Rambling about a somewhat routine classical language

A forum for all topics related to constructed languages
Post Reply
User avatar
Nmmali
rupestrian
rupestrian
Posts: 16
Joined: 01 Apr 2019 06:17
Location: The United States

Rambling about a somewhat routine classical language

Post by Nmmali »

Preface
I have hitherto dedicated my efforts to a language called Dhaujai. Over the past year, I sought to combine Japanese and Romantic grammatical patterns, which lead to wild variations in head directionality, morphology, and syntax in general. In short, the latter is a mess and I am done with it. Incidentally I’m also done with trying to handle the world, so I’ve put the Dhau people on an island in the corner of the map, on the backburner, perhaps in perpetuity, and shall now commence, from scratch, at a new and more feasible starting point - that of an ancient and entrenched river-valley civilization, which can form the bedrock of new languages and cultures further down the line. In any case, for those purposes for which I need to have documents that none other can read, it is best to distance the language from a European paradeigm of inflection.
What does one do when he encounters a corpse in an RPG? Loot it. I hate letting perfectly good words go to waste, especially seeing as they are already securely lodged in my brain, and, as such, I shall strip the late Dhaujai of its numbers and basic vocabulary, and adjust them to the phonology below described.
Here we go.
The language may be variously named Old Cortar (stress the first syllable, like quarter), Old Cottar, Old Cortari or Old Cottari, and, being a ‘classical’ sort of language, it is no longer widely spoken as an L1, so we will direct our attention to the dialect of Kipatra, codified in the region’s Late Antiquity, and whose pronunciation is typically taught to scholars as the standard (like Alexandrian Greek pronunciation being extended to Attic).
Old Cortar Phonology
Plosives: /p/, /t/, /ʈ/, /k/; /b/, /d/, /ɖ/, /g/.
Fricatives: /s/, /ʂ/, /x/, /h/; /z/, /ʐ/.
Affricate: /ts/, /ʈʂ/.
Nasals: /m/, /n/, /ŋ/.
Tap: /ɾ/
Approx. /j/.

/ä/, /i/, /u/, /e/, /o/, /ɜ/
Lazy romanization (Criticize to your heart’s content): <p, t, tr, k; b, d, dr, g; s, sr, x, h; z, zr; ts, tsr; m, n, ng, r, y>, <a, i, u, e, o, eo>. Retroflex geminates, <ttr, ddr…>. If it becomes necessary, an apostrophe <t’r> will indicate that two letters are not a digraph, although I don’t think it will be necessary with the kind of assimilation that I foresee.
Before Cortar was written, there was a /φ/, but it scooted back to [h].
The articulation at /t/, /ts/ and /d/ is dental.

In the Oldest Cortar, (called Law Cortar, because the laws are still recorded in this dialect) the vowels were [a], [ i ], [ u ], [ e ], [o], all of which had long variants, and [œ:], and [ɔ:], which were exclusively long, having been diphthongs even earlier. The [œ:] unrounded into [ε:], taking with it the diphthong [ai], which assumed the form of the same monophthong. [ɔ:], likewise subsumed [au], Vowel length was lost, and in the process [ɔ] unrounded to [ʌ]. [ʌ] centralized to [ɜ]. [ε] merged with it later in Kipatra. [œ], and [ʌ, ɔ] had, again, descended from diphthongs whose glides started as back vowels, and in the orthography remained noted as such because ancient chanters pronounced them in hiatus; the scribes recognized only five so-called ‘elemental’ vowels: [ɜ] is not indicated as a monophthong in the internal orthography.
The above occurred as the result of a shift that regularized the distribution of stress, governed by the position of long vowels, so that the distinction between long and short vowels vanished altogether, a process complete by the time of the Kipatra period.

