Gada language - a isolate in Sweden

A forum for all topics related to constructed languages
User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gadish - a isolate in Northern Sweden

Post by Omzinesý »

A new attempt on adpositions

Adpositional phrases can appear as 1) preposition noun+NOM, 2) noun+GEN postposition, or 3) preposition noun+NOM postposition.


1) preposition noun+NOM

Location prepositions
Is 'at'
Aj 'from'
Ha 'to'
Ran 'through'
Prepositions of location are usually used in generic contexts.

Other prepositions
Pejja 'without'
Etc.


2) noun+GEN postposition

Heza 'inside'
Hezi 'into'
Hezta 'from the inside'
Hezti 'through the inside'

Dossa 'on the surface'
Dossi 'onto the surface'
Dosta 'from the surface '
Dosti 'over the surface '

Erpa 'near X '
Erpi 'to the area near to X'
Erpta 'from the area near to X '
Erpti 'through the area near to X '

Etc.


3) preposition noun+NOM postposition.
The postposition can also be seen as an adverb. It can also appear somewhere else than after the noun if there is the preposition too.

Is X heza 'inside'
Ha X hezi 'into'
Aj X hezta 'from the inside'
Ran X hezti 'through the inside'

Prep + Postp construction is used when
- a relative clause follows the noun
- the location is emphasized/focused


Postpositions can also be used as pronouns. They are formed with the same -u as definite nouns.

Hezu 'in it'
Hezuj 'into it '
Heztu 'from its inside '
Heztuj 'through its inside '

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gadish - a isolate in Northern Sweden

Post by Omzinesý »

A attempt for a new vowel system. It has some Germanic feel.

y:, i:, u: <ü, i/ij, o>
ʏ, ɪ, ʊ <y, i, u>
ø:, e:, o: <ö, e, å>
ɐ, ɑ: <ä, a>


The long vowels only appear in open syllables or stressed, final syllables with only one coda consonant. In unstressed syllables they though lose their length and differ from the sort ones only by quantity.
The short vowels appear in all closed syllables and in all stressed syllables.

So the short and long vowels can both appear in stressed open syllables (or stressed, final syllables with only one coda consonant).

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gadish - a isolate in Sweden

Post by Omzinesý »

I revive this project. It's old enough so I can redo it completely.

I think southern Sweden could be a better place actually.
Pitch accent, uvular r and velar fricative are more natural there.
Of course, the language having saved there is more improbable, maybe I can ignore it.

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gadish - a isolate in Northern Sweden

Post by Omzinesý »

Omzinesý wrote:
15 Jul 2019 17:06
Consonants

p t t͡ɕ k <p t k/kj k>
b d g <b d g>
m n ŋ <m n ng>
f s x h <f s ch h>
v z ʝ <v z j>
l r~ɹ ʀ~ʁ <l r rh>

Voiceless plosives are aspirated word-initially and in the beginning of a stressed syllable. In those positions, the voiced plosives can be a bit less voiced.
/ t͡ɕ k/ are more clearly phonemes than those in Swedish. That is because the softening is blocked in verbs, especially.
/ŋ/ is always geminated.
Sound change /ʂ/ => /x/ is quite resent. The phoneme usually appears in Germanic/Swedish loan words that had *sk in the beginning. Gadaish still has consonant clusters /sk/ and /sj/ on syllable boundaries.
All consonants except /v z ʝ/ can be preaspirated in the coda position. It's analysed as an h+C cluster.
/r~ɹ ʀ~ʁ/ are fricatives rather than trills.
I am still struggling with the two "trills".
I'd like to have both appear in consonant clusters/Cɾ/ and /Cʁ/. If there was a sound change r => ʁ in the area, they should probably have merged in this lang as well.

Maybe /ʁ/ is marked with <r>, so I can have <ar> for /ɑ:/, a German style pronunciation.
/ɾ/ could derive from /d/, like the flap in English, and also be written with <d>. <pd, td, kjd, bd, dd, gd> could actually be an interesting orthographic feature.
But then the language should have gone through many changes reasently:
CVdV
CVɾV
CɾV
Reduction of first vowels is especially impbable, because they are usually stressed in the area.

User avatar
Aszev
admin
admin
Posts: 1381
Joined: 11 May 2010 05:46
Location: ꜱᴇ

Re: Gada language - a isolate in Sweden

Post by Aszev »

The change from [r] to [ʁ] started in the 18th or 19th century, so seeing as your speakers belong to a separate linguistic community, it's not unreasonable that they would keep their distinction. There were pockets of [r]-using dialects even in the South.

