Ruritanian

[brblues]

Hi everyone! I’ve decided to revamp my conlang sketch Mbath into a conlang called Ruritanian, and the lang is now rooted in an althist world whose major difference from “here” is the existence of the ethnic group of the Ruritanians in the Balkans. Sometime in the 18th century or so, the territories settled by them, roughly comprising Slavonia (today part of Croatia), Vojvodina (part of Serbia), the Romanian Banat, and a swathe of adjoining land in what is today Southern Hungary, became part of the Habsburg Empire. The course of the broad strokes of history is unaltered until 1849, when the uprisings of that year are successful in Austria, last not least due to the very active support by the Ruritanians. This leads to the formation of the United States of Greater Austria and Transleithania, whose constituent member nations are granted far-reaching autonomy, but keep the Habsburg emperor. I’ll write more about the althist in conworlding posts – there’s also a lot I’m not decided on yet, for instance small questions such as if there will be magic or not :D

But now for the language itself - this is an a-priori, so it's placed in the Balkans, but not genetically related to any of the surrounding languages. I will still some Balkan sprachbund features gradually though, I think.

This first post will solely deal with the phonology.

Inventory



Orthography

/m/ <m> /n/ <n> /ɳ/ <ṇ>
/p/ <b> /ph/ <p> /mb/ <mb> /t/ <d> /th/ <t> /nd/ <nd> /ʈ/ <ṭ> /k/ <g> /kh/ <k> /ᵑg/ <ng>
/s/ <s> /ʃ/ <ṡ> /ʂ/ <ṣ> /ħ/ <h>
/ɭ/ <l> /ɻ/ <r> /j/ <j,i> /w/ <w,u>
/i/ <i> /y/ <y> /u/ <u> /o/ <o>
/ɛ/ <e> /ɒ/ <a>

There may be some tweaks to this, plus etymological spellings, once I've more firmly established historical processes and the place of the language in its area.

Allophones and synchronic sound changes

- In onset, the prenasalisation of the voiced stops is not realised by all speakers, except in careful speech; when following a consonant in a cluster, it is never realised; and when following a vowel, it is realised as the corresponding nasal consonant
- Aspirated consonants are only realised with audible aspiration in coda position; however, an aspirated onset raises vowels as per the following pattern:

Code: Select all

ɒ > ʌ; o > ɯ; ɛ > e

- The above vowel raising pattern also applies to syllables whose onset contains /ʃ/
- Unstressed [ʌ] is commonly realised as schwa, but not by all speakers
- Unstressed /i/ is realised as [ɨ], which is in turn lowered to schwa unless it carries secondary stress (constituents of compounds, incorporated nouns and some other morphological elements of verbs may carry secondary stress)
- Word-final schwa can be deleted
- /w/ and /j/ in coda can only follow the vowel directly, and they form the following diphthongs with the vowels:

Code: Select all

j: [ɒ,ʌ]+ /j/ > [aɪ̯]; [o,u,ɯ] + /j/ > [ɔɪ̯]; [ɛ,e] + /j/ > [eɪ̯]; [i,y,ɨ] + /j/ > [i:] if stressed; otherwise [ɪi̯] 
w: [ɛ,e] + /w/ > [eu̯]; [i,y,ɨ] + /w/ > /ju/; [ɯ,u] + /w/ > [u:] if stressed, [u] otherwise; [o,ʌ,ɒ] + /w/ > [aʊ̯] if stressed; otherwise [oʊ̯]

- /w/ is realised as [v] in onset; speakers in the “High South” area realise syllable-final /w/ as [f]
- Consonant assimilation rules, in sequence:
1. Geminate plain consonants block assimilation
2. Plain consonants assimilate next to aspirated consonants if possible (i.e. if they have an assimilated counterpart)
3. Plain consonants assimilate next to retroflex consonants if possible (i.e. if they have a retroflex counterpart)

Phonotactics

Ruritanian has a maximum syllable structure of CCCVCC.
In a three-consonant cluster, which can only occur in onset (except for syllable boundaries, where these rules don’t apply)) the third consonant must be a liquid or semivowel, and the sonority hierarchy principle must be followed:
Stop – Fricative – Nasal - Liquid – Semivowel
The reverse of this hierarchy applies in the coda.
Clusters of two stops, fricatives or nasals can occur in onset only (again, with the exception of syllable boundaries).


