Old Hieratic Tarandim, NP: Syntax

[Ares Land]

So, an old topic on the language formerly known as Kangrim attracted unexpected but much appreciated interest.

I was kind of bothered with that old version and, coincidentally, I had to work out a new version for the Middle Seas thread. So, here's how the language looks like now...

Old Hieratic Tarandim is the oldest attested language in the Tarandim family. It was -- in fact, it still is the language of scripture, magic and theology. The existing corpus is big on religious myth and bronze-age city-states accounting. For an idea of what its speakers were like, you can check out the Middle Seas thread over here.

The native name is Ušim Quaŋram Ṭarandim which means, well, 'Old Hieratic Tarandim'. Its speakers just called it quaŋeram ṭendah, 'hieratic, priestly language'.


A sample sentence:
Mɔšindim, quene he Hower ham mɔrɛgɛ ogɛmmɔraŋ ṭakummoŋaṭe Avva-he-Qeran.

mɔ<ši>nd-im / quen-e he ho-wer mɔrɛ-gɛ o-gɛ-<m>mɔ<ra>ŋ-im ṭa-kummoŋ-a-ṭe Avva he Qeran

wisdom.ABST<his>REL / king-PL that.idea red-ANIM.PL that human-ALL 4SG-ALL-<HAB>-rule<JUSS>-that THUS-decree-PRF-REMOTE Fountain that Qeran

The Fountain, Qeran, in His wisdom, appointed Red Kings to rule mankind.

This sentence provides examples of some of the key characteristics of the language:

• OVS word order is fairly common - more accurently, the unmarked order is patient-verb-agent
• It is head-marking; note the agent and patient prefixes: o-gɛ- in ogɛmmɔšiŋ 'that they decree'.
• There's a fair bit of infixing and other non-concatenative morphology, with for instance the infixes -ši- 'his', -ra 'jussive', vowel change: howɛh.
• Non-concatenative morphology also includes vowel changes: ovva 'flow (habitually)' > avva 'fountain'
• Nominal morphology and agreement is based on noun classes (20 of them). Three of them occur in that sentence: ABST (abstractions), ANIM (animate), THUS (verb phrases and nominalizations).

Verbs.

Stative verbs

The key distinction in the OHT verb system is that between stative and transitive verbs.
Stative verbs are intransitive and marked with patient markers:

gagammo vɛʾssɔmbɔr
monster ANIM.PAT<HAB>sleep-PAST.
The monsters were sleeping then.


Describing the exact scope of both categories is out of the scope of these short notes. But for now, we may say that stative verbs are often translated as adjectives.(*)

samɔh
1s.AGT-great
I am great.

(*)There's no such category as adjectives in the Tarandim languages; English adjectives may be translated as stative verbs, nouns or adverbials.


Transitive verbs

Transitive verbs are marked for agent and patient. Agent and/or patient prefixes may be freely omitted though, or more accurately, implied.

Let's run through a few examples:

hiṭaquen
1s.AGT-judge<HAB>
I judge; I am a judge.

Gie he Ṭagas hišṭaquen
Taqse 3s.POSS-people 1s.A-ANIM-judge
I judge the people of Ṭagas.

The 3rd person singular agent is usually unmarked:
šaṭaquen
∅-ša-ṭaquen
3p.PAT-3s.AGT-judge
He judges them.

But a '4th person' prefix marks a break in the animacy hierarchy:

Ve ðɔndar si he ṭohueqqɔm ewer ovṭasquende he vɔṭes.

ve ðɔn-dar si he ṭo-hue<q>qɔ=m ewer o-v-ṭasquen-de he vɔ-ṭes.

and time-COUNT.PL then HEAD-eat<HAB>=REL this-ANIM.PL 4.AGT-ANIM.PAT-judge<JUSS>-REMOTE that giant.llamas-HEAD.PL

And in those days the giant llamas will judge those who ate them.

Patient marking
The patient is marked with a prefix:

sa-sɔŋ
1s.P-sleep
I'm sleeping

Patient prefixes take an alternate 'shortened' form when following a subject prefix.
kɔssɔsim

kɔ-s-sɔsim

FOOD-1s.PAT-soothe

It soothes me.

Third person patients are marked with a noun class prefix.

Huaṭar ṭaṭara, touḫinder ṭɔvasara.

hua-ṭar ṭa-ṭa-ra, touḫinder ṭɔ-vasa-ra

day-TIME.PL TIME-count-PFR/ wall.BODY.PL BODY-build-PERF

The days are counted / the walls are built

Aspect
Unmarked verbs are typically habitual or imperfective:

veṭaḥ

write

he writes, he’s writing.

The perfective is marked with -a:
Veṭaha

veṭah-a

write-PRF

He has written.

Or -ra
Hueqɔra

hueqɔ-ra

eat-PRF

He has eaten.

