Zeko-Romance development thread

A forum for all topics related to constructed languages
User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Zeko-Romance development thread

Post by Zekoslav » 14 Aug 2019 14:06

1. INTRODUCTION

This isn't, I confess, my first Romance conlang: that honor would go to Occitan, which I accidentally reinvented while trying to invent a Gallo-Romance language somewhat more conservative than French...

Anyway, while reading about Vulgar Latin, I discovered that a 1st declension genitive ending -ÆS was attested in inscriptions but didn't survive anywhere. I thought it would be cool to invent a Romance language which did preserve it, and by necessity it would have to preserve other cases as well. I decided to keep it realistic: Old French and Old Occitan preserve a nominative vs. oblique system, and Romaninan preserves a nominative/accusative vs. genitive/dative system. Combining the two gives us a case system of nominative, accusative and genitive/dative, while introducing the lost 1st declension genitive ending -ÆS gives us the possibility of not merging the genitive and the dative. Since Romance languages in general merge the genitive and the dative, I decided to merge them in the plural while keeping them distinct in the singular.


2. GENERAL FEATURES

A Romance language which preserves cases will by necessity have to be peripheral, so as not to be overrun by the variety of Latin named "Proto-Western-Romance" by Wikipedia and "Late Imperial Lingua Franca" by Salmoneus: we're looking at another Sardinian or Romanian.

Must-haves:

- The variety of Latin which it derives from must preserve final /s/.

- The variety of Latin which it derives from must be isolated early from other varieties of Latin.


Shoulds:

- The language should be in contact with a language which also preserves cases, such as Greek, Primitive Irish or Germanic.


Mustn't-haves:

- Whatever caused Romanian to loose final /s/, this language mustn't have.


3. POSSIBLE LOCATIONS

So far, the Black Sea, Ireland and the Canaries have been suggested.

Canaries: Opportunity of early settlement and isolation, similarity to African Romance which is similar to Sardinian, but not much contact with other languages and not much is know about the indigenous language.

Ireland: Alt-historically, it would be interesting if one of the early medieval Irish kingdoms was Romance-speaking. Linguistically, a proper (not bogolang) Romance language on Irish substrate would be interesting. However, the language probably wouldn't survive up to modern times.

Black Sea: Greek, Sarmatian and Gothic neighbours sound fun. However, I'm not sure what kind of settlement would be needed to romanize the area. When the migrations hit they might flee to the Caucasus or to the Carpathians.

What does everyone think?
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

User avatar
Shemtov
runic
runic
Posts: 3094
Joined: 29 Apr 2013 04:06

Re: Zeko-Romance development thread

Post by Shemtov » 14 Aug 2019 23:50

Could a minority language in Hungary work? Maybe in Pannonia? It could have started as a transition between Dalmatian and Daco-Romance, and then got Hungarian and German influence
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien

User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Re: Zeko-Romance development thread

Post by Zekoslav » 15 Aug 2019 12:10

I think these two loose final /s/, which would make my main idea impossible: if it preserves cases, it would preserve those preserved in Romanian. Anyway, I'll soon post some morphological musings which may make this decision easier.
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

User avatar
qwed117
mongolian
mongolian
Posts: 4455
Joined: 20 Nov 2014 02:27

Re: Zeko-Romance development thread

Post by qwed117 » 15 Aug 2019 18:12

Zekoslav wrote:
14 Aug 2019 14:06
2. GENERAL FEATURES

A Romance language which preserves cases will by necessity have to be peripheral, so as not to be overrun by the variety of Latin named "Proto-Western-Romance" by Wikipedia and "Late Imperial Lingua Franca" by Salmoneus: we're looking at another Sardinian or Romanian.
A lot of Romance dialects retain weird casing systems and are not peripheral. I was reading up on a variety of, I believe Italian, that maintains -ibus endings and uses it as a partitive (for example "ferrus -> *ferribus > ferriu)
Spoiler:
My minicity is Zyphrazia and Novland
What is made of man will crumble away.

User avatar
Shemtov
runic
runic
Posts: 3094
Joined: 29 Apr 2013 04:06

Re: Zeko-Romance development thread

Post by Shemtov » 15 Aug 2019 19:33

Zekoslav wrote:
15 Aug 2019 12:10
I think these two loose final /s/,
I don't think Dalmatian did, so in an early stage, the language as a transition between Dalmatian and Daco-Romance, that just is like Dalmatian in that regard. An earlyish date of isolation is possible, if it became surrounded by Hungarian as soon as the Hungarians settled the area.
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien

User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Re: Zeko-Romance development thread

Post by Zekoslav » 15 Aug 2019 20:46

The only half-decently attested dialect of Dalmatian was a variant spoken in the 19th century, and it seems to have had all the hallmarks of (broadly understood) Eastern Romance: loss of word-final /s/ and plurals derived from the Latin nominative. The numbers in this Swadesh list are particularly indicative of the loss of word-final /s/. This could have happened anytime but since Italian (including Gallo-Italian) and Romanian loose word-final /s/ as well it may as well have been early (this is in fact a problem with a Pontic Romance language: it would have to be settled from an area which preserves word-final /s/).

