Sound Change Quickie Thread

[----]

I believe that the German r is more of an approximant than a trill in most cases, though. Very few German speakers use the trill, many can't even pronounce it.

[Qwynegold]

Nasal+stop clusters.

Was this directed at me? You can get implosives out of that?

Btw, what the hell has happened to the Correspondence Library? Now it's just an annoying website that sells computers.

[WeepingElf]

Btw, what the hell has happened to the Correspondence Library? Now it's just an annoying website that sells computers.

Your link is wrong; it is now here.

[Qwynegold]

Btw, what the hell has happened to the Correspondence Library? Now it's just an annoying website that sells computers.

Your link is wrong; it is now here.

Oh, thanks!

[Nortaneous]

Nasal+stop clusters.

Was this directed at me? You can get implosives out of that?

Btw, what the hell has happened to the Correspondence Library? Now it's just an annoying website that sells computers.

Yes. Quoting is a major pain on Android, so.

You could also get them from voiced stops; voiced stops tend to vary with implosives in langs that have proper voiced stops, e.g. French. And even some dialects of English.

[Qwynegold]

Nasal+stop clusters.

Was this directed at me? You can get implosives out of that?

Btw, what the hell has happened to the Correspondence Library? Now it's just an annoying website that sells computers.

Yes. Quoting is a major pain on Android, so.

You could also get them from voiced stops; voiced stops tend to vary with implosives in langs that have proper voiced stops, e.g. French. And even some dialects of English.

Aha, nasals are good then. That way I can get rid of unwanted clusters.

[Grunnen]

I saw a number of threads here of people who were creating future descendants of English, and that got me thinking. So now I am working on a future descendant of Dutch, as that's probably easier for me. I've come up with a number of sound changes of which I'm not sure if they are plausible.

The current phoneme inventory of (my kind of) Dutch could be seen as:

consonants

Code: Select all

 	         labial	alveolar	palatal	(u)v[u/e]lar
plosives	 p b      t d                  k
fricatives  f v      s z                  χ
nasals	   m        n                    ŋ
liquids 	 ʋ        l                    ʀ
semivowels  w                   j 

vowels

Code: Select all

       front fr.round back
high	i ɪ   y ʏ      u
mid    e ɛ   ø        o ɔ         ə
low    a              ɑ
diphs  æi̯   œy̯      ʌu̯

Now I have a stress shift that causes some /ə/'s to end up in stressed syllables, and Dutch doesn't allow that. So I could say that a /ə/ simply becomes an /ʏ/ (as has happened before in Dutch) in that position, but that's not very interesting. At this stage the phoneme inventory given above is still pretty much valid. So now I have this:

ə/ɐ/_(Cn)A, A=a,ɑ,ɛ,ɔ Cn=any number of consonants.
ə/ɪ/_(Cn)I1, I1=ɪˌɪː
ə/i/_(Cn)I2, I2=i,e
ə/ʊ/_(Cn)U, U=u,o,ø
ə/ʏ/_(Cn)Y, Y=y,ʏ,ʏː,ʊː
ɐ/ɔ/_P, P=any labial
ɐ/ɛ/_T, T=any alveolar and j
ɐ/a/_K, K=any velar/uvular
(Yes, here there are long high lax vowels)

Somewhat later in the development of the language stress differences become less pronounced. Would this be plausible under those conditions?:
ə/u/_P, P=any labial
ə/i/_T, T=any alveolar and j
ə/a/_K, K=any velar/uvular

Then another question

Later in the development of the same language I have this sound change:

Code: Select all

wC/P/_
jC/J/_
ɥC/Q/_

C=b d f χ h j k l m n p ʀ s t ʋ w ŋ
P=b b f ʃ f ɥ c ʋ m m p ʎ f p ʋ w ɲ
J=d ɟ s ʃ ʃ j c ʎ n ɲ t ʎ ʃ c l ɥ ɲ
Q=d d s ʃ s ɥ c l n n t ʎ s t l ɥ ɲ

Q develops by first labialising and then palatalizing the labialised forms.

And finally:
voiceless stops in coda position change to glottal stops with a secondary articulation depending on the poa of the original stop.
p/ʔʷ/_(Cn)., .=syllable break, Cn=any number of consonants.
t/ʔ/_(Cn).
c/ʔʲ/_(Cn).
k/ʔˠ/_(Cn).