Other allomorphs include -ŋa (replacing the preceding consonant):
tiqɔm 'he takes' > tiqɔŋa 'he took'
and -ša (likewise, replacing the preceding consonant):
kɔpus, 'he crosses it' > kopuša 'he crossed it'.

Meanings include (but are not limited to) perfective, experiential or resultative.
hiveṭara ṭɔṭe išvɔraŋah

hi-veṭa-ra ṭɔ-ṭe iš-vɔ<ra>ŋah

1s-write-PRF THUS-give 2s.PAT-study<JUSS>

I have written (it) so that you study (it) / I have written so study! (it)

gataša.

gata-ša

beget-PERF

He has children.

Ṭelemma siŋewɔra, ve queŋaš vɛna ṭembɔquaŋera Avva-he-Qeran.

Ṭelemma si-ŋewɔra / ve queŋ-aš vɛna ṭem-bɔ-quaŋer-a avva he qeran.

Ṭelemma PLACE-inside-PRF / and king-ANIM.PL ANIM-by 3P-ANIM-propitiate-PRF source that Aqran ANIM-for.

They had entered Ṭelemma, and they sacrificed their king to the Source-Aqran.

The perfect of stative verbs is a bit more difficult; but to put it shortly, it implies that the described state is the result of a previous action:
Ve he ŋammo ṭɔšɔwera ṭɔðaŋa

ve he ŋammo ṭɔ-sɔwe-ra ṭɔ-ða-ŋa

and that palace BODY-blue-PERF BODY-gold-PERF

And the palace was (painted) blue and (its walls were plated with) gold.

The habitualis marked by reduplication. With disyllabic verbs, the second consonant is reduplicated

tiqɔm > tiqqɔm
he takes > he always takes

Monosyllabic verbs reduplicate the first consonant instead; an epenthetic vowel is added when needed, when there is no prefix:

mɔŋ > momɔŋ
he rules > he's a ruler
gemɔŋ > gemmɔŋ
he rules all > he's an hegemon

[/b]The stative habitual conveys the implicature that the described state is essential, not episodic:
ṭɔðam > ṭɔððam
it’s (plated with) gold / it's golden > it’s made of gold.

vɛṭoma > vɛṭṭoma
he’s quiet > he’s a quiet one.

[bradrn]

Looks excellent so far! I’d be interested to see more of this.

A couple of questions:

A sample sentence:
Mɔndišim, quene he Hower ham mɔrɛgɛ ogɛmmɔraŋ ṭakummoŋaṭe Avva-he-Qeran.

mɔ<ši>nd-im / quen-e he ho-wer mɔrɛ-gɛ o-gɛ-<m>mɔ<ra>ŋ-im ṭa-kummoŋ-a-ṭe Avva he Qeran

wisdom.ABST<his>REL / king-PL that.idea red-ANIM.PL that human-ALL 4SG-ALL-<HAB>-rule<JUSS>-that THUS-decree-PRF-REMOTE Fountain that Qeran

The Fountain, Qeran, in His wisdom, appointed Red Kings to rule mankind.

You gloss the word Mɔndišim as mɔ<ši>nd-im — the infix appears to have moved between the gloss and the sentence. Is there any reason for this, or is that just a typo?

The key distinction in the OHT verb system is that between stative and transitive verbs.
Stative verbs are intransitive and marked with patient markers … Transitive verbs are marked for agent and patient.

Given that this just seems to be an intransitive/transitive distinction, is there any reason why the intransitive verbs are called ‘stative’ rather than ‘intransitive’?

(Also, this appears to be an unsplit ergative-absolutive system — is that correct? The use of the term ‘stative’ makes it sound like it could be an active-stative system instead, but it doesn’t seem like this language uses that.)

But a '4th person' prefix marks a break in the animacy hierarchy:

Ve ðɔndar si he ṭohueqqɔm ewer ovṭasquende he vɔṭes.

ve ðɔn-dar si he ṭo-hue<q>qɔ=m ewer o-v-ṭasquen-de he vɔ-ṭes.

and time-COUNT.PL then HEAD-eat<HAB>=REL this-ANIM.PL 4.AGT-ANIM.PAT-judge<JUSS>-REMOTE that giant.rodent-HEAD.PL

And in those days the giant rodents will judge those who ate them.

I’m not sure I understand what you mean by this — could you elaborate further on what you mean by ‘a break in the animacy hierarchy’?

The perfective is marked with -a:

A possible typo: you say this is the perfective, but all your examples seem to be perfect rather than perfective.

[Ares Land]

Looks excellent so far! I’d be interested to see more of this.

Thanks!

You gloss the word Mɔndišim as mɔ<ši>nd-im — the infix appears to have moved between the gloss and the sentence. Is there any reason for this, or is that just a typo?

Yes, thanks for catching this one. The error is in the first line: the correct form is mɔšindim, as glossed.

Given that this just seems to be an intransitive/transitive distinction, is there any reason why the intransitive verbs are called ‘stative’ rather than ‘intransitive’?