As a solution for the d>r thing, you could have stress on the second syllable, and then a reduction of the first syllable as a way of adapting into the Germanic stress pattern. Of course, such a change would need to be old, so it might not fit for that reason.

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

Aszev wrote:
29 May 2020 14:25
The change from [r] to [ʁ] started in the 18th or 19th century, so seeing as your speakers belong to a separate linguistic community, it's not unreasonable that they would keep their distinction. There were pockets of [r]-using dialects even in the South.

As a solution for the d>r thing, you could have stress on the second syllable, and then a reduction of the first syllable as a way of adapting into the Germanic stress pattern. Of course, such a change would need to be old, so it might not fit for that reason.
You are very right. That is not a good development either.
It seems both of them must be there before "[r] to [ʁ]" in Swedish. How boring.

I think I'll make some degree of Proto-Gada first.

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

There is a medio-passive voice in Gada. In monosyllabic verbs, it is marked by tones. Earlier medio-passive marker thus disappeared and the tone of a polysyllabic word is left.

fi ót 'to take'
fi òt 'to go'

fi klán 'to find'
fi klàn 'to be there'

fi lýn 'to speak'
fi ly`n 'to think' (to speak to oneself)

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gadish - a isolate in Northern Sweden

Post by Omzinesý »

Omzinesý wrote:
29 Jul 2019 11:36
A new attempt on adpositions
[...]
3) preposition noun+NOM postposition.
The postposition can also be seen as an adverb. It can also appear somewhere else than after the noun if there is the preposition too.

Is X heza 'inside'
Ha X hezi 'into'
Aj X hezta 'from the inside'
Ran X hezti 'through the inside'
[...]
I think I'm going to follow the strategy 3) but the postpositions are just relational nouns. 'Inside' is thus just "hez".

"is trasva" thus means 'in the house'. But the location can be specified "is trasvaj hez" 'inside the house" literally: 'in the houses's inside'

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

Now I have developed the sound changes of vowels much more thoroughly.

Vowels

Old Gada has the mere 5-vowel system with two diphthongs. Length is not phonemic.

/i/, /u/
/e/, /o/
/ɐ/
/ɐi/, /ɐu/


Development of stressed vowels

Middle Gada vowel inventory has eight phonemes. Length is not distinctive but conditioned by openness of the syllable. /ø:/ does not appear in closed syllables. Midlle Gada lacks a high back-vowel.

/y/ [y:][ʏ], /i/ [i:][ɪ]
/ie/, /uo/
/ø:/, /e/ [e:][e], /o/ [o:][o]
/ɐ/ [ɐ:][ɐ]

Modern Gada vowel inventory has twelve phonemes. Length is not distinctive. That of /y/, /u/, /ø/, /o/, and /e/ is conditioned by the coda. The phoneme is pronounced long before a weak coda and short after a heavy coda. Long vowel phonemes (those without short allophones), i:, ɑ:, yø, ie, and uo, can only appear in syllables with a weak coda (or in open syllables). Short vowel phonemes (those without long allophones), ɪ and ɐ, can appear in all syllables.

/y/ [y:][ʏ], /i:/, /ɪ/, /u/ [u:][ʊ] <y>, <ij>, <i>, <u>
/yø/, /ie/, /uo/ <yø>, <ie>, <uo>
/ø/ [ø:][ø], /e/ [e:][e], /o/ [o:][o] <ø>, <e>, <o>
/ɐ/, /ɑ:/ <a>, <aa>


Between Old and Middle Gada, diphthongs are monophthongized, /e/ and /o/ are diphthongized when the syllable is open, /u/ is fronted to /y/, and vowel length is conditioned by openness of the syllable and high vowels gain near high allophones in closed syllables.

Between Middle and Modern Gada, a new /u/ is gained, conditioning of length (of some vowels) is changed to weakness of the coda and allophones of /ɐ/ and /i/ become phonemic, and some vowels are rounded in some morphological forms and some word-final vowels disappear.


An open syllable has no coda. A closed syllable has a coda.
A heavy coda has at least two consonants and at least one of them belongs to the preceding syllable. All other codas, also lack of one, are weak. So, open syllables are seen as a subtype of weak codas.