then the glottal stop disappears, affecting the previous vowel in various ways:
Vʔ/V:/_
the second part of the diphthongs below is nonsyllabic.
tense high vowels
iʔʷ/iu/_
iʔʲ/i:/_
iʔˠ/iɛ/_
yʔʷ/yu/_
yʔʲ/y:/_
yʔˠ/yœ/_
uʔʷ/u:/_
uʔʲ/ui/_
uʔˠ/uɔ/_
lax high vowels
ɪʔʷ/ɪʊ/_
ɪʔʲ/ɪ:/_
ɪʔˠ/ɪɛ/_
ʏʔʷ/ʏʊ/_
ʏʔʲ/ʏ:/_
ʏʔˠ/ʏœ/_
ʊʔʷ/ʊ:/_
ʊʔʲ/ʊɪ/_
ʊʔˠ/ʊɔ/_
tense mid vowels
eʔʷ/eʊ/_
eʔʲ/e:/_
eʔˠ/eɐ/_
øʔʷ/øʊ/_
øʔʲ/ø:/_
øʔˠ/øœ/_
oʔʷ/o:/_
oʔʲ/oɪ/_
oʔˠ/oɐ/_
lax mid vowels
ɛʔʷ/ɛʊ/_
ɛʔʲ/ɛɪ/_
ɛʔˠ/ɛ:/_
ɔʔʷ/ɔʊ/_
ɔʔʲ/ɔɪ/_
ɔʔˠ/ɔ:/_
low vowels
ɑʔʷ/ɑo/_
ɑʔʲ/ɑɛ/_
ɑʔˠ/ɑ:/_
aʔʷ/aɔ/_
aʔʲ/ae/_
aʔˠ/a:/_
Sorry it is such a long and unstructured list, but I don't know how to put it more straightforward.
To anyone who came this far reading this: Thanks a lot!
So, do you think this is plausible?

[Astraios]

Well, don't just stand there with a bunch of sound changes, show us some future-Dutch words!

[Grunnen]

Well, don't just stand there with a bunch of sound changes, show us some future-Dutch words!

straat "street" /stʀat/ becomes /ʰtʀa:/
pompen "to pump" /'pɔmpə(n)/ becomes /fɔb/
binden "to bind" /'bɪndə(n)/ becomes /bɪnt/
ritselen "(ehh)" /'ʀɪtsələ(n)/ becomes /ʀɪ:zɔw/
raspen "to grate" /'ʀɑspə(n)/ becomes /ʀɑʰp/
water "water" /'ʋatəʀ/ becomes /ʋasi:/
melden "to report" /'mɛldə(n)/ becomes /mɛp/
geeuwt "yawns" /χewt/ becomes /ɣɪ:p/ (well, unless the inflectional paradigms get levelled out)
zout "salt" /zʌut/ becomes /zaɔ/
acceleratie "acceleration" /ɑksɛləʀatsi/ becomes /ɑ:'slaʀɐzi/
with the sound changes the way I have them now

and I will be able to respond in about five hours, possibly earlier. Sorry for my bad planning...

[spats]

Well, don't just stand there with a bunch of sound changes, show us some future-Dutch words!

straat "street" /stʀat/ becomes /ʰtʀa:/
pompen "to pump" /'pɔmpə(n)/ becomes /fɔb/
binden "to bind" /'bɪndə(n)/ becomes /bɪnt/
ritselen "(ehh)" /'ʀɪtsələ(n)/ becomes /ʀɪ:zɔw/
raspen "to grate" /'ʀɑspə(n)/ becomes /ʀɑʰp/
water "water" /'ʋatəʀ/ becomes /ʋasi:/
melden "to report" /'mɛldə(n)/ becomes /mɛp/
geeuwt "yawns" /χewt/ becomes /ɣɪ:p/ (well, unless the inflectional paradigms get levelled out)
zout "salt" /zʌut/ becomes /zaɔ/
acceleratie "acceleration" /ɑksɛləʀatsi/ becomes /ɑ:'slaʀɐzi/
with the sound changes the way I have them now

and I will be able to respond in about five hours, possibly earlier. Sorry for my bad planning...

Cool stuff!

What's the final phoneme inventory?

Are preaspirated stops at the beginning of a word realized as such? Do they instead become normal aspirated stops? Do they combine with the previous word?

Is the distinction between /ɪ:/ and /i:/ phonemic? Do they ever merge? Is there an /e:/ as well? When is contrastive length phonemic? Does VʰC contrast with V:C?

Does the language now allow voiced finals, or are they realized as voiceless?

How does the loss of final stops (in some contexts) affect verb conjugation? Are final stop clusters simplified and then lost, or are they kept? If they're lost, do those verbs just become irregular, or do all verbs fall under a new paradigm?

[Grunnen]

Hmm, somehow doubled my post...

[Grunnen]

Cool stuff!

