Gozhaaq Azure

=Gozhaaq Azure (not finished)= Gozhaaq Azure [gɔ.ˈd͡ʒaːk' ˈa.zu.ɾe] is fictional language, spoken by the fictional Azura Folk. This tribe is inspired by the Gerudo Folk, a folk of the Legend of Zelda series. I will not go into detail in terms of culture, since the purpose of the page is simply to have a summary of the conlang.

Introduction to Gozhaaq Azure
Gozhaaq Azure is a highly agglutinative (head final) language, having almost only suffixes. In additon, it is has a rich derivational morphology, allowing to derive from a root nouns, adjectives and verbs. But before we go into morphology, I will talk about Phonology, so that you can read the Glossings.

Consonants
* = Allophones

Note: ([j]) is possible variant of /ʎ/ and ([ɾ]) is a possible variant of /r/. There are no strict rules which form to use in which enviroment. The speaker can decide which one to use.

Vowels
The existing vowels are /i/, /ɔ/, /a/, /u/ and /ɛ/

Script
Almost all IPA symbols correspond exactly to the script in Gozhaaq Azure, of course with some Exceptions:


 * d͡ʒ => 
 * ʎ and j => 
 * r and ɾ => 
 * k =>
 * ɔ => 
 * ɛ => 

Note: There is a special handwriting script and an ancient script, but those will not be a topic on this page.

General Phonology Rules
As you have seen in the table, there two allophones, which are generated by following rules:


 * h -> ç / _ i
 * d͡ʒ -> ʃ / n_[+kons,-voiced]
 * there is actually only one word where this rule can be applied: Nzhita (under) => [nʃta]. Very weird word, its preserved from ancient Gozhaaq Azure.

Syllable Structure
Allowed Structure:


 * CV
 * CVV
 * CV/n/
 * (NC*V)
 * * only voiced consonants, mostly  and 

Morphology
As you have seen, the phonology of Gozhaaq Azure is rather simple and not that interesting. Morphology on the other hand is more interesting in this language.

The root
a root in Gozhaaq Azure is firstly a lexicon entry which can be used as a simple noun (This is used in plain conversation). Some examples:


 * "Va" meaning water
 * "ya" meaning arrow
 * "hala" meaning sky
 * "Ne" meaning continent
 * "qame" meaning gold or money

The ending vowel indicates the gender of a root (with exceptions). There are two genders: Female (mostly ending in "e") and Male (mostly ending in "a")

Determiners
"Va" itself is a root and a noun at the same time. Normaly in formal speech Determiners have to be used. There are the definite determiners (DEF) and the indefinite Determiners (INDEF). They are, of course, all suffixes. Both types of determiners correspond to the gender of the root, by having "-aq", the masculine DEF (M.DEF or MASC.DEF), "-eq", the feminine DEF (F.DEF or FEM.DEF). Then of course "-an" (F.INDEF) and "-en" (M.INDEF).

you get following meanings by attaching the determiners to the roots (Example: "ne" and "va"):


 * "Va" is MASC, therefore: "Vaaq" meaning "the water"
 * "Vaan" meaning "a water" is also possible, but means something like a drop (of water), a puddle, or a sea
 * "ne" is FEM, therefore: "Neeq" meaning "the continent", or with INDEF, "Neen" meaning "a continent"

Plural Suffix
The Plural Suffix "ni" is in the same Domain as the determiners which means that the plural suffix and the determiners cannot coexist. Example ("ne" = continent, "qame" = gold/money):


 * "Neni" meaning "continents"
 * "qameni" meaning a lot of pieces of gold

Case Markers
The next domain is the domain of case markers, which come after the determiner domain. The Case markers are:


 * ga (NOM)
 * gi (ACC)
 * gu (DAT)
 * go (GEN)
 * ge (LNK)
 * ge is productive linker, being also a coordinator

There is an phonological rule due to a syllable structure clash (example "va" = water):


