Naswiyan Thread

[KathTheDragon]

This is (going to be) a thread about Naswiyan. I'll try to make it interesting, and try not to forget to keep posting stuff. Something something unending praise.

Contents

Phonology
Nominal Morphology
Noun Phrases
Noun Phrases (cont.)
Clitics
Pronouns
Pronouns in Noun Phrases
Adverbs
Verbs:

1. Aspect
2. Mood
3. Nominal Forms
4. Root Classes
5. Root Classes (cont.)
6. Valency
7. Unmarked Arity Change
8. Marked Arity Change

Derivation:

1. Verb to Verb
2. Others

Clauses:

1. Introduction

[KathTheDragon]

Phonology

There's little to remark upon here, since my focus was more heavily on morphology and syntax.

Consonants:

	Bilabial	Dental	Alveolar	Post-Alveolar	Alveopalatal	Velar	Uvular	Glottal
Pulmonic Stop	p	t	ts <c>	tʃ <č>	tɕ <ć>	k	q	ʔ
Ejective Stop		t' <d>	ts' <z>	tʃ' <ž>	tɕ' <ź>	k' <g>	q' <ġ>
Fricative	ɸ <f>	θ <þ>	s	ʃ <š>	ɕ <ś>	x <ħ>	χ <ḫ>	h
Approximant		l	r		j <y>	w
Nasal	m	n				ŋ

Vowels:

	Front	Central	Back
High	i		u
Mid	e	ə	o
Low		a

Phonotactically, syllables are maximally CCVC, where onset clusters may be OO, ON, OR, OY, NR, NY, RY (O = stop/fricative, N = nasal, R = r/l, Y = w/y). Word-final syllables, however, may also end in one of the clusters OO, NO, RO, RN. Word-medial clusters may only be two consonants long, and both syllabify into the onset of the following syllable if possible, otherwise the first consonant syllabifies into the coda of the previous syllable. E.g. ma.kwi, hir.rət

Vowels in open syllables lengthen and show a quality difference wrt their short allophones: /i/ [ɪ~iː], /u/ [ʊ~uː], /e/ [e~ɛː], /o/ [o~ɔː], /a/ [æ~ɑː]. /ə/ is forbidden in open syllables, and whenever it would occur, it's replaced with /a/. Stress is also placed on the left-most long vowel in the stress domain, or the first vowel if all vowels are short.

There's a phonemic contrast between initial ʔ- and Ø-, in that the latter gains a phonetic initial consonant based on the previous word: if it ends in a consonant, that consonant is copied (e.g. hžan aliku [htʃ'æn ˈnɑːliːkuː]); if it ends in a front vowel, it gets prothetic [j] (e.g. aki im rit [ˈʔɑːkiː ˈjɪm rɪt]); if it ends in a back vowel, it gets prothetic [w] (e.g. su un imi [ˈsuː wʊn ˈniːmiː]; if it ends in /a/ or this is the first word in the phrase (the boundaries of which I haven't settled on), it gets prothetic [ʔ] (e.g. ina ano ul [ˈʔiːnɑːʔɑːnɔː ˈwʊl]).

[KathTheDragon]

Nominal Morphology

Noun morphology is pretty straightforward, with nouns inflecting for singular and plural number, and having either common or neuter gender (which is only distinguished morphologically in the singular). Nouns also belong to one of three stem classes (imaginatively named A, B, and C), distinguished by how they treat the gender markers and the plural suffix, as well as whether or not the stem is modified in inflection.

Class A nouns show no stem changes*, and do not have overt gender marking in the absence of any suffixes. Their plural is marked with -ən, and when the gender markers reappear (e.g. when the suffix pronouns are attached to mark possession) they are -o for common gender and -e for neuter. Examples are cur "a son", curən "sons", curoḫ "his son"; nat "a forest", natən "forests", nateḫ "his forest".

Class B nouns syncopate their last vowel* when any suffix is present, and is like class A with respect to gender markers and the plural suffix. Examples are parək "a loaf of bread", parkən "loaves of bread", parkoḫ "his loaf of bread"; ufeš "a horn", ufšən "horns", ufšeḫ "his horn".