Hmm, this should've been written more clearly. The distinction is mostly semantic/aspectual: stative verb describe states, not processes. There's no change involved, so there's no real notion of agent/patient. A good comparison is to adjectives in IE languages.

(Also, this appears to be an unsplit ergative-absolutive system — is that correct? The use of the term ‘stative’ makes it sound like it could be an active-stative system instead, but it doesn’t seem like this language uses that.)

Hmm, I think I'll cover this in detail in the next post.
Basically, stative verbs are always intransitive (with patient markers), all dynamic verbs are ambitransitive: any verb will allow an agent prefix or a patient prefix or both.
So in the end, the system is active-stative.

For instance, 'he runs' is rɛťa, which also means 'he chases'.

But a '4th person' prefix marks a break in the animacy hierarchy:

I’m not sure I understand what you mean by this — could you elaborate further on what you mean by ‘a break in the animacy hierarchy’?

Agent marking doesn't distinguish noun classes; in fact it doesn't even mark plural for non-humans. The agent is most often the noun phrase that is highest up the animacy hierarchy; when the agent is 'less animate' than the patient, the prefix o- is used (instead of null marking as is usual). This is the case in the sample sentence (llamas are lower in the hierarchy than the persons they judge).

Oh, by the way, I caught and fixed an error in that sentence: vɔṭes are giant llamas, not giant rodents.

A possible typo: you say this is the perfective, but all your examples seem to be perfect rather than perfective.

I'll have to come up with better examples then Both labels would be equally valid, the range of meanings cover both perfective and perfect.

[bradrn]

Thanks for the clarifications!

But a '4th person' prefix marks a break in the animacy hierarchy:

I’m not sure I understand what you mean by this — could you elaborate further on what you mean by ‘a break in the animacy hierarchy’?

Agent marking doesn't distinguish noun classes; in fact it doesn't even mark plural for non-humans. The agent is most often the noun phrase that is highest up the animacy hierarchy; when the agent is 'less animate' than the patient, the prefix o- is used (instead of null marking as is usual). This is the case in the sample sentence (llamas are lower in the hierarchy than the persons they judge).

Minor clarification: does this mean that it uses a direct-inverse system?

[Vardelm]

Really like the sample sentences! They're quite evocative of some kind of religious text. This is something I'd like to do with my own grammars as they become more developed.

Stative verbs

The key distinction in the OHT verb system is that between stative and transitive verbs.
Stative verbs are intransitive and marked with patient markers:

What happens with verbs like "see" or "hear"? Are those just not considered stative or do they use a prepositional phrase or something for the object being perceived?


Compared to Kangrim, am I seeing correctly that much (all?) of the infixation was replaced with reduplication of the 1st or 2nd consonant?

With that said, I find the habitual use of statives ("implicature that the described state is essential, not episodic") is a really cool touch.

[Travis B.]

I'm amused by the giant llamas.

[Ares Land]

Minor clarification: does this mean that it uses a direct-inverse system?

Yes! I'm not always very good at labelling the different constructions... I've worked on the Tarandim language for oh, about ten years, often tracing changes from the proto-language and trying to figure out the result... so often enough, I instinctively know the rule but never really bothered to write it down. In a sense it's a little bit like documentating a natlang. I think xxx and elemtilas were talking about something like that in the plausibility thread.

What happens with verbs like "see" or "hear"? Are those just not considered stative or do they use a prepositional phrase or something for the object being perceived?

They're transitive verbs, but when talking about the state of being seen, the verbe is changed to its stative form (through a vowel change V > a in the first syllable)
kɔrum 'to see' > karum 'to be visible'
viṭeniha 'he saw us, we were seen' > veṭaniha 'people hear us, we can be heard, we're audible'.

Compared to Kangrim, am I seeing correctly that much (all?) of the infixation was replaced with reduplication of the 1st or 2nd consonant?

No, the reduplication was there already in the old version, but I decided it was an habitual. Actually there's more infixing and vowel change, I had to save these for another post
(I didn't really keep track of the changes, but the language is now more consistently prefixing, the order of affixes make more sense, and the diachronic causes of internal vowel change are different).

[Vardelm]

They're transitive verbs, but when talking about the state of being seen, the verbe is changed to its stative form (through a vowel change V > a in the first syllable)
kɔrum 'to see' > karum 'to be visible'
viṭeniha 'he saw us, we were seen' > veṭaniha 'people hear us, we can be heard, we're audible'.

Ah, that makes sense! Simple & nifty!

No, the reduplication was there already in the old version, but I decided it was an habitual. Actually there's more infixing and vowel change, I had to save these for another post

Gocha. Looking forward to it. I'm a sucker for infixation.

---

On a side note, after reading this last night, I was thinking about verbs stems in one of my 'langs (Devani, from the old board). I've wanted to find some way for the stems to alternate form for the main division in verbs, similar to Arabic's perfective/past & imperfective/present stems. Your use of reduplication got me thinking of that, and I found the example of Quileute on Wikipedia. I think something like that plus a few other changes will work well for me, so thanks for the inspiration!