Thanks

What's the final phoneme inventory?

consonants

Code: Select all

ʰp p b ʰt t d   c ɟ   ʰk k g
f       s       ʃ     χ
m       n       ɲ     ŋ
ʋ       l       ʎ     ʀ
w               j	

vowels

Code: Select all

       front fr.round back
high	i ɪ   y ʏ      u ʊ
mid    e ɛ   ø        o ɔ         ə
low    a              ɑ

all vowels except /ə/ have both a long and a short variant, which contrast phonemically (due to loss of final voiceless consonants and compensatory lengthening). There is also an extensive list of diphthongs, that are listed in my first post on this. These diphthongs are very recent and there will probably be some mergers in the future, as some are very alike.

Are preaspirated stops at the beginning of a word realized as such? Do they instead become normal aspirated stops? Do they combine with the previous word?

I haven't decided on that yet, but I think they will be realised as such. Perhaps only if the word is said emphatically. The preaspiration will be applied following a word ending in a vowel. I don't think they will be realised as normal aspirated stops.

Is the distinction between /ɪ:/ and /i:/ phonemic? Do they ever merge? Is there an /e:/ as well? When is contrastive length phonemic? Does VʰC contrast with V:C?

This distinction is phonemic, as are the other length distinctions. In the vowel changes I made the assumption that the Dutch "long" vowels aren't actually "long" but tense, so long /ɪ:/ doesn't (necessarily) merge with /i:/. In certain unstressed syllables they merge to (and thus with) /i/. The other high vowels show a parallel pattern.
There is indeed an /e:/ (as I mentioned implicitly earlier in this post).
Long vowels indeed contrast with short vowels followed by preaspirated stops. Long vowels followed by preaspirated stops are however fairly rare.

Does the language now allow voiced finals, or are they realized as voiceless?

The language does allow for voiced finals. These are the result of an unvoiced stop becoming a glottal stop and merging with a preceding nasal. The old voiced stops that suffered from auslautverhartung have (probably, not sure yet) (all?) become voiceless phonemically.

How does the loss of final stops (in some contexts) affect verb conjugation? Are final stop clusters simplified and then lost, or are they kept? If they're lost, do those verbs just become irregular, or do all verbs fall under a new paradigm?

Yes, that is indeed very interesting. Where final clusters of voiceless consonants occur, the first will become a glottal stop, but the second is prevented from turning into one by being preceded by a glottal stop. This glottal stop is then treated as preglottalisation on the second stop and has turned into preaspiration. If there would be no levelling in the verb conjugations, things get really complicated, with some five or perhaps more paradigms with multiple subparadigms... So I am thinking of ways to simplify matters somewhat.[/quote]

[TaylorS]

Ooh, Future Dutch! Nice!

[Grunnen]

Ooh, Future Dutch! Nice!

Thanks, I hope so indeed

[spats]

vowels

Code: Select all

       front fr.round back
high	i ɪ   y ʏ      u ʊ
mid    e ɛ   ø        o ɔ         ə
low    a              ɑ

all vowels except /ə/ have both a long and a short variant, which contrast phonemically (due to loss of final voiceless consonants and compensatory lengthening). There is also an extensive list of diphthongs, that are listed in my first post on this. These diphthongs are very recent and there will probably be some mergers in the future, as some are very alike.

With the length distinction, this is an enormous vowel system with a lot of very similar vowels. I count 27, not including diphthongs; the largest attested systems tend to be between 17 and 20 - and many (though not all) have other features like nasalization and murmuring to help out. I'm not saying it's impossible, but it seems likely that a number of mergers would happen before you got to this point.

If there would be no levelling in the verb conjugations, things get really complicated, with some five or perhaps more paradigms with multiple subparadigms... So I am thinking of ways to simplify matters somewhat.

I'm going to guess that there would be significant leveling, at least with less-commonly-used verbs. Can you give what the conjugation table would be for a few of the more common Dutch verbs with different endings, assuming no leveling? How might you level them?

[Grunnen]

vowels

Code: Select all

       front fr.round back
high	i ɪ   y ʏ      u ʊ
mid    e ɛ   ø        o ɔ         ə
low    a              ɑ

all vowels except /ə/ have both a long and a short variant, which contrast phonemically (due to loss of final voiceless consonants and compensatory lengthening). There is also an extensive list of diphthongs, that are listed in my first post on this. These diphthongs are very recent and there will probably be some mergers in the future, as some are very alike.

With the length distinction, this is an enormous vowel system with a lot of very similar vowels. I count 27, not including diphthongs; the largest attested systems tend to be between 17 and 20 - and many (though not all) have other features like nasalization and murmuring to help out. I'm not saying it's impossible, but it seems likely that a number of mergers would happen before you got to this point.

It is one of the very last sound changes very suddenly leads to all these different vowels. So there hasn't really been much time to merge the vowels. Before that change, there are only a few instances of long /I:/, /Y:/ and /U:/, no other long vowels and no real diphthongs. Would that make it plausible? I assume the next step would indeed be to merge a fair number of them.