To summarize the above:
Vowel length distinctions lost in early-to-mid antiquity.
/ɔ:/, /au/ > /ɔ/> /ʌ/ > /ɜ/
/œ:/, /ai/ > /ε/ > /ɜ/ (the last after /ʌ/ > /ɜ/ is finished).
None of this is at all obvious in the internal orthography.

Phonotactics.
(C1)V(C2).
If that isn’t the most vanilla syllable structure you’ve ever seen…
It’s actually just a little more complicated. Only the following are permitted in the coda: /t/, /k/, /n/, /ŋ/, /s/ and /ɾ/, except that everything but the affricates may form geminate consonants.
The coda rules are because of the following transformations:
• The non-sibilant fricatives kicked the bucket in the coda around Early Antiquity.
• The voiced plosives simply devoiced in this context. Though, for what it’s worth, in some dialects, not Kipatra, they lengthened the vowels before them, and then dropped altogether.
• The voiceless retroflex plosive merged with the flap in this position in mid-antiquity.
• /ʂ/ > /s/ in this environment.
• The affricates weren’t permitted in the coda to begin with.
• If the approximant were ever in the coda at all, it was conflated with the diphthongs immediately.
• The remaining labials were delabialized here and became /n/ and /t/.
There may have been a time when one could have inserted the yod as a glide after C1 when the latter happened to be a stop, but it was long gone by the time of the Kipatra period, leaving no trace of having been there.
Internal Orthography.
The Cortars have been in a strange state of digraphia. They actually sparked writing for the whole Eastern hemisphere – they invented a typical ancient mixed logography. It was written with a stylus or incised into wares. It accepted only straight lines.
The Cortars exported the idea of writing to the Pai islanders to the north. The latter quickly pulled together their own logographic script. Somewhat like the Egyptians, their script appears fully formed, and within a few centuries of the establishment of the Cortar palace script. In a rather cliché move, the Pai liked to keep their writings secret for their priests, oracles and whatnot. Trickles of it got down to their merchants, and, at length, those invented an alphabet of it. By the end of Antiquity, the idea of an alphabet had made its way back down to the Cortars. They have since had a state of digraphia that isn’t unlike the Hangul ~ Hanja situation in early modern Korea. The alphabet, of course, has the reputation of being a peasants’ script, and the logography… you get the picture.
The other way the Pai islanders corrupted the otherwise pristine Cortar script was the introduction of bark paper in mid-antiquity. As a result, the Cortars learned to write in cursive with a brush or pen. At Kipatra, there was no standard brush script, which was typical of letters and palace chancellery hand, but was never to be used in the recording of serious literature; that was the exclusive domain of the rigid forms of the classical palace script, now written with a reed pen. This practice has continued over the centuries, but Old Cortar as a literary standard is generally written in the logography, because alphabetization was not yet widespread until very late in the classical period.
Additionally, the above ‘serious’ literature predictably prefers archaic spellings when phonetic elements are involved (conjugations and the like), whereas the chancellery hand had license to adapt to the contemporary phonology. The classical palace script never did, however, indicate vowel length.
The Meaty Parts.
Old Cortar is not an ambitious project. It is loosely head final and agglutinative. The order of the verb phrase will lean towards VO. We’re aiming for an Ergative-Accusative alignment.
Nouns in the absolutive are unmarked, whereas the ergative has an enclitic ‘tsi’, which devolves to ‘-ssi’, ‘-si’ , or ‘-s’ in the Kipatra dialect: intervocalic affricates broke apart - /ts/ fragmented into /t/ and /s/ at syllable boundaries, and the t was assimilated as a geminate consonant. With regards to the ‘tsi’ ~ ‘-‘s’ spectrum, you’ll end up with a mixed bag in the literature. In fact, there are probably going to be a few times when lazy scribes don’t even indicate an ‘ -‘s’ due to the already clunky nature of the script, mostly in the sort of ‘hieratic’ chancellery hand.
Nga-ssi e-tuyeo-n rer-(0).
Dog-ERG PERF-bite-TRANS. Man-ABS.
The dog bit the man
also
Nga’s e-tuyeõ rer. (high-class spoken in Kipatra; nasalization was rampant but not considered canonical, nor widely indicated in writing).

