Mestian

General Information
Mestian (natively Mástā) is the language of some of the dragon-herding peoples in the central lowlands and on the Dragonforge. It is part of the lowlands Sprachbund. It is part of the Adaric language family and is a distant relative of Sarhan.

Speaking from a mixed diachronic and synchronic point of view, Mestian is a heavily fusional language descended from a heavily agglutinating stage. Many of the previously agglutinative components of the now-fusional morphemes are still vaguely evident and partially identifiable, although many have long since merged beyond transparency. Its tonal system descends from a mixture of stress accent and the disappearance of /*h/: most dipping long vowels come from a vowel that was lengthened by the loss of that /*h/.

Mestian is natively written most frequently in the indigenous Dragon Imperial alphabet, specifically modified to suit the needs of the language. It is written with full stress marks and is used as the template for orthographies of the languages recorded and described by Mestians.

Vowels
Mestian features distinctive stress and a pitch accent system: short vowels can either be stressed, marked with an acute accent such as in <ú>, or unstressed (unmarked), while long vowels can have either a peaking or dipping intonation when stressed. Dipping is marked with a tilde diacritic, such as <ũ>, and peaking is marked with a circumflex, such as <û>. The dipping and peaking diacritics replace the macron of long vowels.

Mestian also allows the formation of diphthongs between two short elements. This is fairly unproductive and exists only in some certain, fixed terms and terminology — possessive pronouns and archaic, semi-unanalysable compounds — but prevents insertion of an excrescent element. These diphthongs are metrically counted as a single vowel and may have both a stressed upshift and pitch contours. The acute is usually written only on the first element, but the tilde and circumflex are instead replaced with a double tilde and double inverted breve that both span the entire diphthong. These are written as: <áu a͠u a͡u>. In lieu of these, in situations where the diacritics are hard to type or not supported by the font used, the double diacritics can be replaced by a simple tilde and circumflex on the first element: <ãu âu>.

Anaptyxis
Mestian features an anaptyctic system that serves to break up consonant clusters. This system operates by inserting an on- or offglide demivowel, most frequently [ɨ̆ ~ ɨ̥̆] that is usually orthographically indicated by an apostrophe < ' >. This glide vowel can frequently transform into preaspiration before voiceless consonants — usually geminates — giving example initial sequences such as <'ppa> [ʰppɐ]; this makes word-initial gemination distinctive.

Graphically speaking, this apostrophe can be called a "superscript iota" as it derives from the grapheme used to represent the full vowel /ɨ/; it is considered neither an independent grapheme nor a phoneme.

Excrescence
To break up vowel sequences, especially those involving tonic, stressed vowels, Mestian employs a simple excrescent process: it inserts sonorants between two vowels, forming clear syllable boundaries. The sonorant inserted between the vowels depends on the quality of the vowels like so:

Some results might overlap, and some situations are undefined: in both cases the excrescent element is idiolectic and open-ended, making excrescence a non-bijective function.

Cluster Simplification
Even though its phonotactics are extremly permissive and lenient towards many sequences, a phenomenon of simplification of internal consonant clusters may readily be observed in the language of many native speakers.

Phonotactics
Mestian phonotactics allow for a high number of complex syllables.

Nominals
The category of nominals in Mestian includes nouns, pronouns and adjectives.

Nouns
Mestian has a moderately complex fusional noun system that distinguishes ten cases and three numbers. Mestian nouns are grouped based on their declension patterns -- the suffixes they take -- and their accentual paradigms -- how their accent shifts when they inflect.

They have five genders: ignic, aquatic, animate, inanimate and neuter. The division into these gender classes is largely arbitrary.

The ten cases of Mestian are:
 * Nominative
 * Accusative
 * Partitive
 * Possessive¹ (inalienable)
 * Possessive² (alienable)
 * Dative
 * Vocative
 * Comitative
 * Lative
 * Locative

Most nouns are regularly categorisable into declension categories — their nominative suffixes are usually indicative of their declension class — but some may be irregular in that they resemble one class but decline as another, or in that they may take suffixes of two classes, either in part or whole.

Use of Cases
Besides the fact that they are required by adpositions, each of Mestian's cases has its own common purpose and use:


 * 1) The Nominative is primarily used to mark subjects of intransitive verbs and agents of transitive verbs.
 * 2) The Accusative is primarily used to mark objects of transitive verbs that are regarded as a unitary whole.
 * 3) The Partitive is primarily used to mark objects of transitive verbs that are not representative of the whole they belong to.
 * These three cases are generally considered the three core cases of Mestian. Other cases are thus considered oblique.
 * 1) The two Possessive cases respectively mark possessors of inalienable and alienable possessions.
 * 2) The Dative is most freuqently used to mark recipients, beneficiaries and related roles.
 * 3) The Vocative is used only as a marker for the nominative argument replacement in imperatives, as well as to denote addressees.
 * 4) The Comitative is used to mark company or other entities associated with an entity.
 * 5) The Lative primarily denotes roles related to motion without regards for direction or course.
 * 6) The Locative primarily denotes roles related to static locations without regards to relative placement.

Accentuation Patterns
Nouns generally fit into one of four large accentual patterns:


 * Short
 * Mobile -- most frequent
 * Static -- very rare
 * Long
 * Mobile -- common
 * Static -- slightly uncommon

While each of the patterns has its irregularities, the most common type of motion is relocation of stress in the direction of extension when the stem gets extended. Mobile long nouns generally either leave behind a long vowel upon relocation (either tone) or geminate the following consonant (only dipping). Static-stressed short nouns are overwhelmingly those that have a stressed short vowel followed by an unstressed long vowel: stress in these cases doesn't shift to the long vowel.

Many long nouns occasionally change their tone between paradigms.

