Cheyadeneen Scratchpad

[evmdbm]

Ok - so this thread needs to be read alongside the Vedreki thread - (http://verduria.org/viewtopic.php?f=3&t=395) or some of the worldbuilding stuff will make much less sense. The languages are different enough that you will be able to follow the grammar of Cheyadeneen in isolation though.

Cheyadeneen is spoken by the majority human population of the Empire of Cheyaden, some 233m people, who together with 167m Vedreki and approximately 40m speakers of Naxaqeen languages make up the total population of the Empire of Cheyaden. It is unrelated, but has extensive borrowings – particularly in the area of artistic and musical vocabulary - from the neighbouring Azdrai language of Vedreki. The calendar was also borrowed from the Vedreki (anyone interested in the calendar?). In turn I anticipate much legal, commercial and governmental vocabulary going the other way from Cheyadeneen to Vedreki. Cheyadeneen is related to the language of Korkeltese, spoken by the human population of the island of Korkelt in the Eastern Ocean, and more distantly from the language of the minority human population in the Tarkentian Empire – Tarnaqi. The standard version of the language described here is the language of the imperial capital, Cheyaden itself, although dialectal variation exists.

Phonology:

Consonants	bilabial	labio-dental	alveolar	velar	post-alv	retroflex	glottal
stop	p,b		t,d	k,g		th,dh
fricative	f,v		s,z	kh	sh		h
affricate					ch, j
approximant	w	r, l
nasal	m	n

Note the retroflex stops, not found in English (but similar to those found in Hindi) and the lack of the uvular series found in Vedreki.
Note the voiced alveolar fricative and the lack of the voiced velar fricative in contradistinction to Vedreki

“r” is generally left unpronounced in word-final positions or just before other consonants. Standard Cheyadeneen is (in contradistinction to standard Vedreki) a non-rhotic dialect in that sense.

As with Vedreki, voice assimilation both progressive and regressive occurs eg a voiced obstruent (stop or fricative) is unvoiced prior to a following unvoiced obstruent. So for example “beg” (six) becomes bekko (sixth) on the addition of the ordinal affix -ko

Unvoiced initial stops are aspirated; elsewhere unaspirated (a rule shared with Vedreki).

“y” and “w” are sometimes included in the consonant table, but “w” is arguably an allophone of “u” and consonantal “y” an allophone of the vowel “y”.

Cheyadeneen is less tolerant of clusters than Vedreki and a number of examples of syncope occur. For example where two stops occur together followed by an approximant the middle stop is deleted. –kwe is the affix indicating “A person who does…” Soldier is therefore kat (war) plus -kwe. The “k” is lost to form katwe (soldier/warrior)

vowels	front unrounded	central unrounded	back unrounded	front rounded	back rounded
high	i		y		u
mid	e		o
low		a

Cheyadeneen has a six vowel system – the main differences with Vedreki being the high back unrounded “y”, the more central location of the “a” and the more closed “mid” position of the “o” as well as the lack of front rounded vowels. There are four diphthongs which are “ei” as in “bait” (spelt “ey” before another vowel), “au” as in “cow”, “ai” as in “aisle” – again spelt “ay” before a vowel, “oe” as in “boy”.

Doubled or long vowels will be marked by a circumflex eg â, but are not common.

Under the influence of Vedreki we find that Cheyadeneen has become a syllable-timed language.

Phonotactics:

Generally each syllable should receive roughly equal stress, consistently with the syllable-timed nature of the language. The maximal syllable structure is (C)(C)V(C)(C)
Velk (fire) CVCC
Dha-hal (blue) CV-CVC
Ha-zak-che (emperor) CV-CVC-CV
Wekh-la (long) CVC-CV
Sten-e CV-V

Syllables will therefore split between consonants. The first syllable of the word will usually, but not always, be of C(C)V structure. We can set out the constraints as

(p, b, k, g, f, v,) (l, r) + V + (r, l) + (Stop)
(s) + (t) + V + (r, l) + (Stop)
C + V + C
C + V + + (r, l) + (Stop)

[Pedant]

Interesting...
So, what’s Cheyaden actually like, as an empire?