If there would be no levelling in the verb conjugations, things get really complicated, with some five or perhaps more paradigms with multiple subparadigms... So I am thinking of ways to simplify matters somewhat.

I'm going to guess that there would be significant leveling, at least with less-commonly-used verbs. Can you give what the conjugation table would be for a few of the more common Dutch verbs with different endings, assuming no leveling? How might you level them?

You mean the verbs you get after applying the sound changes, but without the levelling?

inf=infinitive
1ev.tt=first person singular present tense
2ev.tt=2s present tense
3ev.tt=3s present tense
mv.tt=1,2,3plural present tense
ev.vt=1,2,3s past tense
mv.vt=1,2,3pl past tense
vd=past participle
td=present participle

The infinitives are given in the current spelling, <oe>=/u/, the diacritics mark tenseness and vowel length: '=long tense, `=lax, ^=long and lax. <u> as the second part of a diphthong is either /u/ or /ʊ/. Sorry for the inconvenient transcription, but I copypasted from an excell sheat I made for myself.
Strong and irregular verbs are marked with *

Code: Select all

Nederlands inf.	1ev.tt	2ev.tt	3ev.tt	mv.tt	ev.vt	mv.vt	vd	td
eten*       é	é	é	é	é	â	á	γeγé	esònt
zitten*	zî	zî	zî	zî	zî	zâ	zá	γezé	zìsònt
zouten	zao	zao	zao	zao	zao	zao	zao	γezao	zafònt
kappen	gào	gào	gàht	gàht	gào	gàht	gàht	γegàht	gàfònt
zoeken*	zoeo	zoeo	zoeht	zoeht	zoeo	zòht	zòht	γezòht	zoegònt
ruiken*	rae	rae	raht	raht	rae	roa	roa	γeroa	rašònt
kijken*	gae	gae	gaht	gaht	gae	gea	gea	γegea	gašònt
lopen*	lou	lou	loht	loht	lou	liu	liu	γelou	lofònt
bieden*	bit	bit	bí	bí	bit	bot	bot	γebot	bidònt
leven	lef	lef	leht	leht	lef	left	left	γeleft	levònt
wagen	wag	wag	waht	waht	wag	wagt	wagt	γewagt	waγònt
kennen	gèn	gèn	gèd	gèd	gèn	gènt	gènt	γegènt	gènònt
komen*	gom	gom	gob	gob	gom	kwàm	kwam	γegom	gomònt
dolen	dóe	dóe	dóe	dóe	dóe	dóet	dóet	γedóet	dôelònt
delen	dîw	dîw	dî	dî	dîw	dîp	dîp	γedîp	dîlònt
boren	bôej	bôej	bôe	bôe	bôej	bôec	bôec	γebôec	bôerònt
deren	dí	dí	dí	dí	dí	dít	dít	γedít	dîrònt
porren	fòj	fòj	fòi	fòi	fòj	fòc	fòc	γefòc	fòrònt
mollen	mòw	mòw	mòu	mòu	mòw	mòp	mòp	γemòp	mòlònt
zwaaien	swaj	swaj	swae	swae	swaj	swac	swac	γeswac	swajònt
geeuwen	γîw	γîw	γî	γî	γîw	γîp	γîp	γeγîp	γîwònt
									
raspen	ràhp	ràhp	ràhp	ràhp	ràhp	ràhp	ràhp	γeràhp	ràhpònt
klotsen	klôs	klôs	klôht	klôht	klôs	klôht	klôht	γeklôht	klôzònt
hupsen	hùus	hùus	hùuht	hùuht	hùus	hùuht	hùuht	γehùuht	hùuzònt
pompen	fòb	fòb	fòmht	fòmht	fòb	fòmht	fòmht	γefòmht	fòmpònt
ronken	ròg	ròg	rònght	rònght	ròg	rònght	rònght	γerònght	rònkònt
planten	plàd	plàd	plàd	plàd	plàd	plàd	plàd	γeplàd	plàntònt
binden*	bìnt	bìnt	bìd	bìd	bìnt	bònt	bònt	γebònt	bìnnònt
zorgen	zòš	zòš	zòht	zòht	zòš	zòšt	zòšt	γezòšt	zòžònt
melden	mèp	mèp	mèu	mèu	mèp	mèp	mèp	γemèp	mèbònt
worden*	wòc	wòc	wòi	wòi	wòc	wòc	wòc	γewòc	wòďònt
werpen*	wê	wê	wê	wê	wê	wí	wí	γewô	wèsònt
									
bezemen	bezoem	bezoem	bezoeb	bezoeb	bezoem	bezoemt	bezoemt	γebezoemt	bezoemònt
ritselen	rîzòw	rîzòw	rîzòu	rîzòu	rîzòw	rîzòp	rîzòp	γerîzòp	rîzòlònt
wateren	wasí	wasí	wasí	wasí	wasí	wasít	wasít	γewasít	wasarònt
									
hebben*	hèp	hèp	hèp	heht	hèp	hàt	hàt	γehàt	hèbònt
zijn*	zaň	bèn	bèd	ìs	zaň	wàs	waj	γeweht	zaňt

And I just realised that I left out the preaspirated consonants in the final consonant inventory.
Yes, I start to feel like working out a good deal of vowel mergers, and then I will start levelling the inflectional paradigms.
I was thinking of getting rid of the present participle. At present it's only really used as adjective, so I could probably work out another way of turning verbs into adjectives, and leave the pp behind.
Really appreciate your feedback spats!