Class C nouns show no stem changes, like class A, but always have overt gender marking, namely -u for common gender and -i for neuter. Their plural suffix is -an. Examples are nuźnu "a fish", nuźnan "fishes"; riki "a thought", rikan "thoughts".

There are a few nouns that make their plural in -e instead of -ən/-an, all of which are inherently dual, e.g. mawuri "an ear", mawure "ears".

A number of nouns can alternate between common and neuter gender, which fall into a few semantic classes. The largest one are verbal abstracts, where the common gender indicates a concrete noun (e.g. þaśat "a text"), while the neuter gender indicates an abstract noun (e.g. þaśat "writing"). Another notable class are nouns that denote a person or thing from a place, e.g. common naswi "a Naswiyan person" vs neuter naswi "the Naswiyan language". Other nouns display a strict animacy contrast, such as common caniyu "a twin" vs neuter caniyi "a copy".

---

* Class A and B nouns ending in a vowel in the singular might seem to violate this rule, but in reality this is not the case - stems ending in a glide show contraction, and a number of class B nouns have identical last two consonants, with the second one being regularly deleted when word-final. The contraction rules are ay > e, aw > o, ey/oy/iy/uy > i, ew/ow/iw/uw > u. Also note that class A nouns ending in -a have an apparent plural suffix -n, and the a is apparently deleted before a gender marker (though it's more accurate to say that contraction is taking place in both cases).

[Vardelm]

The common vs. neuter gender is cool. I like how it can be applied in different ways. Haven't seen a system like it before.

[KathTheDragon]

I took a fair bit of inspiration from the Hittite gender system for it, particularly in a few cases of alternating gender.

[KathTheDragon]

Addendum on proper nouns

Since proper nouns in Naswiyan are generally meaningful words and compounds, they can in principle already have plural morphology. Thus, in order to pluralise a proper noun, the particle lan is used, e.g. lan Kamare "people named 'Kamare'". This can come with notable changes in meaning, giving such alternations as Naswi "the Naswiyan language", lan Naswi "Naswiyan dialects".

[KathTheDragon]

Clitics

I want to talk about pronouns, but before I can talk about them with regards to syntax I need to talk about clitics, since virtually all pronouns show clitic behaviour in certain situations.

There are two broad types of clitic in Naswiyan, termed "free" and "bound". Bound clitics (marked with - like affixes) are those whose domain is a single word, and necessarily attach to that one word. Examples I've already used are lan "plural of proper nouns", a bound proclitic, and -ḫ "3sg common suffix pronoun", a bound enclitic. The remaining suffix pronouns are the remaining bound enclitics, while the inventory of bound proclitics also includes ħo "emphatic topic marker", nu "no, not, negative marker", and ya "vocative marker".

Free clitics (marked with =) are those whose domain is a clause, and when there are only proclitics or only enclitics present, they stack up on the first non-clitic word. When both proclitics and enclitics are present, however, all the clitics end up before the first non-clitic word, forming a single clitic chain. For example:

ina

not=

qafla

be_hungry

-ḫ

-3S.C

"He is not hungry"

qafla

be_hungry

-ḫ

-3S.C

śu

=3S.N

"He is hungry for it"

ina

not=

śu

=3S.N

qafla

be_hungry

-ḫ

-3S.C

"He is not hungry for it"

The free proclitics include the aspectual-modal markers hipa "irrealis", kuš "yet, but rather, mirative", re "interrogative", tak "realis indicative", and ʔiye "habitual-iterative", as well as art "thus, so", ina and inona, both markers of clausal negation, and šan "quotative, marker of direct speech". These are grouped into five slots in the clitic chain:

1. šan
2. art, re
3. tak, hipa, kuš
4. ina, inona
5. ʔiye

Further restrictions on co-occurence are that tak and ʔiye may not occur together, and re cannot occur with any slot 3 clitics.

The free enclitics are much more straightforward, as the only one not related to pronouns is iś "subordinator", which also comes first in the enclitic chain. iś is the only enclitic with restricted co-occurence outside of its slot, namely that it cannot occur with šan. Following iś are the "dative" pronouns, combinations of the preposition na "for, to" with a personal pronoun, but only when this functions as the recipient of a ditransitive verb. Lastly are the dependent pronouns, which serves as a good segue into the post on pronouns.