[quinterbeck]

“r” is generally non-rhotic, unlike Vedreki

Rhotic is basically a fancy word for r-like so I have no idea what you mean by this. How can r be not like r?*

As with Vedreki, voice assimilation both progressive and regressive occurs eg a voiced obstruent (stop or fricative) is unvoiced prior to a following unvoiced obstruent. So for example “beg” (six) becomes bekko (sixth) on the addition of the ordinal affix -ko

Are there rules for when voice assimilation is progressive or regressive?

*wow deep

[evmdbm]

Rhotic is basically a fancy word for r-like so I have no idea what you mean by this. How can r be not like r?*

Do you pronounce r in park? If so, you speak a rhotic dialect of English. If not, you speak a non-rhotic dialect. That's it, as I understand things (https://en.wikipedia.org/wiki/Rhoticity_in_English. It's the basically the same here. When r is at the end of a Cheyadeneen word you generally don't pronounce it. I suppose I could just have said word-final r and r before another consonant are not pronounced, but I didn't

Are there rules for when voice assimilation is progressive or regressive?

Probably... eventually... it's an iterative process as I come up with more words - and there aren't many yet.

Interesting...
So, what’s Cheyaden actually like, as an empire?

Well some of this is under the Vedreki scratchpad - the map and so on. But the empire is centred on the Cheyadeneen archipelago which is a volcanic chain of islands at the edge of the North Erelese, Khardan and Cheyadeneen plates in the Inner Sea, bordered to the south by the continent of Erelan, north west by the southern end of the continent of Khardan and north east by the continent of Naxaq (or Nashak in Cheyadeneen). The main human population centres, which include the capital Cheyaden are in the west and south of the largest island, also called Cheyaden. The empire also extends along the western seaboard of Naxaq. While the Cheyadeneen and Vedreki live cheek-by-jowl in the same Empire I want to keep the languages in separate threads, but I do need to say something, I guess here. Vedreki aren't human and live in large-ish bands and are divided into five castes. Rule of the Vedreki therefore comes under the Council of Cities; there are six cities - Black City (Moris Udrek), White City (Arven Udrek), Red City (Krad Udrek), Green City (Ikat Udrek), Blue City (Lun Udrek) and Yellow City (Tilpit Udrek) and while each individual city council has five noble caste bands and two warrior caste bands the Council of Cities has one (noble) representative from each city and a chairman. Rule is oligarchical and the Council has sole power to make laws affecting solely the Vedreki.

The Cheyadeneen by contrast live in nuclear families and have half a democracy. I say half, because while they have universal suffrage, elections to a unicameral Parliament, and an elected Prime Minister with control over domestic policy, the emperor keeps tight control over foreign and defence policy, the security forces and the military - nothing exercises a ruling imperial royal family like staying the ruling imperial royal family! The emperor also has the ultimate say over anything that affects both human and Vedreki communities - environmental law or conflict of laws rules relating to human-Vedreki relations. Those imperial level laws are enacted via a decree of the imperial council (essentially the equivalent of the British Privy Council) but there will be prior consultations with both the Cheyadeneen Government and Parliament and the Council of Cities. The imperial council is basically the Chairman of the Council of Cities, Prime Minister and whoever else the emperor sees fit to appoint, but usually amounts to a securocracy (heads of the intelligence services, army, foreign and defence ministries) alongside a shifting organisation of other ministry heads - environment, inter-species relation etc. The current emperor, Tamukhis XIII has though delegated much of the day-to-day running of things to the Crown Prince, Khoser and he is currently exercised by the liberal human obsession of the moment - should we impose abolition of caste on the Vedreki. For caste see http://verduria.org/viewtopic.php?f=3&t=395#p15824

Vedreki politics tends to revolve around which families (bands) are up or down and manoeuvring for spots on the various councils as they open up. Human politics revolves around the typical political parties - from the royalist (and peculiarly named) Crimson Square Party (crimson being the imperial colour) through the equally right-wing, but more libertarian and newer Nationalist Party to the left wing Progressive Party or Social Action Party and the (surprise surprise) banned Republican Action Party, which mounts occasional terrorist attacks.

