Zompist Bboard Again

Noun Phrases

We'll take a quick break from verbs to talk about noun phrases, which really should've been done immediately after the post on noun morphology but oh well it's here now.

Noun phrases come in three types, possessive, appositive, and connective, which will be tackled in reverse order. All three types are, on the surface, formed by placing the two nouns side-by-side, but as we'll see, it's a little more complex when the nouns are themselves phrases.

Connective noun phrases rather simply connect two nouns that are perceived to have a close relationship, e.g. su se "man and woman". Whether the nouns are joined with "and", "or", or something else is very dependent on context.

Appositive noun phrases are basically used to denote the same object with two different nouns, and the effect is by-and-large that of modifying a noun with an adjective or (non-genitive) determiner. The two nouns are generally required to agree in number and gender, and is the only time that nouns that normally have invariable gender can change it, e.g. ćanuk pyačetu "my daughter, the fox" vs pyačeti "fox". The major exception to the agreement rule is that pluralistic singulars (e.g. collectives or nouns denoting groups) do not pluralise, e.g. katən þiħne "7 lives" (literally "lives, a group of 7 things").

Possessive noun phrases are used (unsurprisingly) to convey possession, but only of the inalienable kind, e.g. sadi-lamok "the dragon's tail". As might be suggested by the hyphen, these phrases are very tightly bound, to where no clitics or apposited nouns can intervene, and they are also combined into a single phonological word, i.e. [ˈsɑːdiːlɑːmok]. Hence, they could easily be analysed as a type of compound, but as both members are freely inflectable, and as Naswiyan possesses another compound type where this is not the case, possessive phrases will be treated as phrases, and not as compounds.
It should be noted that when the first noun in a possessive phrase is also the first noun in an appositive phrase, the possessive phrase takes precedence and the apposited noun comes after the possessor, e.g. sadi-lamok lum "the dragon's long tail". This, naturally, creates ambiguity as to which noun in the genitive belongs to the apposition, and when it cannot be inferred by agreement (as in the example here) it is left up to context.

A frequent semantic subtype of the possessive is the partitive, where the possessor is plural (or a pluralistic singular) and the possessed denotes a part or parts of it, e.g. ne nat "a tree of the forest". This example embodies one of the typical forms of the partitive, where the possessor is the collective of the possessed. The other is where the possessed is an indefinite pronoun, e.g. nuġi muratən "any of the logs".

I assume disambiguation between the 3 types is mostly based on context? Or is there more info coming when you place them into phrases that allows you to tell which type you're dealing with?

It's contextual, yeah. Originally I didn't even distinguish possessive phrases graphically, but I realised that was silly since they have different phonological and syntactic behaviour.

Noun Phrases (part 2)

This should probably have been included at the end of the previous post, but it can go in its own post.

While the inalienable genitive is expressed through possessive phrases, alienable genitives are expressed with the noun ne, which can be rendered as "which belongs to". This is actually a derivative of the adverb na "for, to", which likewise can also be used to express possession, e.g. ćaŋsət talu na mutuk "this carving belongs to my mother". ne takes a possessor complement exactly like na does (in fact, it forms a possessive phrase with its complement, though it isn't hyphenated) and the phrase stands in apposition to the possessed:

hiži

food-ꜱɢ.ɴ

ne

of{ꜱɢ.ɴ}

fasa

wildcat{ꜱɢ.ɴ}

"the wildcat's food"

Being a noun, ne can inflect:

ŋusən

bird-ᴘʟ

nayən

of-ᴘʟ

rik

king{ꜱɢ.ᴄ}

"the birds of the king"

ne can also be used without an overt possessed noun, in which case it means "that/something which belongs to":

nayən

of-ᴘʟ

ktunyən

human-ᴘʟ

"what belongs to humans"

Pronouns in Noun Phrases

Pronouns, both personal and non-personal, have a number of syntactic features when used in noun phrases that set them apart from normal nouns. This post will detail these deviations, and any construction not mentioned should be assumed to follow regular nominal syntax.

