Good to know. Thanks!Tristan Radicz wrote: ↑25 Sep 2018 21:10A small note about positional palatalisation: in South Slavic it correlates with the distinction of yers' reflexes - in Bulgarian (sans part of the Western dialects) and Macedonian, the two reflexes are different (front and back respectively) and there is palatalisation in position before front vowels (lost in Macedonian), while Western South Slavic merged the two yer vowels and bears no traces of such palatalisation.
Below I am slowly adding to a huge list of features that my Pannonian language will share. This list will be expanded upon periodically.
- t and d are lost before l as in most South Slavic, including Central Slovak.
- The desinence of the Slavic nasal front vowel is "broad" (likely [æ] or [ɛ]). It will likely merge with /a/ before a palatal consonant, otherwise it becomes /e~ie/.
- The desinence of the Slavic nasal back vowel is "narrow." It's highly likely based on the evidence from Slavic loanwords into Hungarian that the Pannonian value was [ũ] (sorry, Slovenes!)
- The non-palatalized syllabic l becomes /u/, merging with the outcome of the nasal back vowel.
- The palatalized syllabic ʎ, as with ʎ in general, depalatalizes fully to l.
- The mid-high vowels eː and oː have diphthongized to ie and uo respectively.
- The pronunciation of /v/ varies from [v] to [ʋ] to [w]
- Stress is retracted from a final short vowel. Slovak and Czech have fixed stress on the initial syllable, and so to does some northern Stokavian dialects near the Hungarian border. I have decided to follow this route. Uniquely, length is preserved despite tone being lost in its place.
- The desinence of the instrumental singular -a stems is simply -u, from earlier -ǫ
- The neuter ending is long
- A productive diminutive suffix is -ek. This is likely due to Hungarian influence, which coincidentally has the same suffix.
- Instead of a morphological passive, the third person reflexive is used in its place.
- Pannonian's negative past tense deviates syntactically from standard WSS: "sem ne" is heard over "nisam."
I've started going past sound changes and toying with some of the noun declensions in Pannonian. Here's an example:
Code: Select all
NOM: -∅ ACC: -∅ GEN: -a DAT: -ü LOC: -ü INS: -em
(Small orthography note: acute accents mark long vowels and diareses mark the fronted versions of back vowels.)
Code: Select all
NOM: hrád ACC: hrád GEN: hrada DAT: hradü LOC: hrádü INS: hradem
Code: Select all
NOM: hradi ACC: hradi GEN: hrád DAT: hradem LOC: hradiéh (uncertain) INS: hradi
Code: Select all
NOM: hradövi ACC: hradöve GEN: hradöva DAT: hradövem LOC: hradöviéh (uncertain) INS: hradövi
Shortening of auslaut vowels. Posterior to the retraction of final stress and the fall of the yers is a general shortening of auslaut vowels, similar to the processes in Czech and Hungarian. This happened around the latter half of the 13th century.
Merger of /x/ and /ɦ/: A sound change that occurred around the 15th century is the merger of /x/ and /ɦ/ (from /g/) into the phoneme /h/. This sound change follows Hungarian around the 14th century which changed /x/ to /h/. Then, /ɦ/ de-voiced to /h/.
Lateral chain shift: First, non-geminate /l/ (which was probably [ɫ]), becomes /u/ before consonants (including in between two other consonants.) This causes a chain shift, depalatalizing /ʎ/ to /l/. This means Pannonian lacks /ʎ/ as a phoneme, similarly to Hungarian.
The velar consonants /k/ and /g/ become /c/ and /ɟ/ before /ɲ/. Therefore, the word for "book" is [ˈcɲihɑ]. Likewise, /s/ and /z/ become close to postalveolar [ʃ] and [ʒ]. Apparently this sounds quite "redneck," which means I've succeeded. :p
Finally, a note on orthography: I'm going to employ two (technically three) orthographies: the first is a gaglolitic orthography in use mainly by the clergy. The second orthography is a Croatian inspired one, except /c/ and /ɟ/ are represented by <cj> and <gj> respectively. The third orthography is something of a mix between Hungarian and Old Czech, whereas <s> is /s/, <sz> is /ʃ/, <zs> is /ʒ/, <ny> is /ɲ/, and so on. In both orthographies, long vowels are represented by acutes, front rounded vowels by diaereses, and long front rounded vowels by double acutes (the so called "hungarumlaut.)" Ill try my hand at translating the "Hej, Sloveni" first, and see which orthography works the best.
I'm somewhat confident I can answer this now. The stress system of Pannonian is characteristically South Slavic. Like the language of the Kiev leaflets and dialects of Slavonian Croatian, Pannonian had retracted stress from a final syllable to the initial syllable, but left intermediate syllables alone. This is evident in the language's autonym, Panúski [pɐˈnuː.ski], where stress remains on the second syllable. Nonetheless, northern dialects of Pannonian (i.e. OTL Sopron/Ödenburg) have indeed fixed stress to the initial syllable, much like their German and Slovak neighbors.