[Ares Land]

Root and pattern morphoogy

OHT, like all Tarandim languages has a root and pattern morphology, somewhat similar to Semitic roots or IE ablaut.

The standard OHT verb root is disyllabic: CVCV(C). A series of pattern then apply to these roots, yielding different forms.

But let's have a closer look. We'll use vɛṭah, 'he writes' as an example.

Vowel change.

We distinguish vowel change from vowel loss here; vowel change always affect the first vowel.

• CaCVC forms a stative : vɛṭah > vaṭah 'it is written, it's writing, it's a written note'.
• CeCVC forms the atelic, a verb form with often idiosyncratic translations. The basic meaning is that there's no endpoint; the completion of the even is unknown, and irrelevant. In this case, vɛṭah > veṭah, 'he writes (letters), he is writing, he's a scribe'
• CaNCVC forms a stative, with an additional meaning of obligation: vanṭah 'it must be written, it's worthy of being written'. (N is a nasal mutation, but here N+ṭ yields the very boring nṭ)
• CoCVC is the dynamic equivalent: voṭah 'he can write, he must write'
• CiCVC forms the internal optative (that is, it marks the agent's wishes) viṭah, 'he wants to write'.

We have of course already mentioned the habitual, formed by reduplication:
vɛṭah > vɛṭṭah 'he writes often'

Infixes

• CVšCVC forms the external optative (used for the speaker's, or another person's wishes) It's also used for not entirely certain future events: vɛṭah > vɛšṭah, 'I hope that he writes', 'I think he'll write'.
• CVraCVC forms the jussive, used for orders, commands, and very probable future events: vɛraṭah 'he shall write', 'let him write'
• CVɓCVC is used for the past deductive; that is, for past event the speaker believes true due to deduction: vɛɓṭah 'he wrote', 'Obviously he wrote' 'He must have written, because...'
• CVbCVC is used for the past reportative, that is, events the speaker knows by hearsay: vɛmbṭah 'they told me he wrote'.

Second vowel deletion

A notable example of second vowel deletion forms a causative:
CVCVC > CVCCua
vɛṭah > vɛṭhua 'to dictate'

Other than that, it's chief use is in forming verbal nouns, when added to a class marker:

vɛṭah > vaṭah > vaṭh- + -ne (class SHARP, sharp, cutting tools) > vaṭhane 'stylus' (with epenthetic a added)

vɛṭh + ṭɔr (class BODY, buildings) > vɛṭhaṭɔr 'library'

Participle can be formed by adding the suffix -m to the CVCC form:
vɛṭham 'writing'.

The CVCC form can be combined with the other patterns:
vɛṭah > vaṭah 'written' > vaṭh- + -ši (class HAND, hand-held objects) > vaṭhaši 'book' (with epenthetic a added)

vɛmbṭham, 'writing (or so I heard)'
viṭham 'wanting to write'

Prefixing with deletion
Not to be confused with person or noun class prefixes, which don't affect the root.

A few common patterns:

• gaCCVC > forms a negative or an opposite stative : gavṭah 'illiterate'
• neCCVC > forms a comitative: nevṭah 'write with'
• ṭeCCVC > forms a reciprocal: tevṭah 'to correspond'
• giCCVC > softens the meaning of the verb: givtah 'to scribble'.
• ṭoCCVC > adds importance to the meaning of the verb: ṭovṭah 'to write scripture'.

You can use the preceding patterns with prefixed-deleted verbs: simply use the prefix vowel as the first root vowel.

ṭovṭah > ṭavṭah 'scripture'
> ṭavṭhaši 'religious book'
> ṭoravṭah 'he shall write (laws/scripture/myths) down'

[Vardelm]

I didn't realize before (maybe you hid it well) that this language had root & pattern morphology. Looks really good to me, and it's different enough from the Semitic languages that it has a different feel. As one example of that, I'd point to the "vowel change" section. Just changing the 1st vowel isn't something I've seen in Semitic languages, but it seems simple & natural.

I do like that each of the patterns has at least one V in there, so that the pattern keeps at least 1 vowel from the root/stem. It reminds me of the "characteristic vowel" in Tolkien's Adunaic.

[Ares Land]

OK, now I'll do something a little different. I tend to cover morphology first, and as I tend to gravitate towards synthetic languages, I kind of run out of steam before getting to syntax.

So this time I'll go through syntax, and comment occasionally on morphology when necessary -- or if any of you have questions, of course.

Usually I try to gloss sample sentences as exhaustively as possible. The results are sometimes a little overwhelming:

Ve ðɔndar si he ṭohueqqɔm ewer ovṭasquende he vɔṭes.

ve ðɔn-dar si he ṭo-hue<q>qɔ=m ever o-v-ṭasquen-de he vɔ-ṭes.

and time-COUNT.PL then HEAD-eat<HAB>=REL this-ANIM.PL 4.AGT-ANIM.PAT-judge<JUSS>-REMOTE that giant.llamas-HEAD.PL

And in those days the giant llamas will judge those who ate them.

