zompist bboard

Would it be possible, in some hypothetical future variety of English, for methathesis to occur thay would switch syllabic sonoranit into the coda of the preceding syllable, e.g.:

castle /kʰæsɫ̩/ → /kʰæɫs/
nation /neɪ̯ʃn̩/ → /neɪ̯nʃ/
acre /eɪ̯ʔkɻ̩/ → /eɪ̯ɻʔk/

Or should I go for something more indirect, e.g.:

castle /kʰæsɫ̩/ → /kʰæsɫə/ → /kʰæɫsə/ → /kʰæɫs/
or castle /kʰæsɫ̩/ → /kʰæsəɫ/ → /kʰæsɫə/ → /'kʰæɫsə/ → /kʰæɫs/?

I've also got another question: what conditional sound changes do people here like the most? I've been trying to put some in my conlang's diachronica, but I don't know which ones to choose.

Metathesis is usually a sporadic change.

mèþru wrote:Metathesis is usually a sporadic change.

While I can't think of examples off the top of my head, it can be regular.

@Knit Tie: I find that reasonably plausible myself.

Zaarin wrote:

mèþru wrote:Metathesis is usually a sporadic change.

While I can't think of examples off the top of my head, it can be regular.

Slavic liquid metathesis was regular as all hell.

@Knit Tie: I find that reasonably plausible myself.

Even the top version, without any schwas?

Knit Tie wrote:

Zaarin wrote:

mèþru wrote:Metathesis is usually a sporadic change.

While I can't think of examples off the top of my head, it can be regular.

Slavic liquid metathesis was regular as all hell.

@Knit Tie: I find that reasonably plausible myself.

Even the top version, without any schwas?

Sure. I understand the motivation for the schwas: it makes the resonants non-syllabic. But the phonemic status of syllabic resonants in English is sufficiently blurry that I don't have a problem buying that they become non-syllabic when moved to an environment where English does not usually have syllabic resonants (i.e., after vowels).

I would expect vocalization.

Nortaneous wrote:I would expect vocalization.

Yeah, that's probably the most plausible option. I do wonder, though, what would the nasals vocalise to? Rhotic will turn into a low vowel, the lateral into a back vowel, but what about the nasals? Nasal vowels? Just /a/?

Knit Tie wrote:I do wonder, though, what would the nasals vocalise to? Rhotic will turn into a low vowel, the lateral into a back vowel, but what about the nasals? Nasal vowels? Just /a/?

I guess the simple answer is "pretty much anything", I mean just look at the different results of the loss of syllabic resonants across Indo-European. I myself would be inclined towards something line a nasalised schwa, but it's really up to you.

-əl either merges into -əw (back when I didn't have a Gameboy but heard people talking about Pokemon I picked up that there was one called /bijdrijəw/) or becomes -uw. The rest seem stable for now, unless you're assuming nonrhoticity, which I wouldn't for future AmEng because sociolinguistics. So it's out of projecting existing trends forward and into pure conshit, so you can do whatever.

Here there's l-vocalization, which I've described before, and some final /n/ elision (which could be interpreted as [n̩] > [ɘ̃]), particularly before vowels, but rhoticity seems quite stable except in certain words with multiple rhotics where one of the rhotics is lost (e.g. caterpillar), which is a feature shared with much of NAE.

How could I achieve that /x/ is lost, but /h/ is preserved, and later becomes /x/? This is admittedly a weird change, one would expect /h/ to get lost easier than /x/; hence, some intermediate steps are probably necessary to do the job. All this would have to go its course in at most 2,000 years. Also, there are labialized and palatalized counterparts of /x/ and a labialized /h/. These labializations and palatalizations must be preserved. Any ideas?

@WeepingElf - Middle English had /x/ and /ç/ as allophones of /h/ in coda and lost them both or, sporadically, merged them with /f/. So perhaps you could restrict them to different positions and then just nuke the one that has /x/? If you want to preserve palatalised and labialised /x/s, I think you could just go /xʲ/ → /ç/ → /ʃ/ and /xʷ/ you can easily change into any of /p/ /b/ /f/ /v/ /w/ or even /ʍ/.

I know of the English case, but in this case, things are different - /x/ and /h/ are different, constrasting phonemes. The protolanguage has these fricatives: /f s x^j x x^w h/. (There is a stop series at each of these POAs except glottal.) Of these, /f/ first becomes /h^w/. What should happen then is in one language that the velar fricatives drop, but the glottals survive as /h/, later to become /x/. In a sister language, the velars merge with the glottals, with the secondary articulations staying intact. (/s/ stays intact in both languages.)

BTW: If you haven't guessed it by now, what I am working on aren't really conlangs (though it has repercussions on my main conlang family); it is a hypothesis about the PIE laryngeals of mine we are currently discussing in the Great PIE Thread at L&L. The two daughter languages mentioned above are, of course, Hittite and Late PIE. People so often have good ideas about what kind of sound changes are plausible and what is attested that I thought they could say something on this, too.

