Ni'óno3íínow - Esseintially just Latin really

[DesEsseintes]

Ni'ónotíín

As some of you may know, I am a great admirer of the Algonquian languages of North America, and in particular of the Plains Algonquian languages - Arapaho, Blackfoot and Cheyenne. The diachronic histories of these languages contain some of the wildest sound changes I've ever come across, and they are a major inspiration for my Híí language family.

As I was working on the changes from Proto-Híí to modern Hííenununóóoþa using Arapaho as a guideline, I had a fun idea: What if I tweaked the sound changes from Proto-Algonquian to Arapaho so that they could be applied to Latin? And so Ni'ónotíín was born.

What is this all about, Des?

It's simple really. I encoded the sound changes as laid out in the Index Diachronica, and then used Mark Rosenfelder's Sound Change Applier (SCA2) to apply them.

The Phonologies

The inventories of Latin and Proto-Algonquian are surprisingly similar.

Latin

/m n/
/p t k kʷ/
/b d g/
/f s h/
/l j w/
/r/

Proto-Algonquian

/m n/
/p t t͡ʃ k/
/θ* s ʃ/
/j w/
/r*/

*some analyses give /ɬ l/ for /θ r/

The main differences are:
- Latin has /f/ while PA has /θ/
- Latin has a voicing distinction
- Latin contrasts /r l/, whereas PA only had a single liquid
- PA has phonemic postalveolars

I decided to:
- treat /f/ as if it were /θ/
- retain the voiced stops and have them undergo changes parallel to those of the unvoiced stops
- merge /r l/
- ignore the postalveolars for now (more on that later)

The Code

I will spoiler this, but here are the rules. I copied them almost directly from the Index Diachronica, but made minor adjustments. The last few rules are basically rewrite rules.

Spoiler:

c/k/_
f/3/_
l/r/_
v/w/_
u/w/C_V
i/j/C_V
V//_#
w//_#
we/o/_
ō/ī/_
o/i/_
ū/ī/_
u/i/_
w//C_[iī]
e/i/#_
3/š/C_
3/š/_C
h//_C
[smnr]/'/_C
w/j/C_
G/n/#_
k//_
g//_
p/k/_
b/g/_
s/n/#_
s/h/_
Kr/3/_
r/n/[#V]_
r/h/C_
L/S/_CC
ā/ō/_
a/o/_
/'/CS_#
ī/ū/B(X)(X)_
i/u/B(X)(X)_
m/w/_B
m/w/B_#
m/b/_F
m/b/F_#
k/č/_F
k/č/F_#
Sn//_#
/h/#_V
ē/ei/j_
n//_j
j//C_
h//_#
ū/uu/_
ē/ee/_
ī/ii/_
ō/oo/_
j/y/_
č/c/_

Sample Results

The result is a language that looks and feels surprisingly like Arapaho. The phoneme inventory is somewhat larger than Arapaho's though, because of the voiced stops:

/n/ n
/t t͡ʃ k ʔ/ t c k '
/b d g/ b d g
/θ s h/ 3 s h
/j w/ y w

/e i o u/ e i o u
+ length (long vowels written double)

lector → net
doctor → dit
focus → 3ii
jocus → nii
districtus → di'3iti
cīvitatem → hiiwitoteb
adoptare → hodikt
opera → hic
secundus → nei'di
fīlium → 3iiyib
fīliam → 3iiyow
pōntem → ci'teb
diēs → dei
vacuum → nouw
mātrem → woo3eb
patrem → ko3eb
solus → nini
amīcum → howuuuw
lingualatīna → ni'onotiin
linguaromāna → ni'oniwoon

What's Next?

There are still several issues I haven't tackled:
- Arapaho has /ʃ/ that developed from PA /θ/ in consonant clusters and PA /t͡ʃ/. The latter doesn't occur in Latin, and /f/ doesn't really cluster in Latin either, so I have to think of a way to derive it
- Latin has clusters that PA simply didn't have, such as pt ct ps, etc., and I haven't really addressed this issue
- I want to derive an Arapaho-style pitch accent from the Latin accent; so far, I've just done this by hand, as in lingualatīna → ni'onotiin → ni'ónotíín, but it should be possible to encode this

---

This is obviously just for fun, and it really helped me gain a better understanding of exactly what happened between Proto-Algonquian and modern Arapaho. Hopefully, I'll be able to devise equally interesting stuff for Hííenununóóoþa.