Let me know if you have any thoughts, I will hopefully get the real deal soon - the grammar of the language, and the althist of the conworld! [/s]

This is no longer up-to-date; please see the later posts for the current incarnation of the language!

[brblues]

Verbs!

Verbs are the morphologically most complex part of Ruritanian and convey not only information about the action, but also about the arguments, such as the grammatical number of arguments as well as the definiteness of the S/O argument (subject of intransitive clauses, object of transitive clauses; never the agent of a transitive clause).

An important thing to note is that, through diachronic vowel harmony processes, verbs generally contain only front vowels, with the notable exceptions of the remote-past affix right in the very first slot of the template (derived from an auxiliary verb that only fused onto the verb quite recently) and vowels resulting from synchronic sound changes. The template looks as follows:



Even ignoring the relative clause markers REL as well as the irrealis marker IRR for now, we can already form some quite impressive verbs, such as:

<Ksijmtsyu!>
ksij-mt-sy-w
/’khʃijmth.syw/
[kʃi:mth.ʃu]
take-PST-2PL-FEM.DEF
“You guys took her/it!”

Or a sentence even:
<Bimtlirnins go mojn Mjandraṡun.>
bi-mth-lir-nins go mon Mjandraṡ-un
/pimth.ɭiɻ.nins ko mon ‘Mjɒnd.ɻɒ.ʃun/
[pimth.ɭəɻ.ɳəns ko mon Mjɒnd.ɻɒ.ʃun]
give-PST-PL.DEF-HEARSAY ACC egg Mjandraṡ-DAT
“You’re said to have given Mjandraṡ the eggs.”

This is no longer up-to-date; please see the later posts for the current incarnation of the language!

[WeepingElf]

This is not a bad conlang, but I must say that it doesn't feel like something from the Balkan Peninsula at all. Things like the retroflex stop, the prenasalized stops, the lack of sibilant affricates, and the whole verbal morphology simply feel out of place in that part of the world. Each part of the world has its own distinctive "linguistic flavour" regardless of relationships, as languages influence each other and develop common traits. I don't see this in your conlang, sorry.

[brblues]

Thanks for your feedback! You've voice some concerns I had myself about simply planting an a-priori in a real-world area without changing the phonology too much. Main reason for that was that I just rather liked the phonology (ofc that could less charitably be called pure laziness...), and it was hard to change any of the moving parts without changing most of what I like about it. The vowel system looks quite at home in the area to me though, while with the consonant inventory I'm just gonna have to suspend disbelief a bit. Regarding the language overall, I feel that the grammar will in the end hopefully not stand out much more in the area than Turkish and Hungarian do now, especially given that the settled area is on the fringe of the Balkan sprachbund itself - though there's contact there with Turkish and Hungarian, from which I will take a good number of loans and calques, in addition to those from the sprachbund itself, plus vowel harmony processes (those are mostly no longer productive though). Grammatical features from the sprachbund itself that are present so far are the renarrative mood (hearsay-past in the verb template) and a genitive-dative merger.

[Omzinesý]

Trying to conlang in a real-world area and to adapt the lang to the area as well as possible can be interesting. But if one does not want to do that, I don't see why he should. Peculiar things can happen and there are very out-of-the-place natlangs, as well.
It is very possible that those out-standing features are the Ruritanian flavor and they should have affected the Balkan Sprachbund. Conlanging doesn't really have rules.
But yes, if one is trying to make a Balkan lang, there could be more Balkan features.

[brblues]

I've been thinking about how to "Balkanise" this lang, and there might be a few options, but none seem ideal to me. Maybe you guys can help me decide!

a) The first option would of course be to simply not do that, and claim "suspension on disbelief", based on the fact that natlangs can be quite out of place areally, too, especially ones from a different language family (Hungarian and Turkish are still quite different from the rest of the languages spoken on the Balkan, after all).

b) There really are a couple of stand-out features phonologically, a few of which could be remedied without changing the overall system too much. These might be:

- Do away with retroflex stops, and instead have a three-way distinction between voiced, plain unvoiced and aspirated unvoiced.