Compounds
Mestian doesn't frequently employ compounds; most of its compounds are fossilised and semi-irregular formations of a slightly earlier stage of the language. Compounding is productive, though not to a great extent. Compounds most usually take the shape A-B.

Mestian noun compounds are generally divided into four types:
 * Tatpuruṣa -- where A modifies, narrows down and specifies B. Left-branching and head-final.
 * Bahuvrīhi -- where A modifies B, where the modified AB is an attribute or possession of an external head. Left-branching and head-final.
 * Dvandva -- where A and B work in tandem and co-ordinate to make a noun with a combined meaning of both. Ambicapital.
 * Āmreḍita -- where A and B are (mostly) identical, expressing iteration or collectiveness. Reduplicative, ambicapital.

Tatpuruṣa compounds are, for the most part, utilised as raw lexical elements. They are, in Sanskrit terms, also aluk-samāsa compounds; their case endings generally remain, frequently reduced or clipped. The majority of such compounds have their modifier element be in a possessive or partitive case: examples are words such as assundúffū (armful), which is built up as assún(u)-dúffū, where the first element, ássam (arm), is in the possessive² case. Such compounds are generally stressed only on its compound heads. If they are made up of a head and, for example, a tatpuruṣa modifier, the head of that tatpuruṣa is thus also accented.

Such compounds most frequently have A be a modifier of B that narrows its field of definition. Certain sets of older compounds might use different definitions of some of their elements. All such compounds decline solely using B.

Bahuvrīhi compounds generally behave and function similarly to tatpuruṣa compounds, but they are specific in that the resulting semantic compound is not indicative of its actual meaning: bahivrīhi compounds generally reference something that possesses the compound. This means that such compounds are very culturally sensitive as they are most often allusions to sociopolitics, history, literature or folklore. Bahuvrīhi compounds are likewise accented only on their heads.

Dvandva compounds are very strongly ambicapital and always have only two elements, the heads, that are optionally linked with the conjunction tā. The first head of a dvandva compound is usually in the lative, while the second head has a normal declension. The semantics of the resulting word are a sum of the semantics of the two heads.

Āmreḍita compounds are ambicapital remnants of a wider reduplicative system that older stages of Mestian used to employ for morphosemantical purposes. They are generally fixed lexical items or expressions, though some semiproductive transparent formations do exist. Such a compound is made up of one duplicated head. The first head of the compound is usually either in the lative or dative cases, although irregular endings and minor stem alternations may appear. The second head of the compound declines normally and carries the stress of the whole compound.

(I) A-Stems
Mestian a-stems make up the first declension class in the language. By definition they are nouns with an <-a> in the nominative, optionally followed by a single consonant. A-stems can be in any of the five genders. The declension class has two subtypes: The two subtypes differ minimally in the suffixes they take: soft a-stems have an allomorph -īt in the locative.
 * Hard a-stems -- ending in a hard consonant followed by <-a>
 * Soft a-stems -- ending in any one of {š ž tš dž j} followed by <-a>

The three main suffix classes are:
 * a-class
 * as-class
 * an-class

The as-class nouns merge the nominative and accusative singulars and plurals, while an-class nouns merge the nominative and comitative cases in all numbers. The most common accentuation patterns of the a-stems are the short mobile-stressed nouns that make up the bulk of its lexical mass, as well as both mobile-stressed and static-stressed long nouns.

Neuter and inanimate nouns merge the two possessive cases.

An example aquatic a-class noun: pírka, pirkássa (fish, aqu):

An example inanimate a-class noun: rhýka, rhykássa (pebble, ina):

An example animate an-class noun: itáran, itarássa (swan, ani):

An example ignic as-class noun: apákas, apakássa (breath, ign):

(II) UR-Stems
Mestian ur-stems make up the second declension class in the language. They are nouns that end in <-ur> in the nominative. They are most often ignic, inanimate or neuter. They do not have subclasses.

Neuter and inanimate nouns merge the two possessive cases.

An example ignic noun: ĩkur, ikkússa (snake, ign):

An example neuter noun: âghur, āghússa (egg/ovum, neu):

(III) Ā-Stems
Mestian ā-stems make up the third declension class in the language. They are nouns that end in <-ā> in the nominative, and are almost exclusively neuter. All nouns of the class merge the two possessive cases, regardless of gender.

The ā-stems come in two subclasses:
 * regular-class
 * džā-class

The džā-class category specifically includes abstract nouns derived from adjectives using the derivational <-džā> suffix. These nouns have their own unique declension that doesn't quite line up with the regular declension of this declension class, although similarities are too great to be ignored.

An example regular neuter noun: qásā, qasášša (sound, neu):

An example regular neuter noun (different stress paradigm): elétā, elétašša (table knife, neu):

An example džā-class neuter noun: tándžā, tandžâšša [tán(es)-džā] (tension, neu):

(IV) Ę/Ą-Stems
The fourth declension superclass in Mestian includes nouns that end in the vowels <-ę> and <-ą>. The super-class has two paradigms, corresponding to the endings, that have parallels in declension and are close enough to be grouped into one category.

An example ę-stem aquatic noun: ýšpę, yšpímmę (tadpole, aqu):

An example ą-stem animate noun: óllą, ollímmą (sky, ani):

(V) Ū/Ī-Stems
The fifth Mestian declension class includes nouns that end in the vowels <-ū> and <-ī>. The two endings do not actually cause a difference in inflection: the only difference is that nouns respectively take <-u-> and <-i-> as thematic vowels.

An example ū-stem aquatic noun: dásū, dasúššą (blackfish, aqu):

An example ī-stem inanimate noun: akãkī, akakkíššą (revelation, ina):

(VI) Nasal Stems
Mestian nasal stems are nouns that end in any of {m n ŋ}, preceded by either a vowel or a single consonant. The category systematically includes nouns ending in <-an> as standard members; many of these overlap with a-stems that end in <-an>. A good number of these nouns are slightly irregular in unpredictable and uncategorisable ways.