So when it's not necessary to get deep into what each affix or pattern does, I'll simplify these ruthlessly:

Ve ðɔndar si he ṭohueqqɔm ever ovṭasquende he vɔṭes.

and times then DEF HEAD-ate those they-ANIM-will_judge DEF giant.llamas

Examples will be taken from the Ṭavṭah-he-Heš, the Tarandim holy book. Or rather, books, it is twice the Bible in length, and comprises myths, history, codes of law, case laws, sentences of the Ṭaga oracles and hymns of praise. Part of it is given in logographic, with an alphabetic gloss.
The Ṭavṭah-he-Heš was compiled during the Tarandim Dark Ages from earlier logographic texts and oral tradition -- its compilation spurred the creation of an alphabet and the beginnings of grammatical analysis: the ṭovṭaver -- the priests responsible for the compilation -- were convinced that proper pronunciation and language use were vital, and their native language had already changed quite a bit.
(How much had it changed ? Well, to use an analogy, it's almost as if American ministers used the King James Version with an IPA transcription on the side to get the 17th century English right.)

OK, that said, let's go on.

Stative sentences.

Stative sentences are used to describe states, as a very rough guide. It expresses a relation of being something, being somewhere or having a quality.

Word order is Topic - Subject - Predicate.

Mɔvimɓse, Eḫergi vɛguenṭah
prideful.PL / Eḫergi ANIM-third
The third prideful one was Eḫergi. / Among the prideful, the third one was Eḫergi.

Avva-he-Qeran morand ṭoganimga.
The Fountain-Qeran greatness THUS-unfathomable.
The Fountain, Qeran's greatness is unfathomable.

kampsṭa ḫɛrti ṭoḫoŋ
my.view road-your THUS-evil
In my eyes, your way is evil.

In all of those cases, the predicate is a stative verb: guenṭam, 'be the third', ganimga 'to be unfathomable', ḫoŋ 'be wrong, be evil'.

Definitions

Let's define these terms a bit: the topic is what the sentence is about; it sets a framework for the following predication. The topic is known to both speaker/writer and listener/reader: for that reason, it's considered definite.
It has either been mentioned before:
Mɔvimɓse, Eḫergi vɛguenṭam => mɔvimɓse, 'the prideful ones'
Or refers to common knowledge:
kampsṭa , 'my view'
Avva-he-Qeran, 'The Fountain, Qeran'

The subject of stative sentence is, quite simply, what is being defined or described by the predicate.

Topic and subject are typically pronouns, nouns, or noun phrases. Verb phrases may be used, but OHT has a strong preference for verbal nouns. But more on this later.

The predicate of stative sentences is the description; the following can be predicates:

Stative verbs:

Gueŋ vɛḫeððɔs , 'the king was wrathful'
king ANIM-<HAB>angry

A NP or pronoun with a pseudo-verb.

Defining pseudo-verbs will be somewhat circular; a pseudo-verb allows an NP to be used as a predicate.
The most common one is =si.

When the subject NP is present, it behaves as a clitic and attaches to the attribute NP:
Qeran sɔs he toḫiŋsi
Qeran sɔs he toḫiŋ=si
Qeran house that wall=be
'Qeran is a citadel'.

But, when the subject NP is omitted, pseudo verbs behave as a separate word and takes patient prefixes:

Sɔs he toḫiŋ kasi
house that wall FEM-be
She is a citadel.

Qeran sasi
Qeran 1.PAT-be
I am Qeran.

Pseudo-verbs don't take any other inflection. To express tense or modality, different pseudo-verbs are used.
For instance: ṭe 'must', sne 'be similar to' and pɔr, 'used to be'.

Moram vere vipɔr, tere rušmer gueŋše visne.
great lord we-PAST / therefore their.ghosts bones we-LIKE.
We used to be great lords and now we are like the ghosts of bones.

Qeran sasi. Sasi. Gonṭe saṭe.
Qeran 1.PAT-be / 1.PAT-be / always 1.PAST-must.
I am Qeran. I am. I always shall be.

More about topics

Subject=topic

Of course the subject can be the topic, and in fact, that's the most common case:

Avva mɔŋɛssi
Avva mɔŋɛh=si
Fountain lord=be
The Fountain is our lord.

Predicate=topic

But what if the predicate is the topic?

If the predicate is an NP with a pseudo-verb clitic, that's fairly easy:

NP1 NP2-si (NP1 is the topic) => NP2 NP1-si (NP2 is the topic)

Ŋɛsoṭah ṭaṭmovansi => ṭaṭmovan ŋɛsoṭassi
Our faith is our wealth. => our wealth is our faith.