- In the wake of the previous change, do away with prenasals and make them simply voiced.

- Possibly: Change the place of articulation of liquids from retroflex to something else.

- Possibly: Add more sibilants.

c) Postulate the current state of the language as a proto-language or early stage of the language at some point relatively soon after its speaker settled in the area, then work it through sound changes very much influenced by neighbouring Balkan languages.

d) Make a completely new language that is adapted to the environment from the get-go. My question there would be: how?

[Davush]

I think there are a number of ways of looking at it...

I suppose, first of all, is to decide how geographically isolated and/or how strong the cultural/linguistic identity of these speakers is, as well as population size. With a strong identity and a big enough population, I don't think it's totally beyond the realms of plausibility for such a phonology to be present where you say – after all, Basque doesn't really look like it fits into the Iberian Peninsula at all yet it has managed to survive.

If you did want it to blend into the surrounding languages a bit more, my personal opinion would be to just remove the retroflex stop and nasal, the pharyngeal (or change to uvular/velar), and change the aspirated/unaspirated distinction to a voiced/voiceless one. The addition of affricates perhaps only in loans could also add a bit more Balkan flavour? Other parts of the phonology might seem a bit 'unusual' for the area, but not completely out of place (Greek has prenasalized stops, and Albanian allows initial NC clusters, as far as I know).

Personally, I like the idea of it being quite different from the surrounding languages while sharing some traits, and it sounds like an interesting scenario.

[Omzinesý]

What about just making the retroflex stop a retroflex affricate? Stop and affricate realizations can even be allophonic. The three series of affricates and sibilants in Slavic can be analysed as dental, retroflex and palatal.
I think Albanian has a velar/dark lateral. Many languages have a palatal lateral.
Modern Greek has prenasalized stops (according to some analyses).

[Omzinesý]

Where did you get this idea from?
I think it might happen as a diachronic process (Somebody else knows more about phonetics and formants.), but as a synchronic description, it sounds odd. At least I haven't seen such a process.

- Aspirated consonants are only realised with audible aspiration in coda position; however, an aspirated onset raises vowels as per the following

[brblues]

Thanks for your replies Davush and Omzinesý!

If you did want it to blend into the surrounding languages a bit more, my personal opinion would be to just remove the retroflex stop and nasal, the pharyngeal (or change to uvular/velar), and change the aspirated/unaspirated distinction to a voiced/voiceless one. The addition of affricates perhaps only in loans could also add a bit more Balkan flavour? Other parts of the phonology might seem a bit 'unusual' for the area, but not completely out of place (Greek has prenasalized stops, and Albanian allows initial NC clusters, as far as I know).

Personally, I like the idea of it being quite different from the surrounding languages while sharing some traits, and it sounds like an interesting scenario.

All these changes sound very reasonable to me and I might go ahead and use them :) ...except, however, for changing to a simple voiced-voiceless distinction. The reason is that that would destroy one specific feature I was going for, namely that two consonant series merge under some conditions (here: aspirated and plain stops in onset), but there is a phonemic difference due to the aspirated consonants raising the vowel (except in vowels that are already high, where it's all merged again). That's why I thought of adding a voiced series in, so I'm at a three-way distinction for stops again. A bit more about the consonant series in the following response:

Where did you get this idea from?
I think it might happen as a diachronic process (Somebody else knows more about phonetics and formants.), but as a synchronic description, it sounds odd. At least I haven't seen such a process.

- Aspirated consonants are only realised with audible aspiration in coda position; however, an aspirated onset raises vowels as per the following

I am not quite sure anymore how I was inspired for it, but the idea is that it actually was a diachronic process: Aspirated onsets raised the following vowel, and later the aspiration was lost, except in coda. It could also be a matter of analysis: the relevant vowel allophones could maybe also be analysed as separate phonemes, and the consonants as allophones. Important to note though that the series are not really fully merged even where it might look like it when regarding the raised vowels as phonemes, since aspirated consonants in the onset of a syllable could still trigger assimilation in a preceding syllable.