I don't think Pannonia is a good location, for this reason and others (even if it boils down to me having helped Ælfwine make his Pannonian Romlang and Slavlang and getting bored of Pannonia because of that).

@qwed117: That's interesting (and I didn't know that!). Italian dialects in general do weird things, like swapping feminine singular and plural endings for each other, marking all verb tenses for gender and not just those containing participles, innovating a fourth gender and having a conditional mood formed from both the Standard Italian perfect of "to have" and Western Romance imperfect of "to have" depending on the person. However, at a first glance this seems to me like an isolated case form having taken on a new function, and what I'm proposing is a more-or-less German-like system. You're right that a language doesn't need to be totally isolated to be interesting (Old French was pretty central and still had cases), but I would be good for it not to be positioned at a major crossroads.

I'm really split between the Black Sea (for potential Greek, Gothic and Sarmatian influence) and Ireland (for a potental outlying Western Romance language). Just want to figure out where my morphology would make sense the most.
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

Ælfwine
roman
roman
Posts: 943
Joined: 21 Sep 2015 01:28
Location: New Jersey

Re: Zeko-Romance development thread

Post by Ælfwine » 15 Aug 2019 22:50

(this is in fact a problem with a Pontic Romance language: it would have to be settled from an area which preserves word-final /s/).
The loss of /s/ was an innovation, not an archaism. A significantly conservative dialect could retain /s/, even if it is geographically where one might find plurals with -i instead. Compare those PIE dialects in the far east that are otherwise "centum" as opposed to "satem."

Also, I don't agree that Pannonia *must* not have -s plurals because Dalmatia hadn't. Friulian is east of Venetian, which aligns with southern Italian in its plural formations, while Friulian forms its plurals with -s. (I've also read that the Friulians initially came from Noricum and were pushed south due to the Slavic and German invasions of the area.) Something to me suggests that the division of Romance dialects was more of a northwest/southeast divide than a pure west/east divide, but that's just speculation on my part.

As for the area itself, I want to see a Black Sea romlang, if only because one of my conlangs is in that area.
My Blog
Current Projects:
Crimean Gothic — A Gothic language spoken in Crimea (duh)
Pelsodian — A Romance language spoken around Lake Balaton
Jezik Panoski — A Slavic language spoken in the same area
An unnamed Semitic language spoken in the Caucus.

User avatar
Shemtov
runic
runic
Posts: 3094
Joined: 29 Apr 2013 04:06

Re: Zeko-Romance development thread

Post by Shemtov » 16 Aug 2019 00:05

Could Noricum work, with HG influence, and be a transition between Dalmatian and Rhaeto-Romance? If it was cut off from Rhaetia at an early date, it could be a minority language of Austria.
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien

User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Re: Zeko-Romance development thread

Post by Zekoslav » 16 Aug 2019 12:01

DECLENSIONAL MUSINGS

This is how I imagine Latin declensions would develop. Since location and sound changes are currently in the brainstorming phase, everything is written as if it was Latin (all caps because it's cool and because it's tradition).


1. General developments

The vocative case merges with the nominative, and the ablative case merges with the accusative. The first is already underway in Classical Latin, and the second is a natural consequence of Vulgar Latin sound changes, particularly the loss of word-final /m/.

The genitive and dative cases remain distinct in the singular, but merge in the plural. The genitive/dative plural ending descends from the Latin genitive plural ending. This is a compromise between the general Romance merger of genitive and dative and the unusual conservatism of my language.


2. Pronominal declension

Noun declension, especially the 1st and 2nd declensions, are partially influenced by the pronominal declension so it's useful to take a look at it first.