 * Vaaq + NOM = *Vaaq ga (CC not allowed)
 * g -> ∅ / k_ (or even C_)
 * Repair: Vaaq a = the water (NOM)
 * Deletion is a common strategy in Gozhaaq Azure, in is even marked in the script: Vaaq'a
 * Vaan + NOM = Vaan ga (allowed)

Advanced domain: Applicative (Benefactive (BEN) and Instrumental (INST))
the Linker "ge" allows to attach "-pa" BEN and "-qo" INST to the stem (Example: Qame):

Qame-eq-e-pa

gold(F)-F.DEF-LNK-BEN

"for the gold"

Qame-eq-e-qo

gold(F)-F.DEF-LNK-INST

"with the gold"

Noun Types
We already discussed the phenomenon, where a nouns final vowel indicates it's gender (Type 1 Noun). This is not always the case, there a lot of nouns which have a lexicalised gender feature (although one could say that this feature is generally lexicalised). Categories can be established by looking at the examples:

(Evidence for the lexicalised gender is the attached Determiner)


 * Zhepe (M) = spear, Zhepe-aq
 * Zhevode (M) = animal, Zhevode-aq
 * Nama (F) = name, Nama-eq
 * Zhovese (M) = kingdom, Zhovese-aq

=> Type 2 Noun: Vowel mismatch


 * Sari (M) = Sword, Sari-aq
 * Zotatu (F) = statue, Zotatu-eq
 * Vi (M) = journey, vi-aq

=> Type 3 Noun: No Gender Vowel underlyingly


 * Zho = Ruler, Zho-aq/eq = King/Queen
 * Vo = Human, Person, Vo-aq/eq =Man/Woman
 * Fo = Child, Fo-aq/eq = boy/girl

=> Type 4 Noun: No Gender Feature inherited, mostly human entities

Type 4 Nouns: Special Properties
Type 4 Nouns are somehow special, since not having a gender feature inherited (presumably like Type 1). Type 4 Nouns work exactly as other Nouns but having special meanings which can be derived due to the already mentioned morphology (Example "Fo" = Child)


 * Fo = child
 * Fo-aq = the boy
 * Fo-eq = the girl
 * Fo-ni = boys and girls/children

But there is more possible, which requieres a re-analysis of the determiners:


 * Fo-a = boy
 * Fo-e = girl
 * Fo-a-ni = boys
 * Fo-e-ni = girls

We established that the determiners have an inherited gender marker, but it seems like we could split up the determiners ("Va" = Water):

Va-a-q

Water(M)-MASC-DEF

But *Vaa is not possible, which leads to 2 theories:


 * 1) There are 2 Classes of Nouns: Class 1 = {Type1,Type2,Type3} and Class 2 = {Type4} or simply Class [-hum] and Class [+hum], although not all human entities are in the [+hum] class.
 * 2) * Class 1 Nouns have their own Determiner, which is -aq,-eq,-an,-en
 * 3) * Class 2 Nouns have Gender markers -a,-e and Determiners -q and -n


 * 1) Same Theory like in 1., but here Class 1 has a Gendermarker which needs a determiner and Class 2 Gendermarker does not need a determiner

Theorie 1 Analysis:

Va-a-q

Water.1(F)-F-DEF

"The Water"

Alternative: Since Class two Nouns don't have Gender features inherited, the brackets with the gendermarker in it in class 1 nouns indicate that this noun indeed is a Class 1 Noun.

Va-a-q

Water(F)-F-DEF

"The Water"

Theory 2:

2 Classes of Gender Markers:

Va-a-q

Water(F)-F.1-DEF

"the water"

Fo-a-q

child-M.2-DEF

"the boy"

I prefer to use theory 1.