[dewrad]

This has strong Hurro-Urartian energy, and it is making me very happy.

[KathTheDragon]

I've never actually looked into Hurro-Urartian. Perhaps I should.

[bradrn]

There are two broad types of clitic in Naswiyan, termed "free" and "bound". Bound clitics (marked with - like affixes) are those whose domain is a single word, and necessarily attach to that one word. Examples I've already used are lan "plural of proper nouns", a bound proclitic, and -ḫ "3sg common suffix pronoun", a bound enclitic. The remaining suffix pronouns are the remaining bound enclitics, while the inventory of bound proclitics also includes ħo "emphatic topic marker", nu "no, not, negative marker", and ya "vocative marker".

Quick question: how do bound clitics differ from affixes? From this description they seem to act exactly like affixes.

[KathTheDragon]

In the case of the suffix pronouns, by virtue of promiscuous attachment and that they are always syntactically full NPs (they never appear in a manner that'd suggest agreement is taking place). The others are etymologically separate words, and are written as such, which is likely skewing my decision. But, I did think of this while I was writing, which is why I switched from marking them with = to - in glosses.

[KathTheDragon]

Pronouns

Pronouns inherit most features of nouns, and add another distinction on top of that: dependency. Dependent pronouns are clitics, being either free or bound depending on their function, which will be detailed as those functions come up. Independent pronouns, on the other hand, obligatorily stand in first position. Dependency is likewise assigned based on function. Personal pronouns are a minor exception, as they have a third variant in the suffix pronouns, which take up some of the roles of bound dependent pronouns.

Personal Pronouns

Personal pronouns distinguish first, second, and third person, singular and plural number, and in the third singular, also common and neuter genders, for a total of seven pronouns. Their suffix, dependent, and independent forms are as follows:

Suffix:

	sg.	pl.
1	-w ~ -k*	-n
2	-c	-čen
3C	-ḫ	-śen
3N	-ś

*-k is used after high vowels, -w otherwise

Dependent:

	sg.	pl.
1	ku	nu
2	cu	činu
3C	ḫu	śinu
3N	śu

Independent:

	sg.	pl.
1	ʔikku	ʔinnu
2	ʔiccu	ʔiččinu
3C	ʔiḫḫu	ʔiććinu
3N	ʔićću

Other Pronouns

All non-personal pronouns inflect as class C nouns with alternating gender, and most of them fail to distinguish dependency morphologically, the sole exception being the interrogative pronoun mu, having two independent forms, formal matu, and informal malu. They do however still show the syntactic effects of dependency, with the exception of remu "which?", which only exists as an independent pronoun.

The remaining pronouns are as follows:

• tu "that"
• talu "this"
• qilu "(an)other"
• zyu "somebody"
• nuwu "nobody"
• nuġu "each, any; every, all"

zyu, nuwu, and nuġu are all indefinite pronouns, which divide up the functional space as follows: nuwu is restricted to direct negation (tak nuwu im "nobody's here"); nuġu is used for free choice (dro čpiti nuġi "choose any door") and as the standard of comparison (žənraḫ her nuġu "she's faster than anyone"); and zyu is used for all other indefinite functions (sar yo zyu "go without anyone", hizyac iś zyu, ḫu saħ na imi "if you know anyone, bring them here", hipa zyi ul "something might be there", tašo zyi "I saw something", re cu ar haž zyi "will you eat anything?").

[dendana]

Super neat overview so far, which is great because I’ve only seen snippets of Naswiyan before...looking forward to more! Couple questions:

• What happens if you have two clitics in the same slot position? Like, for slot 2, “Is he thus hungry?”

• When do you use the suffix/dependent/independent pronouns?

[KathTheDragon]

What happens if you have two clitics in the same slot position? Like, for slot 2, “Is he thus hungry?”

You don't. In the specific case of art, it's used to mark sentences in a natural progression (e.g. steps in a recipe), so it wouldn't ever appear in a question even if it could. "thus, so" may not have been the best choice for translation.

When do you use the suffix/dependent/independent pronouns?