Hope that somebody likes this. Feel free to ask for clarifications, make corrections, suggestions, etc.

[DesEsseintes]

First Refinement

I ran the numbers through the changes, and here are my initial results, with added accent marks:

1 ūnum → híínib
2 duo → d
3 trēs → 3éé
4 quattuor → yótt
5 quīnque → híí'
6 sex → né
7 septem → nékteb
8 octō → hít
9 novem → níib
10 decem → déeb

As you can well imagine, I wasn't too happy about how the number two turned out.

There is a rule V//_# that word-final vowels drop regardless of length or quality, but that's obviously not going to work for monosyllables. So I've refined it to this:

V//V(C)(C)C_#

That gives me much more pleasing results for monosyllables:

duo → di'
sī → nii
me → be'
ne → ne'
quī → hii
quā → yoo

Going back to the numbers, I'm not happy with 4 yótt either, but I'm going to have to decide what to do with stop clusters in general before deciding on that one.

Another minor note, I added a line at the beginning of the code:
x/ks/_

[Creyeditor]

Hey, I reall like this idea
What about treating Latin /k/ as PA /tS/ and Latin /kʷ/ as PA /k/?

[DesEsseintes]

Hey, I reall like this idea
What about treating Latin /k/ as PA /tS/ and Latin /kʷ/ as PA /k/?

Glad you like it.

I was actually thinking the opposite, having /kʷ/ → /tS/...

[zyma]

amīcum → howuuuw

Overlong vowels are allowed?

- Arapaho has /ʃ/ that developed from PA /θ/ in consonant clusters and PA /t͡ʃ/. The latter doesn't occur in Latin, and /f/ doesn't really cluster in Latin either, so I have to think of a way to derive it
- Latin has clusters that PA simply didn't have, such as pt ct ps, etc., and I haven't really addressed this issue

Just an idea that might not actually work at all since I don't know how strictly you want to stick to using only documented/reconstructed/etc. Arapaho sound changes:

You could have /pt kt ps/ > /ft kθ fs/ > /θt kθ θs~θt/, etc. or alternatively /pt kt ps/ > /pθ kθ pθ/, etc. or some combination of those.

- I want to derive an Arapaho-style pitch accent from the Latin accent; so far, I've just done this by hand, as in lingualatīna → ni'onotiin → ni'ónotíín, but it should be possible to encode this

If you don't mind my asking, what do you mean by this? Do you mean you have a way you want to do it, but you're still looking for a way to add that to the list of sound changes that can be applied by the website?

(Also, I would have guessed "Ni'ónotíín" was a corruption/descendant of a term like "Neolatin" at first, but I like this etymology better.)

This is obviously just for fun, and it really helped me gain a better understanding of exactly what happened between Proto-Algonquian and modern Arapaho. Hopefully, I'll be able to devise equally interesting stuff for Hííenununóóoþa.

Hope that somebody likes this. Feel free to ask for clarifications, make corrections, suggestions, etc.

I like it. It does seem like a very fun project, and I'm glad to hear it's also helping you, so to speak, with Hííenununóóoþa. I look forward to seeing what you come up with in the future!

Going back to the numbers, I'm not happy with 4 yótt either, but I'm going to have to decide what to do with stop clusters in general before deciding on that one.

Maybe this could be one of the situations that could help resolve the "dilemmas" involving post-alveolar sibilants and dental non-sibilants?

Also, I very much like how the numbers turned out, particularly nékteb.

Hey, I reall like this idea
What about treating Latin /k/ as PA /tS/ and Latin /kʷ/ as PA /k/?

Glad you like it.

I was actually thinking the opposite, having /kʷ/ → /tS/...

I think either way could work, but don't quote me on that.

[thetha]

Arapaho doesn't have /ʃ/, it has /s/ which derived from previous *ʃ

[DesEsseintes]

Arapaho doesn't have /ʃ/, it has /s/ which derived from previous *ʃ

Oops, my bad. I was referring to the /ʃ/ at the intermediate stage which derived from PA /t͡ʃ/ and /θ/, and yields /s x/ in modern Arapaho.

[Lambuzhao]

Just an idea that might not actually work at all since I don't know how strictly you want to stick to using only documented/reconstructed/etc. Arapaho sound changes:

You could have /pt kt ps/ > /ft kθ fs/ > /θt kθ θs~θt/, etc. or alternatively /pt kt ps/ > /pθ kθ pθ/, etc. or some combination of those.