[finlay]

Are those abbreviations dutch?

Just glancing through, I notice that PL.PRES and 1S.PRES and INF are always the same, as are 2S.PRES and 3S.PRES, although sometimes they're all the same. I think you can certainly get away with a major exception for 'be', but I'm not sure about 'have', since it just seems a bit odd that the 2S.PRES merges with the 1S instead of 3S. I'd expect it to level out.

As for the past tense and participles, I do notice that you have marked some strong vowels, but I don't see a way of predicting the "weak" vowels – it seems that you'd have to learn a new past tense for every single verb. So maybe decide on a regularised way of adding a suffix to the infinitive or something – this is what people mean when they talk about levelling and analogy. Like maybe -t.

Then you've got the participles, of which one is unpredictable becaues you can't tell which stem to add ɣe- to (sometimes it's the past tense itself, sometimes it's from a different root, although you're more consistent in using the past tense here), and the other is just plain unpredictable. It all kinda depends how "crazy" you want your language to be.

[Grunnen]

Are those abbreviations dutch?

Yes.

Just glancing through, I notice that PL.PRES and 1S.PRES and INF are always the same, as are 2S.PRES and 3S.PRES, although sometimes they're all the same.

Indeed

I think you can certainly get away with a major exception for 'be', but I'm not sure about 'have', since it just seems a bit odd that the 2S.PRES merges with the 1S instead of 3S. I'd expect it to level out.

Okay, I see the pattern, seems like something that would happen.

As for the past tense and participles, I do notice that you have marked some strong vowels, but I don't see a way of predicting the "weak" vowels – it seems that you'd have to learn a new past tense for every single verb. So maybe decide on a regularised way of adding a suffix to the infinitive or something – this is what people mean when they talk about levelling and analogy. Like maybe -t.

No, the weak vowels get a bit messed up, though it's not the case there's absolutely no predictive value in the stem vowel. I thought I would at least add -ht to past tense forms ending in a vowel, so perhaps I will take -ht as the past tense ending and apply it to the infinitive (or 1ps, which is the same). In regular verbs, -ht is very often the ending the past tense forms get. For stems not ending in a vowel I think I will use -aht instead of -ht. /a/ is probably the vowel most often preceding -ht in verbs descending from regular Dutch verbs.

Then you've got the participles, of which one is unpredictable becaues you can't tell which stem to add ɣe- to (sometimes it's the past tense itself, sometimes it's from a different root, although you're more consistent in using the past tense here), and the other is just plain unpredictable. It all kinda depends how "crazy" you want your language to be.

So the past participle could be formed by ɣe- + stem + -ht. Incidentally, that looks a lot like what we got now: ge- + stem + t/d.

Thanks a lot.

[spats]

The infinitives are given in the current spelling, <oe>=/u/, the diacritics mark tenseness and vowel length: '=long tense, `=lax, ^=long and lax. <u> as the second part of a diphthong is either /u/ or /ʊ/. Sorry for the inconvenient transcription, but I copypasted from an excell sheat I made for myself.
Strong and irregular verbs are marked with *