These chunks of nominal morphology should probably be analyzed as suffixes, given that they concatenate: the genitive can serve as a substantive, and as such its suffix, ‘i’, will, for instance take the ergative, ‘i-tsi’ & ‘i-ssi’ (but, of course, there’s nothing underlying the absolutive):
Nga-(0) John-i-(0)
Dog-ABS John-GEN-ABS.
The dog is John’s.

John-i-’s e-tuyeo-n-a-rik.
John-GEN-ERG PERF-bite-TRANS-1P OBJECT-DIM.
John’s (dog) nipped me.

M’ u nga-’s e-tuyeo-n John-i(-0).
CONJ. 3P.DEM Dog-ERG PERF-bite-TRANS John-GEN-ABS.
And that dog bit John’s (dog)!

That’s quite enough dog biting. This is a good point to state that there is frequent ellipsis of all phrases, save that of the verb, and to highlight that there is no copular verb. And since we’re talking about this kind of noun marking, let’s talk about what happens when there needs to be a nominalized phrase that’s not a participle. There are two strategies – they involve an old catch-all referential noun ~ pronoun.
Fishing for dinner:
U zrat ipi.
3P DEM-(ABS) Fish OK/acceptable.
That fish is OK.

Mapik-i tat r’-i (dak) y’ utun.
Bridge-GEN below LOC-GEN (NOMINALIZER)-ABS NEG so.
Or
Dak y’ utun, mapik-i tat r’-i.
NOMINALIZER1-ABS NEG. so, Bridge-GEN below-LOC-GEN1.
The one under the bridge isn’t.

The latter involving the rightwards shift is perhaps more common in speech, particularly when the nominalized phrase gets long and clunky, but both see a lot of use. As you can see, the nominalizer may be omitted in the first construction, but obviously not in the second. Postpositional phrases can take the genitive, particularly to modify nouns.
Determiners, being mostly the demonstratives, cannot stand alone, but all must take the same dak in the absence of something to modify. Strangely, the numbers act like determiners in this respect.

Uttringa?
INTER.
“How many (do you want)?”

Set eon dak. U dak.
One DIM. NOMINALIZER-ABS, DEM. NOMINALIZER-ABS.
“Just one. That one.”

Returning to the examples further up, the negative is not analyzed as a prefix, but rather a defective adjective or adverb because, although nouns and adjectives can predicate sentences, their syntax is derived from the nominal system and with the verbal system was never conflated; in fact, the negation cannot directly modify a verb, but must take a periphrastic construction of the form ‘not well’. Returning to the dogs for a moment:
U dak-(0) ya nga-(0).
That NOMINALIZER.-ABS not dog-ABS.
That’s no dog.
Y’ik (0)-i-waran-ne nga-i tun.
(Not well)NEG IMPERF.-3P-bark-TRANS. dog-GEN. like.
It doesn’t bark like a dog.

This leads to special issues, because we are left in a position in which there is no way to indicate aspect on a construction of the linking form N1 N2, or “Nga John-i”. In the literary standard it may occasionally be thus resolved:
Nga-(0) e-mi-(0) John-i-r.
Dog-ABS PERF-exist-INTRANS John-GEN-LOC.
Lit. “The dog was at John’s”.

You’ll notice that this has the unfortunate side effect of being ambiguous in some contexts: Was the dog John’s, or was it in his bedroom? For all we know, that’s a substantive genitive.

Gayas-(0) e-mi kik-i trazo-r.
Place-ABS PERF-exist King-GEN agora-LOC.
“The place was King’s square”, or “The place was at (i.e. within) King’s square”.