An example ignic noun: trąkún, trąkúnnu (small duck, ign):

An example animate noun: awãraŋ, awarhúnnu (muscle fibre, ani):

(VII) L-Stems
Mestian l-stems are nouns that end in <-l> in the nominative; many of them are irregular and have complex stem alternations.

An example l-stem animate noun: tâlhęl, tąllêlle (blink, ani):

(VIII) Alveopalatal Stems
Mendian alveopalatal stems are nouns that end in an alveopalatal consonant in the nominative, namely one of {š ž tš dž}.

An example alveopalatal-stem inanimate noun: béč, béčella (boulder, ina):

(IX) Consonantal Stems
Mestian consonantal stems are generically nouns that end in consonants other than {m n ŋ l š ž tš dž}, while also excluding those that end in <-ur> (but not nouns that generically end in <-r> preceded by another vowel).

Example consonant-stem animate noun: dârab, dârabašša (fowl, ani):

Pronouns
The pronominal system of Mestian only has dedicated pronouns for first and second persons; third person pronouns as such don't quite exist as their role is taken up by demonstratives that all have full declensions.

First and Second Persons
Both the first and second person pronouns in Mestian have emphatic and unstressed forms for all case-number combinations. The first person pronoun lacks a vocative (due to semantics), and both persons lack the possessive cases, compensating with dedicated possessive pronouns.

The first person pronoun is:

The second person pronoun is:

First and Second Person Possessives
Mestian has two sets of first and second person possessives: a stressed form that is used as a pronoun and is independent, and an unstressed clitic form used attributively.

The first person possessive is:

The segment sequence marked MB is realised as a /nd/ between vowels, but /mb/ utterance-initially and after consonants. It is otherwise a normal, albeit underspecified, segment sequence.

The second person possessive is:

Third Person
The third person pronominal system is more open: all the pronoun types follow one declension pattern.

The proximate pronoun dínaj (this one) serves as a good example:

All other third person pronouns decline the same. Some common ones:
 * dívaj -- that one (that you can see)
 * dínaj -- this one (that you can see)
 * kévaj -- that one (that you cannot see)
 * kénaj -- this one (that you cannot see)
 * âssaj -- the previously mentioned one
 * ŷlaj -- the same one
 * ûttaj -- relative pronoun
 * gẽraj (gheránna) -- subordinate head

The third person pronouns all optionally take a gender determiner that depends on the gender of that which they refer to. These determiners aren't added when people are referred to. The determiners are:

These determiners cause stress shift and are attached directly onto the inflected pronoun.

Numbers
Numbers in Mestian are an intermediate category: they act as nominal modifiers, akin to adjectives, but always decline like fifth declension nouns (except for tîwa, which is a first-declension noun), without regard to gender. Nominalised and pronominalised numbers can take any gender, as appropriate to what they refer to.

Mestian uses a duodecimal (base 12) counting system. It constructs multiples of higher positions than one in a compound fashion: they are shaped like dvandva compounds of a radix number in the lative singular, optionally without its final vowel, and a declinable position multiplier that follows it. Numbers built as sums of different radix-multiplier multiples are written sequentially in an ascending order (ones, then twelves, and so on) either with a linking tā, or naked.

Adjectives
Mestian nominal adjectives follow a similar paradigm pattern as do nouns: they are divided into declensions and inflect for number and case. Adjectives have two more dimensions: they also inflect to indicate the gender of nouns they stand with and have degrees of comparison. Neuter nominative and lative singular adjectives can also be used as adverbs and thus modify both verbs and other adjectives.

A full adjective citation is made up from anywhere from two to ten principal parts, indicating gender forms and accent alternations. Some adjectives merge some of their genders and thus do not make a full distinction. Gender in adjectives is most commonly indicated by the adjective taking a different, gender-appropriate declension: some declensions are linked to one gender while others may hold adjectives of several different genders or multiple genders at once.

Comparison of adjectives is handled with a derivational process: comparatives and superlatives are handled by accentually weak suffixes that are attached to the stem of the adjective. Adjectives distinguish positive, comparative, superlative and excessive degrees:

Adjectives generally follow stem patterns: many adjectives can be grouped into principal part families or classes based on similar patterns. The most common classes are:

(Ia) AK-Stems
Mestian ak-stem adjectives generally make up the bulk of inanimate and neuter adjectives. They closely follow the nominal a-stem declension, distinguishing the possessive cases even when the nouns they stand with do not. An example ak-stem adjective: sânnak, sānnássa (cruel, ina/neu):

(Ib) AS-Stems
Mestian as-stem adjectives make up the plurality of Mestian ignic adjectives. They closely follow the nominal as-class declension. An example as-stem adjective: ũlas, ullássa (beige/cream, ign):

(IIa) U-Stems
Mestian u-stem adjectives generally contain ignic and animate adjectives. They share some suffix shapes with ū-stems and ur-stems. Example u-stem adjective: tánu, tanúšša (tense, ign):

(IIb) US-Stems
Us-stem adjectives are exclusively ignic. They are often somewhat irregular as they often derive from syncopated or contracted adjectives and nouns of other classes. Their declension follows several patterns, and as such aren't specifically tied to one single declension class. Most adjectives of this class exist only in it or exist in only it and within u-stems.