Code: Select all

Nederlands inf.	1ev.tt	2ev.tt	3ev.tt	mv.tt	ev.vt	mv.vt	vd	td
eten*       é	é	é	é	é	â	á	γeγé	esònt
zitten*	zî	zî	zî	zî	zî	zâ	zá	γezé	zìsònt
zouten	zao	zao	zao	zao	zao	zao	zao	γezao	zafònt
kappen	gào	gào	gàht	gàht	gào	gàht	gàht	γegàht	gàfònt
zoeken*	zoeo	zoeo	zoeht	zoeht	zoeo	zòht	zòht	γezòht	zoegònt
ruiken*	rae	rae	raht	raht	rae	roa	roa	γeroa	rašònt
kijken*	gae	gae	gaht	gaht	gae	gea	gea	γegea	gašònt
lopen*	lou	lou	loht	loht	lou	liu	liu	γelou	lofònt
bieden*	bit	bit	bí	bí	bit	bot	bot	γebot	bidònt
leven	lef	lef	leht	leht	lef	left	left	γeleft	levònt
wagen	wag	wag	waht	waht	wag	wagt	wagt	γewagt	waγònt
kennen	gèn	gèn	gèd	gèd	gèn	gènt	gènt	γegènt	gènònt
komen*	gom	gom	gob	gob	gom	kwàm	kwam	γegom	gomònt
dolen	dóe	dóe	dóe	dóe	dóe	dóet	dóet	γedóet	dôelònt
delen	dîw	dîw	dî	dî	dîw	dîp	dîp	γedîp	dîlònt
boren	bôej	bôej	bôe	bôe	bôej	bôec	bôec	γebôec	bôerònt
deren	dí	dí	dí	dí	dí	dít	dít	γedít	dîrònt
porren	fòj	fòj	fòi	fòi	fòj	fòc	fòc	γefòc	fòrònt
mollen	mòw	mòw	mòu	mòu	mòw	mòp	mòp	γemòp	mòlònt
zwaaien	swaj	swaj	swae	swae	swaj	swac	swac	γeswac	swajònt
geeuwen	γîw	γîw	γî	γî	γîw	γîp	γîp	γeγîp	γîwònt
									
raspen	ràhp	ràhp	ràhp	ràhp	ràhp	ràhp	ràhp	γeràhp	ràhpònt
klotsen	klôs	klôs	klôht	klôht	klôs	klôht	klôht	γeklôht	klôzònt
hupsen	hùus	hùus	hùuht	hùuht	hùus	hùuht	hùuht	γehùuht	hùuzònt
pompen	fòb	fòb	fòmht	fòmht	fòb	fòmht	fòmht	γefòmht	fòmpònt
ronken	ròg	ròg	rònght	rònght	ròg	rònght	rònght	γerònght	rònkònt
planten	plàd	plàd	plàd	plàd	plàd	plàd	plàd	γeplàd	plàntònt
binden*	bìnt	bìnt	bìd	bìd	bìnt	bònt	bònt	γebònt	bìnnònt
zorgen	zòš	zòš	zòht	zòht	zòš	zòšt	zòšt	γezòšt	zòžònt
melden	mèp	mèp	mèu	mèu	mèp	mèp	mèp	γemèp	mèbònt
worden*	wòc	wòc	wòi	wòi	wòc	wòc	wòc	γewòc	wòďònt
werpen*	wê	wê	wê	wê	wê	wí	wí	γewô	wèsònt
									
bezemen	bezoem	bezoem	bezoeb	bezoeb	bezoem	bezoemt	bezoemt	γebezoemt	bezoemònt
ritselen	rîzòw	rîzòw	rîzòu	rîzòu	rîzòw	rîzòp	rîzòp	γerîzòp	rîzòlònt
wateren	wasí	wasí	wasí	wasí	wasí	wasít	wasít	γewasít	wasarònt
									
hebben*	hèp	hèp	hèp	heht	hèp	hàt	hàt	γehàt	hèbònt
zijn*	zaň	bèn	bèd	ìs	zaň	wàs	waj	γeweht	zaňt

Okay - for most verbs, it looks like the present 2ps, 3ps, and past all or mostly fall together (especially with vowel mergers). This suggests that this language will probably find some other way to express past tense, perhaps with a periphrastic construction. Since French and German both do it using to be/to have + plus a participle, and Dutch already has such a construction, it makes sense it would probably do the same thing.

You could even keep a simple past form for "to be" and "to have", just to be tricky.

Regarding pronunciation, there are some odd things - if you have <-mht>, how is that realized? <-nght>?

Most verbs have a 1ps/plural vs. 2ps/3ps distinction; that might persist as an artifact, or it might not. It might be that all fall in with one or the other. Or some distinction gets generalized to most verbs. I could see -t (which mutates a final nasal) being the standard ending for 2ps and 3ps, maybe even spelled -t in all cases (<gomt> instead of <gob>, for example). A few common verbs will remain irregular, along with groups of verbs that share a paradigm (a common final alternation or vowel alternation between forms).

I actually expect that it won't lose the gerund and participle forms, since they're so distinct, though there may be some simplification of inflectional paradigms. Of course, it's completely up to you.

Finally, I want to say how interesting it is to see a system that in many ways is becoming like English - mostly isolating but with a lot of really organic irregularity and some interesting inflectional leftovers.

[Grunnen]

With the length distinction, this is an enormous vowel system with a lot of very similar vowels. I count 27, not including diphthongs; the largest attested systems tend to be between 17 and 20 - and many (though not all) have other features like nasalization and murmuring to help out. I'm not saying it's impossible, but it seems likely that a number of mergers would happen before you got to this point.