Firstly, no pronoun can be directly used in a connective phrase, as nouns can be, so instead a dummy plural pronoun is inserted, and the two (or more) pronouns are replaced by the dependent pronouns. It should be noted that this construction can only express the "and" sense of connection; for the other meanings, paraphrases are required. To illustrate, see świdin ku cu "your and my book", žašŋo činu cu ḫu "I waited for you and her", and ʔiććinu talu tu holanu "My friends are him and them". These examples also demonstrate how the person of the dummy pronoun is determined by the pronouns being connected: if a 1st person pronoun is present, the dummy pronoun is 1ᴘʟ; if a 2nd person pronoun is present, the dummy pronoun is 2ᴘʟ; if only 3rd person or non-personal pronouns are present, the dummy pronoun is 3ᴘʟ.

Secondly, all pronouns that are second in apposition are dependent, and cliticise to the first non-clitic word of the noun (phrase) that it's in apposition to: su tan se "that man and woman", where the plural tan agrees with the connective phrase su se "man and woman".

Lastly, personal pronouns as possessors always use the suffix pronouns, regardless of alienability. Thus, sadi-lamok "the dragon's tail", and sadik "my tail", but hiži ne lamok "the dragon's food", hižik "my food". Personal pronouns can be used as the complement of ne (as suffix pronouns, since the complement is syntactically a possessor) but this never has the meaning of possession: naśət nayu "work for my benefit".

Just bumping to say I haven't forgotten about this thread. I've been on holiday and then I've been struggling to get the current post I'm working on into a state where I think it's comprehensible.

KathTheDragon wrote: ↑Sat Aug 08, 2020 8:32 am I've been struggling to get the current post I'm working on into a state where I think it's comprehensible.

Funny how what seems like it should be a simple explanation turns into a convoluted mental nightmare!

Looking forward to more, whenever it comes.

Well, half the trouble isn't that it's too complicated, it's figuring out how to present the information so it flows well. What order do you present ideas in, how do you chain ideas together, and so on.

Verbs (part 6 of 8)

Valency

First of all, I'm using "valency" in a wider sense than simple arity, where the following two verbs, despite both having one argument, are of differing valence:

ḫara

roar\ᴘꜰᴠ

lamok

dragon{ꜱɢ.ᴄ}

"The dragon roared"

mara

be_happy\ᴘꜰᴠ

im

in

lamok

dragon{ꜱɢ.ᴄ}

"The dragon was happy"

In the first, the subject is unmarked, while in the second, the subject is obligatorily the complement of the adverb im. These "marked" arguments (which do not necessarily use im) are on the surface identical to adverb phrases in all respects, but unlike normal adverb phrases, they cannot be moved before the verb without a clefting construction (which will be explained in detail later), nor can other adverbs be placed before them:

*im lamok mara
"The dragon was happy"

*mara ḫnaharu im lamok
"The dragon was happy yesterday"

We'll begin by looking at each arity separately, and then the ways verbs can change arity in the next post.

Avalent Verbs

Avalent verbs have no arguments, and there are (presently) only two verbs that are always avalent: yti "rain", and žis "snow". These both use infinitive 4, which is exclusive to avalent verbs.

Intransitive Verbs

Intransitive verbs have only a single argument, which can be "marked" or not. The relation between intransitive verb and non-marked subject is reminiscent of a possessive phrase: the subject is phonologically bound to the verb, and a personal pronoun uses its suffix form: guza tukuk [ˈk'uːts'ɑːtuːkʊk] "my brother coughed", rəsso "I was ill". Intransitive verbs use one of infinitive 1, 2, or 3, depending on whether they're agentive, patientive, or stative respectively: kara "cry", ćiweḫ "fall", tpać "remain".

All intransitive verbs with marked subjects are stative verbs of emotion or feeling, and use im, like mar "be happy" above.

Transitive Verbs

Transitive verbs have two arguments, a subject which is like a non-marked intransitive subject, and an object, which may or may not be marked. Just as the relationship between verbs and non-marked subjects is like a possessive phrase, that between verbs and non-marked objects is like an appositional phrase, and in particular, pronouns use their dependent forms: šgulaḫ ku "he betrayed me". Transitive verbs use either infinitive 1 or 3 depending on whether or not they're stative: non-stative verbs use infinitive 1 (qafəč "cut"), stative verbs use infinitive 3 (lya "permit").

Marked objects are more diverse than marked subjects, with no unifying form or semantics: gaʔaya katti imu "I like the kitten", təmgaḫ ħat čpitti "he passed through the doorway", þaćsan ar harwafacceḫ "we celebrated her hatchday".