An example us-stem adjective: ĩmus, īmbíllu (bubbly, ign):

(IIIa) EJA-Stems
Mestian eja-stem adjectives include aquatic, ignic and neuter adjectives. They are for the most part derived adjectives, but also contain the majority of aquatic adjectives. They follow a relatively simple declension pattern. An example eja-stem adjective: gháreja, ghárajšša (black, aqu/ign):

(IIIb) ESA-Stems
Esa-stem adjectives are primarily ingnic, although there exists a significant inanimate and neuter minority. They follow a relatively simple declension pattern. A sample esa-stem adjective: zýlesa, zýlišša (blind, ina/neu):

(IIIc) EŠIS-Stems
Mestian ešis-stems are a special class of gender-agnostic adjectives. They are primarily derived and participal. An example ešis-stem adjective: galéšis, galemmášša (incoherent, ign/aqu/ani/ina/neu):

Verbs
As the concept of clear-cut morphemes is a bit more loose in Mestian verbs, this section is divided into a theoretical and practical section. This is due to the complex morphophonological alternations that exist between morphemes, that's evolved out of a highly agglutinative and isolating system.

Theory
Mestian verbs are quite complex: they inflect for person of multiple arguments, aspect, telicity, tense and mood. They can take on multiple sequential clitics that further narrow their meaning and extract diverse additional semantic information.

Verb Shapes
All Mestian verbs follow a basic shape pattern that determines the order of their morphemes. A full Mestian verb consists of the following components:

These are, in order:
 *  -- preverb; usually prefix or postposition
 *  -- Slot A, one of the two slots for personal agreement
 *  -- the root of the verb
 *  -- the verb's morphosemantic characteristic suffix; can contain more than one such suffix
 *  -- Slot B, the other of the two slots for personal agreement
 *  -- the inflection suffix slot

Person Agreement
The verbs can take four different kinds of personal agreement affixes, grouped into two groups:
 * Primary
 * Nominative agreement affixes
 * Secondary
 * Accusative agreement affixes
 * Partitive agreement affixes
 * Oblique agreement affixes

The positioning of the affixes is complex: there is a strong tendency for the primary affix to be in Slot B, while the secondary affixes are more closely bound to Slot A. Their position is mostly conditioned by grammatical voice.

Both the primary and secondary affixes have four persons and three numbers. These are:
 * Persons
 * First person (speaker)
 * Second person (addressee)
 * Third person proximate (prominent)
 * Third person obviate (sidelined)
 * Numbers
 * Singular
 * Dual
 * Plural

Mood and Modality
Mestian modality is primarily dependent on modal affixes and mood. It makes a distinction between three classes of moods:
 * 1) Realis moods
 * 2) *Indicative
 * 3) *Gnomic
 * 4) Irrealis moods
 * 5) *Subjunctive
 * 6) *Volitive
 * 7) *Potential
 * 8) *Interrogative
 * 9) Deontic moods
 * 10) *Imperative
 * 11) *Hortative
 * 12) *Permissive

Mood, polarity and tense are inflected together with the same affix. Some moods have different behaviours in the negative: these are primarily the prohibitive (from the permissive) and imprecative (from the hortative); they have irregular semantics but regular morphology. Not all mood-tense combinations exist.

Aspect and Telicity
Aspect and telicity are closely connected in Mestian. The combinations of telicity and aspect are:
 * 1) Telic
 * 2) *Inchoative
 * 3) *Momentane
 * 4) *Delimitative
 * 5) *Terminative
 * 6) *Iterative
 * 7) Atelic
 * 8) *Defective
 * 9) *Momentane
 * 10) *Progressive
 * 11) *Frequentative

The aspect combinations line up fairly evenly:

Aspects are either an inherent property of the root or are marked in the morphosemantic suffix slot.

Voice
Mestian verbs can have one of four (two simple and two 'augmented') voices: The applicative is extremely limited and generally no longer productive in the language; it exists in semantically shifted applicatives and similar fosilised constructions. The active and passive are generally indicated by order of person affixes, while the mediopassive requires a dummy pronoun (remnant of older reflexive construction); applicatives used to be formed by infixation and reordering of person affixes, and these processes have by large been thoroughly irregularised.
 * Simple
 * Active
 * Passive
 * Augmented
 * Applicative (†)
 * Mediopassive

Tense
The Mestian tense system is deceptively simple: there are only three tense-like forms extant in the language, alongside a few fossilised remnants:
 * Present
 * Nonpresent
 * Timeless
 * Telic
 * Present Telic (†)
 * Future Telic (†)
 * Future (†)

The telic tenses remain in a few verbs and are generally unproductive. The future is mostly preserved in hortatives.

The present and nonpresent are the primary temporal distinctions in the language; the nonpresent is the result of the merger of the future and an unattested past tense. The timeless is a tensified aorist that doesn't encode temporal information as such.

The distribution of tenses against aspects is somewhat skewed -- the atelic aspects occur more frequently in the present, whereas telic ones are more frequent timeless or nonpresent -- and some tenses are found only in some aspects -- telic tenses can only have telic aspects, and the future can have only atelic aspects.

Tenses are usually marked either by way of a morphosemantic suffix, or in the inflection slot.

The distribution of tenses is closely linked to moods:

Praxis
The practical section of Mestian conjugation deals with the morphology of verbal inflectional morphemes.

Tense and Aspect
The tense-aspect-mood (TAM) system of Mestian is partially derivational: there exist various fusions -- as well as various degrees of fusion -- of the three categories marked by one or more affixes that can sometimes carry semantic baggage.