Code: Select all

 MASCULINE:                        FEMININE:                         NEUTER:
|-----|-----------|-----------|   |-----|-----------|-----------|   |-----|-----------|-----------|
|  -  |    sg.    |    pl.    |   |  -  |    sg.    |    pl.    |   |  -  |    sg.    |    pl.    |
|-----|-----------|-----------|   |-----|-----------|-----------|   |-----|-----------|-----------|
|  N  | ILL-E     | ILL-Ī     |   |  N  | ILL-A     | ILL-Æ     |   |  N  | ILL-UD    | ILL-A     |   
|-----|-----------|-----------|   |-----|-----------|-----------|   |-----|-----------|-----------|
|  A  | ILL-UM    | ILL-ŌS    |   |  A  | ILL-AM    | ILL-ĀS    |   |  A  | ILL-UD    | ILL-A     |   
|-----|-----------|-----------|   |-----|-----------|-----------|   |-----|-----------|-----------|
|  G  | ILL-ĪUS   |           |   |  G  | ILL-ÆIUS  |           |   |  G  | ILL-ĪUS   |           |   
|-----|-----------| ILL-ŌRUM  |   |-----|-----------| ILL-ĀRUM  |   |-----|-----------| ILL-ŌRUM  |   
|  D  | ILL-Ī     |           |   |  D  | ILL-Æ     |           |   |  D  | ILL-Ī     |           |   
|-----|-----------|-----------|   |-----|-----------|-----------|   |-----|-----------|-----------|
Since the earliest times, there was a tendency to replace the pronominal dative singular ending -Ī by the nominal endings -Ō (m./n.) and -Æ (f.). Feminine -Æ instead of -Ī appears already in pre-classical Latin (and is inherited by the Romance languages), while masculine/neuter -Ō instead of -Ī appears only in late Latin. This is certainly in order to avoid gender syncretism, which only appears in the pronominal declension. The feminine genitive singular ending -ÆIUS instead of -ĪUS arises from simple analogy: -Ī : -ĪUS = -Æ : -ÆIUS.

Most Romance languages (Sardinian looks as if it didn't, but I haven't been able to find a historical grammar of Sardinian to verify) later spread masculine/neuter genitive and dative endings -UIUS and -UĪ from the relative pronoun CUIUS and CUĪ to other pronouns, replacing -ĪUS and -Ī. I'm leaning to not having this change, especially if Sardinian didn't have it, since the aforesaid influence of the pronominal declension on the nominal declension is less convoluted that way: with the change, pronominal m./n. Dsg -Ī would have to replace nominal m./n. Dsg -Ō before it itself is replaced by -UĪ, which limits the time available for -Ō > -Ī. Without the change, no problem.


3. First declension

Code: Select all

 FEMININE:
|-----|-----------|-----------|
|  -  |    sg.    |    pl.    |
|-----|-----------|-----------|
|  N  | -A        | -Æ        |
|-----|-----------|-----------|
|  A  | -AM       | -ĀS       |
|-----|-----------|-----------|
|  G  | -ÆS       |           |
|-----|-----------| -ĀRUM     |
|  D  | -Æ        |           |
|-----|-----------|-----------|
An otherwise lost genitive singular ending -ÆS is preserved, eliminating the syncretism between the genitive and the dative in the singular.

Ideas for further development:

- Prolonged contact with Greek might introduce masculine first declension nouns in -AS.

- The loss of word-final /m/ will cause a merger of the nominative and the accusative in the singular. The two cases may remain distinct in the plural (it's possible) or they may merge by analogy (it happened in Old French). If they merge, would the nominative or the accusative be the more likely winner? Accusative won in Old French, but that's probably because it was unmarked in the Old French nominative vs. oblique (< accusative) case system, so the nominative may win in my language.


4. Second declension

Code: Select all

 MASCULINE:                        NEUTER:
|-----|-----------|-----------|   |-----|-----------|-----------|
|  -  |    sg.    |    pl.    |   |  -  |    sg.    |    pl.    |
|-----|-----------|-----------|   |-----|-----------|-----------|
|  N  | -US       | -Ī        |   |  N  | -UM       | -A        |
|-----|-----------|-----------|   |-----|-----------|-----------|
|  A  | -UM       | -ŌS       |   |  A  | -UM       | -A        | 
|-----|-----------|-----------|   |-----|-----------|-----------|
|  G  | -ĪS       |           |   |  G  | -ĪS       |           |
|-----|-----------| -ŌRUM     |   |-----|-----------| -ŌRUM     |
|  D  | -Ī        |           |   |  D  | -Ī        |           |
|-----|-----------|-----------|   |-----|-----------|-----------|
I imagine, if cases were preserved, declension would eventually be simplified and regularized. For that reason I decided to borrow the masculine/neuter dative singular ending -Ī from the pronominal declension: It reduces case allomorphy across declensions (-Ī is now the Dsg of 2nd, 3rd and pronominal declensions) and makes the 2nd declension more parallel to the 1st declension (both now have syncretism of Dsg and Npl). Once the syncretism of Dsg and Npl is established, the addition of a final -S to the genitive singular by analogy to the first declension simply presents itself (even if I got the idea to do it before considering other changes, the borrowing of m./n. Dsg -Ī actually came later!): Gsg = Dsg/Npl + -S in both the first and second declensions.