Ditransitive Verbs

Ditransitive verbs have three arguments, a subject, an object, and a recipient. The subject and object are as in transitive verbs (including that the object may or may not be marked) while the recipient is always marked by na. The classic example of a ditransitive verb is yat "give": yta lamok świdi na ktuni "The dragon gave the book to the human". Ditransitives show the same distribution of infinitive 1 and 3 as transitives: stative verbs have infinitive 3, while non-stative verbs have infinitive 1 (tafər "offer to", žnaḫ "smell of").

Ditransitives with a marked object are (almost) exclusively limited to causatives (more on which later) derived from transitive verbs with marked objects, such as suġi "cause to like" from ġe "like" (sġaya holik imu na katti "my friend made me like the kitten").

Quick question: in your initial example mara im lamok, what arity is mara? And what valency is it, if that’s different?

mar is an unary verb, the subject being im lamok. It's analogous to e.g. dative experiencers in European languages. I don't have a formal way to talk about the valency other than to state "it has a single argument which is marked with im". Under the terminology I'm using, arity is the number of arguments, valency is how those arguments are coded (what adverb, if any, accompanies them).

Verbs (part 7 of 8)

Changing Arity

There are a number of ways that a verb can change its arity, which can either be marked (e.g. ćan "be sitting" → sućon "seat smn."), unmarked (fiḫen "melt (intr)" ~ faḫən "melt (tr)"). When the change is unmarked, it's not possible to determine which member of the pair is more basic, so they're best treated as an inherent capacity for the verb to alternate between two arities. Marked changes are much clearer, as the verb with the marking is derived from the verb without.

Verbs with Alternating Arity

A formal detail to point out is that alternating verbs in principle have two infinitives, one per arity. In many cases they end up being the same, such as lma "be free" ~ lma "be free from", but in the rest they are distinct, such as ħilu "cook (intr)" ~ ħalo "cook (tr)". These can be connected with the semantics of the alternation: lma deletes its transitive object (ʔiye lamyo "I am free", ʔiye lamyo rik "I am free from the king"), while ħilu~ħalu deletes its transitive subject (ħalwa nag "the meat cooked", ħalwo nag "I cooked the meat").

Avalent~Intransitive Verbs

Three verbs alternate between avalent and intransitive: sŋi "be hot", tŋi "be warm", and źir "be cold". When avalent, they are weather verbs and use infinitive 4, and when intransitive, they are verbs of feeling, and so have a marked subject marked with im, and being stative they use infinitive 3.

Intransitive~Transitive Verbs

There are two types of alternating intransitive~transitive verb: those that delete their subject (called labile), and those that delete their object (called ambitransitive). Like transitive verbs, ambitransitive verbs always use either infinitive 1 or 3 depending on whether or not they're stative, as stativity doesn't change between the transitive and intransitive forms. Labile verbs, on the other hand, all use infinitive 2 in their intransitive form, by virtue of retaining the patient argument, rather than the agent (e.g. saġa~siġi "break").

Ambitransitives are particularly rich in marked objects, such as rak "think": rəkko "I'm thinking", rəkko ar žyic "I'm thinking about what you did". Labile verbs, meanwhile, are very rarely marked, with þaḫ~þiḫ "seem, look like" being the only example: þəḫḫac ke mutuk "you look like my mother", þəḫḫa ke śeteḫ "it seems that he's leaving".

Transitive~Ditransitive

There are three kinds of alternating transitive~ditransitive verb: those that delete their subject (analogous to labile verbs), those that delete their object (analogous to ambitransitive verbs), and those that delete their recipient. Only the latter exist as morphologically simple verbs, such as pasəd "forget, forgive" (pasdo wišsec "I forgot about hurting you", pasdo no wišsec "I forgave myself for hurting you"), while the first two both only exist as causatives (and so will be covered in the section on causatives). Much the same infinitive alternations are found with transitive~ditransitive verbs as with intransitive~transitive verbs: those that delete their subject alternate between infinitive 2 and infinitive 1, while those that delete their object or recipient always have either infinitive 1 or infinitive 3 depending on whether or not they are stative.