Pronouns
Gozhaaq Azure is a Pro-Drop Language but Pronouns can still be used, i.e for emphasis:


 * *very rare, not common

Possesion
To achieve a possesive meaning, Gozhaaq Azure uses a genitive Construction (example: Qaza = house)

Yo go   Qaza

I   GEN  House(M)

"My House"

Fo-a-q-o                    Qaza

Child-M-DEF-GEN   house(M)

"The Boy's house"

Interestingly, following constructions are also possible:

Yo go   Qaza-a-q/n

I   GEN  House(M)-M-DEF/INDEF

"A/The House of mine"

Fo-a-q-o                    Qaza-a-q/n

Child-M-DEF-GEN   house(M)-DEF/INDEF

"A/The House of the boy"

Adjective Domains
Adjectives Are derived with the suffix "-bi", which is attached to the stem (Example "i" = good, ~the goodness)

i-e-q = the good

i-bi = good

With this information, we can now say:

i-bi                   Fo-e-q

good-ADJ      Child-F-DEF

"the good girl"

This works only when the ADJ is in the position as modificator, it works differently as a predicate:

Predicate Domain
Adjectives have their own Tense markers:


 * -ta for present Tense
 * -ata for past tense
 * -reta for future tense

Example:

Fo-a-q-a                          i-bi-ta

Child-M-DEF-NOM        good-ADJ-PRES

"the boy is good"

Just for the record, here some adjectives (with the ADJ Suffix)


 * atarabi - new
 * bologabi - big
 * semobi - small
 * difubi - difficult
 * ezabi - easy
 * omobi - heavy
 * hazhabi - fast

Fo-e-q-a                              semo-bi-ata

Child-F-DEF-NOM              small-ADJ-PST

"the girl was small"

There are more Suffixes that are possible in this domain but they need an Auxilliary Verb to be grammatical. Since I didn't introduce you to verbs, this section will be desripted later.

Verb Domains
To get a Verb from a root, Gozhaaq Azure, like for Adjectives, uses a special Suffix "-v", which is a verbalizer. But there is no infinitive structure, so only attaching "-v" is ungrammatical. Therefore, after the verbalizer, comes the TMA Morphology:

Some Roots, which are commonly derived to verbs


 * Nemu - sleep

Note: I do not write the gender of the noun, because it is irrelevant in this section.
 * hazha - run
 * Fora - Question -> to ask
 * Bere - drink
 * te - have
 * se - be

The Verbalizer will be written here in the table as well.

This Table also contains the Negation Suffix ("-na", "-ne", "-no", "-a"), because the negation suffix alters depending on the suffix which preceded the negation:

Of course, Negation is not mandatory, it is used only if you want to negate the sentence. This table can be analysed in terms of syncretism, but I won't do a analysis of this here.

Example (hazha = run): Remember: it's a Pro-Drop Language

Hazha-v-ise

run-VBLZ-2SG.PRES

"you run"

Hazha-v-ise-na

run-VBLZ-2SG.PRES-NEG

"you didn't run"

Hazha-v-irizan

run-VBLZ-2/3PL.COND.PST

"you (pl)/they would have run"

Hazha-v-irizan-a

run-VBLZ-2/3PL.COND.PST-NEG

"you (pl)/they wouldn't have run"

Passivation
There is the Passive Suffixe "-zho" which is attached after the negation (if there is no negation, then of course it is placed after the TMA-Morphology)

Example: Fora - Question

fora-v-a-(na)-zho

Question-VBLZ-1/3SG.PST-(NEG)-PASS

"He/She/It/I was (not) asked"

AUX-Domain
Auxilliaries are also considered as Suffixes, but to attach them to Mainverbs, a special Auxilliary Linker is used. Some Auxilliaries are written here:

Auxlliary Linker = "-te", mostly shortened do "-t" = 
 * ere - want
 * ode - can
 * use - must/have to
 * efe - CAUS
 * the Causative Affix is in the same domain as Auxilliaries

Example:



Hazha-v-i-t-ere-v-asamo-ne

run-VBLZ-1/3SG-AUX.LNK-want-VBLZ-1PL.PST-NEG

"We didn't want to run"