[brblues]

I have revised the whole thing somewhat and removed the retroflex nasal, stop and lateral, making those palatal, added an alveolar lateral, and moved the pharyngeal to velar. There might have been some other changes I've now forgotten about, plus some cosmetic ones.

The new system looks as follows:

/m/ <m> /n/ <n> /ɲ/ <ṅ>
/p/ <b> /ph/ <p> /mb/ <mb> /t/ <d> /th/ <t> /nd/ <nd> /c/ <ḋ> /ch/ <ṫ> /k/ <g> /kh/ <k> /ᵑg/ <ng>
/s/ <z> /ʃ/ <s> /x/ <h>
/t͡s/ <c> /ʈʂ/ <c̣>
/l/ <l> /ʎ/ <lj>
/ɻ/<r> /j/ <j,i> /w/ <w,u>
/i/ <i> /y/ <ü> /u/ <u> /o/ <o>
/ɛ/ <e> /ɒ/ <a>

I then also replaced the retroflex assimilation rule with a palatal assimilation rule:

3. Alveolar consonants assimilate to their palatal counterpart when preceding a palatal consonant

This is no longer up-to-date; please see the later posts for the current incarnation of the language!

[brblues]

I've tried that new phonology and orthography out on two previously posted short sentences:

<Ksijmtzüu!>
ksij-mt-sy-w
/’khʃijmth.syw/
[‘kʃi:mth.ʃu]
take-PST-2PL-FEM.DEF
“You guys took her/it!”

<Bimtlirnins go moṅ Mjandrasun.>
bi-mt-lir-nins go moṅ Mjandraz-un
/pimth.liɻ.ninʃ ko moɲ ‘Mjɒnd.ɻɒ.ʃun/
[pimth.ləɻ.ɳənʃ ko moɲ Mjɒnd.ɻɒ.ʃun]
give-PST-PL.DEF-HEARSAY ACC Egg Mjandras-DAT
“You’re said to have given Mjandras the eggs.”


And another sentence to test out the new orthography and phonology:

<Gbjun ṅelhndem go mi-Koṅ kzümtm?>
gbjun ṅel-h-nde-m go mi-Koṅ kzü-mt-∅-m
/kpjun ‘ɲɛl.ndɛm ko mi‘khoɲ khsymth.m/
[kpjun ‘ɲɛl.dɛm ko mə‘kɯɲ ksym.təm]
soon eat-IRR-1PL-MASC.DEF.DO ACC REL.MASC-sheep bring-PST-2SG-MASC.DEF.DO
“Are we soon eating the sheep you brought?”

This is no longer up-to-date; please see the later posts for the current incarnation of the language!

[Creyeditor]

The first word looks a bit Nilotic to me. Maybe it's just diacritics and the initial <Ks>.

[Nortaneous]

I suppose, first of all, is to decide how geographically isolated and/or how strong the cultural/linguistic identity of these speakers is, as well as population size. With a strong identity and a big enough population, I don't think it's totally beyond the realms of plausibility for such a phonology to be present where you say – after all, Basque doesn't really look like it fits into the Iberian Peninsula at all yet it has managed to survive.

It just hasn't kept pace with developments in Iberian Romance - it doesn't look too far off from earlier Spanish.

(Greek has prenasalized stops, and Albanian allows initial NC clusters, as far as I know)

Some dialects of Italian also allow NC clusters. Maltese is a better example, but it shows signs of influence from surrounding languages.

There are plenty of zones of influence that don't correspond to national languages - it could be possible to use those. Front rounded vowels are well-attested in Slavic dialects (Slovene, Rusyn, etc.), but the national languages all happen not to be based on varieties with front rounded vowels. It's also possible to take tendencies that already exist and accelerate them a little for alien effect - I have an IE conlang outline somewhere that's set somewhere in northern Europe and has phonemic ejectives - which is a little far out from Europe, but the path of development is *kk > *ˀk > kʼ (cf. Icelandic kk > ʰk, English k# > ˀk, etc.) followed by loss of some initial unstressed vowels making it possible for them to appear word-initially.