If the predicate is a stative verb, we have to go through a nominalization first:

NP V => V-NOM NP-si

Eḫergi vɛguenṭah => Guenṭam Eḫergisi
Eḫergi ANIM-third => third.one Eḫerg-be
Eḫergi was the third one. => The third one was Eḫergi

How about the reverse transformation?
Conceivably, we could do either:
V-NOM NP-si => NP V-NOM-si (that is, given that the nominalization is a NP, moving it back to the predicate position with the appropriate pseudo-verb)
Guenṭam Eḫergisi => Eḫergi guenṭamsi
third.one Eḫergi-be => Eḫergi third.one=be
or: V-NOM NP-si => NP V (that is, simply reversing the previous transformation)
Guenṭam Eḫergisi => Eḫergi vɛguenṭah

In practice, both options are found in our corpus, but there's a preferrence to the second one, especially if the nominalization is fairly transparent. That is, if the noun moved to the predicate position can be transformed into a verb in a fairly trivial fashion, that stative verb will be used instead of a =si construction.

Topic=pronoun

What if the topic is a pronoun? Well, let me introduce you to a new friend, the topicalizer pseudoverb a:

NP pron-si => pron-a NP-si
Qeran sasi. => Saa Qeransi
Qeran 1.PAT-be => 1.PAT-TOP Qeran-be
I am Qeran => Me, I am Qeran

By the way, that has the implication that OHT has no independent pronouns at all, only pronominal prefixes attached to a verb or pseudoverb (titles and honorifics may also be used).

How not to topic.

The topic can simply be omitted, if it was mentioned earlier or is otherwise understood:
Geŋam ṭogansguɔ.
presence THUS-unbearable.
The (Fountain, Qeran's) presence is unbearable.
This is the beginning of an hymn of praise so the topic is evidently the Fountain, Qeran.

However, given that the topic is in the first position, in the absence of an overt topic, the subject can always be understood to be the topic. That is a problem, because the topic is understood to be a definite NP. What happens if the subject is indefinite, or is just being introduced.

For instance:
Mɔr he ḫiŋ hisrɛppɔr.
Mɔr he ḫiŋ hisrɛh=pɔr.
man ART youth prophet=was
The young man was a prophet.

Implies that we already know who the young man is.

If he's just being introduced to the narrative, than we use another pseudoverb: geŋ
Mɔr he ḫiŋ=geŋ hisrɛh=pɔr.
man ART youth=exist prophet=was
A young man was a prophet / There was once a young man who was a prophet.

[Vardelm]

Just a suggestion on the glossing: add the gloss tag around each word rather than the whole sentence. That breaks things up & makes it a lot more readable. Granted, it's a pain in the arse!

Ve ðɔndar si he ṭohueqqɔm ewer ovṭasquende he vɔṭes.

ve

and

ðɔn-dar

time-COUNT.PL

si

then

he

that???

ṭo-hue<q>qɔ=m

HEAD-eat<HAB>=REL

ever

this-ANIM.PL

o-v-ṭasquen-de

4.AGT-ANIM.PAT-judge<JUSS>-REMOTE

he

that

vɔ-ṭes

giant.llamas-HEAD.PL

And in those days the giant llamas will judge those who ate them.

I'll try to get through this when I'm not working.

[Ares Land]

Just a suggestion on the glossing: add the gloss tag around each word rather than the whole sentence.

Ah thanks!

Moving the subject.

The subject can be moved after the predicate:

Vɛŋɔveŋa hogar.

vɛ-

ANIM-

ŋɔveŋ

scream

a

PFV

hogar

wolf

A wolf screamed.

What's the difference with the presentative construction with =geŋ?
Moving the subject back is often done when it's not the focus, that is, when it doesn't relate to the listener/reader's expectations: it's neither contrastive, nor answering a question.

Compare:
Hogargeŋ vɛŋɔveŋa

hogar

wolf

geŋ

exist

vɛ-

ANIM-

ŋɔveŋ

scream

a

PFV

Possible translations:
That was a wolf that screamed.
Once there was a wolf that screamed.

Overt topic marking

The topic could be separated from the rest of the sentence by a pause. This can be identified in hymns, which are accompanied with basic rythm and pitch notation.

For instance:
Qeran, sɔs he toḫiŋsi

The topic could also be marked with the pseudo-verb =a that, as we've seen, is mandatory with pronominal topics:

Qerana, sɔs he toḫiŋsi
Qeran-TOP house of wall-be

And, finally, it could be marked by the clitic =ŋe
Qeraŋŋe, sɔs he toḫiŋsi
Qeran-TOP house of wall-be

Oh, by the way, what's the difference between the pseudo-verb =a and the clitic ŋe?
The one difference is that ŋe can't take pronominal prefixes.

*ŋea => saa 'as for me'

[WeepingElf]

This all reads very interesting. A rich and beautiful language it is you have created (or are still creating). Rock on!

[Ares Land]

This all reads very interesting. A rich and beautiful language it is you have created (or are still creating). Rock on!

Oh thanks! I'm still working on it... Hopefully, there's still a fair bit of syntax coming in

[Ares Land]

Locative sentences.