So, I worked out a fair number of mergers:

Code: Select all

ɪ/e/_, ɪ:/i:/_ 
a/ɑ/_ ɑ:/a/_ (long to short)
ɛ/a/_ ɛ:/e:/_
ʏ/ø/_ ʏ:/y:/_
œ/ɑ/_
ʊ/o/_ ʊ:/u:/_
ɔ/ɑ/_ ɔ:/o:/_

iu/y:/_ yu/u:/_ ui/y:/_
iɛ/e:/_ yœ/ø:/_ uɔ/o:/_
ɪʊ/iw/_ ʏʊ/yw/_ ʊɪ/uj/_
ɪɛ/je/_ ʏœ/je/_ ʊɔ/wo/_
eʊ/ø:/_ øʊ/o:/_ oɪ/ø:/_
eɐ/e:/_ øɐ/ø:/_ oɐ/o:/_
ɛʊ/au/_         ɔʊ/au/_
ɛɪ/ai/_         ɔɪ/ai/_
aɔ/au/_         ɑo/au/_
ae/ai/_         ɑɛ/ai/_

So this leaves me with 15 vowels (excluding the two diphthongs):

Code: Select all

i y u
e ø o
which can be both long and short, and

a ɑ ə
which can only be short

And two diphthongs:
au ai

Are these changes OK and the resulting vowel system plausible?

[Timmytiptoe]

Will anything interesting happen to the "zijn aan het"-construction? In my speech, it tends to get absorbed into the verb, and I'd expect that to continue.

For example:
Ik ben aan het eten
[ɪk bɛnan tetə]
Jij bent aan het schieten
[jɛi bɛnan tsχitə]
Hij is aan het lopen
[hɛisan dlopə]
Wij zijn aan het douchen
[ʋɛizənan duʃə]

[Grunnen]

Will anything interesting happen to the "zijn aan het"-construction? In my speech, it tends to get absorbed into the verb, and I'd expect that to continue.

For example:
Ik ben aan het eten
[ɪk bɛnan tetə]
Jij bent aan het schieten
[jɛi bɛnan tsχitə]
Hij is aan het lopen
[hɛisan dlopə]
Wij zijn aan het douchen
[ʋɛizənan duʃə]

I hadn't thought of that one yet, but it's certainly an interesting idea. I very much like the way progressives can be formed in Dutch, for example the zit te + inf., lig te + inf. etc., so funny things will probably happen there. I'll keep this in mind!

[Grunnen]

Okay - for most verbs, it looks like the present 2ps, 3ps, and past all or mostly fall together (especially with vowel mergers). This suggests that this language will probably find some other way to express past tense, perhaps with a periphrastic construction. Since French and German both do it using to be/to have + plus a participle, and Dutch already has such a construction, it makes sense it would probably do the same thing.

Indeed. I thought I could use the inf/1ps form for the 2&3ps as well, leaving the past tense form distinct. I was planning on using "to be/to have + plus a participle" for past tense formation, but wanted to keep the current past tense forms for something else. Currently, people often use them to ask for confirmation on something they agreed upon with others: We gingen toch om zeven uur eten? "We were going to eat at seven, right?". I thought that with a few steps expanding its use, I could deploy it as a form that indicates that the information provided in the sentence is supposed to be common knowledge, or has already been mentioned in the conversation. But I don't know if such a system would be plausible...

You could even keep a simple past form for "to be" and "to have", just to be tricky.

Such things are really fun.

Regarding pronunciation, there are some odd things - if you have <-mht>, how is that realized? <-nght>?

How would that become <-nght>? I'm not sure, myself, I think pronouncing a nasal followed by a preaspirated stop isn't difficult... So I thought -mht would simply be pronounced as a m followed by a preaspirated t. Perhaps devoicing the last bit of the nasal. That would be something that would follow rather naturally I think.

Most verbs have a 1ps/plural vs. 2ps/3ps distinction; that might persist as an artifact, or it might not. It might be that all fall in with one or the other. Or some distinction gets generalized to most verbs. I could see -t (which mutates a final nasal) being the standard ending for 2ps and 3ps, maybe even spelled -t in all cases (<gomt> instead of <gob>, for example). A few common verbs will remain irregular, along with groups of verbs that share a paradigm (a common final alternation or vowel alternation between forms).

Yes, that's what I was thinking.
What I had worked out is the following:
All singular present tense forms become the same.

Code: Select all

1ps 2&3ps result
VV VV VV
VV Vht VV
VS Vht VS
VN VD VD
VJ V(V) V(V)
VVS VVht VVS
VD VNht VD
VNT VD VNT
VT V(V) V(V)

V=vowel
S=fricative
N=nasal
D=voiced plosive
J=glide

I actually expect that it won't lose the gerund and participle forms, since they're so distinct, though there may be some simplification of inflectional paradigms. Of course, it's completely up to you.