IMO the main problem is the pʰ/p contrast - it's too far south for that. Plosive phonation is highly areal. Retroflexes are also unlikely, but r > ɹ as an allophonic process is common in Albanian and sporadic in Turkish, so that's probably fine - and then the retroflexes could be from coalescence of r + coronal, which I assume is already the case because there's no /ⁿɖ/. (But then why isn't there a /ʈʰ/ either?)

The reason is that that would destroy one specific feature I was going for, namely that two consonant series merge under some conditions (here: aspirated and plain stops in onset), but there is a phonemic difference due to the aspirated consonants raising the vowel (except in vowels that are already high, where it's all merged again). That's why I thought of adding a voiced series in, so I'm at a three-way distinction for stops again.

Voiced consonants can also affect vowel height (and then merge into voiceless consonants or whatever) - this is common in SEA but would be a little weird here, tho maybe not much weirder than English and Turkish having the exact opposite sound changes around vowel length and consonant voicing (Turkish *V:T > VD)

[brblues]

I've had a think about things, and decided to try the diachronic route, starting my language around the year 1000AD with an inventory were similar to that of Late Common Slavic, from what I could glean from wiki articles. The dentals I took from Greek, which I think had them around the time.

[Vlürch]

https://i.ibb.co/xmBxzgd/protophon.png

That's a fun phonology, just a bit weird with that /ø/. Usually if a language has one front rounded vowel, it'll be /y/, and if a language has /ø/, it'll also have /y/. Not sure if that's an actual universal, though, and there's no reason why you couldn't handwave it even if it is since some languages just have weird phonological quirks; I'd assume /y/ lowering to /ø/ isn't unheard of, even though I can't think of any examples off the top of my head, so that could be a possible justification for its presence if you feel like you want a justification.

Anyway, I really like that it has /ð/ and a contrast between velar stops and fricatives.

[Omzinesý]

Hopi has /ø/ without /y/.

[brblues]

I think the /ø/ was a bit of a misstep, or rather a step out of tune. I want the language to diachronically develop rounded front vowels, and for some reason already included one of them there.

The main route I wanted to go for the front rounded vowels was to have the labialised vowel round [ i] and [e] to [y] and [ø], then at some point lose labialisation. A second route is fronting of the rounded back vowels after palatalised consonants, and then there could be an eventual third way where rounded back vowels are fronted in a kind of umlaut.

[brblues]

[screwed something up while editing, please view the next post}

[brblues]

I've been working on the protolang for Ruritanian a bit for some time now, calling it Old Ruritanian. This is spoken by the Ruritanian people in the northern Balkans area around the year 1000, so roundabout the latest stage of Common Slavic, from which I've taken cues about the phonology and also some sound changes. I will post a bit about the conworlding/althist in the relevant section soon, too.

I have now drawn up the sound changes on the way to Ruritanian. This is all still preliminary and not set in stone.

The phonology of Old Ruritanian I'm working from is this:


how to do exponents on keyboard

However, /v/ and /f/ are added in quite soon, the interdental fricatives disappear very quickly. In the code for sound changes I use a for [ä] and o for [ɔ] for simplicity.

The very first change is Slavic liquid metathesis, and the interdental fricatives disappear.

Spoiler:

Code: Select all

[C](')[V]r > [C]r(')[V]

[C](')[V]lj > [C]lj(')[V]

[C](')[V]l > [C]l(')[V]

θ > f / _#, _.
θ > s

ð > d / #_
-ð

Example word: (City name) <Mursapolis> [mur.'sa.po.lis} > [mru.sa.po.lis]


Then back vowels are fronted after palatalised consonant, and word-final consonants devoice. There's more liquid changes as the palatalised lateral goes fully palateral lateral, and the non-palatalised counterpart changes to [r]. The latter is a Balkan sprachbund change cited on the relevant wiki page as:

change from l to r in Romanian, Greek and very rarely in Bulgarian and Albanian.