This will be addressed as those situations come up (pretty sure I said so)

[KathTheDragon]

Adverbs

Adverbs in Naswiyan are exclusively modifiers of the clause (and not necessarily the verb, as Naswiyan has non-verbal clauses as well). Adverbs may either occur alone or with a nominal complement, in which case they function like prepositions, or even as conjunctions if the complement is a nominal verb.

There are 15 primary adverbs, which are:

• her "up(wards)", "above, upon; more than"
• śir "down(wards)", "under, below; less than"
• ar "forwards", "to(wards), at; on, at"
• iḫ "backwards", "(away) from; off (of)"
• im "here", "(with)in, inside"
• ul "there", "outside"
• ħat "everywhere", "through(out), during"
• hol "nearby", "around, near (to)"
• yal "far away", "far from"
• i "together", "(together) with, and"
• yo "apart", "without"
• re "maybe*", "or*"
• na "**", "for, to, at"
• in "**", "by, due to"
• ke "similarly", "like, as"

* re basically expresses options and optionality, so as an adverb, it expresses that the clause is but one of several options (even when indicative), and as a preposition/conjunction, it introduces an alternative option
** I can't think of English expressions for na and in, but they express purpose and cause respectively.

Secondary adverbs are generally formed with the suffix -a, which can not only be attached to nominals, but also the primary adverbs as well to get meanings closer aligned with the prepositional meanings, e.g. ima "inside", ara "that way". With nominals, the meaning of the secondary adverb depends on the semantics of the nominal in question, e.g. sadi "tail" → sada "at the back"; haru "daytime" → harwa "in the daytime". Pronouns as a rule produce adverbs of manner: mu "what" → ma "how". The adverbs ḫna "before" and sda "behind" are irregular variants of ḫuna "at the front" and sada "at the back".

A second (much more minor) adverbialiser is the suffix -ya, which produces time adverbs generally from other adverbs (e.g. ḫna "before" → ḫnaya "in the future") but it can also be attached to nominals (e.g. nuwu "nothing" → nuya "never").

[KathTheDragon]

Verbs (part 1 of 8)

Verbs are complicated!

Not including combinations with clitics, verbs distinguish aspect (perfective and imperfective), mood (indicative, subjunctive, optative, and imperative), as well as a category called "perfect" which is semantically a bit nebulous, and so will be treated in combination with the aspects, which it combines with all three of. There are additionally four nominalised categories: the nominal verb, the active and passive participles, and the infinitive/verbal noun.

Aspect

The primary distinction is between perfective and imperfective, the latter of which is marked by doubling the last consonant of the verb root (more on which later). The clitic ʔiye adds a third aspect, habitual-iterative, and can only combine with perfective verbs. The aspects also contribute relative tense to the verb (though it can still be overridden by explicitly tensed adverbs such as imya "now"), with the perfective being relatively past, and the imperfective and habitual being relatively present.

The aspects interact with aktionsart in a fairly predictable way; in particular, the imperfective of stative verbs (such as lan "to be heavy") marks an on-going state without regard to its bounds, that of telic verbs (such as laməś "to clean") focusses on the process leading up to the culmination, while for punctual verbs (such as fayən "to blink") it is iterative. The interaction with the perfect is a bit more complex. The perfective perfect generally marks the end of the event, as well as the consequent state (if there is one). The imperfective perfect typically marks an interrupted version of the regular imperfective with the implication that under normal circumstances the event would have continued; this means that it's most often used in or with clauses expressing circumstance. The habitual perfect is more straightforward, marking a past habitual analogous to "used to" in English, though explicitly with the nuance that the event no longer takes place at the present time.

Time for some glossed examples:

gama

be_wrong\PFV

holiy

friend

-o

-SG.C

-c

-2S

"Your friend was wrong."

kmamra

burn\IPFV

mur

wood

-i

-SG.N

in

by

kumər

fire

"The wood is burning from the fire."

ʔiye

HAB=

guza

cough\PFV

-ḫ

-3S.C

ħat

through

harw

day

-ən

-PL

nuġ

all

-an

-PL

"She coughs every day."