I thought along those lines, shifting to an equivalent fricative, if available. But, like shimo, I don't know a lot about PA sound changes.

Does another related Algonquian dialect have latin-like consonant clusters?
If so, what environments in PA did they come from?
Maybe you could reverse-engineer the Latin consonant clusters that way, if the results don't seem too wonky.

On another note,

You could level /pt kt/ > [tt] basically what you did with numbers 4, 7, 8.

---

/ks ps/ > ??

I'd stray from using NOM.SG as a base form, to weed out final /x ps/. Mebbe use ACC/ABL:
cantatrix -> cantatricem
princeps -> principem

What's left? Some S-perfects & attendant PST.PTCPs (lexi, lapsum, etc)?
This list has most of the 'usual suspects' (scroll almost halfway down)
http://www.quia.com/jg/639135list.html



Capsa? There's a nasalized alt.form campsa. Mebbe capsa -> campsa -> *camsa -> ??

In fact, mebbe make

/ks ps / -> /ns ms/ ???

And, for that matter,

/kt pt/ -> /nt mt/ (though you already seem to have come up with a viable solution)


Ex? You could always use the alt.form 'e'.

[Lambuzhao]

BTW, great, great idea IMHO.

I only wish I could be more help, but I only know a thimbleful of Lenni-Lenape, and I did this comparison on 'turtle' words/isoglosses in the Algonquian languages.
http://www.cbbforum.com/viewtopic.ph ... in#p189036

But I say that this is the basis for a most interesting thought experiment/

[thetha]

Arapaho doesn't have /ʃ/, it has /s/ which derived from previous *ʃ

Oops, my bad. I was referring to the /ʃ/ at the intermediate stage which derived from PA /t͡ʃ/ and /θ/, and yields /s x/ in modern Arapaho.

oh. Well here's an idea: in Proto-Algonquian the sequences *(ti) *(θi) always changed to *ci *ši. Since you're using f as a replacement for θ maybe you could have f > š before front vowels and get it from that.

As an example, you could have *filius > šiliih > sinii.

[kanejam]

this
is
amazing
holy
crap
...
seriously
just
wow

Hey, I reall like this idea
What about treating Latin /k/ as PA /tS/ and Latin /kʷ/ as PA /k/?

Glad you like it.

I was actually thinking the opposite, having /kʷ/ → /tS/...

This is such an amazing project, and /kʷ/ → /tS/ seems like such a bogolang sort of change just to line up the phonologies. If anything I would go for the Romance route, k/tS/_[jie], kʷ/k/_/a. Of course you don't need to do that either, some branches didn't palatalise at all. I can't wait to see how you Arapahoise the grammar

jocus → nii
opera → hic
solus → nini

Proof that Arapaho can't possibly be a natlang, only a conlanged could have come up with changes like these!

[qwed117]

this
is
amazing
holy
crap
...
seriously
just
wow

Hey, I reall like this idea
What about treating Latin /k/ as PA /tS/ and Latin /kʷ/ as PA /k/?

Glad you like it.

I was actually thinking the opposite, having /kʷ/ → /tS/...

This is such an amazing project, and /kʷ/ → /tS/ seems like such a bogolang sort of change just to line up the phonologies. If anything I would go for the Romance route, k/tS/_[jie], kʷ/k/_/a. Of course you don't need to do that either, some branches didn't palatalise at all. I can't wait to see how you Arapahoise the grammar

jocus → nii
opera → hic
solus → nini

Proof that Arapaho can't possibly be a natlang, only a conlanged could have come up with changes like these!

If anybody saw this for PIE, the definite answer would be "That never happened"

[Bristel]

Rock on! I also thought the name was derived from neolatin, but lingualatīna makes a lot of sense too.

[Ahzoh]

jocus → nii
opera → hic
solus → nini

Wha? What? How?

[DesEsseintes]

Thanks for all the feedback, people! I'm glad to see that others are enjoying this.

Overlong vowels are allowed?

Overlong vowels do occur, yes, mostly as a result of elided /k g/. They are however treated as belonging to different syllables, and adding accent marks makes this clearer:

amīcum → howúúuw

Speaking if accents, I know exactly how I want to apply the Arapaho pitch accent, and hopefully I can explain it soon. I just meant that I hadn't thought of how to encode it.

much useful stuff...