Note that there is the first Person/third Person marker in the main verb, which doesn't contain any semantic information, since the AUX contains the TMA information for the Subject. Therefore the mentioned semantically empty marker is considered as a default marker, which purpose is to be a bridge from the verbalizer domain to the Linker domain, which is the same domain as the Passivation Domain. This is not en elegant solution, some agglunative languages like quechua nominalize the mainverb and add an AUX to have an AUX construction. There is no reason why this couldn't work for Gozhaaq Azure, therefore here an alternative contruction:



Hazha-t-ere-v-asamo-ne

run-AUX.LNK-want-VBLZ-1PL.PST-NEG

"We didn't want to run"

Since run is already nominalized (because a root is a Noun already) there is no need for nominalization. Hence this alternative construction would be more plausable, I accept both ways (but prefer the second)

Comparative/Superlative
Now that we haved talked about verbs, we can now finally talk about the COMP/SUP, where COMP is in the Domain of Predicate Tense markers of adjectives. The Suffix for the comparative is "-ra", and the superlative, which is attached to the comparative is "-zha". Examples:

qorasee-bi-ra

beautiful-ADJ-COMP

"more beautiful"

qorasee-bi-ra-zha

beautiful-ADJ-COMP-SUP

"most beautiful"

Since the comparative is in the Domain of Tense markers, the tense markers are now blocked, so that a Auxilliary Verb ("se-" - to be) is needed. To form a sentence in the comparative, a comparing object is needed, by added the comparing suffix/particle "yora" after the comparing object, you need a LNK to attach it to the COMP-Object. Example:



Pala-a-q-e-yora, Qaza-a-q-a qorasee-bi-ra se-v-i

Palace(M)-M-DEF-LNK-COMP house(M)-M-DEF-NOM beautiful-ADJ-COMP be-VBLZ-1/3SG.PRES

"In comparison to the palace, the house is more beautiful

It is debatable, whether "yora" is a suffix or a particle. It's likely a suffix, since all other morphology, like INST or BEN is blocked.

Animated Substantive Construction
Since we know that all forms are derived from the root, which is noun, it doesn't necessarly mean that this root is already an animated noun. Take for example the root "tataqa" which means "fight" as in "the fight". But how do you derive an animated noun like fighter?

To do that, there is a complex derivation. I will derive the already mentioned example "tataqa":

The affix for animated nouns is "-i", it is in the domain before the gender marker. Therefore, firstly this affix is attached to the root:

tataqa -> tataqa-i

Secondly, the gender marker has to be attached, so that the speaker know the gender:

tataqa-i -> tataqa-i-a/e

Here we get a clash: In the Underlying form, there are 3 vowels in the row, which is not allowed in Gozhaaq Azure. In this language, such phenomenons are fixed by using deletion or vowel chaning into a glide vowel. Since a glide vowel is considered to be consonant, the clash is fixed. In this construction, the chaning method is used:

tataqaia -> tataqaya

tataqaie -> tataqaye

now this type of nouns are behaving like nouns of Type 1. Therefore, to attach determiners, the gender marker is obligatory:

tataqaya -> tataqayaaq = the male fighter

tataqaye -> tataqayeen = a female fighter

This analysis gives evidence to the cyclic derivation of the language, because phonology is typically the last domain in the derivation steps, but here you have phonological repairing inside of some morphological operations.

Here is the output when a Plural suffix is used


 * tataqayani = male fighters
 * tataqayeni = female fighters
 * tataqaini = fighters

As you can see, in the third example, phonology did not have to repair the clash because there was no clash from the beginning.

Glossing this kind of thing is very hard, but this would be an appropriate approach:

tataqa-y-a-a-q

fight-ANI.SUBS-M-M-DEF

"the male fighter"

unfortunately, this glossing would imply that a gender marker can attach itself to the same gender marker, which could make it recursive. However, the analysis from before shows the word has gone through a reanalysis. Due to the underlying structure, this gloss is still the best way of glossing the beforementioned phenomenon.

Syntax
Gozhaaq Azure is an SOV Language.