Code: Select all

Old Gada		Middle Gada	Modern Gada 
i (open syllable) 	=> i:		(weak coda) = i: 
i (closed syllable) 	=> ɪ 		(all codas) = ɪ
e (opens syllable)	=> ie		(weak coda) = ie
e (closed syllable)	=> e 		(weak coda) = e:  
					(heavy coda) => e
ɐi			=> e:  		(weak coda) => e:
ɐ (open syllable) 	=> ɐ:		(weak coda) => ɑ:
ɐ (closed syllable)	= ɐ		(all codas) = ɐ
u (open syllable)	=> y:		(weak coda) => y:
u (closed syllable)	=> ʏ		(weak coda) => y:
					(heavy coda) => ʏ, sporadically ø
eu 			=> ø:		(weak coda) => ø:
o (open syllable)	=> uo		(weak coda) = uo
o (closed syllable)	=> o		(weak coda) => u: 
					(heavy coda) => ʊ, sporadically o 
ɐu 			=> o: 		(weak coda) = o:
Edit: It'spossible that I take off ai => e: It is not needed for Modern Gada.
Other sound changes in Modern Gada
/i/, /ie/, and /e/ are rounded before /u/, like /au/, yielding /y/, /yø/, and /ø/. That mostly happens when definite forms of nouns are formed, and thus applies stressed syllables only if they are word-final and open.


Development of unstressed vowels

Between Old Gada and Middle-gada
ɐ / _# => Ø (i.e. word-final ɐ disappears)
(C)VCV.CV(C) => (C)VC.CV(C) (i.e. deletion of open middle syllable), sometimes blocked
i, u /_C => ɪ, ʏ (.i.e. non-word-final high vowels become near high)

Between Middle Gada and Modern Gada
o, u / _# => ɐ (i.e. word-final o and u merge and become ɐ)
i/ɪ, e, ɐ _# => ʏ/y, ø, o (in formation of Definite forms of Nouns)
Word-final /u/ sporadically appears


In unstressed syllables, Modern Gada thus has seven vowels.

/y/ [y][ʏ], /i/ [ɪ], /u/ [u[ʊ]
/ø/, /e/, /o/
/ɐ/
Last edited by Omzinesý on 04 Jun 2020 15:03, edited 1 time in total.

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

ih 'in, on, at' (location)
ha 'to' (goal)
aj 'from'
haaj 'through, via'

ez 'with'
pejja 'without'

ha traasvaj myb 'towards the house' ~ 'to the direction of the house'
aj traasvaj myb 'from the direction of the house'


I'm also changing the verb meaning 'to have' to liez, which is historically li 'to be' ez 'with'

I'm still considering the distinction between Dative case and preposition ha. But at least now, I think ha codes goal and Dative codes recipient only.
Last edited by Omzinesý on 07 Jun 2020 15:32, edited 1 time in total.

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

Habeo construction and predication of nouns of characteristics

In Old Gada, "Adjectives" were pure nouns. They were abstract and uncountable.
sakaun 'beauty'
tar 'length'

They appeared in Genitive-Dative to modify nouns.
sakauni billo 'a beautiful girl' (girl of beauty)
tari billo 'a tall girl' (a girl of length)

To predicate a characteristic, one has to use Habeo Construction, which is formed by the possessor in Genitive-Dative and the copula.
Billoj il skaun. (The girl is beautiful.) (The girl has beauty.) (To the girl [there] is beauty.)
Billoj il tar. 'The girl is tall.' (The girl has length.) (To the girl [there] is length.)


In Modern Gada, nouns of characteristics still modify nouns as genitive modifiers.
skauni billa 'a beautiful girl' (girl of beauty)
tari billa 'a tall girl' (a girl of length)

Nouns of characteristics can also be countable and name an object with that characteristic.
gi skån, skånva 'a beauty, the beauty'
gi tar, tarva 'a tall person, the tall person'

They can also appear in copula clauses. They have an article, however.
Billå rha gi skån. 'The girl is beautiful.' (The girl is a beauty.)
Billazy garha skånzi. 'The girls are beautiful.' (The girls are beauties.)
Billå rha gi tar. 'The girl is tall.' (The girl is a tall one.)

An alternative way is to use nouns of characteristics in Habeo Construction. It differs from that of Old Gada in that it has a transitive habeo-verb lez 'to have'. It derives from old copula verb il 'is' sis 'was' and preposition ez 'with'.
Billå lez skån. 'The girl is beautiful.' (The girl has beauty.)
Billå lez tar. (The girl has length.)