I don't have much to offer, except this looks way cool! I love the Anatolian aesthetic, and the stuff you've been doing with arity and valency is very nice!

Verbs (part 8 of 8)

This has been way too long coming and I've been putting it off for no good reason at all.

Marked Arity Change

Marked arity changes are unlike their unmarked counterparts, as mentioned above. There's no one type of marking either, as the passive uses a bound clitic, the detransitive uses a free clitic, and the causative uses a root-level affix.

Passive

The passive serves to mark the agent of a verb as background information, allowing it to be omitted, and optionally reintroduced as the complement of the adverb in:

qafča

cut\ᴘꜰᴠ

ħulu

cook

kasi

cheese

"The cook cut the cheese"

qafča

cut\ᴘꜰᴠ

-t

-ᴘᴀꜱꜱ

kasi

cheese

"The cheese was cut"

qafča

cut\ᴘꜰᴠ

-t

-ᴘᴀꜱꜱ

kasi

cheese

in

by

ħulu

cook

"The cheese was cut by the cook"

Note that a reintroduced agent is an adjunct rather than an argument, and as such can be fronted for emphasis, or freely reordered with other adjuncts:

in

by

ħulu

cook

qafča

cut\ᴘꜰᴠ

-t

-ᴘᴀꜱꜱ

kasi

cheese

"It was the cook who cut the cheese"

qafča

cut\ᴘꜰᴠ

-t

-ᴘᴀꜱꜱ

kasi

cheese

ḫnaharu

yesterday

in

by

ħulu

cook

"The cheese was cut yesterday by the cook"

The passive suffix cannot co-occur with suffix pronouns, due to its recent historical origin as one of them, so a pronominal passive subject remains as a dependent pronoun, as in the active verb:

lśa

lick\ᴘꜰᴠ

cu

=2ꜱ

lamki

dragonet

"The dragonet licked you"

lśa

lick\ᴘꜰᴠ

-t

-ᴘᴀꜱꜱ

cu

=2ꜱ

"You were licked"

The passive contrasts with labile verbs, as they completely remove their agent from being involved in the action. Compare čapta čpiti "The door opened (on its own, perhaps the latch is loose)" with čaptat čpiti "The door was opened (by someone intentionally)".

Detransitive

Similarly to the passive, the detransitive indicates that the object is background information, and can thus be omitted. Unlike the passive, though, it cannot be reintroduced as an adjunct:

þasya

write\ᴘꜰᴠ

-ḫ

3ꜱ.ᴄ

þśiyan

word-ᴘʟ

her

on

świdi

book

"He wrote some words in the book"

þasya

write\ᴘꜰᴠ

-ḫ

3ꜱ.ᴄ

ŋa

=ᴅᴇᴛʀᴀɴꜱ

her

on

świdi

book

"He wrote in the book"

The detransitive marker is a free clitic occupying the same slot as the dependent pronouns, hence follows all other clitics in the clitic chain:

art

thus=

kuš

ᴍɪʀ=

ina

ɴᴇɢ=

nac

=to-2ꜱ

ŋa

=ᴅᴇᴛʀᴀɴꜱ

yta

give\ᴘꜰᴠ

-ḫ

3ꜱ.ᴄ

"And yet he did not give to you"

The detransitive does not contrast with ambitransitive verbs, so the detransitive is unavailable for that class: *məllo ŋa "I am speaking".

Causative

The causative is used to add an external agent that causes the action to take place. There are actually two ways of constructing the causative: the s-causative, where the root is prefixed with s, and the ža causative, where the verb is made into a complement clause of the verb ža "do". It should also be noted that s-causatives are treated as if they were verb roots for the purposes of inflection and derivation, which is not the case with ža causatives. Also note that s-causatives have a variant type of infinitive 1 with u instead of a as its vowel, e.g. sukru "insult" (← kara "cry").

In the s-causative, the old subject is demoted to the complement of na, with the external agent taking its place:

nərra

learn\ɪᴘꜰᴠ

-n

-1ᴘ

kur

sun

ulya

then

"We were learning about the sun"

snanra

ᴄᴀᴜꜱ-learn\ɪᴘꜰᴠ

nan

=to-1ᴘ

mutuk

mother-ꜱɢ.ᴄ-1ꜱ

kur

sun

ulya

then

"My mother was teaching us about the sun"

In the ža causative, no such adjustment takes place, as the external agent becomes the subject of ža:

žya

do\ᴘꜰᴠ

mutuk

mother-ꜱɢ.ᴄ-1ꜱ

nərra

learn\ɪᴘꜰᴠ

-n

-1ᴘ

iś

=ꜱʀ

kur

sun

"My mother made us learn about the sun."