Thank you, Lambuzhao! Dealing with clusters will be my next undertaking, and you've made it clearer what I'm up against.

oh. Well here's an idea: in Proto-Algonquian the sequences *(ti) *(θi) always changed to *ci *ši. Since you're using f as a replacement for θ maybe you could have f > š before front vowels and get it from that.

As an example, you could have *filius > šiliih > sinii.

Thanks for that, Teddy. This was so staring me in the face and yet I just didn't see it. I've incorporated this already, and the set of rules dealing with it is:

tj/č/_
fj/š/_
t/č/_[iī]
f/š/_[iī]
f/3/_

Results are as follows:

nātiōnem → noo3ííneb
fīlius → sííyi
fīlium → sííyib

I have a dilemma here, though: What to do with /d/ before /i j/? I don't think I want a voicing contrast in the fricatives...

This also means that I don't need to have /kʷ/ → /t͡ʃ/. /kʷ/ simply becomes /w/ when k disappears, and soon after goes the way of all glides to /j/.

---

Several posters have expressed surprise at the somewhat extreme changes that occur. I want to reiterate that I haven't exaggerated anything; this is actually pretty much what happened between Proto-Algonquian and Arapaho.

[gestaltist]

Very entertaining. Subscribing.

[kanejam]

I have a dilemma here, though: What to do with /d/ before /i j/? I don't think I want a voicing contrast in the fricatives...

You could go the Spanish route and have it become /x/. According to the Wikipedia page, Arapaho contrasts /x h/ which you don't seem to have at the moment. It might also give you something interesting to do with /l/ (ie /lj/ -> /x/) rather than just merge it unconditionally with /r/.

Several posters have expressed surprise at the somewhat extreme changes that occur. I want to reiterate that I haven't exaggerated anything; this is actually pretty much what happened between Proto-Algonquian and Arapaho.

I know it's what happened but it doesn't make it any less wacky; I remember staring at the cognate set on the Wikipedia page several months back and thinking 'how?'. I've done the same with Albanian and several Austronesian languages.

[zyma]

Overlong vowels do occur, yes, mostly as a result of elided /k g/. They are however treated as belonging to different syllables, and adding accent marks makes this clearer:

amīcum → howúúuw

Ah, thanks for the explanation! Do overlong vowels/sequences of three of the same vowel in a row occur in Arapaho as well?

[DesEsseintes]

Overlong vowels do occur, yes, mostly as a result of elided /k g/. They are however treated as belonging to different syllables, and adding accent marks makes this clearer: amīcum → howúúuw

Ah, thanks for the explanation! Do overlong vowels/sequences of three of the same vowel in a row occur in Arapaho as well?

Yes, they do. As an example, the word for cat is wo'óoó.

[DesEsseintes]

The Two Palatalisations

I've worked out what I'm going to do about the palatalisations. Here I'm going to give a brief account of how they work.

The First Palatalisation

Reconstructions of Proto-Algonquian include a phoneme *č. However, this appears to be almost entirely an allophone of *t before *i. I don't know whether *tj clusters occurred in PA (I suspect they didn't), but I have decided to apply this rule to tj clusters in Latin (I treat /i/ in position C_V as a glide). So we get these rules:

tj → č
t → č / _{i ī}

I have decided to have Latin /d/ turn into the corresponding fricative in the same environments, yielding:

tj dj → č š
t d → č š / _{i ī}

Note that at a later stage č → 3 unconditionally before the second palatalisation kicks in, while š remains intact until later.

Note also that PA *θ underwent palatalisation to š, and I chose to have the same happen to Latin f.

The Second Palatalisation

This occurs after the shift of p (b) → k (g) (PA only had *p of course, the b comes from Latin). It affects the newly formed velar stops in the same environments as the dentals in the First Palatalisation. Again I chose to have the voiced stop become the fricative.

kj gj → č š
k g → č š / _{i ī}

Death of a Shibilant

As we have seen, š developed at different stages for different reasons, from f d and g, but there is no š in modern Arapaho or Ni'ónotíín. That's because s had already vanished at the time of the second palatalisation, and š rushed in to save the day, going to s before front vowels and x before back vowels.

Three Examples

I was going to give several examples, but it's getting late and I want to post this, so here goes:

bōnus → bīnis → gīnis → gīnih → šīnih → sííni
fīliā → šīnjō → šīyō → sííyoo
hōdiē → hōšē → hīšē → híísee