[quinterbeck]

Do you pronounce r in park? If so, you speak a rhotic dialect of English. If not, you speak a non-rhotic dialect. That's it, as I understand things (https://en.wikipedia.org/wiki/Rhoticity_in_English. It's the basically the same here. When r is at the end of a Cheyadeneen word you generally don't pronounce it. I suppose I could just have said word-final r and r before another consonant are not pronounced, but I didn't

Right, I see. That's quite a specialised use of the term! And you would usually see it used to describe accents of English rather than the phoneme itself. Crosslinguistically, rhotic simply describes a sound as being r-like (https://en.wikipedia.org/wiki/Rhotic_consonant) e.g. in French, ʁ is a rhotic.

[mèþru]

That article used to include examples of such dialectal variation in other languages; it's not just an English phenomenon

[Salmoneus]

No, but it's a property of languages, not of phonemes. "Non-rhotic" just means "isn't R" (used of phonemes) or "doesn't pronounce the R" (of dialects).

What evm appears to be talking about is some sort of sandhi rule, which should probably go in its own section...

[evmdbm]

No, but it's a property of languages, not of phonemes. "Non-rhotic" just means "isn't R" (used of phonemes) or "doesn't pronounce the R" (of dialects).

What evm appears to be talking about is some sort of sandhi rule, which should probably go in its own section...

I'm not talking about how to pronounce "r" - whether it's a trill or an approximant or whatever - so yes it is different. That said even assimilation seems like sandhi (not that I really think I understand sandhi). syn-pathy = sympathy is apparently an example of internal sandhi https://en.wikipedia.org/wiki/Sandhi. On that basis beg= six, beg-ko = bekko (sixth) is internal sandhi too.

Different topic. I have been playing around with alphabets. A question. Cheyadeneen has a number of consonants and vowels not found in Vedreki - the retroflex stops th and dh; affricates ch and j, along with h, w and y. Vedreki has the voiced velar fricative gh, and uvular obstruents q, and qh not found in Cheyadeneen along with its front rounded vowels also not found in Chayadeneen. Is a common alphabet plausible? I did have an idea that the Vedreki might adopt the Cheyadeneen alphabet or the other way round - haven't decided which way yet.

[Xwtek]

Is a common alphabet plausible?

Absolutely. Latin has been used to write a language with extremely diverse phonology from English, to Spanish, to Yoruba, to Indonesian, to Vietnamese, to Nahuatl, etc. Like Latin, you have to devise digraph or diacritic.

[evmdbm]

Cheyadeneen can be considered as a generally SVO language (or maybe AVO is better given the ergative alignment) and many of the corollaries are also found: prepositions rather than postpositions, postnominal genitives etc. Word order is more fixed in Cheyadeneen than in Vedreki because of the way in which a verb, as we will see, can for instance take two absolutive arguments. I'll do verbs next.

Cheyadeneen nouns come in two numbers: singular and plural and have no grammatical gender. There are four cases in an ergative-absolutive alignment. These are ergative, absolutive, dative and genitive. The ergative, dative and genitive suffixes are common to all nouns with a couple of additional rules – if the absolutive (unmarked) singular ends in a vowel which is different from the suffix vowel an epenthetic “y” is inserted; if the same, only the consonant is added. No absolutives end in “y”, although some follow the replace “i” with “y” rule before another vowel. There is no definite or indefinite article. Adjectival declension, allowing for case and number agreement is identical.

Singular

Abs	Erg	Dat	Gen
zilda (star)	zildat	zildak	zildassen
che (king)	cheyat	cheyak	cheyassen
velk (fire)	velkat	velkak	velkassen

Plural

Abs	Erg	Dat	Gen
zildayun	zildayos	zildayuk	zildayussen
cheyun	cheyos	cheyuk	cheyussen
velkun	velkos	velkuk	velkussen

Adjective comparisons are formed with –ai for the comparative and –is for the superlative.
Wekhla = long; wekhlai = longer; wekhlis = longest

Chegabai wekhlai ta Sevegabai
Kings river-abs, long-comp-abs sg than North river-abs
The Kingsriver is longer than the North river

Chegabai wekhlis gabai ve Cheyadenak
Kings river-abs, long-super-abs sg, river-abs-sg, in Cheyaden-dat
The Kingsriver is the longest river in Cheyaden

Ps. the verb to be has no present tense.