The s-causative is only used with intransitive and transitive verbs (but not intransitive verbs with marked subjects), while the ža causative is mainly used with ditransitive verbs and intransitive verbs with marked subjects, but can in principle be used with any verb.
Where both causatives are available, as in the examples above, the s-causative tends to be more lexicalised than pure causation, whereas the ža causative strictly expresses causation.

Another interesting point about the s-causative is that when it is made from a stative verb, the causative has change-of-state semantics, and is also labile, with its intransitive counterpart using infinitive 2: ġnas "be healthy" → suġnos "cure (of an illness)" ~ siġnes "recover (from an ailment)".

I've gone back and edited the opening post with a table of contents.

KathTheDragon wrote: ↑Sun Aug 30, 2020 7:33 am I've gone back and edited the opening post with a table of contents.

Nice! Simple & effective. I may play copy-cat.

By all means.

Derivation (part 1 of 2)

The boundary between inflection and derivation is very blurry in Naswiyan, as most formations are highly productive, or else fossilised. A good case is the participles, which on the one hand are used to express certain kinds of relative clause (e.g. su wurnu "the man who fights against me"), but on the other are also used to denote agent and patient nouns (wurn "enemy, opponent").

Verb to Verb

The s-casuative is the most common verb-to-verb derivative, and is covered in a fair amount of detail in this post.

The pluractional is formed by duplicating the last two consonants of the root, e.g. daġən "lose" → daġnaġən "lose multiple times". The meaning of the pluractional is generally that the verbal action occurs multiple times in some connected manner, in a way that depends a lot on the plurality of its arguments. With a plural (whether grammatically or semantically) subject, it often indicates that each member of the subject performed the action separately:

duma

arrive\ᴘꜰᴠ

-n

-1ᴘ

"We arrived (as a group)"

dwamuma

arrive~ᴘʟᴜʀ\ᴘꜰᴠ

-n

-1ᴘ

"We arrived (individually)"

But it can be modified by adverbs to alter how the subject is split up:

dwamuma

arrive~ᴘʟᴜʀ\ᴘꜰᴠ

-n

-1ᴘ

cana

two-ᴀᴅᴠᴢ

"We arrived in pairs"

With a singular subject (or a plural subject modified to be represented as a unified group) and a plural object, it tends to indicate that the subject performs the action on each member of the object, fairly parallel to the case of a plural subject:

kamra

burn\ᴘꜰᴠ

-w

-1ꜱ

nan

tree-ᴘʟ

"I burned the trees (all in one go)"

kmaramra

burn~ᴘʟᴜʀ\ᴘꜰᴠ

-w

-1ꜱ

nan

tree-ᴘʟ

"I burned the trees (individually)"

kmaramra

burn~ᴘʟᴜʀ\ᴘꜰᴠ

-w

-1ꜱ

nan

tree-ᴘʟ

cana

two-ᴀᴅᴠᴢ

"I burned the trees in pairs"

With a singular subject and a singular object (or no object at all), it tends to indicate that the subject performs the action iteratively:

ŋda

sleep\ᴘꜰᴠ

-ḫ

-3ꜱ.ᴄ

talaharwa

today-ᴀᴅᴠᴢ

"I slept today"

ŋdəŋda

sleep~ᴘʟᴜʀ\ᴘꜰᴠ

-ḫ

-3ꜱ.ᴄ

talaharwa

today-ᴀᴅᴠᴢ

"I slept several times today"

Derivation (part 2 of 2)

Verb to Noun

The use of the participles as agent and patient nouns is touched upon here.

The instrumental is generally used to denote a tool or other noun used to perform an action (þaśa "write" → maþśuwi "pen", ŋas "fly" → maŋosi "wing"), but it can also have locational semantics (ka "live" → makwi "home, homeland"). It is always neuter and class C. Formally, it has the following patterns when strong: 2lit maCCi, 3lit maCCuCi, 4lit maCCoCCi. The only deviation worth noting is that weak verbs change their final ʔ to w. Everything else can be determined using the general principles laid out in this post.