I will certainly keep the past participle, but the present participle isn't used that much in Dutch. Not in forming verb tenses (well, except an archaic continuous that worked pretty much identical to the English one, my grandmother sometimes still uses it... but she's 94 years of age).
It is still used to turn verbs into adjectives though.

Finally, I want to say how interesting it is to see a system that in many ways is becoming like English - mostly isolating but with a lot of really organic irregularity and some interesting inflectional leftovers.

Yes, that's indeed what seems to be happening. But that's the way we're heading anyway. Slightly quicker than our neighbours to the east, but slower than our neigbours to the west

[spats]

I thought I could use the inf/1ps form for the 2&3ps as well, leaving the past tense form distinct. I was planning on using "to be/to have + plus a participle" for past tense formation, but wanted to keep the current past tense forms for something else. Currently, people often use them to ask for confirmation on something they agreed upon with others: We gingen toch om zeven uur eten? "We were going to eat at seven, right?". I thought that with a few steps expanding its use, I could deploy it as a form that indicates that the information provided in the sentence is supposed to be common knowledge, or has already been mentioned in the conversation. But I don't know if such a system would be plausible...

...

All singular present tense forms become the same.

...

That sounds fine. I'm not convinced the distinction between the 1ps and 2ps/3ps forms would go away, though. English doesn't need it's -s in the 3ps, but it keeps it around out of pure nostalgia.

That doesn't mean you can't still keep the old simple past <-t> forms around to use for mood-marking too. In your example, the question is asked in the 1pp, which wouldn't normally have the -t form; having the -t form would mark it as requiring confirmation. And of course, a few verbs which are irregular now might still retain an irregular form (though you'd expect them to wear away/be leveled by analogy like the English subjunctive over time).

What do the pronoun series look like?

[Grunnen]

I thought I could use the inf/1ps form for the 2&3ps as well, leaving the past tense form distinct. I was planning on using "to be/to have + plus a participle" for past tense formation, but wanted to keep the current past tense forms for something else. Currently, people often use them to ask for confirmation on something they agreed upon with others: We gingen toch om zeven uur eten? "We were going to eat at seven, right?". I thought that with a few steps expanding its use, I could deploy it as a form that indicates that the information provided in the sentence is supposed to be common knowledge, or has already been mentioned in the conversation. But I don't know if such a system would be plausible...

...

All singular present tense forms become the same.

...

That sounds fine. I'm not convinced the distinction between the 1ps and 2ps/3ps forms would go away, though. English doesn't need it's -s in the 3ps, but it keeps it around out of pure nostalgia.

In Dutch at the moment there is a tendency for auxiliary verbs to develop a uniform singular form in the present tense, which isn't necessarily the same as the root of the verb. So if I were to merge the 1ps forms, it would be just the extension of this pattern. ex:
"kunnen" -> "kan"
"willen" -> "wil"
I hope I can find some reason to make this collapse plausible. Would you know one?

That doesn't mean you can't still keep the old simple past <-t> forms around to use for mood-marking too. In your example, the question is asked in the 1pp, which wouldn't normally have the -t form; having the -t form would mark it as requiring confirmation. And of course, a few verbs which are irregular now might still retain an irregular form (though you'd expect them to wear away/be leveled by analogy like the English subjunctive over time).

That sounds interesting!

What do the pronoun series look like?

After simply applying the sound changes, this is what we'll get:

subject pronouns

Code: Select all

ɪk "I" becomes je
jæi "you" jaj
jə "you" jə
y "you" y
hæi "he" haj
i "he" i
zæi "she" zaj
zə "she" zə
hɛt "it" he:
ət "it" i:
ʋæi "we" ʋaj
ʋə "we" ʋə
jʏli "you" jøli
zæi "they" zaj
zə "they" zə
 

object pronouns

Code: Select all

mæi "me" maj
mə "me" mə
jʌu "you" jaw
jə "you" jə
y "you" y
hɛm "him" hem
əm "him" um
har "her" haj
dər "her" di:
hɛt "it" he:
ət "it" i:
ɔns "us" ɑns
jʏli "you" jøli
hʏn "them" høn
zə "them" zə

possessive pronouns

Code: Select all

mæin "my" maɲ
mən "my" min
jʌu "your" jaw
jə "your" jə
zæin "his" zaɲ
zən "his" zin
har "her" haj
dər "her" di:
no word for "its" in Dutch
ɔns "our" ɑns
jʏli "your" jøli
jə "your" jə
hʏn "their" høn

So now there's a pronoun that can mean "he" or "her": haj. That's a bit tricky, but it shouldn't lead to ambiguity. There is currently a development in Dutch replacing zij/ze (3pp, subj) with hun (3pp, obj), which is still very much stigmatised. But it's also spreading. So I think I'll include that.