The tenses are marked using a huge variety of suffixes and alternations: the timeless has four classes, the nonpresent has six and the present has eighteen different unique classes. The present classes are:


 * Type 1: Morphosemantic presents (seven classes)
 * Type 1a: -n- nasal root-infixed presents
 * Type 1bp: Primary -ešte- augmentations
 * Type 1cp: Primary -ŋ- stems
 * Type 1dp: Primary -ŋ- augmentations
 * Type 1bs: Secondary -ešte- augmentations (ablauting)
 * Type 1cs: Secondary -ŋ- stems (ablauting)
 * Type 1ds: Secondary -ŋ- augmentations (ablauting)
 * Type 1e: -lh- extensions
 * Type 1f: -r/rh- extensions
 * Type 1gp: -w- ~ -j- alternating extensions
 * Type 1gs: -u- ~ -i- ablauting
 * Type 2: Doubly-kinetic presents (three classes)
 * Type 2an: -n- nasal root-infixed, nasal-prefixed presents
 * Type 2ak (†): -n- nasal root-infixed, velar-prefixed presents
 * Type 2as (†): -n- nasal root-infixed, s- prefixed fientives
 * Type 2bs: -(a)gha- augmented, s- prefixed denominal presents
 * Type 2bw: -(a)gha- augmented, w- prefixed detelic presents
 * Type 3: Doubly-static presents (five classes)
 * Type 3ap: Primary -s- root-infixed, -aw- ~ -aj- augmentations
 * Type 3as (†): Secondary -s- root-infixed, -aw- ~ -aj- stems (ablauting, irregular)
 * Type 3b (†): -esse-, -gh- suffixed augmentations (ablauting, irregular)
 * Type 3c (†): -(a)šše-, -gh- suffixed augmentations (ablauting, irregular; suppletive)
 * Type 3d: -ŋ- augmented, -gh- suffixed causatives (ablauting; semantically shifted)
 * Type 3e: -zd- augmented, -d- suffixed perfectives
 * Type 4: circumfixed (two classes)
 * Type 4a (†): -s- prefixed, -gh- suffixed
 * Type 4b: -s- prefixed, -d- suffixed
 * Type 5: irregular
 * - ablauting raw stems
 * - primary raw stems
 * - suppletive
 * - irregular raw
 * - irregular pseudo-augmented

The nonpresent classes are:
 * Type 6: Morphosemantic nonpresents (five classes)
 * Type 6a: Primary -axi- ~ -axu- stems
 * Type 6b: Primary -rh- stems
 * Type 6cp: Primary -gha- stems
 * Type 6cs: Secondary -gha- stems (ablauting)
 * Type 6d (†): Primary -zda- stems
 * Type 6e (†): Primary -zda- augmentations
 * Type 7: irregular
 * - ablauting raw stems
 * - primary raw stems
 * - suppletive
 * - irregular raw

The timeless classes are:
 * Type 8 (four classes)
 * Type 8a: Primary -s- augmentations
 * Type 8b (†): Primary -žd- stems
 * Type 8c (†): Primary -žd- augmentations
 * Type 8d (†): Primary -žd- extensions

The distinguishing feature forms of these classes are:
 * infixed
 * augmented
 * stem
 * extended
 * prefixed
 * suffixed
 * primary vs. secondary

Infixations are single consonants inserted into a verb's root to provide inflectional information. Stem additions are morphemes added directly to roots to form stems onto which morphosemantic suffixes are added. Augmentations are inserted at the end of the morphosemantic slot. Extensions attach to either stem additions or augmentations, both already previously extant as active parts of the inflectional system. Prefixes and suffixes are, respectively, in the prefix and suffix slots; temporal prefixes are usually closer to the root than particles or postpositions, whereas suffixes are usually the first in the suffix slot. The difference between primary and secondary inflecions is that secondary inflection involves ablaut as part of its inflectional process; this ablaut is irregular and varies from verb to verb.

Verbs are (primarily) arbitrarily sorted into classes; verbs can be in several classes at the same time, with semantics that may or may not have since shifted.

The aspect inflection morphemes have four different shapes, based on whether they are surrounded by consonants or vowels:

Aspect affixes are attached to the verb, with a general tendency to attach to the telic momentane and atelic progressive when deriving/changing aspect. Many verbs have differing telic and atelic stems that may diverge quite wildly from one another. Additional affixes exist, but they usually carry semantic, derivational information along their aspectual information -- the ones in the table are primarily inflectional.

Person and Voice
Mestian is pecular in that it inflects for voice using the order of affixes for person marking in the verb. Thus, person marking and voice are intrinsically linked. The distinction between the active and passive is determined by the ordering of affixes into the two slots:

Here the red letter represents the primary affix and the black letter the secondary affixes. Applicatives attach a <-žde-> as augmentations onto the moprhosemantic slot, after the last morpheme (including temporal augmentations) and use the same affix order as passives. Mediopassives use the order of indicatives, whilst requiring the reflexive pronoun <ãmaj> to inflect and come after the verb.

Mestian verbs agree with the nominative, accusative, partitive and oblique arguments with unique affixes, in turn grouped into primary (nominative) and secondary (accusative, partitive, oblique). Verbs can take at most three at a time; they cannot have both the accusative and partitive at the same time. The secondary affixes, when occuring together, can fuse into one unseparable and partially opaque hyper-affix, with a less frequent separate form. The person affixes have four different shapes, based on whether they are surrounded by consonants or vowels.

To improve legibility and reduce table size and number, the following matrix is used:

Mood and Polarity
Grammatical mood in Mestian has a simplistic realisation: it is plainly marked with an inflectional suffix at the end of the verb. The modal suffix is conflated with polarity.

Mestian moods are divided into three morphophonological classes:
 * D-Moods
 * Indicative
 * Gnomic
 * Potential
 * Interrogative
 * S-Moods
 * Subjunctive
 * Permissive
 * K-Moods
 * Volitive
 * Imperative
 * Optative

The mood affixes have two forms: a postconsonantal and postvocalic form:

In the event that the interrogative needs to be stacked with another mood, that mood's regular affix is used, and is then followed by a cliticised =jā / =jāla. This formation is used most frequently with the gnomic and subjunctive; the simple interrogative is considered to be indicative in nature.