Noun to Noun

The collective and singulative are formed with the same suffix, -ət (or -at when applied to a class C noun). The singulative sense is only used with mass nouns, and nouns inherently denoting groups, while the collective is used with any other noun: wikər "grass" → wikrət "blade of grass", tuwiḫ "(set of) claws" → tuwiḫət "claw", isi "person" → isyət "community". The two formations are distinguished by their gender: singulatives are common, while collectives are neuter. The suffix is class A when applied to a class B or C noun, and class B when applied to a class A noun. A particular specialisation of the singulative is found when applied to numbers, producing ordinals: hira "(a group of) 3" → hirrət "3rd". The collective is also specialised when applied to infinitives, producing neuter abstract nouns, with common gender variants with concrete meaning: dayəš "confess" → dešət "confession", ḫašən "blow" → ḫašnət "jet of air".

The nisbe (a term ultimately taken from Arabic linguistics) is a very general derivational suffix. The general form is -i, but it gains an intrusive t when applied to a class C noun, or when immediately following another nisbe suffix: lamok "dragon" → lamki "hatchling; egg" → lamkiti "oval, ellipse", mažnuḫi "nose" → mažnuḫti "nostril". Like the collective/singulative suffix, it is class A when applied to a class B or C noun, and class B when applied to a class A noun. It can also be applied to adverbs, both primary and secondary. Nisbes from secondary adverbs are always class A, while nisbes from primary adverbs are unpredictably class A or B. Semantically, the nisbe is very complex, and its sense varies wildly depending on the semantics of its base. The general sense is "that associated with X", but it frequently becomes more specific in more regular ways. Some of the more common classes of nisbe will be listed here.

• from an adverb: "someone/thing that is X" (hol "close by" → holi "companion")
• from an infinitive: "ability to X, to be Xed" (ŋad "sleep" → ŋadi "tiredness", la "count" → le "countable")
• from a location: "person/thing from X" (nasu "valley" → Naswi "Naswiyan person; Naswiyan language")
• from an animal: "young X, infant X" (kata "cat" → katti "kitten", isi "person" → isyi "child, infant")
• from a number: "the numeral X" (þiħne "(a group of) 7" → þiħnati "the numeral 7")

Adverbialisers

There are two adverbialising suffixes, -a which creates general adverbs whose semantics are broadly locative or similative (haru "day(time)" → harwa "in the daytime"), and -ya which creates specifically temporal adverbs (nuġi "everything" → nuġya "always"). -a can be applied to any noun or primary adverb, in the latter case modifying the adverb's meaning in a less-than-predictable way (im "here" → ima "inside", hol "nearby" → hola "again"), while -ya can only be applied to pronouns and locative/directional adverbs (nuwi "nothing" → nuya "never", ar "forwards" → arya "soon").

Clauses (part 1 of ?)

Introduction

The primary distinction between clauses in Naswiyan is between verbal and non-verbal clauses. Verbal clauses are headed by a verb, while non-verbal clauses are not:

ʔiye

ʜᴀʙ=

ana

dwell\ᴘꜰᴠ

aki

moon

im

in

rit

sky

"The moon dwells in the sky"

aliku

deer-ꜱɢ.ᴄ

tu

that-ꜱɢ.ᴄ

ar

towards

nat

forest

"That deer is heading towards the forest"

Non-verbal clauses may be further subdivided into nominal and existential clauses. Nominal clauses are those that feature two independent noun phrases, while existential clauses feature only one:

inona

ɴᴇɢ=

ćanuk

daughter-ꜱɢ.ᴄ-1ꜱ

šugol

traitor

"My daughter is not a traitor"

hipa

ɪʀʀ=

reyən

river-ᴘʟ

im

here

"There might be rivers here"

As can be seen, existential clauses very frequently feature adverbs, but these are not required:

tak

ᴅᴇᴄ=

guyeħ

danger

"Danger!"

Similarly, nominal clauses may feature adverbs:

surśoc

teacher-ꜱɢ.ᴄ-2ꜱ

ku

=1ꜱ

sdaharwa

tomorrow

"I will be your teacher for tomorrow"