Naswiyan Thread

[Ares Land]

That's very nice! May I ask who are the speakers and what they're like?

[KathTheDragon]

The speakers are a group of about a thousand magic-using dragons living in a town hidden in the Scottish highlands. The backstory's really only relevant in the roleplays involving my character, who comes from exactly this community.

[bradrn]

The speakers are a group of about a thousand magic-using dragons living in a town hidden in the Scottish highlands. The backstory's really only relevant in the roleplays involving my character, who comes from exactly this community.

Maybe this is a silly question, but: if the speakers are non-human, then why does Naswiyan seem so similar to human languages? (Although I do see an easy resolution — you could say that the dragons learnt language from the humans! Although that somehow feels a bit unsatisfying.)

[Vardelm]

Maybe this is a silly question, but: if the speakers are non-human, then why does Naswiyan seem so similar to human languages? (Although I do see an easy resolution — you could say that the dragons learnt language from the humans! Although that somehow feels a bit unsatisfying.)

Or this is their communication translated into something humans can understand.

[KathTheDragon]

Because I'm too lazy to try to make something truly non-human, and my dragons already have very human-like cognition (because I'm literally trying to play one as a character and non-human cognition is hard to portray). Under those conditions, why shouldn't their language be human-like?

[KathTheDragon]

Verbs (part 2 of ?)

Mood

As mentioned earlier, Naswiyan possesses indicative, subjunctive, optative, imperative, mirative, and interrogative moods. The mirative and interrogative are conveyed with the particles kuš and re respectively, as mentioned earlier. The subjunctive and optative are expressed by prefixes u- and hi- on the finite verb (and otherwise jointly through the irrealis particle hipa). The imperative uniquely has its own stem, which distinguishes perfective and imperfective, but cannot take the perfect suffix.

Nominal Forms

The two participles and the nominal verb basically have the function of a relative clause, where the participles are direct relative clauses (antecdent = subject, "the man who saw me") and the nominal verb is an indirect relative clause (antedecent ≠ subject, "the man who I saw"). Another (optional) use of the passive participle is to express relative clauses where neither subject nor object is expressed, and the antecedent is some other component ("the man who is given to").

The infintive, meanwhile, is a very defective form, not requiring any arguments to be expressed, and simply expresses the action itself. Arguments can still be expressed, of course, through the two genitive constructions (more on which later).

I've been kinda tired all day so please ask me for clarification on anything here.

Superceded by this post.

[bradrn]

Because I'm too lazy to try to make something truly non-human, and my dragons already have very human-like cognition (because I'm literally trying to play one as a character and non-human cognition is hard to portray). Under those conditions, why shouldn't their language be human-like?

That makes more sense — thanks!

[sasasha]

Because I'm too lazy to try to make something truly non-human, and my dragons already have very human-like cognition (because I'm literally trying to play one as a character and non-human cognition is hard to portray). Under those conditions, why shouldn't their language be human-like?

Norse dragons were generated from humans who were cursed or whose greed turned them into dragons, e.g. Reginn, Fáfnir. Perhaps your dragons might have started out as humans, too?

[KathTheDragon]

As it happens, no. These dragons evolved long before humans did.

[Jonlang]

Not much to add, but I'd just like to point out that an abandoned name for one of my conlangs is Nasawian /nas.ˈaʊ̯.jan/.

[KathTheDragon]

I haven't forgotten this, just had a rough couple of days lately. More to come

[KathTheDragon]

Verbs (part 2 of 8)

Mood & Mood-Adjacent

Mood is expressed through the prefixes u- "subjunctive" and hi- "optative", as well as the particle hipa which covers both. The prefixes only occur on the finite verb, and exclude the particle tak when used, as it necessarily marks the indicative. hipa is only used in clauses with no finite verb.

The subjunctive is the general irrealis mood, marking most non-real events. One core use of the subjunctive is to express conditions:

u-

ꜱʙᴊᴠ-

tašo

notice\ᴘꜰᴠ-1ꜱ

iś,

=ꜱʀ

cyo

greet\ᴘꜰᴠ-1ꜱ

"If I see him, I'll greet him."

As a rule, it does not mark the future; only uncertain futures can be in the subjunctive. Certain futures (or, as least, futures perceived to be certain) are always in the indicative.

The optative, meanwhile, marks a desire of the speaker:

hi-

ᴏᴘᴛ-

kyac

live\ᴘꜰᴠ-2ꜱ

i

and

waþ

succeed\ɪɴꜰ.1

"Live long and prosper."