Spoiler:

Code: Select all

[B] > [F] / j_

vc > vcl / _#

lj > ʎ
l > r

Example word: (City name) <Mursapolis> [mru.'sa.po.lis} > [mru.'sa.po.ris]


Then I rounded unrounded vowels after labialised consonants, and the labialisation on the consonant disappeared. I also made ɣ disappear.

Spoiler:

Code: Select all

w(')i,w(')e,w(')a > y,œ,œ
ɣ > g / #_
ɣ > j

Example word: (City name) <Mursapolis> [mru.sa.po.ris] no changes here


Then there is the same umlaut that Albanian has as per wiki:

Albanian: back vowels are fronted before i in the following syllable.

Spoiler:

Code: Select all

[B] > [F] / '_ & _([C])([C]).([C])([C])i  

Example word: (City name) <Mursapolis> [mru.sa.po.ris] no changes here either

After that follow some more vowel changes, mostly self-motivated. [e] is raised if stressed and otherwise lowered, and the lowered outcome splits into short and long [a] depending on whether it's in a closed or open syllable, respectively.

The reverse happens with [ɔ], which is raised unless stressed, and otherwise lowered. [a] is rounded and backed after labial consonants.

Spoiler:

Code: Select all

e > i / '_
e > ɛ

ɛ > a / _[C]
ɛ > a: / _., _# 

o > u ! '_

o > ɒ

a > ɒ / [m,p,b,w](')_

Example word: (City name) <Mursapolis> [mru.'sa.po.ris] > [mru.'sa.pu.ris]

The next changes first add a bit more Balkan flavour, and then all unstressed short [a] is deleted. Then i is lengthened if stressed, and otherwise I made it "palatalise if appropriate" (i.e. I just changed it to j, and will then check manually if the resulting cluster sounds like it would be palatalised or if I should have a short i pronounced there...).

Spoiler:

Code: Select all

u > vu / [V](.)_

w > v

-a ! '_, _:

'i > 'i:
i > j ! _:

Example word: (City name) <Mursapolis> [mru.'sa.po.ris] > [mru.'sa.purjs]

The next step is more of a revolution then a simple change. Vowels are first lengthened in all open syllables, and then all short vowels are deleted. Then the glottal stop and fricative and stop also disappear (in coda those had mostly prevented lengthening of the vowel of the syllable).

Spoiler:

Code: Select all

-. ! _*[V]
[V] > [V]: / _., _#,_[vc].


-ʔ
-h

-[V] ! _:,'_
-:

Example word: (City name) <Mursapolis> [mru.'sa.purjs] > [mru.'saprjs]


Then, all the palatalised consonants result in mostly postalveolar consonants, and [r] also undergoes more changes.

Spoiler:

Code: Select all

sj,tj,dj > ʃ,tʃ,dʒ

r(.)j > ʃ
r > t / _#
r > n / _.

k(.)j,g(.)j > c,ɟ

Example word: (City name) <Mursapolis> [mru.'saprjs] > [mru.'sapʃs]

This cluster, at least more pronounceable than the last one, still makes me consider having a general rule that clusters with the same place of articulation be simplified. Or I could do that as a sporadic change only. Opinions?


Then there is still regressive voicing assimilation, cf. Hungarian.

Spoiler:

Code: Select all

[vc] > [vcl] / _(.)[vcl]
[vcl] > [vc] / _(.)[vc]

Example word: (City name) <Mursapolis> [mru.'sapʃ(s)] No changes here, this is the final form.


It's a start I guess, and I have also got some ideas about which grammatical changes to implement from Old Ruritanian to Modern Ruritanian. I will still see about how everything will be realised exactly, where to put epenthetic vowels, when to vocalise consonants etc. Still, even then the end result is only a raw draft, and I will probably add some more changes to get to a nice phonology I want for the "modern" stage of the language (19th century).
What will eventually come out will look different to what I'd originally planned, but similarities will remain, and it will be fun to get there


EDIT: I have added a sample word - will add more as I go on!