[evmdbm]

The Cheyadeneen verb inflects for tense, mood, aspect and voice; the anti-passive being morphologically marked through reduplication. Cheyadeen also has a separate periphrastic passive voice, meaning that it has three distinct voices plus the imperative. Unlike Vedreki which inflects for gender and number, Cheyadeen verbs have different inflections for each person and number, meaning that – as in Latin – pronouns are not normally included. Contrasting with Vedreki which at best only partially encodes aspectual information in the verb via periphrastic perfect tenses, there is a strict distinction in Cheyadeneen between aspects: the imperfective which indicates a continuous, habitual or uncompleted action and the perfective which indicates the completion of an action. There is only one present tense which is inherently imperfective and the future imperfective is formed with an auxiliary. This means that while the past perfect in Vedreki will pretty much always correspond to the past perfective in Cheyadeneen, the simple past in Vedreki might, depending on context, be translated as either past imperfective or past perfective. There are five indicative tenses and three subjunctive tenses. Since there is no future subjunctive it is redundant to refer to the past imperfective subjunctive; it is simply called the imperfective subjunctive. Questions are formed with a particle: vek, cha or chanut depending on whether the expected answer is neutral, yes or no respectively.

Infinitives end in
-az
-oz

Because we're going to be here a long time otherwise, this post will concentrate on the indicative. We'll do the subjunctive and the two non-active voices - anti-passive and passive in two future posts. The really interesting stuff that I might have wrong comes when we get to different voices and means of relativisation.

Present tense

Infinitive	sakaz - to run	soroz to drink
1ps	sakai	soroe
2ps	sakau	soru
3ps	sakan	sorun
1ppl	sakat	sorut
2pppl	sakada	soruda
3ppl	sakadat	sorudat

Past imperfective

infinitive	sakaz	soroz
1ps	sakia	soria
2ps	sakias	sorius
3ps	sakian	soriun
1pppl	sakiat	soriut
2pppl	sakiada	soriuda
3pppl	sakiadat	soriudat

Past Perfective

infinitive	sakaz	soroz
1ps	sakadai	sorusoe
2ps	sakadau	sorusu
3ps	sakadan	sorusun
1ppl	sakadem	sorusum
2ppl	sakad	sorusuda
3ppl	sakadet	sorusudat

Future Perfective (simple)

infinitive	sakaz	soroz	toraz
1ps	sakelmai	sorolmoe	torai
2ps	sakelmau	sorolmu	torau
3ps	sakelman	sorolmun	toran
1ppl	sakelmat	sorolmut	torat
2ppl	sakelmada	sorolmuda	torada
3ppl	sakelmadat	sorolmudat	toradat

The reason why I have added toraz (to be) in the future is because the future imperfective is formed with the future tense of "to be" and the infinitive

torai sakaz = I will run (yes, this is a blatant steal from Russian)

Personal pronouns are not often used, but in the third person the fact that the language has an ergative-absolutive alignment comes to the fore. Cheyadeneen is morphologically ergative, but has a nominative-accusative syntactic alignment, like Basque. Transitive verbs will always have a subject in the ergative and an object in the absolutive so

The man is drinking the water
Steneyat sorun vod
Man-erg sing drink-pres ind 3ps water-abs sing

Where a transitive verb drops the object the subject remains ergative in the active voice - but can be converted to absolutive in the anti-passive.
The man is drinking.
Steneyat sorun.
Man-erg drink-pres ind 3pd

In this formula you are stressing the fact that the man is doing something... as opposed to nothing.

Intransitive verbs, however, take an absolutive argument as their subject.
The man is running
Stene sakan.
Man-abs sg run-pres ind 3ps

The man died.
Stene nifusun
Man-abs sg die-past ind 3ps

Negation is indicated via the use of the word nut, meaning no or not.
The man did not die
Stene nifusun nut.
Man-abs sg, die-pres ind 3ps, not.