Contrasting with the optative is the imperative, which is built on a unique pair of stems (for perfective and imperfective aspects; the perfective can again be combined with ʔiye to form the habitual). Where the optative marks a desire, the imperative marks a demand, and has both 2nd and 3rd person forms (though only distinguished in the singular). These forms are:

	sg	pl
2	-Ø	-ən
3C	-u
3N	-i

Lastly, the interrogative and mirative are formed from the particles re and kuš respectively, together with a non-imperative clause. The interrogative is (naturally) used to mark questions, both polar and non-polar, while the mirative marks things that are surprising or unexpected to the speaker. When used with the subjunctive, it can in particular to be used to express doubt.

[sasasha]

I have no expertise to add anything. But FWIW I think there's a lot of cohesion and restraint at work here; it feels very natural, as if I'm reading a description of a natlang. It's a personal judgment, but many an overview of a conlang (including my own) reveals a kitchen sink of affixes/clitics which have dizzyingly varied and exotic forms. I feel you are hitting a sweet spot, where yours seem 'just right' to create comfort, naturalism and internal logic, as well as interest.

May I ask how Naswiyan is being used in your (I'm presuming) D&D/Pathfinder/other RPG campaign? From what you've said you're playing a Naswiyan speaking dragon. Does the language come into your gameplay at all?

Another question: has Naswiyan been affected by language contact with any of the other languages that hang around the Scottish Highlands?

[KathTheDragon]

Verbs (part 3 of 8) - Nominal Forms

Participles

Naswiyan has two participles, active and passive. Fundamentally, they're nouns expressing "one who Xs" and "one who is Xed", thus making them analogues to direct (subject as pivot) relative clauses. Each participle expresses aspect, though they're limited to perfective and imperfective, since they cannot combine with ʔiye. Additionally, they cannot take the perfect suffix, and as a result the perfective participles also have resultative overtones, especially the passive.

The passive participle can also be used when neither subject nor object are expressed in the relative clause, and the pivot is some other phrase:

su

man{ꜱɢ.ᴄ}

ytitu

give\ᴘᴘᴛᴄᴘ.ɪᴘꜰᴠ-ꜱɢ.ᴄ

nayoḫ

to-ɴᴍᴢ-ꜱɢ.ᴄ-3ꜱ.ᴄ

"The man who is being given to."

Note the resumptive pronoun, -ḫ, which is obligatory in this construction. This is the only available use with intransitive verbs:

šin

lake{ꜱɢ.ɴ}

nfiki

teleport\ᴘᴘᴛᴄᴘ.ᴘꜰᴠ-ꜱɢ.ɴ

aryeś

to-ɴᴍᴢ-ꜱɢ.ɴ-3ꜱ.ɴ

"The lake which was teleported to."

The active participle can also take a complement, representing the object of the verb, and when the complement is a personal pronoun, it takes the suffix form (otherwise dependent):

ħulalu

cook\ᴀᴘᴛᴄᴘ.ɪᴘꜰᴠ{ꜱɢ.ᴄ}

nag

meat{ꜱɢ.ɴ}

"The one who's cooking the meat."

The participles are also frequently lexicalised as agent and patient nouns, e.g. šagwəl "betray" → šugol "traitor", daro "choose" → driwi "choice, option". In these lexicalisations, it's typically the perfective participle used.

Nominal Verb

The nominal verb is a more direct nominalisation of the verbal clause, and in general preserves the argument structure exactly, as such featuring heavily in complement clauses:

ar

to

čaptiś

open\ɴᴍᴢ.ᴘꜰᴠ-ꜱɢ.ɴ-3ꜱ.ɴ

"... so that it opens."

It's based on the same stems as the finite verb, but being a nominal it also inflects for gender and number (defaulting to neuter singular, as here). It additionally cannot take mood prefixes, though it can take the perfect suffix.

The nominal verb can also be used as an indirect (not subject as pivot) relative clause. If the object is the pivot, then it can simply be omitted, whereas if some other phrase is the pivot, it is replaced by a resumptive pronoun. The subject can never be the pivot in this kind of clause, thus drawing a contrast with the active participle, and it can never be omitted, contrasting with the passive participle.

Infinitive

The infinitive is the generic verbal noun, expressing just the action itself. As such, it cannot take any TAM marking, though it can still have arguments added:

galse

drink\ɪɴꜰ-ꜱɢ.ɴ

-ḫ

-3ꜱ.ᴄ

ne

of{ꜱɢ.ɴ}

re

water{ꜱɢ.ᴄ}

"His drinking of the water."

It's often found in contexts where it's directly dependent on another verb, and can inherit all its TAM features (and commonly its subject) from that verb, making it a very common verbal complement:

re

ɪɴᴛ=

uwanyac

ꜱʙᴊᴠ-like\ɪᴘꜰᴠ-2ꜱ

śode

read\ɪɴꜰ-ꜱɢ.ɴ

-ś

-3ꜱ.ɴ

"Do you want to read it?"