Location is expressed using pseudo-verbs or relational nouns


Pseudo verbs.

Just like the pseudo-verbs we've already covered attach as clitics, or are combined with pronominal prefixes.

Gɔmegtas, Gɔmemɔh, Ṭaŋŋarah Wa he Ṭoŋqeh.

gɔmegtas

all.father

/

gɔmemɔh

all.mother

/

ṭaŋŋarah

maker

wa

water

he

ART

ṭoŋ

life

=qeh

=OVER

All-Father, All-Mother, Maker are over the Water of Life.

Hege via Avvamar.

he-

ART-

ge

all

vi

1PL.PAT

-a

TOP

avva

Fountain

mar

out.of

We're all coming out of the fountain.

Ve Avva saseš, ve Avva saso.

ve

and

avva

fountain

sa

city

-seš

with

ve

and

avva

fountain

sa

city

-so

-under

And the Fountain was with the city, and the Fountain was under the city.

Relational nouns.

Most locatives, though, are relational nouns, most of which used for body parts.

A very literal translation will help: you don't say 'in front of the palace' in OHT, but:
ŋammo gaŋŋuoš

ŋammo

palace

gaŋŋuo

face

-š

3SG.POSS

The palace its face

Tɛŋueŋ paŋas, Tɛŋueŋ higos, geŋŋem tsuvos, Dasrevve gaŋŋuos.
Tɛŋueŋ my right hand, Tɛŋueŋ my right hand, rope my feet, Lord Jaundice my face.
Tɛŋueŋ (ie. Hell) is to my left, Hell is to my right, my feet are on a rope, and Lord Jaundice is in front of me.

Topic, Subject and Predicate.

A NP with a locative pseudo-verb can be used as a predicate; the sentence is then structured that way:

Topic Subject NP=pseudo-verb

As in:
Ṭaga Avva kɔnsɔsso.

Ṭaga

Topic

Avva

Subject

kɔnsɔsso

Predicate

In Taga,

Topic

the Fountain

Subject

is under the cave.

Predicate

A relational phrase, that is a noun phrase headed by a relational noun can likewise be used as a predicate.

Guorses hisrer he kɔnsɔs ŋevɔrši.
birth.ceremony oracles the cave womb-its
During birth ceremonies, oracles are in the cave's womb
During birth ceremonies, oracles are in the cave.[/i]

Si tere Mossa gueŋši he ŋammo sumeš
Then therefore Mossa king-its the palace arse-its
And the king of Mossa was behind the palace.

Relational phrases are never followed by the locative proverb =si. Other pseudoverbs may be used:

Guorses hisrer he kɔnsɔs ŋevɔršiṭe.

guorses

birth.ceremony

hisrer

oracles

he

ART

kɔnsɔs

cave

ŋevɔr

womb

-š

3S.POSS

=ṭe

=must

During birth ceremonies, oracles must be in the cave.

The locative phrase as topic

Locative phrases are natural topics, in that they set the stage, so to speak, for action or description. For this reason, they are most commonly found, not as predicates, but as topic.
If no suitable verb can serve as predicate, the existential geŋ is used.

Sa giš mɔr ḫiŋŋeŋ

sa

city

gi

gate

-š

3S.POSS

mɔr

man

ḫiŋ

young

=geŋ

exist

In the city, there was a young man.

We'll not in passing that the choice of relational nouns can be non-intuitive. For instance gi is almost always preferred, and may mean 'from', 'at', 'in', 'inside' when a city-state is mentioned.

Ve ŋammo gueŋu gueŋ he huŋaŋŋeŋ

ve

and

gueŋ

king

-u

-1PL.POSS

sume

arse

š

3.POSS

gueŋ

king

he

ART

huŋaŋ

enemy

=geŋ

-EXIST

And behind our king was the enemy king.

Non-locative relationals
For completeness's sake, I'll mentioned that some nouns behave as relationals, though they are not locatives. The most important of these is veḫem'name'.

Qeran veḫessa.
My name is Qeran, I am Qeran.

Sa Ṭaga veḫešša

sa

city

Ṭaga

Ṭaga

veḫem

name

=ša

3S.POSS

The city is called Ṭaga.

[Ares Land]

Dynamic sentences

Dynamic sentences require a dynamic verb. Dynamic verbs and sentences shows an action, which may be physical, perceptual or mental, occuring over a span of time.

The unmarked word order in dynamic sentences is Patient - Verb - Agent :

He sa ṭensgavɓor mɔšše Tsuggɔn.

he

ART

sa

city

ṭen-

3P.AGT-

s

PLACE

-

gavɓo

plunder

-r

-PFV

mɔš

descendant

-š

-his

-e

-PL

Tsuggɔn

Tsuggɔn

The sons of Tsuggɔn plundered the city.

A topic may be placed in first position (Topic - Patient - Verb - Agent)

Mardis, mɔse ṭevkesara he Maťandom.