Derivation
Mestian, as a very inflected language, also employs a large amount of derivational strategies that either map stems from word class to word class or semantically redefine them inside their own word class. Mestian derivation is very productive and predominantly regular, though some outlier irregular forms, analogical levellings and backformations have indeed been inherited from older phases of the language.

Roots from which all stems are derived are traditionally typeset in smallcaps, while their produce is consistent with inline text.

Noun-to-Noun
The most common derivational strategy in Mestian is class internal noun-to-noun mapping. This is usually done through the use of diversely productive noun suffixes that attach to noun roots and stems, extending them to form new, semantically different stems. The suffixes used in derivation are usually further unanalysable, though some may be relict compounding elements — usually stemming from analogised tatpuruṣa compounds — fossilised as suffixes.

The most characteristic traits of derived nouns are that they are accentual wholes — that is, they have only one accent locus that functions the same as in regular, direct nouns — and that they are stronlgy right-branching and head-initial — productive prefixes are morphologically less bound than suffixes. Stems of derived and underived nouns are functionally identical: both take affixes in the same way.

Underived nouns are a bit of a misnomer: they are directly nominalised roots that take nominal affixes in a straightforward fashion. Most such nouns come from monovalent roots — i.e. roots that haven't given other words in word classes directly — and may most accurately be called direct nouns. Such nouns take inflectional suffixes directly onto the root that acts as their stem:


 * - "jade" > tšál(i)s-Ø, tšals-ášša (rarely, tšass-ášša) "jade, white jade"
 * - "swim, float, be piscine" > pírk-a, pirk-ássa "fish"

Some nouns might look like they were directly derived from roots, yet have fossilised, unproductive affixes. This is especially common with alveopalatal nouns, such as:


 * - "rocky" > bét-š-Ø, bet-š-élla "boulder"
 * - "use wing" > sór-(o)š, sor-š-élla "dragon wing"

Loaned nouns also fall under the category of direct nouns for consistency's sake, though they generally have no actual native Mestian root they are linked to:


 * dârab-Ø, dârab-ašša "fowl̈" < Mendian dãrab "fowl"

Some nouns, due to their morphology rendered opaque by the tooth of time, may appear to be either direct or derived when they aren't:


 * - "leaf" > î, ij-ášša "leaf"; appearing as if it has a fossilised suffixed -j-
 * - "climb, ascend" > fĩ, fij-ášša (occasionally, fĩšša); from *fij-iẖ, reanalogised as class IX (as opposed to its original Vī)
 * - "flat surface" > sĩ, sĩ-šša (never *sijášša/*sījášša); from *siẖ-iẖ, reanalogised as a root noun whereas it is actually a class IX noun, analogised from a class Vī noun

Actual Mestian derivational processes function much alike general inflection on a morphophonetic level: stress shifts caused by derivational pronouns work the same, and cause the same stem alternations, as do shifts from inflectional morphemes. They are extremely productive on roots and direct nouns: nearly any suffix can be applied to a minimally derived noun, the only constraint being reasonable semantics. Such derivational suffixes are pluripotent, being able to attach to and extend a wide variety of stems:


 * pírk-čī > pírkč-ī, pirkč-í-ššą "fish scales" (occasionally contracted to pírčī, dropping the -k-)
 * ãd-čī > áčč-ī, ačč-í-ššą "rice paste" (with voicing assimilation and assimilatory lengthening from -dč-)
 * pirk-ápnā > pirk-ápn-ā, pirk-apn-á-šša "fish grounds"
 * pĩš-ápnā > pišš-ápn-ā, pišš-apn-á-šša "orchard"
 * ĩk-ápnā > ikk-ápn-ā, ikk-apn-á-šša "snake den, snake spawning grounds"
 * î-ára > ij-ár-a, ij-ar-á-ssa "treetop, tree crown"

Further specification is somewhat less productive, with productivity dropping as the noun grows in size through repetitive derivation:


 * šar-ja-k-jurtuqur > šarjak-jurtúq-u-r, šarjak-jurtuq-ú-ssa "book-bringer" (rare)
 * î-ar-ápnā > ijar-ápn-ā, ijar-apn-á-šša "canopy" (not frequent)
 * tûr-ku-čī > tūrkú-č-ī, tūrku-č-í-ššą "timber/plank from a large tree"
 * tûr-ku-č-ikkā > tūrkuč-íkk-ā, tūrkuč-ikk-á-šša "that which is characterised by large tree planks; hardwood floor from large tree timber" (chiefly literate, very rare)

At a certain point, completely grammatical sequences of derivational suffixes stop reading as usable forms, though their meanings are intuitable for most native speakers:


 * tûr-ku-č-ikk-jurtuqur > **tūrkučikkjurtúq-u-r, **tūrkučikkjurtuq-ú-ssa "hardwood floor bringer" (parsable, artificial and never used)
 * pĩš-apn-a-čī > piššapná-č-ī, piššapna-č-í-ššą "produce of orchards, fruit" (extremely technical, pointlessly deriving a noun from itself as the meaning goes fruit > orchard > fruit)

Adjective-to-Adjective
Mestian adjective-to-adjective derivation is less diverse and less productive, and usually spans not more than two chained derivational morphemes. As with nouns, adjectives can be divided into derived and direct adjectives, based on their relation to the root from which they derive. Due to the peculiarities of the adjective system and its relation to noun classes, a lot of roots give both direct and derived adjectives:


 * - "tense" > tán-u, tan-ú-šša "tense, springy" ign. adj.
 * tan-u-lus > tánul-u-s, tanúl-u-šša "very tense" ign. adj.
 * tan-tas > tánt-a-s, tant-á-ssa "tense" cA. adj.