Labile verbs have two infintives, encoding the two valencies, such as ladəŋ "entertain", lideŋ "laugh". The transitive infinitive is termed "infinitive 1" while the intransitive infinitive is termed "infinitive 2". Infinitive 1 is also the infinitive used by all transitive verbs (e.g. kaya "attack"), and intransitive verbs encoding an agentive action (e.g. sapət "jog"), while infinitive 2 is also used by intransitive verbs encoding a patientive action, such as ħimi "occur". There are two more infinitives as well: infinitive 3, used by stative verbs (e.g. pkas "be clean"); and infinitive 4, used by the small class of impersonal verbs (e.g. yti "rain").

[KathTheDragon]

I have no expertise to add anything. But FWIW I think there's a lot of cohesion and restraint at work here; it feels very natural, as if I'm reading a description of a natlang. It's a personal judgment, but many an overview of a conlang (including my own) reveals a kitchen sink of affixes/clitics which have dizzyingly varied and exotic forms. I feel you are hitting a sweet spot, where yours seem 'just right' to create comfort, naturalism and internal logic, as well as interest.

Thanks!

May I ask how Naswiyan is being used in your (I'm presuming) D&D/Pathfinder/other RPG campaign? From what you've said you're playing a Naswiyan speaking dragon. Does the language come into your gameplay at all?

Backstory, whenever my character names something, and sometimes when she insults people.

Another question: has Naswiyan been affected by language contact with any of the other languages that hang around the Scottish Highlands?

So far I've only made note of loanwords from Proto-Celtic, such as ktuni "human" < *gdonyos, or parək "bread" < *baragos, which also dates back to a time when dragons (and the precursor to Naswiyan) were spread fully over the British Isles. From around the early-to-mid first millennium, humans started massacring dragons in Europe indiscriminately, so there'd be no sustained human contact from that point onwards.

[WeepingElf]

I am not sold at all to the dragons thing (but I understand that your interest is in dragons, and that is OK to me), but the language you are developing for them is very beautiful, interesting and well-done!

[KathTheDragon]

Verbs (part 4 of 8)

Root Classes

Verbs have the clearest instantiation of the consonantal root system at the heart of Naswiyan morphology. Verbal roots can be divided into a number of formal classes based on several properties of the root, one of which is the number of consonants. The smallest roots have only 2 consonants, and the longest have 4 (though secondary stems can have up to 6). The other properties are as follows:

• "Strength": if the final consonant is ʔ, the root is called "weak"; if y or w, "semi-weak"
• "Integrity": if the second-to-last consonant is ʔ, the root is called "hollow"; if y or w, "semi-hollow"; if it is the same as the final consonant, "geminate". Roots may not end in ʔʔ, yy, or ww.

If a root is none of (semi-)weak, (semi-)hollow, or geminate, then it's called "strong".

It should be noted that ʔ very rarely appears as such, due to being more of a marker that there is no consonant in that slot (but the slot still exists). Word-initially (as other positions are already covered), it disappears before a vowel (e.g. ašra "fear" for *ʔašra), while before a consonant it's replaced by the vowel a, e.g. ašar "to fear" for *ʔšar. When a mood prefix is involved, however, instead an epenthetic glide appears before a vowel (uwašra "would fear" for *uʔašra), while the ʔ is simply absent before a consonant (ula "would be big" for *uʔla)

I'll give only a full paradigm for the strong verbs, as the forms in the other classes can be derived fairly straightforwardly through a fairly small set of rules that shall be given shortly. Vowels marked as ə change to a in open syllables. In terms of noun class, active participles are normally class A, passive participles are always class C (and are given here in the common gender), and the infinitives are all normally class A. The exceptions are as follows: infinitive 1 and 2 are class B in 3-consonant roots; the 2-consonant imperfective active participle is class B (so its stem is regularly CuCC₂-); any other active participle that would end in a cluster that is not a valid coda adds an epenthetic ə and is also class B (e.g. kumər "smth that burns").

2-consonant

	ᴘꜰᴠ	ɪᴘꜰᴠ
finite	CC-	CəCC₂-
ᴀᴘᴛᴄᴘ	CuC	CuCa
ᴘᴘᴛᴄᴘ	CiCu	CCiC₂u
ɪᴍᴘ	CaC	CCa(C₂-)
ɪɴꜰ.1	CaC
ɪɴꜰ.2	CiC
ɪɴꜰ.3	CaC
ɪɴꜰ.4	CiC

3-consonant

	ᴘꜰᴠ	ɪᴘꜰᴠ
finite	CəCC-	CCəC₂C-
ᴀᴘᴛᴄᴘ	CuCC	CuCəC₂C
ᴘᴘᴛᴄᴘ	CCiCu	CCaC₂iCu
ɪᴍᴘ	CCaC	CəCC₂aC
ɪɴꜰ.1	CaCəC
ɪɴꜰ.2	CiCeC
ɪɴꜰ.3	CCaC
ɪɴꜰ.4	CCiC