The man isn’t drinking water; he is drinking wine.
Steneyat sorun nut vod. Sorun zeth.
Man-erg sg, drink pres ind 3ps, not. Drink pres ind 3ps, wine-abs sg.

[evmdbm]

Subjunctive: The inflections are common to both conjugations so only sakaz is given.

The subjunctive is used in a number of situations. It is used in reported or indirect speech. “He said that…” and in subordinate clauses indicating wish or request. “I asked him to do X” It is also used in conditional sentences “If X were true…” It is used in purpose constructions. “I did it in order to…” and in first and third person imperatives.

I asked him to read the letter
Bopradai tahak yomen ulia
Ask-past ind perf 1ps, that, read-pres subj 3ps, letter-abs

I am ordering him to read it
Lurathai tahak yomen sat
Order-pres ind 1ps, that, read-pres subj 3ps, it-abs

In practice this is always done with the present subjunctive as use of a past subjunctive would imply the impossible – ie that you were ordering him to have already done it (if the perfective subjunctive).

I said that I would read it
Arusoe tahak yome sat
Say-past ind perf 1ps, that, read-pres subj 1ps, it-abs

I will say that I have read it
Arolmoe tahak yomomai sat
Say fut perf ind 1ps, that, do-perf subj 1ps, it-abs

Here different tenses are used depending on the relationship of the act reported to the time at which it is reported.

If he had been in Cheyaden he would have been safe
Tidh toman ve Cheyadenak toman arkas
If, be-perf subj 3ps, in, Cheyaden-dat sg, be-perf subj, safe-abs sg

The perfective subjunctive indicates the completed nature of the circumstances. Contrary to actual past fact (ie he was not safe…)

If he were to be in Cheyaden he would be safe
Tidh en ve Cheyadenak, en arkas
If be-pres subj 3ps in Cheyaden-dat sg, be- pres subj 3ps, safe-abs sg

If he were in Cheyaden he would be safe (but he isn’t and he’s not)
Tidh tomysten ve Cheyadenak, tomysten arkas
If be-imp subj 3ps in Cheyaden-dat, imp subj 3ps safe-abs sg

The present subjunctive indicates a possible future condition, whereas the imperfective subjunctive would indicate something contrary to present fact.

I went to Cheyaden in order to see the emperor
Idadai ke Cheyadenassen tahak blete hazakche
Go-past perf ind 1ps to Cheyaden-gen that see-pres subj 1ps emperor-abs sg

The present subjunctive indicates that relative to the time of the main action (here going to Cheyaden) the subordinate clause verb was in the future. Consequently the use of the imperfective or perfective subjunctive makes no sense in the context of a purpose clause.

Let’s go to Cheyaden
Idet ke Cheyadenassen
Go-pres subj 1ppl to Cheyaden-gen

Let him go to Cheyaden!
Iden ke Cheyadenassen
Go-pres subj 3ps to Cheyaden-gen

sakaz	present	imperfective	perfective
1ps	sake	saky	sakomai
2ps	sakes	sakyste	sakomau
3ps	saken	sakysten	sakoman
1ppl	saket	sakystet	sakomam
2ppl	sakeda	sakysteda	sakomada
3ppl	sakedat	sakystedat	sakomadat

[evmdbm]

As noted, other than the active voice, Cheyadeneen has both an antipassive and a passive voice. The antipassive deletes the absolutive object and “promotes” the ergative subject to the absolutive. It is formed morphologically by reduplication.

The man drinks
Stene sosorun
Man-abs sing drink-pres-ind AP 3ps

There is an emphasis difference here. In the antipassive the emphasis is on the unknown/generic/indefinite nature of what is being drunk. By contrast if we say Stenayat sorun we emphasise the fact the man is drinking.

- The king isn’t doing anything, is he?
Cheyat khulan emo chanut

- The king is reading.
Cheyat yoman.