Cross-Noun-Adjective Derivation
Derivational processes that map nouns to adjectives, or adjectives to nouns, are fairly common in Mestian: they are more productive than adjective-to-adjective derivation, though not nearly as much as noun-to-noun.

Noun-to-adjective derivation generally makes adjectives of all genders: the split between adjectives of differing genders doesn't generally carry over to derived adjectives. Such derivation works with both direct and derived nouns indiscriminately: the only limit to the size of the stem is reason. Direct noun-to-adjective derivation obviously produces only derived adjectives:


 * pírkč-i-lu > pirkčí-lu, pirkči-lú-šša "fish-scaly, pertaining to fish scales"
 * ikén-lu > ikél-lu, ikel-lú-šša "nocturnal, nightly"
 * tšáls-tas > tšást-a-s, tšast-á-ssa "bejaded, bearing jade" (with cluster simplification -lst- > -st-)

Syntax
Mestian is very strongly a head-initial language, with the natural word order varying between VSO and VOS. Verbs always come first in verb phrases, while noun phrases can have their heads be in almost any position amongst their modifiers, with a strong preference towards being initial. The model of syntax used for constructing Mesian syntax is a form of constituency grammar with elements of a constructional grammatical framework that assumes a sharp, but not binary, gradient between the lexicon and syntax.

Noun Phrases
Mestian noun phrases (NPs) are patterned after a typical lowlands phrase model: despite its relative freedom in word order, Mestian has a very strong preference for head-initial NPs. Mestian NPs are made up of a head and a sequence of modifiers. The head of a NP can be a noun, a nominalised adjective, a nominalised number, a third person pronoun, a RC or an NP, and modifiers can be adjectives, NPs in possessive cases, possessive pronouns, numbers, NPs in cases/with adpositions or APs.

The sequence of modifiers is built of blocks of consistent sequences that can be swapped around one for another but cannot have the order of their contents altered. The blocks are:
 * shape-size-colour
 * moral.opinion-quality.opinion-age
 * ethnicity/nationality-material
 * number
 * unclassified adjective types

Possessive NPs and pronouns always come directly after the head, regardless of where the other blocks are located. Blocks can either all follow or precede the head. The general tendency is for numbers to come the farthest from the head: if the NP is head-initial, numbers will tend to be the last block in the sequence; if it is head-final, they will tend to be the first block in the sequence.

Mestian noun phrases have a wide range of complexity:
 * trąkúlut
 * two drakes
 * a simple NP made up of one nominal
 * rhykúra lorą
 * our pebbles
 * a NP with a head and a possessive
 * pírka kénir gháreja
 * this one's black fish
 * a NP with a head, a modifier and a possessive
 * sỹlut āžapnírā zylúsut ūlísut
 * the elklands' two blind does
 * a NP with a head, a modifier, a possessive and a number

Conjunctive Noun Phrases
Mestian employs a fairly robust conjunction model: all noun phrases joined by a NP-linking conjunction are considered one NP and can thus be modified by modifiers together as a single unit. For disambiguative purposes, all such NPs can be modified from either direction; all the NPs that make up such an NP must be modified so that the modifiers of an individual NP cannot be confused for a modifier of the whole conjunctive NP — this usually means that the first element will be modified by a postfixed modifier, while the second one will be modified by a prefixed modifier. Example phrases:


 * kurašára tā zôdan
 * the bones and blood
 * a simple conjunctive NP with two simple NPs made up of one nominal each
 * lĩją tā sỹlur mulúlu
 * the black deer and doe
 * a conjunctive NP modified by one adjective
 * lĩją mulúlu tā mulúlu sỹlur
 * the black deer and the black doe
 * a conjunctive NP made up of two NPs with one head and one modifier each

Partitive Phrases
Partitive NPs are a special subset of generic NP-NP phrases. They are usually made up of two main components: a modifier in the partitive case and the head that does not take a special case. Such constructions often feature nouns such as lilîldžā or líndžā (most of, plenty of) that denote a chunk that represents a subset of the modifier. The modifier is most commonly plural or is already a collective or mass noun. Such constructions are often head-final, to better facilitate the modification of the head. An example phrase:
 * ijélas líndžā
 * plenty of leaves

Calendar Constructions
Calendar constructions are a special subset of partitive NPs that specifically deal with indicating slices of time based on a calendar system. They're made up of two parts: a time specifier and a system descriptor. The time specifier is made up of an interval noun in the lative, such as éli (year) or tẽrkur (calendar day), and a number; as Mestian generally fails to distinguish ordinals from cardinals, these are usually supplanted by the lative or locative of the number respectively. The system descriptor is always in the partitive; it specifies the item relative or relating to which the specifier counts time — it can be any NP, from some as generic as sírū (life) to those as specific as Zortághrjan (the Dragon Imperium). Example phrases:
 * tẽrke pârīt sirúššą
 * one day of life
 * elíččē pâr-tā-urulíddžgarríččē ptaghrúnnu
 * the thirty-seventh year of the reign of ...

Verb Phrases
Mestian verb phrases (VPs) are built fairly simply: they are made up of a head in the form of a verb, and modifiers in the form of AdPs. VPs are divided into finite and non-finite VPs; finite VPs are fairly rigidly head-initial, and non-finite VPs can be either head-initial, head-medial or head-final, with a tendency towards a head-initial configuration. Non-finite VPs are essentially NPs with AdP modifiers. They are derived from finite VPs in ways that do not always imply a straightforward derivation path; they may also take part in clause-like transformations and possibly take arguments. Sample VPs:
 * tâwęškasī
 * do I begin to think?
 * a simple finite VP without any complements
 * skaghnyrškát firúraqan
 * I begin to wait for you while burning
 * a finite VP with an adverbial complement
 * hejpásī taur kíris lakáqan kiri
 * my looking at you while decorating you
 * a complex set of non-finite VPs, where one is a nominalised head and the other is an adverbial modifier

Non-finite VPs that act as NPs can take case and number markings by either directly inflecting or by taking a third person pronoun (usually <ŷlaj>) as a head which then inflects, and acting as its (unmarked) modifier. Such phrasal constructions are strongly head-initial.