This way there's a nice amount of morphological change to the original Latin declensions.


5. Third declension

Code: Select all

MASCULINE/FEMININE:               NEUTER:
|-----|-----------|-----------|   |-----|-----------|-----------|
|  -  |    sg.    |    pl.    |   |  -  |    sg.    |    pl.    |
|-----|-----------|-----------|   |-----|-----------|-----------|
|  N  | -IS       | -ĒS       |   |  N  | -E        | -(I)A     |
|-----|-----------|-----------|   |-----|-----------|-----------|
|  A  | -EM       | -ĒS       |   |  A  | -E        | -(I)A     |
|-----|-----------|-----------|   |-----|-----------|-----------|
|  G  | -IS       |           |   |  G  | -IS       |           |
|-----|-----------| -(I)UM    |   |-----|-----------| -(I)UM    |
|  D  | -Ī        |           |   |  D  | -Ī        |           |
|-----|-----------|-----------|   |-----|-----------|-----------|
Except that the nominative singular of most nouns would be regularized to -IS, there isn't any additional morphological change set in stone at the moment, since the results would necessarily interact with sound changes: Does short I become /i/ as in Sardinian, or /e/ as in other languages?

Ideas for further development:

- A genitive/dative plural ending -ĒRUM really wants to come into being since the existing ending is bound to get wrecked by sound changes. It could be a survival from the 5th declension (if it's remotely realistic) or an analogical creation after the 1st and 2nd declension endings (if it isn't).

- Prolonged contact with Greek might introduce masculine third declension nouns in -ES. However, that ending would likely be replaced by -IS since a) Greek η eventually became /i/ and b) Appendix Probi shows that there was a strong tendency to replace -ES by -IS in inherited words as well.

- The 2nd declension nominative plural ending -Ī will definitely spread to 3rd declension masculines. It might spread to feminines if nominative and accusative plural stay distinct or if the nominative ending replaces the accusative one, but not if the accusative ending replaces the nominative one (so feminines would be either Npl -Ī, Apl -ĒS, or both -Ī or both -ĒS).

- Feminine 1st declension nouns characteristically have no -S in the nominative singular, while feminine 3rd declension nouns have one. Old French preserves this situation very well, but my language may as well copy the 1st declension and get rid of the final -S in 3rd declension nominative singulars of feminine nouns. This would be especially likely if 1st declension masculines in -AS appear in Greek loanwords so that final -S becomes characteristically masculine and no final -S characteristically feminine.

- That syncretism of Nominative and Genitive singular, both -IS, is nasty! How do I get rid of it? My two scenarios depend on which development of vowels i choose:

a) Sardinian style, short I becomes /i/.

Third declension genitive and dative would become -is and -i, merging with the corresponding 2nd declension endings, and the nominative plural would, as said, also be -i by analogy. Nominative singular would be -is and accusative singular would be -e. An analogical replacement of Nsg -is by -es is imaginable (2nd declension -u, -us, 3rd declension -e, -is > -es), but contrary to the existing tendency of replacing -ES by -IS. I concluded before that Greek wouldn't be much of a help because of it's own sound changes.

b) Mainland style, short I becomes /e/.

I'd get nominative singular -es and accusative singular -e from -IS and -EM by sound changes, negating the problem mentioned above. Dative singular and nominative plural would be -i, and genitive singular would be -es. To get rid of the syncretism between the Nsg and the Gsg, I could just borrow the Gsg -is from the 2nd declension, since the dative singular and the nominative plural would already be the same. This analogy seems to me more probable that the one in option a), so I'm leaning not to have the sardinian development of vowels.

What do you all think? [:D]
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

IEPH
cuneiform
cuneiform
Posts: 112
Joined: 23 May 2016 00:40

Re: Zeko-Romance development thread

Post by IEPH » 19 Aug 2019 15:41

Zekoslav wrote:
14 Aug 2019 14:06
This isn't, I confess, my first Romance conlang: that honor would go to Occitan, which I accidentally reinvented while trying to invent a Gallo-Romance language somewhat more conservative than French...
Not to nitpick, but Occitan is NOT a conlang, but a natural language with real speakers.

shimobaatar
korean
korean
Posts: 11686
Joined: 12 Jul 2013 23:09
Location: PA → IN

Re: Zeko-Romance development thread

Post by shimobaatar » 19 Aug 2019 16:17

IEPH wrote:
19 Aug 2019 15:41
Zekoslav wrote:
14 Aug 2019 14:06
This isn't, I confess, my first Romance conlang: that honor would go to Occitan, which I accidentally reinvented while trying to invent a Gallo-Romance language somewhat more conservative than French...
Not to nitpick, but Occitan is NOT a conlang, but a natural language with real speakers.
Yes, indeed, but I don't believe "Occitan is a conlang" was the claim being made. It sounds to me like Zekoslav was dissatisfied with the extent to which their first Romlang ended up resembling the natural language Occitan, so here they're humorously referring to that conlang by the same name as the natural language it resembled and speaking of accidentally reinventing Occitan.