Che yoyoman would have a different emphasis and given the question expects the answer that he’s not doing anything would seem peculiar. But it makes sense if the question were
- What is the king reading?
Khenar cheyat yoman
what-abs, king-erg sg, read-pres ind 3ps

- I don’t know, but he is reading (something)
Nut zinoe, e (che) yoyoman (emo)
Not, know-pres ind 1ps, but (king-abs sg) read-pres ind AP 3ps, (something-abs sg)

A second use of the antipassive, retaining the absolutive object, but changing the subject from ergative to absolutive case is to allow for relativisation. Cheyadeneen has a restrictive rule that only absolutive (unmarked) arguments can be relativized. Relative clauses are examined below.

The passive voice was developed late under the influence of the Vedreki passive. Essentially the Cheyadeneen saw that it had uses and developed something to fit the gap. The agglutinative nature of the Vedreki verb with the passive infix -Vgh- is difficult to fit into the Cheyadeneen inflectional structure, so it developed periphrastically with the auxiliary laraz, which separately means to become, and the past participle. This allows for the dropping of the ergative subject altogether and the raising of the absolutive object to the subject position. The ergative can be reintroduced with jav + dative. Unlike in Vedreki the passive allows for a dative shift; this forms a verb with two absolutive arguments, which is common in the antipassive so nobody saw the problem, and it allows for relativisation of an otherwise dative argument. It is not therefore possible to say in Vedreki

Lenkiek akhtieghes kitieba
Priest-nom sg, give-past ind pass masc sg, book-acc sg
The priest was given the book

But it is possible to say in Cheyadeneen
I gave the book to the priest
Dakadai napis thenwayak
Give-past perf 1ps, book-abs sg, priest-dat sg

The priest was given the book
Thenwa laradan dakus napis (jav eyak)
Priest-abs sg, laraz-past perf ind 3ps give-past part, book-abs sg (by I-dat)
Alternatively
The book was given to the priest:
Napis laradan dakus thenwayak
Book-abs sg, laraz-past perf ind 3ps, give past part, priest-dat sg

The king ate the fish
Cheyat sedadan viltu
King-erg sg, eat-past perf ind 3ps, fish-abs sg

The fish was eaten by the king
Viltu laradan sedus jav cheyak
Fish-abs sg, laraz-past perf ind 3ps, give past part, by, king-dat sg

The book was given to the priest by the king
Napis laradan dakus thenwayak jav cheyak
Book-abs sg, laraz-past perf ind 3ps, give past part, priest-dat sg, by, king-dat sg

Now part of the reason why I care about this is because I discovered by chance there are languages where the there are serious restrictions on the role an antecedent (in fact it seems usually a postcedent) can play in the relative clause. So I pinched this and a device from Basque and cam up with this ....

Complications are caused by the Cheyadeneen rule that the relativized argument in the relative clause must be absolutive; such clauses are noted by a clitic -(n)at on the end of the verb in the relative clause. So

a) The man who ate the fish saw me
Viltu sesedadanat steneyat bletadan eika
Fish-abs sg, eat-past perf 3ps AP, rel clause, man-erg sg, see past perf 3ps, I-abs
Here the anti-passive is required otherwise the relativised argument must be ergative, which is banned.

b) The man who I saw walked across the road
Bletadainat stene shefusun da raluak
See-past perf 1ps (rel clause), man-abs sg, walk-past perf 3ps, across, road-dat-sg

c) I hit the man whom I had seen earlier
Chelet bletadainat stene nikusoe
Earlier, see-past perf 1ps (rel clause), man-abs, hit-past perf 1ps

d) I hit the man who had spoken
Arusunat stene nikusoe
speak-past perf 3ps, man-abs sg, hit-past perf 1ps

Have I got this right? There are some constructions I think that can't work at all. I can put datives into the absolutive by having a passive in the relative clause, but things like (e) can't work in Cheyadeneen although they do in Vedreki

e) The king, whose laws are just, spoke
(Vedreki version) Mudrel, bekat sakonnae oneb natnae malenknae, flakhnus
King-nom sg, that law-nom-pl, he-gen, be-pres-pl, just-nom pl, speak-past masc sg

If I do carry on with this - rather than just giving up and having an English-like relative pronoun - how do other languages deal with this. Would we simply have to say? "The king spoke. His laws are just."