Adverbials can be formed with preposition ez 'with'.
Y pehni ez skån. 'S/he sings beautifully.' (S/he sings with beauty.)

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

Aszev wrote:
29 May 2020 14:25
As a solution for the d>r thing, you could have stress on the second syllable, and then a reduction of the first syllable as a way of adapting into the Germanic stress pattern. Of course, such a change would need to be old, so it might not fit for that reason.
How probable would d => r be as the onset of a stressed syllable?
Maybe through d => ð => r
Spanish d => ð, at least, doesn't care about stress.

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

Omzinesý wrote:
04 Jun 2020 11:38
ih 'in, on, at' (location)
ha 'to' (goal)
aj 'from'
haaj 'through, via'

ez 'with'
pejja 'without'

ha traasvaj myb 'towards the house' ~ 'to the direction of the house'
aj traasvaj myb 'from the direction of the house'


I'm also changing the verb meaning 'to have' to liez, which is historically li 'to be' ez 'with'

I'm still considering the distinction between Dative case and preposition ha. But at least now, I think ha codes goal and Dative codes recipient only.
'with' the comitative meaning is expressed with word haaz 'company'. Its etymology is the same as Finnish kanssa, the same Germanic word as Hansa alliance. It is a noun whose semantic head is expressed as its possessor.

ih ozzi-zy-j haaz
in friend-PL.DEF-GEN company
'with the friends.'

It can also appear with other prepositions:

ha ozzizyj haas 'to the friends' (to their company)
aj ozzizyj kaas 'from the friends'

It's possible that ez 'with' is a more used in Old Gada and nowadays only appears to form adverbs.
lez 'has', sisez 'had' is also lexicalized.

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

Gada has two or three (depending on how you count them) accents (word intonations).

The terms "Acute", "Gravis" are used in Swedish grammars, so I adopted them. They are not the most intuitive ones for Gada (or even for Swedish).

Gada accents resemble those of Southern Swedish. That is they have only one peak per word.

Different accents are not marked in Gada orthography unless there is an evident problem with inferring the meaning from the context.

"Acute Accent"
The tonal peak appears on the first more of the first syllable. Because typical stressed syllables in Gada, like in Swedish, are at least two morae, Acute ends up being a slightly falling tone one the stressed syllable. The other syllables have a low/neural tone.
Words that were monosyllabic in Old Gada have Acute Accent.

ot /óot/ 'to take' (<= OGada aut)
suhk /sʊ́hk/ 'home' (<= OGada sok:)

"Gravis Accent"
The tonal peak appears on the second mora of the first syllable. Because typical stressed syllables in Gada, like in Swedish, are at least two morae, Acute ends up being a slightly rising tone one the stressed syllable.
Words that were polysyllabic in Old Gada have usually Gravis.

òt /oót/ 'to go' (<= OGada auta, where -a is Old Gada Middle-Voice marker)
suhk /sʊh́k/ 'to live somewhere)' (<= Ogada sokka)

"Second-Syllable Accent"
Second-Syllable Accent can be seen as an instance of Gravis. It also has its peak on the second mora, but the first syllable having only one mora, the second mora appears in the second syllable.
In Old Gada, those words had their stress on the second syllable and the first syllable was short, open and unstressed. In Modern Gada, many of those initial syllables are reduced, but some still persist, often along a reduced variant. The most productive type of words with Second-Syllable Accent are plural forms of verb.

tisi /tisí/ 'baby'
gagha /gaʁá/ 'they are'
tule /tulé/ ~ tlie 'father'

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

Modern Gada has the three cases mentioned above: Nominative, Dative-Genitive, and Essive.

It has the three basic prepositions: ih (location), aj (source), and ha (direction).

The word order is genitive/adjective noun, which implies to older SOV word order.

Old Gada thus should have had more postpositions or case suffixes. The locational nouns in Modern Gada do not however have remarkable case inflection.

Are ih, aj, and ha old postpositions that have just moved to prepositions or could they have been gained some other way or even borrowed (ih could well derive from Germanic in)?

Ideas?

User avatar
Omzinesý
runic
runic
Posts: 2830
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Gada language - a isolate in Sweden

Post by Omzinesý »

Number system is still not ready. Just a note for myself

1 gij
2 sat
3 pos
4 jaki
5 bø

Post Reply