User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Re: Zeko-Romance development thread

Post by Zekoslav » 19 Aug 2019 19:40

shimobaatar wrote:
19 Aug 2019 16:17
IEPH wrote:
19 Aug 2019 15:41
Zekoslav wrote:
14 Aug 2019 14:06
This isn't, I confess, my first Romance conlang: that honor would go to Occitan, which I accidentally reinvented while trying to invent a Gallo-Romance language somewhat more conservative than French...
Not to nitpick, but Occitan is NOT a conlang, but a natural language with real speakers.
Yes, indeed, but I don't believe "Occitan is a conlang" was the claim being made. It sounds to me like Zekoslav was dissatisfied with the extent to which their first Romlang ended up resembling the natural language Occitan, so here they're humorously referring to that conlang by the same name as the natural language it resembled and speaking of accidentally reinventing Occitan.
*thumbs up*

You're completely right, that was an attempt at some self-deprecating humour. I actually kept revising that language untill it became sufficiently distinct from any natural language, but it's still a typical Western Romance language and I wanted to do something wilder this time.
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

Clio
sinic
sinic
Posts: 211
Joined: 27 Dec 2012 23:45

Re: Zeko-Romance development thread

Post by Clio » 21 Aug 2019 17:32

I believe it's been said that there are two kinds of novice Romlangers: those who reinvent Occitan, and those who reinvent Catalan.

EDIT: I should say something substantive. Zekoslav, you've mentioned a few times adopting Greek third declension nouns in -ης. Did you know that in Greek many of these nouns were analogized into the first declension (so acc. -ην etc.)? It could be interesting to experiment with different ways of handling those loanwords.
Niûro nCora
Getic: longum Getico murmur in ore fuit
scratchpad

User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Re: Zeko-Romance development thread

Post by Zekoslav » 21 Aug 2019 19:59

Clio wrote:
21 Aug 2019 17:32
I believe it's been said that there are two kinds of novice Romlangers: those who reinvent Occitan, and those who reinvent Catalan.

EDIT: I should say something substantive. Zekoslav, you've mentioned a few times adopting Greek third declension nouns in -ης. Did you know that in Greek many of these nouns were analogized into the first declension (so acc. -ην etc.)? It could be interesting to experiment with different ways of handling those loanwords.
Yea, real Romance languages are so varied that it's probably impossible to come up with a completely original development: in all likelihood some Occitan or Rhaeto-Romance dialect has already done even worse [;)].

As it concerns Greek nouns, I was aware than most nouns in -ης were actually first declension, while only a few were third declension. When it comes to borrowing them into Latin, Classical Latin did it in a very learnèd way, keeping the original declension and even some original suffixes. I was thinking my romlang speakers would do it in a more straightforward way, adapting the nouns into the declension pattern which they resemble the most. So Greek first declension Nom -ης, Acc -ην resemble Latin third or fifth declension Nom -ēs, Acc -em (if I choose the Romanian development of vowels as I'm currently inclined to do, then even the more common third declension ending Nom -is would give the adequate result).

So Greek analogising third declension nouns in -ης after the first declension actually suits my purposes very well! As I've mentioned, given enough loanwords, the Greek pattern of s-ed masculines -ᾱς, -ης and s-less feminines -ᾱ/-α, -η might get copied by Latin, so that third declension masculines have Nom -es < -IS, -ĒS, Acc -e < -EM, but feminines have Nom = Acc -e < -EM. I know masculines in -ᾱς were rare in Attic and the derived dialects of Greek, but they would include some common Christian names like Lucas and Matthias and names can and do influence declension.
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Re: Zeko-Romance development thread

Post by Zekoslav » 28 Aug 2019 12:45

1. VOWELS

It's generally known that there were three main developments of the Classical Latin vowel system:


a) "Sardinian" system: I, Ī > i; E, Ē, Æ > ɛ; A, Ā > a; O, Ō > ɔ; U, Ū > u

b) "Romanian" system: Ī > i; I, Ē > e; E, Æ > ɛ; A, Ā > a; O, Ō > ɔ; U, Ū > u

c) "Mainstream" system: Ī > i; I, Ē > e; E, Æ > ɛ; A, Ā > a; O > ɔ; U, Ō > o; Ū > u