ŋaʔasa

fly\IPFV

-na

-PRF

-w

-1S

iś

=SR

szəlna

CAUS-be_tired\PFV

-w

-1S

"I had been flying when I got tired."

[Vardelm]

Looks good so far. Sounds like it will take a few posts to really get a sense of the verbal system. #waiting....

One possible critique on phonology: in the past, I've been advised that having both / tʃ tɕ / might be a bit close to have both. If you have a natlang for precedent, then completely disregard this.

[Kuchigakatai]

Something something unending praise.

I do like your work so far.

There are two broad types of clitic in Naswiyan, termed "free" and "bound". Bound clitics (marked with - like affixes) are those whose domain is a single word, and necessarily attach to that one word. Examples I've already used are lan "plural of proper nouns", a bound proclitic, and -ḫ "3sg common suffix pronoun", a bound enclitic. The remaining suffix pronouns are the remaining bound enclitics, while the inventory of bound proclitics also includes ħo "emphatic topic marker", nu "no, not, negative marker", and ya "vocative marker".

Free clitics (marked with =) are those whose domain is a clause, and when there are only proclitics or only enclitics present, they stack up on the first non-clitic word. When both proclitics and enclitics are present, however, all the clitics end up before the first non-clitic word, forming a single clitic chain.

I sometimes wonder whether there was a point where Late Egyptian speakers felt that the preverbal suffix-holding words (jw, the modal > gnomic ḫr, the negator bn), once they became unable to be followed by a subject NP, attracted clitics because of their prosodial weight, that is, because they began bearing just enough secondary stress for it. So in a way they were "part of the verb", enough to attract the clitics towards themselves, but not too much, or they'd cling as proclitics onto the verb, as if Late Egyptian bn (or Naswiyan ina) had a prosody like:

ˌbn twj ḥr ˈsḏm
NEG 1SG.CLITIC on hear
'I'm not hearing'

or in your conlang: ˌina =śu ˈqafla-ḫ...

Diachronically I know this is useless because Old and Middle Egyptian had constructions like tm [""suffix"" pronoun or subject noun phrase] [verb phrase] already, but I'm just saying, synchronically as a mental model, at some point, maybe...

** I can't think of English expressions for na and in, but they express purpose and cause respectively.

'For this reason' and 'because of this'/'due to all this'? (I suppose 'therefore' originally had the meaning of either adverbial na or in, but then we're not talking about modern English anymore...)

One possible critique on phonology: in the past, I've been advised that having both / tʃ tɕ / might be a bit close to have both. If you have a natlang for precedent, then completely disregard this.

I think it could work if /tɕ/ is phonetically much more sibilant than /tʃ/. As if the contrast was, to put it in an exaggerated way, "[c]" (but still an affricate) vs. "[tʃ͡sʃ͡s]".

[bradrn]

One possible critique on phonology: in the past, I've been advised that having both / tʃ tɕ / might be a bit close to have both. If you have a natlang for precedent, then completely disregard this.

But lots of languages have /t͡ʂ t͡ɕ/, which really isn’t all that different. (And there’s even some with /t͡ʂ t͡ʃ t͡ɕ/!)

[KathTheDragon]

One possible critique on phonology: in the past, I've been advised that having both / tʃ tɕ / might be a bit close to have both. If you have a natlang for precedent, then completely disregard this.

My original justification was that this was made to be spoken by my dragon OC, who has a much longer mouth than humans.

I sometimes wonder whether there was a point where Late Egyptian speakers felt that the preverbal suffix-holding words (jw, the modal > gnomic ḫr, the negator bn), once they became unable to be followed by a subject NP, attracted clitics because of their prosodial weight, that is, because they began bearing just enough secondary stress for it. So in a way they were "part of the verb", enough to attract the clitics towards themselves, but not too much, or they'd cling as proclitics onto the verb, as if Late Egyptian bn (or Naswiyan ina) had a prosody like:

ˌbn twj ḥr ˈsḏm
NEG 1SG.CLITIC on hear
'I'm not hearing'

or in your conlang: ˌina =śu ˈqafla-ḫ...

Interesting idea, if I followed it correctly. I just woke up as I write this, so I might be a bit slow.