4-consonant

	ᴘꜰᴠ	ɪᴘꜰᴠ
finite	CCəCC-	CCaCəC₃C-
ᴀᴘᴛᴄᴘ	CuCəCC	CuCCəC₃C
ᴘᴘᴛᴄᴘ	CCaCiCu	CCəCC₃iCu
ɪᴍᴘ	CəCCaC	CCəCC₃aC
ɪɴꜰ.1	CaCCəC
ɪɴꜰ.2	CiCCeC
ɪɴꜰ.3	CəCCaC
ɪɴꜰ.4	CəCCiC

[quinterbeck]

There are four infinitives?? Are ɪɴꜰ.1 and ɪɴꜰ.2 the same as those you mentioned in relation with labile verbs earlier? Howcome you didn't mention 3 and 4?

[KathTheDragon]

Because I realised I really should've mentioned them earlier. I've extended the paragraph at the end of Verbs part 3.

[KathTheDragon]

Verbs (part 5 of 8)

Weak Verbs

The transformation from strong to weak is pretty straightforward. Finite verbs and passive participles both change the last radical to y (e.g. √spʔ → sapya "left alone", spiyu "someone who is left alone"), while active participles change it to "a" (best illustrated by example, supa "someone who left alone"), which is still class A (with plural supan). 2-consonant verbs are slightly different, as exampled by √zʔ → zu "someone who knew" (pl. zuwən), zuʔa "someone who knows" (pl. zuʔan). The remaining forms (infinitives and imperative) simply drop the last radical, and likewise act as class A nouns. It should be noted that infinitive 1 and infinitive 2 change their last vowels to a and i respectively.

Semi-Weak Verbs

The most obvious aspect of semi-weak verbs is monophthongisation, which is very common across the paradigm, but as it generally follows the usual rules, no mention will be made of most forms. The forms most deserving of attention are the finite imperfective of 2-consonant verbs (e.g. √ry "watch" → riya "is watching", exemplifying a general rule of the finite forms that ə plus glide becomes a high vowel, rather than a mid vowel), and active participles (aside from the perfective participle of 2-consonant verbs), where the final glide vocalises to a high vowel, and the noun class changes to B (e.g. √klw "search for" → kulu "one who searched for", pl. kulwən).

Hollow Knight Verbs

The main feature of the hollow verbs is that in the imperfective forms, there is always a surface ʔ, while there isn't in all the other forms, as it is dropped by various methods. The perfective and infinitives are easier to analyse than the imperfective, so we'll start there. When the ʔ comes between a vowel and a consonant in that order (i.e. VʔC), then the ʔ is simply dropped (e.g. √dʔs "dance" → dasa "danced", *dəʔsa), while in the reverse (i.e. CʔV) it becomes an a and contracts with the following vowel (e.g. desi "something that was danced", *dʔisi). Intervocalically, the ʔ is dropped and the two vowels contract (e.g. das "to dance", *daʔəs).

The imperfectives in part follow the same rules, except for the aforementioned surface ʔ. This is generally a means of avoiding homophony with the corresponding perfective forms, as can be found by following the above rules with the imperfective forms. This results in the following imperfective forms for 3-consonant verbs: finite daʔasa "is dancing", active participle duʔas "one who is dancing", passive participle daʔisi "something that is being danced", imperative daʔas "be dancing". Hopefully it isn't too hard to extrapolate the rules for longer (quasi-)roots.

Semi-Weak Hollow Verbs

These are derived by starting from semi-weak verbs instead of strong verbs when following the rules for hollow verbs. All that needs saying is that the perfective active participle of 3-consonant verbs is e.g. √ǵʔy "please" → ġuwi "one who pleased" (pl. ġuyən), √tʔw "remove" → tuwu "one who removed" (pl. tuwən).

Semi-Hollow Verbs

The only point of note is that, as mentioned for the semi-weak verbs, ə plus glide in the finite forms becomes a high vowel, instead of the regular mid vowel, e.g. √dyš "admit to" → diša "admits to".

Weak Semi-Hollow Verbs

Much like the reverse combination of glide and ʔ in roots, these combine the features of weak verbs and semi-hollow verbs. This needs no comment.

Semi-Weak Semi-Hollow Verbs

Again, very little of note.

Geminate Verbs

The general rule of geminate verbs is that a word-final single consonant is dropped (and like in the weak infinitives, final ə and e become a and i in infinitive 1 and 2 respectively), and a word-final geminate consonant changes the second one to a (or epenthesises a between the two consonants before dropping the second, whichever description you like more). These dropped consonants naturally reappear in suffixed forms, e.g. √mll "speak" → mula "one who spoke", pl. mullən. Aside from that, there are no complications.