BTW2: I've just got an idea. The glottal fricatives harden to glottal stops; then the velar fricatives get lost; in a third step the glottal stops become uvular stops, which yet later fricativize. Does that make sense?

Are there sound changes besides dropping the last consonant or vowel of a word that could happen to the last syllable of a word?

I was thinking since /k/ in the proto-language already voices to [g] in intervocalic positions, and there's free variation between the stop form and fricatives [x~ɣ] (voicing still dependent on environment), it would be fair to say that [ɣ] became the predominant pronunciation intervocalically, especially at the ends of words. (ETA: This would be syllable initial but in the final syllable of a word.)

Similar sound changes might be applied to the other stops but I wanted to see if it was feasible for this change specifically first. Thanks!

Both fortition (final fortition in German, Dutch, Polish, Russian, etc.) and lenition (English s > z finally in many words) frequently occur word-finally or, more broadly, syllable-finally.

Travis B. wrote:Both fortition (final fortition in German, Dutch, Polish, Russian, etc.) and lenition (English s > z finally in many words) frequently occur word-finally or, more broadly, syllable-finally.

This would be syllable initial but in the final syllable of a word. I hope that clears up what I meant.

LinguistCat wrote:This would be syllable initial but in the final syllable of a word. I hope that clears up what I meant.

One could have lenition of unstressed onsets combined with an trochaic word structure where the final syllable of a polysyllabic word is always unstressed.

Travis B. wrote:

LinguistCat wrote:This would be syllable initial but in the final syllable of a word. I hope that clears up what I meant.

One could have lenition of unstressed onsets combined with an trochaic word structure where the final syllable of a polysyllabic word is always unstressed.

Yes that is more than doable and should work quite nicely. Thanks.

LinguistCat wrote:Are there sound changes besides dropping the last consonant or vowel of a word that could happen to the last syllable of a word?

I was thinking since /k/ in the proto-language already voices to [g] in intervocalic positions, and there's free variation between the stop form and fricatives [x~ɣ] (voicing still dependent on environment), it would be fair to say that [ɣ] became the predominant pronunciation intervocalically, especially at the ends of words. (ETA: This would be syllable initial but in the final syllable of a word.)

Similar sound changes might be applied to the other stops but I wanted to see if it was feasible for this change specifically first. Thanks!

Final consonants can debuccalize, as in English and some other languages. Final vowels can lengthen/break (presumably under stress) or shorten/reduce (as in English).

Speaking of stress, I'd like to do some stress-based lentition/fortition of consonants too. How does /w/ → /gʷ/ → /g/ → /ɣ/ and /j/ → /ʒ/ in the onset of stressed syllables and /m/ → /w̃/ → ∅ in codas of unstressed syllables sound? If I go full-bore with the stress-consonant interaction and divide the entire phonology into two sets of phonemes, one of which can only be founf in unstressed syllables and one in stressed ones, would that be too much?

Knit Tie wrote:Speaking of stress, I'd like to do some stress-based lentition/fortition of consonants too. How does /w/ → /gʷ/ → /g/ → /ɣ/ and /j/ → /ʒ/ in the onset of stressed syllables and /m/ → /w̃/ → ∅ in codas of unstressed syllables sound? If I go full-bore with the stress-consonant interaction and divide the entire phonology into two sets of phonemes, one of which can only be founf in unstressed syllables and one in stressed ones, would that be too much?

g > ɣ is the sort of change I would expect in the onsets of unstressed syllables (especially when intervocalic) or in codas, not in the onsets of stressed syllables (or at least not at the start of an utterance).

You could have g > ɣ in all syllables instead.

Knit Tie wrote:divide the entire phonology into two sets of phonemes, one of which can only be founf in unstressed syllables and one in stressed ones, would that be too much?

Then they might be considered allophonic sets rather than two phonemic sets.

In a small consonant inventory

/m n p t k s j r w/

would it be reasonable for nasal+stop clusters to become nasal+/j/ or at least more generally nasal+approximant?

LinguistCat wrote:In a small consonant inventory

/m n p t k s j r w/

would it be reasonable for nasal+stop clusters to become nasal+/j/ or at least more generally nasal+approximant?

The general patterns I have seen is that nasal+stop clusters voice the stop and preserve stop-ness and rather voiced stops (which you could have allophonically) tend to lenite when by themselves.

Travis B. wrote:

LinguistCat wrote:In a small consonant inventory

/m n p t k s j r w/

would it be reasonable for nasal+stop clusters to become nasal+/j/ or at least more generally nasal+approximant?

The general patterns I have seen is that nasal+stop clusters voice the stop and preserve stop-ness and rather voiced stops (which you could have allophonically) tend to lenite when by themselves.

Yes, but I'm trying to go in a different direction than Japanese (which basically what you said at an early point) and looking for ways to include more nasals in general. So I was wondering if this looked reasonable, even if not the most likely. If not I'll just make them long nasals.