However, this paper points out that certain areas of southern Italy which were thought to have the Sardinian (southern Lucania, northern Calabria) or the Romanian (central Lucania) system actually have systems of their own:


d) "Lucanian" (pseudo-Sardinian) system: I, Ī > i; Ē > e; E, Æ > ɛ; A, Ā > a; O > ɔ; Ō > o; U, Ū > u

Basically, a quality distinction appears only in mid vowels. Later, metaphony (Romance term for umlaut) changes /e/ and /o/ into /i/ and /u/, and even later /e/ and /o/ merge with /ɛ/ and /ɔ/ giving the impression of the Sardinian development, if not for metaphony.


e) "Calabrian" (pseudo-Romanian) system: Ī > i; I, Ē > e; E, Æ > ɛ; A, Ā > a; O > ɔ; Ō > o; U, Ū > u

Basically, a quality distinction appears for both mid vowels and for the front but not for the back high vowel. Later, metaphony changes /e/ and /o/ into /i/ and /u/, and even later /e/ and /o/ merge with /ɛ/ and /ɔ/, giving the impression of the Romanian development, if not for metaphony.


This newly discovered variety tells us that a) every non-Sardinian variety of Romance has some quality distinctions and b) the variety of quality distinctions is greater than previously thought, suggesting the possibility that even more variety existed. This gives me an impulse to create my own vowel system!

One idea is to follow the "Calabrian" system, but introduce a quality distinction for both front and back high vowels. However, instead of short U merging with long Ō as it did in the "Mainstream" system, there would be a chain shift Ū > ʉ; U > u. Later, /e/ and /o/ would merge with /i/ and /u/ (possibly due to Greek influence, as has been suggested for the areas of southern Italy where the same development happened) and the result would be a vowel system of /i/, /ɛ/, /a/, /ɔ/, /u/ and /ʉ/. However, I'm not entirely sure if I like the consequences of that, especially for the verb system (the second and third conjugations would just merge with the fourth).


2. THE FATE OF THE NEUTER GENDER

The same author has suggested that the Eastern Romance treatment of the neuter gender as masculine in the singular and feminine in the plural appeared due to some specific morphological reasons, namely the spread of the nominative plural ending -Æ(C) from those pronouns where it was the same for feminine and neuter nouns to other pronouns and to the definite article: that is, Italian le braccia 'DEF-N.pl arm-N.pl' would come from ILLÆC BRACCHIA instead of from ILLA BRACCHIA.

If I don't include this development, then the neuter gender would likely end up merging with the masculine, maybe providing an alternative "mass plural" ending -a, and if the feminine gender merges Nom and Acc then the Acc plural ending would likely survive. However, if I include this development then the neuter gender would survive and if the feminine gender merges Nom and Acc then the Nom plural ending would likely survive since it would be shared with the neuter gender, which would already use it for both Nom and Acc plural (unless something weird and complicated happens).

whaddaya think? [:O]
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Re: Zeko-Romance development thread

Post by Zekoslav » 16 Oct 2019 12:01

Three new ideas:

1. I'm preserving hic, haec, hoc. It will be reinforced with *ECCU < ecce eum and become first an anaphoric pronoun with no spatial deixis, later the definite article.

This will have two important consequences: 1. provide a pronominal accusative ending in /n/ instead of the usual zero ending: /k(w)ʊn/ < *ECCU HUNC and /kwan/ < *ECCU HANC. 2. provide the abovesaid prerequisites to treat neuter as masculine in the singular and feminine in the plural.


2. I'm dropping final /t/ and /nt/, keeping only final /s/ and rarely (in proclitics) /n/. It's consistent with southern Italian and Romanian which would be the language's closest relatives and with Greek which would be it's adstrate and superstrate.

I'm in doubt whether to save the distinction between the third person singular and plural like Italian and Romanian (mostly) did, or to merge them like Romanian (in the first conjugation) did, but for all verbs and all tenses.


3. I'll do a pseudo-Romanian vowel system where short i merges with long e, and other short vowels with their long counterparts, but only secondarily, with the distinction being possible preserved through some kind of metaphony.
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

Zythros Jubi
sinic
sinic
Posts: 376
Joined: 24 Nov 2014 17:31

Re: Zeko-Romance development thread

Post by Zythros Jubi » 31 Oct 2019 04:21

How do you like the vowel system of Tundrian and its evolution?
Tundrian is the only descendant of Latin that preserved the distinction between original short and long vowels until quite late in its history (9th-10th century AD). Although this distinction has by now disappeared in pronunciation, Tundrian spelling has retained it to this day, as can be seen from the following examples of stressed vowels:

Latin ĕ > Tundrian e [ɛ] ~ [e]. Examples: perdo ['pɛrdu] 'I lose'; ped ['ped] 'foot'
Latin ē > Tundrian ei [\i]. Example: teila ['tila] 'cloth'
Latin ĭ > Tundrian i [\i]. Example: fid ['fid] 'faith'
Latin ī > Tundrian iy [øĭ]. Example: viyta ['vøĭta] 'life'
Latin ŏ > Tundrian o [ɔ] ~ [o]. Examples: porta ['pɔrta] 'door'; vole ['volə] 'he wants'
Latin ō > Tundrian ou \[\u]. Example: soul ['sul] 'sun'
Latin ŭ > Tundrian û [\u]. Example: pûtz ['puts] 'well' (n.)
Latin ū > Tundrian u [y]. Example: mur ['myr] 'wall'

It should be noted that the development of Latin /a/ was not affected by its original quantity:

Latin ă > Tundrian a [ɑ] ~ [a]. Examples: Lat. părtem > part ['pɑrt] 'part'; Lat. mălum > mal ['mal] 'bad'
Latin ā > Tundrian a [ɑ] ~ [a]. Examples: Lat. pāscō > pasco ['pɑsku] 'I feed (animals)'; Lat. portāre > portar [pɔr'tar] 'to carry'
Notice the closed/open syllable distinction for Latin a and (historically short) e, o, as well as the breaking of ī and fronting of ū; from the example words listed by him, it seems that <u> can appear in both closed and unstressed syllable but <iy> cannot in either.
The majority of verbs with root -i- in an open syllable belong here, e.g.: avricar (to shelter): avriyco; castigar (to punish): castiygo; critar (to shout): criyto; dêsirar (to desire): dêsiyro; filar (to spin, file): fiylo; guidar (to guide): guiydo; invitar (to invite): inviyto; mirar (to watch): miyro; pizar (to pound, crush): piyzo; suïcidar-se (to commit suicide): me suïciydo; tirar (to pull): tiyro; etc.
Quite a few verbs of this type do not alternate, however, e.g.: abdicar (to abdicate): abdico; citar (to quote): cito; fricar (to rub): frico; hûmiliar (to humiliate): hûmilio; plicar (to fold): plico; all verbs in -izar, e.g. bautizar (to baptize): bautizo; and some others.
Many learned verbs with -i- in an open stem-final syllable belong in a different category of irregular verbs, and never stress the -i-, e.g.: agitar (to agitate): ágito; êditar (to edit): édito; etc.
When the -i- is in a closed syllable, it never alternates. Examples: bindar (to bind): bindo; brillar (to shine): brillo; piccar (to sting): picco; etc.
Something Romanian:
Tundrian had a palatalization sequence unknown in other western Romance languages, although something similar has occurred in Romanian. Essentially, dental consonants were palatalized before Latin stressed ĕ and ĭ (and, sporadically, also before stressed ŏ and ŭ). In the case of dental stops, there was a further development to affricates. Examples:

Latin tĕrra > tzerra [''tsɛra] 'earth'
Latin dĕcem > dzeç [''dzetʃ] 'ten'
Latin nĭvem > nhive [''ɲivə] 'snow'
Latin sĭtis > xit ['ʃit] 'thirst'
Latin lĕvō > lhevo ['ʎevu] 'I lift'
Latin sŭrdus > xûrd ['ʃurd] 'deaf'
Lostlang plans: Oghur Turkic, Gallaecian Celtic, Palaeo-Balkanic

User avatar
Zekoslav
sinic
sinic
Posts: 287
Joined: 07 Oct 2017 16:54

Re: Zeko-Romance development thread

Post by Zekoslav » 31 Oct 2019 10:09

Tundrian is interesting, and for a time I considered doing something similar. There'd be ī i ē e > i e e ɛ and ū u ō o > u ʊ o ɔ, then ʊ ɔ > u o by metaphony, then u ʊ > ʉ u, then e o > i u. So the distinction between ū and u would be preserved as a distinction between ʉ and u, and o would later merge with u. But I didn't quite like the grammatical outcomes of merging e with i. Right now I'm not entirely certain about anything except that I'll likely merge i and ē. As for palatalisation, I'm thinking about doing something unique, such as /sj/ > /st/.
Languages:
:hrv: [:D], :bih: :srb: [;)], :eng: [:D], :fra: [:|], :lat: [:(], :deu: [:'(]

A linguistics enthusiast who would like to make a conlang, but can't decide what to call what.

- Tewanian languages
- Guide to Slavic accentuation

Salmoneus
MVP
MVP
Posts: 1684
Joined: 19 Sep 2011 19:37

Re: Zeko-Romance development thread

Post by Salmoneus » 31 Oct 2019 20:44

Regarding Lucanian as 'pseudo-Sardinian', with a 'fourth' vowel outomce: there is of course another way to see this.