Adjective Phrases
Mestian adjective phrases (APs) are fairly ruidmentary in their makeup: they are made up of a head in the form of an adjective or AP, and modifiers in the form of adverbs or AdPs. They are almost exclusively head-initial

Comparative Adjective Constructions
Constructions that involve adjective comparison are a special form of adjective phrase and are generally grouped into a concept termed comparative adjective phrase (CAP). A CAdP is made up of one comparative adjective and its complement. The complement may either be a noun phrase or nominal relative clause on one hand — making it a direct complement that takes the dative case on its head and all modifiers that refer to the head — or an adverbial clause on the other — making it an indirect complement.

An example direct complement:


 * íčīleja pirkúri
 * more human than the fish(es)

An indirect comparative complement may be introduced only with the interrogative adverbs such as sîxis (how much) and sîkkari (to what extent). This adverb is then always preceded by the conjunction īka (than).

Adverbial Phrases
Mestian adverbial phrases (AdPs) resemple APs in their makeup: they are made up of a head and modifiers, both of which can be either adverbs or other AdPs. They are exclusively head-initial.

Copular Constructions
Unmarked Mestian copular constructions are fairly fixed in their makeup: the most common copula is the particle ūmi, followed by its predicate in the lative, and then its subject in the nominative. This gives simple copular constructions of the shape <ūmi >.

Such copular constructions are in essence clauses without a functional verb; they are as such much more fixed in their structure. Topicalisation for the purposes of emphasis still follows the structure fairly well.

Copular Topicalisation
Copular topicalisation is a modification of the standard copular construction template. While the core template itself remains unchanged by the process, with a standard copula-lative-nominative order, it introduces an anomalous RC that comes before the construction, is introduced by kévaj acting as a relative head, and is made up of a malformed copula construction that involves the copular particle eza.

If the predicate of the main clause is topicalised, kévaj takes on the lative case and the relative clause then has an embedded nominative; if the subject is topicalised, the reverse applies. This fronting process essentially extracts the topicalised argument out of the copular construction, replaces it with ŷlaj appropriately inflected.

Adjective Decopularisation
Adjective copula constructions undergo decopularisation: they function completely without any copula, or rather involve a zero copula. In such situations, the AP goes before its NP and takes the lative case and neuter gender, regardless of the NP's case and gender; it, though, does agree with the NP in number.

Main Clauses
The Mestian main clause (MC), or the independent clause, is roughly equivalent to the top-level structure in a Mestian sentence. It's made up of at least one VP and optional NPs that correspond to arguments encoded in the VP. MCs can additionally be made up of copular constructions instead of having a regular VP.

On an order scale, Mestian MCs tend to be head-initial, with the VP coming first and its complementary phrases after it. Partitives tend to be more prominent than nominatives, which, in turn, are more prominent than accusatives; this gives an elementary sort of VSO~VOS word order, where the order of elements tends to vary based on what cases the arguments take. As Mestian is fairly highly marked morphologically and encodes little information solely using syntactical configurations, it allows fairly free word order where the most prominent or emphasised element may be pushed closer to the beginning of the clause or sentence.

Independent Interrogatives
Independent interrogatives in Mestian generally take the structure of the simple independent clause: the word order is essentially the same between the two. Simple independent interrogatives take the shape of a regular independent clause with the verb in the interrogative. Interrogatives that take interrogative pronouns merely use them in the same location they'd normally use a non-interrogative.

Subordinate Clauses
Mestian subordinate clauses (SCs) are a type of clause that cannot occur on its own and is instead part of a main clause. They are made up of two parts: a subordinate head, the pronoun gẽraj inflected for the clause's case, and the corpus, or the actual clause itself. SCs themselves don't feature any unique structure or exhibit a particular word order, although Mestian SCs generally take the subjunctive in a greater number of cases. They are most frequently head-initial, with the head being followed by a verb-initial corpus, and usually are put last in the sentence. They cannot be modified externally.

Relative Clauses
Mestian relative clauses (RCs) are a type of subordinate clause that functions as a NP head. They themselves are made of two parts: a relative head in the form of an NP, and a modifier clause. The modifier clause (MC) is thus said to be headed by the relative head (RH). MCs come in two different versions: a normal nominal MC (NMC), and a possessive MC (PMC).

NMCs, directly referring to the RH, are introduced by the relative pronoun, ûttaj, that inflects corresponding to its role in the sentence; as MCs are very starchly head-initial, this then forces prepositions to come right after the pronoun and act as postpositions.

PMCs, referring to possessions of the RH, are instead formed as transformations of possessive NP head-modifier pairs where the modifier becomes a RH and the head then becomes a part of the MC. They are introduced with the relative pronoun ûttaj that instead agrees in case and number with the head of the original NP (aka referent), now transformed into an NP constituent of the clause. This referent is then linked to the MC with a possessive pronoun that agrees in gender and number with the RH and takes the corresponding possessive case.

Verbs in RCs undergo deranking in certain contexts: if they refer to a possible or hypothetical situation, or carry adverbs of uncertainty, renarration, infirm evidentality and their like, they are transformed by the following mapping function:

Tense-mood combinations that do not map cleanly (primarily those involving irregular remnant tenses) either switch tense to a cleaner and more regular form that is then able to undergo such a reductive transformation, or force a suppletive stem replacement. Main clauses that feature imperatives, hortatives and interrogatives cannot undergo relativisation.