on the ends though because many romance language speakers' languages have a vowel on the end of those words, it's just they usually also have a gendered ending (so spanish o/a endings give way to a neuter e); and in the case of French there is just an optional schwa. So in a way interlingua already has wild card letters, but it's like French. We can't decide whether to speak them or not. The default is to always pronounce them but I find it challenging. Another thing is initial /st/ or clusters of three letters which are okay for English speakers but hard for romance speakers. My natlang is English so I struggle to make consistent vowels. The less complex the syllables the easier I can make the vowels consistently. None of this would be so hard if I had a gold standard to emulate for pronouncing interlingua, which I imagine is kind of challenging for a speaker of any romance language. Someone has to prove me otherwise.

My natlang is Spanish, and I'm Spanish. "Romlang" usually means "constructed romance language, a romance conlang".

interesting. Do you find interlingua texts at all difficult to pronounce, or aesthetically clunky?
http://soundcloud.com/mcdutchie/declara ... versal-del how natural-sounding is this pronunciation? I really would rather learn to speak it fast than slow. English murders it's vowels for speed and that won't fly I don't think for a romance language.

Depends - Portuguese is more exposed to Spanish than Spanish to Portuguese, and they're so related that many times a person already knows the word and thinks of the Spanish one, so it's hard to tell. I may be a bit polluted because I already speak Portuguese. People here normally realize São paulo as /sao paulo/, but I'd be inclined to say that has influence from orthography (c.f. also Instituto camões as /instituto kamoes/. But I guess people hearing Portuguese sim would hear sin rather than si, so yes (more n than m).

I may be able to get away with pronouncing it like that, then. Maybe plus a homorganic nasal in the coda.

[Thry]

interesting. Do you find interlingua texts at all difficult to pronounce, or aesthetically clunky?
http://soundcloud.com/mcdutchie/declara ... versal-del how natural-sounding is this pronunciation? I really would rather learn to speak it fast than slow. English murders it's vowels for speed and that won't fly I don't think for a romance language.

How am I helpful though? I already speak English, Portuguese, Spanish and a bit of Catalan (which has lots of final consonants), and can read to some extent Italian and French, as well as pronounce German (which I hope to learn in the future), so I don't find difficulties in pronunciation. However, most monolingual Spanish speakers do have trouble with final consonants and initial-word epenthesis, which is logical.

As a biased Spanish speaker, though, I can tell you that when I looked up interlingua for the first time the first things that came to my mind were "why double consonants? there's no need for them at all - they make it harder to learn for people with languages that lack geminates and have to learn the patterns from scratch", and of course the "liquid s" thing, es- wouldn't harm recognizability.

Aesthetically, it looks a bit artificial, especially regarding verbs and nouns (you gotta miss the gender and verbal agreement), but that's it; much more natural-sounding than Esperanto and nothing extraordinary, the words are very recognizable. About the pronunciation, what can I say? It's an auxlang, it depends on who speaks it. I guess I'd relate it to Italian if spoken ideally.

I may be able to get away with pronouncing it like that, then. Maybe plus a homorganic nasal in the coda.

But what were you asking this for, borrowings? If that is the case, you gotta consider how Portuguese nasals vowels are some in fact nasal diphthongs, bem is "be~i", dão is "dãw", etc.

[LinguistCat]

Would it be possible for a language that has /tk/ as an allowable internal cluster to simplify it, except in proper names? Or would names that have it have to be borrowed from another language with this cluster?

[Tropylium]

Well, proper names may resist orthographic reforms, so you could get that result in a number of proper names as a spelling pronunciation.