Personally, I suspect that early Romance should be seen less as a tree-like series of binary splits, and more as the incomplete spread of a number of shifts through a dialect continuum.

So far as I can see, all the Romance vowel systems can be explained as a result of a simple chain of shifts (ignoring irregularities, analogies, different rules for pretonic,unstressed or final vowels, and the diphthongs, etc):

Starting point in Vulgar Latin: AEIOU all have a length distinction. There is a secondary quality distinction, in which short EIOU are slightly lower than long (there may have been an equivalent distinction in A, or a backing distinction, but this is always lost so we don't know). There is also metaphony: long and short E and O become higher before a high vowel (Romanian backing and lowering suggest that there may have been much more pervasive metaphony that was not phonemicised in most languages).


Shift A: loss of vowel length (and eradication of secondary quality distinction in the process)
Shift B: raised E: and O: heighten further, to a similar height as long I: and U:
Shift C: short I sinks below raised E:, to a similar height as unraised E:
Shift D: short U sinks below raised O:, to a similar height as unraised O:
Shift E: raised E and O heighten further, to a similar height as unraised E: and O:
Shift F: raised E and O heighten further, to be even higher than unraised E: and O:
Shift G: raised E and O break
Shift H: raised E and O, and E: and O:, merge with unraised versions



So, if we look at which shifts occur where:

Sardinian: A (> E)
Early length loss makes the following changes irrelevent to Sardinian.

Lucanian: B > A (> H?)
Basically the same vowel system, except that length loss doesn't fully take effect until the raising of the heightened long mid vowels has already taken place - so that when length is lost, these merge with the high vowels

Calabrian: B > C > A (> H?)
Again very similar, but now the length loss has been delayed further, allowing the widespread Imperial shift C to creep in. As a result, the length loss merges short I and long (unraised) E:

Western: B > C > D > A > H
Here anothe shift has snuck in before length loss: short U lowers to mirror short I. Raising of low-mid E and O is of minor importance and distinction, so does not become phonemic.

Central Italic: B > C > D > E > A
As in Western, except that metaphonic raising of low-mid E and O is exaggerated, leading to merger with high-mid vowels and phonemicisation of metaphony.

Southern Italic: B > C > D > E > F > A
The same, except that the metaphonic raising is even more extreme, so no merger occurs when length is lost. Instead, the semi-high resulting vowels end up broken to simplify the vowel system.

Romanian: C > A > H
The same as Calabrian, except that the more extreme western metaphonic raising never occurs (however, metaphonic lowering and backing are exaggerated, perhaps to produce a similar effect). Even so, the combination of C and A should phonemicise metaphony; however, actual minimal pairs are very few, and the distinction is small, so it's no great surprise that the barely-distinguished phonemes just merge.


This probably oversimplifies things, but I think you see my point: that we don't have to assume breaking apart into different 'systems' from the beginning, but only have to assume that different shifts reach different areas in different orders.

Salmoneus
MVP
MVP
Posts: 1684
Joined: 19 Sep 2011 19:37

Re: Zeko-Romance development thread

Post by Salmoneus » 31 Oct 2019 23:45

Anyway, one idea I've had is to further scrunch up the timeline by having the lengthening of vowels in open syllables begin before the above changes are all completed. That gives you potentially up twenty vowels to play with - or up to twenty-eight including raising metaphony (and excluding lowering). And then even if there's radical simplification, it needn't be in the most straightforward way.


One system that intrigued me was something like the following (only looking at back vowels, and using ò for raised-o, ô for originally long o, and ó for raised ô):


o - o - o
ò - o - o
ô - ô - o
ó - u - u
u - u - u
û - u - u
o: - o: - o:
ò: - ô: - o:/ou - o:/û:
ô: - ô: - o:/ou - o:/û:
ó: - u: - ue - ui
u: - û: - y:
û: - û: - y:

So, to start with, metaphony only applies to long vowels, which are EITHER in an open syllable, OR originally long (but the latter then lose their length). u/û merge early. Then o: from the high-mid vowel (or raised low-mid vowel) breaks before a sonorant, and eventually rises further.

Hence:
focum > fouc (/fu:k/)
foco:s > fóca (fo:k/) (ending from neuter)
po:mum > puim
po:ma > pouma
multum > mult
multo:s > mult
cohortem > cort
cohorte:s > curta
flo:rem > flour
flo:re:s > fluira
formo:sum > formuis
formo:so:s > formósa
hominem > houm (or hóm by analogy?)
homine:s > houmi

etc.

Post Reply