Unlike its South Slavic neighbors to the west and south, Pannonian had lost almost all traces of tone. Some remnants of the previous tonal system can be found in words with long vowels, however this is no longer opaque due to additional sound changes that altered vowel length and analogy. (For example, the long vowel in Panúski is from compensatory lengthening from the lost nasal consonant in Proto-Slavic *Panǫnski.) I'll reserve another post for Pannonian length when the analogy is wrought. Zekoslav had gratefully obliged to help me code some of the tougher Pannonian accent changes. Once that is done, I will manually go through each declension and apply analogy where needed. This is especially necessary due to the unpredictable alternations in vowel length that my changes have caused.
I wonder whether a romlang set in Bohemia is realistic, and how to justify its internal historyÆlfwine wrote: ↑25 Aug 2018 22:59Or the nasal consonant could suggest an uncoupling. Although the length of Hungarian contact isn't too long, merely that a lot of changes happened in a short time span.Zekoslav wrote: ↑25 Aug 2018 18:35Concerning Vencsellő: another possibility is that it was borrowed at a sufficiently early date for the local language to preserve a later depalatalised /tɕ/ ~ /tsʲ/. The preservation of the nasal vowel does suggest early borrowing.Ælfwine wrote: ↑25 Aug 2018 17:16That's possible, I thought that the town of Vencsellő might've been too far to be affected by a South Slavic change.
Alternatively, I thought we could do something like Hungarian and have the outcome of these clusters be /cç/ and /ɟʝ/ respectively, but I thought that would've been too "archaic" for a non-peripheral area.
I'm guessing our length system will be long on accented syllables, short on non-accented ones (or something like that?)
Concerning length and accent: natural Slavic languages are, as usual, much more complicated than that: preservation of length depends on tone, the number of syllables in the word, distance from the accented syllable, and the same can be said for lengthening of short vowels (which additionally tend to be sensitive to consonants closing the syllable, if these exist). With respect to the exact rules, Czech, Slovak and South Slavic each have different rules.
I was thinking about how Pannonian and Pelsodian could influence each other when it comes to accent, and I got some interesting ideas. But, to see how feasible they are, I'd have to know what the distribution of accent in Pelsodian is: Do proparoxytones exist? How many final vowels are preserved?
All of this is looking very promising, so keep up the good work!
Hrrm. Very well then. I'd like to see your ideas.
I've taken it as an areal feature that would lose unstressed /i/ and /u/ word finally, while preserving /a/. Though concerning stress, I briefly had the idea of syncoping pretonic unstressed vowels in certain environments, shifting stress to the front of many words. Had I placed this romlang in the Czech Republic proper (Bohemian), that's certainly something I would have done. Regardless of what happens, I think the situation with stress would become more chaotic and unstable, as some words remain proparoxytone, some do not. Much like Russian in that case.
Doesn't Hungarian have /ʎ/ (ly)?Ælfwine wrote: ↑16 Jan 2019 05:34Lateral chain shift: First, non-geminate /l/ (which was probably [ɫ]), becomes /u/ before consonants (including in between two other consonants.) This causes a chain shift, depalatalizing /ʎ/ to /l/. This means Pannonian lacks /ʎ/ as a phoneme, similarly to Hungarian.
Hate to break you to this, but it seems Jan von Steerbergen beat you down to it:
As Shimo said, Hungarian <ly> is /j/. For the sound changes in Pannonian, I recall some Kajkavian dialects that do something similar.
Just assume the marcomannic wars were successful.
No, Prekmurje Slovene is.
That's similar to standard Slovene; IIRC some Hungarian dialects also pronounce <ly> as /l/, but Hungarian <ly> is mainly from loanwords, isn't it? (Native *lj mostly changed into <gy>)
For my R-Turkic conlang, it seems that *lj was uncommon in Old/Common/Proto-Turkic and I'm hesitating whether to introduce /ʎ/ into it by borrowing.
Proto Turkic *kȫĺek became Hungarian kölyök (Wiktionary), so /ʎ/ might be plausible even in native vocabulary.
This aside, *jẹ̄miĺč became Hungarian gyümölcs. The /l/ was lost in both Oghuz languages and Chuvash (Turkish yemiş, Chuvash śimĕś), so your conlang (what's its name again?) might preserve the /l/ in such clusters.
By the way, how was *ạ *ẹ *ö *ü probably pronounced? Did Proto-Turkic have an ATR vowel harmony like Tatar and Bashkir, or a palatal harmony like Turkish? It seems that *ạ *ẹ were perhaps diphthongs *aɨ *ei?