At this point, I feel like this problem is actually more often resolved by a perfective participle of some kind. It’ll take the form of ‘N1 that has V’ed, N2’. The aspect of the linkage remains functionally in the imperfect or present, but one of the noun phrases has another aspect embedded in it. I have yet to fully outline the participial system, but it will probably involve a periphrastic construction with an auxiliary taking the actual nominal inflection but with some defunct infinitival form taking the aspectual and valency affixes, probably like:

Eok-i e-ama-r gayas-(0), kik-i trazo-(0).
ACT.PART-GEN PERF-go-INF.INTRANS place-ABS, King-GEN agora-ABS.
“The place that we have gone is King’s Square.”
Literally, “The place of those that do the having gone, the King’s Square.”

Eok-i e-ropi-n post-(0), sro dak-(0).
ACT.PART-GEN. PERF-shatter(archaic)-TRANS. Post-ABS, 1P.DEM. NOMINALIZER-ABS.
“The post of him that does the having eff’ed it up, this one.”
“This one is the post I screwed up.”

As you can see, only the subject would be directly modified by eok. If it modifies anything else, eok acts substantively and takes the genitive.
I mean, I could also do a synthetic attributive conjugation, but that would be boring:
E-ama-r-tir-i gayas…
PERF-go-INTRANS.INF-ATTR-GEN. place-ABS,
Where the -r- infix was actually an older, original attributive and the -tir- is some long-gone noun. The latter is maybe safer but, why? I could make things so much more complicated.
I could also have both, actually. Either way, there has to be an underlying attributive conjugation somewhere, because it’s not nominally derived. In the case of ‘eok’ it’s an irregular form that asserted itself as the exclusive auxiliary through an emphatic construction of some kind. The problem with this is that we might expect an auxiliary to come at the right side of the verb, given the existing head directionality and whatnot. The excuse is that the lexical chunk, being an infinitive, originally filled the object position, and that is why the semantic object will end up taking the genitive, because the object slot is already occupied. Conveniently, that puts infinitive in the absolutive, requiring no marking. That should streamline the grammaticalization process. But if it isn’t nominally derived, is it a participle? It might be better analyzed as an attributive auxiliary. Yet, it bears nominal inflection…
Sorry. I was talking about the nominal system before following that bunny trail. Anyway, that’s enough for your target practice today. Shoot ‘em up.
User avatar
Omzinesý
mongolian
mongolian
Posts: 4079
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Rambling about a somewhat routine classical language

Post by Omzinesý »

There is so much stuff in the message that it's a bit hard to read. And you are talking about complex things without explaining its parts, but so conlangs develop.

If the consonant inventory lacked /p/ or /g/, would it be more interesting or "natural"?

Double case marking is interesting, especially with zero Absolutive.

Did you somewhere explain the syntax of the core cases?

I, by the way, like your intrest in syntax before morphology.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Nmmali
rupestrian
rupestrian
Posts: 16
Joined: 01 Apr 2019 06:17
Location: The United States

Re: Rambling about a somewhat routine classical language

Post by Nmmali »

Omzinesý wrote: 01 Aug 2021 23:43 There is so much stuff in the message that it's a bit hard to read. And you are talking about complex things without explaining its parts, but so conlangs develop.
Yeah; I will take this into account in any future posting.
If the consonant inventory lacked /p/ or /g/, would it be more interesting or "natural"?
Possibly? I feel like if either were to vanish it would be a process of redistributing them into allophones that later become phonetic, right? I suppose I could do away with /p/ because I already did away with the bilabial fricative - in fact that would be more reminiscent of the phenomenon from Japanese which inspired it. /p/ > /φ/ > /h/...?
Did you somewhere explain the syntax of the core cases?
No. I'll put up a post-it so I do it this week.
I, by the way, like your intrest in syntax before morphology.
Thanks. I feel like its sort of easier, because then contractions and the like come more naturally, but I digress lest this reply descend into more incoherent rambling.
Post Reply