Mardis

Mardis

vorver

warrior.PL

ṭe-

3P.AGT-

v-

ANIM-

kesar

gut

-a

-PFV

he

ART

maťan

south

dom

son.PL

As for Mardis, the Sons of the South disemboweled his warriors.

Patient or agents may be topicalized by placing them in sentence-initial position:

Tɛŋueŋ Mɔŋsier

The lords of Tɛŋueŋ

/

Ḫɛršiver he Guɔŋŋɔhueŋ, he Vagɔhueŋ

their messengers the White Screamer, the Child Screamer

/

ṭevkuŋuɔŋga

they summoned them

Tɛŋueŋ Mɔŋsier

Agent - Topic

/

Ḫɛršiver he Guɔŋŋɔhueŋ, he Vagɔhueŋ

Patient

/

ṭevkuŋuɔŋga

Verb

He Guɔŋŋɔhueŋ vɛṭaguo Ŋevaŋve

he Guɔŋŋɔhueŋ

the White Screamer

kaṭaguo

he sends her

Ŋevaŋve

Lord Many Insects

he Guɔŋŋɔhueŋ

Patient - Topic

kaṭaguo

Verb

Ŋevaŋve

Agent

As for the White Screamer, Lord Many Insects sent her out.

**

A brief cultural interlude. The White Screamer and the Child Screamer are owls (perhaps the barn owl and the screech owl). Owls were considered messengers of hell, as in Mesoamerica, or in Twin Peaks. It's a common cross-cultural trope, not a very surprising one if you've ever been surprised by an owl at night.
The Lords of Hell, such as Lord Jaundice or Lord Many Insects aren't gods or demons, but creatures, just like human beings, though still scary and shadowy figures, as they're responsible for death and disease. The mythic ancestors Kenṭevɛh and Ṭerɔvkah defeat them in an early story arc of the Ṭavṭah and free their prisoners, and so human beings and animals earned the right to be reincarnated. The Tarandim still believe in reincarnation, which they view positively, unlike most Earth cultures. The Lord of Hell themselves die and are reborn.
Several mythological figure in the Ṭavṭah attempt to make a deal with the Lords of Hell. It never ends well.

[Vardelm]

The unmarked word order in dynamic sentences is Patient - Verb - Agent :

Not sure why, but I love (what is essentially) OVS order.

He sa ṭensgavɓor mɔšše Tsuggɔn.

s

PLACE

Is this essentially a sort of classifier agreeing with "city"?

[Ares Land]

Not sure why, but I love (what is essentially) OVS order.

Well, it is nice to have an excuse to play with it!

s

PLACE

Is this essentially a sort of classifier agreeing with "city"?

Yep. OHT has a noun class system, with somewhere around 22 classes. (I mean, there are 22 now, but I still reserve the right to decide that there are more.

Basically:

• All nouns belong to one of those classes. They may end with a corresponding noun class suffix, though it should be kept in mind that each suffix has several allomorphs, and many words don't bear these. (For instance sa has no suffix at all)
• All verbs, stative or dynamic, agree with their patient in class with a class prefix, with slightly different forms depending on whether the class marker is the first or the second prefix on the verb.
• Verbs also agree with the agent, but agent prefixes use a different system, with direct-inverse marking.
• Agreement is based on morphology when the noun has an ending typical of a particuliar noun class, semantic when it doesn't. (And by 'semantic' I mean "learners have to learn the noun class by rote, because the semantics haven't been clear to anyone since the fall of Taga").
• Using the prefix ṭa- is never incorrect, except for animates. Using the animate prefix -v- is never incorrect either if you can contrive the patient to be animate in some way. (For instance, sa ṭevgavɓor for 'looted the city' would be correct; it's the inhabitants they've stolen from after all) Possibly different use reflect semantic nuances or dialectal variations that are now lost.
• Priests, scholar and magicians who learn OHT nowadays ascribe metaphysical meanings to the classes, and indeed to all aspects of the language. Any master worth his salt could contrive an hour's worth of exegesis from the -s- prefix you quoted above.

Here's the current list of noun classes, with the abbreviations I use:

classes.png (106.96 KiB) Viewed 7503 times

[Vardelm]

Yep. OHT has a noun class system, with somewhere around 22 classes. (I mean, there are 22 now, but I still reserve the right to decide that there are more.

Cool. The list looks good. Can't say why, but it looks/feels a little different than other ones I've seen. Yeah, there's similarities to other NC systems, but this has it's own vibe going. Nice work.

• All verbs, stative or dynamic, agree with their patient in class with a class prefix, with slightly different forms depending on whether the class marker is the first or the second prefix on the verb.
• Verbs also agree with the agent, but agent prefixes use a different system, with direct-inverse marking.

DAMMIT! You beat me to the punch w/ a conlang that has noun classes & direct-inverse marking! I haven't seen other conlangs that have that combo, although it seems like there must be at least 1 out there. Definitely looking forward to see this in action!!!