#proto-da

[vilikemorgenthal]

Lexember #14 : maka & kəni

Conlang: Proto-Da

Second time I use the same conlang in a row. So much for diversity, BUT I had a crucial need to develop vocabulary for this particular language, as the poor size of it is an impediment to the tailoring of interesting example sentences. Or derived words. So today we get TWO words!

The word

maka /mækæ/, vta.: “drink (humans)”

kəni /kənɪ/, vta.: “drink, lap, sip (animals)”

Verbs in Proto-Da have what is called a “preferred gender”, that determines how the arguments will sport their oblique stems or not.

For example, both verbs here are transitive animate, meaning that an unmarked animate argument will be understood as the agent/subject, and an unmarked inanimate argument as the patient/object. If an inanimate must be subject, it’ll display its oblique stem. Same for an animate object.

Obviously, it is rare that verbs meaning “to drink” take an animate object. Maka however can take inanimate subjects: holes in the ground, clefts, rifts “drink” water when it flows in there. Like humans, that is, in one big gulp.

Examples

Kud jum maka būm

water 2s drink blood

“You drink water, and blood”

Kənitə nāŋwitənta bə

drink-INFR COLL-puppy this

“This litter of puppies seem to have drunk”

Derived words

As I haven’t thought of deverbal affixes for now... I’ll pass.

(One sure thing though, verbal stems cannot be used as nouns as in English or Chinese. As Proto-Da syntax is nonlinear, no way word order could serve to discriminate the uses!)

[vilikemorgenthal]

Lexember #13 : ŋūbū

Conlang: Proto-Da

I’ve not started that much on their civilisation. They must have been hunter-gatherers, and/or fishers; that means that a lot if not all of their agricultural vocabulary will have to come from Suneric languages. But we’re not there yet, nowhere near.

The word

ŋūbū /ŋubu/, obl.-m, Ʌ gender: “child (as of age)”

This word does not describe interfamilial relationships. It describes young humans who can walk on their own (as opposed to wadənta “baby, suckling”) but have not passed the adult initiation rites (thus becoming tīmū and wad, man and woman; children are weakly gendered). They get the “upwards” grammatical gender, because they are seen as sentient enough to “look up at the sky” (one of the prerequisites for grammatical gender assignation).

It may also be applied to any sentient being that is short enough.

Example

Wibə ŋin gau ŋūbūm ŋūbūm

see 1s hit child-OBL child-OBL

“I see children hitting each other”

Derived words

bāŋūbū, obl.-m, Ʌ gender: “group of children, playing children”

jūpiə ŋūbūŋā, obl.-m, Ʌ gender: “river child (spirit of a drowned child)”

wāp ŋat ŋūbū, obl. wāpîn ŋat ŋūbūm, Ʌ gender: “Dog Person (mythical creature)”

[vilikemorgenthal]

Lexember #10 : kəg

Conlang: Proto-Da

First Lexember for it! It’s the language of the western indigenous people of the continent the Suner people set their sails to. Until contact with the latter, they had a stone-age tech level, much “less” than the bronze (or iron) of the southern newcomers. In the *present day*, it has given the Da family of languages, a major one next to Suneric.

The word

kəg /kəg/, stative verb: 

be regular, cyclic

be reliable, expected

There are no adjectives in Proto-Da; instead, stative verbs are used, and there are no way to distinguish an attribute from a predicative use, either morphologically or syntactically.

Examples

Bāwad būmā kəgdə

PL-woman blood-POSS be.regular-INFER

“Menstruation is a regular phenomenon” (inferred by someone having no firsthand experience of it, like a man)

Ŋat bə kəg jāg

be DEM be.reliable fire

“This is a well-maintained fire”

Derived words

kəg māt, oblique kəg mātîn, Ʌ (up) gender: “moon (cyclic sun)”

kəgdə būm, obl. kəgdə būmmîn, V (down) gender: “menstruation (in male speech)”

[vilikemorgenthal]

Lexember this year

Last year, I participated for the first time with my as-of-then brand new conlang, Ɣu. I used this language extensively in my diary, and even on a twitter account for a short while. But as of late, I lost the spark and new words are added only when a translation challenge comes up, that is, not so often; and the grammar work is even stiller.

However, I ‘m not working on any conlang now that would provide me with an endless supply of motivation throughout the month. I never was the type for one project, ever spending my minds on various endeavour

So I thought, this year, I will not lexember with only one conlang, rather I will coin one word each day for whichever of my conlangs I’m feeling best about in the morning.

These are:

Ɣu

Proto-Suner

Proto-Da

(new!) Paatara

Worbon

Who knows what else

With such a variety of languages to lose oneself in, I will do my best to provide the reader with a short recap of every language’s background before each post.

[vilikemorgenthal]

Proto-Da: an introduction

My renewed interest in Proto-Suner must not prevent me from working on the other languages of the Ambetau project. A total of four proto-languages needs to be designed for the “continent” the focus will be on, spoken at roughly the same time diegetically, plus a set of closely related dialects for the Groundrippers of Thulé and their simpler kin, that does not need the same time depth.

I already showed a bit of Proto-Da here; my sketch hasn’t gone much farther, but I thought I could already share the details of what was done for now.

The Da language family is restricted to the southwestern peninsula and the mainland south of the northwestern peninsula, however its extent was never threatened by the Suneric languages of more “civilized” peoples, for reasons I have yet to explain. Its general morphosyntactical makeup is a partial mystery for me as well, though I imagine a tendency towards monosyllabism in at least one modern language. We will see how it fares when I set the diachronical machine on.

Phonology

Voiceless stops: p t k

Voiced stops: b d g

Voiceless fricatives: s h

Voiced fricatives: z

Nasals: m n ŋ

Semivowels: j w

Vowels come in pairs, labelled tense/relaxed. A vowel harmony system prevents both sets to occur in a root together, but some suffixes and most of the prefixes do not not adhere to the harmony.

ā/a /ɑ æ/

ī/i /i ɪ/

ū/u /u ʊ/ (after /w/, realized [o ɔ] and written ō/o, as in mwōd ‘not’)

î/ə /ɨ ə/

Four falling diphthongs, that count as lax in roots but can go either way in suffixes, complete the picture: au ai iə uə.

The syllable structure is C|N(j,w)V(C), where the onset must be filled by a consonant; if it’s a nasal, it can be followed by j or w (ŋwit ‘bitch’). The fricatives s z h have a severely restricted distribution, only as word-final codas.

Nouns

There are four genders in Proto-Da, that are the result of applying two different sets of binaries to the referents: the (somewhat classical) animate/inanimate, and the up/down. No need to explain the former; but what does the latter mean?

Simply put, up nouns (hereby Ʌ) are those whose referents are seen as “staying in the sky” like clouds, stars, birds; and those whose referents are “reaching to the sky” like trees, fire and superiors mammals (humans, dogs). Down nouns (hereby V) “stay on the ground” like critters, stones, water, small plant life, or “reach to the ground” like rain, blood and baby animals.

Each gender of a pair is applied in conjunction with one of the other pair to a noun, giving the four combinations Ʌ an., Ʌ in., V an., V in. Those combinations determine the when of the use of the augmented stem with verbs.

The augmented stem of the noun is used as the dispreferred argument of the verb. It will be cleared in the section on verbs. It is also used as a base to add some suffixes to, like -tā/ta: ŋwit ‘bitch’ ŋwitənta ‘puppy’.

Inanimate words: -mîn/-mən. Bī ‘oak tree’ > bīmîn

Animate words in a consonant: -în/-ən. Jum ‘you’ > jumən

Animate words in a vowel: -m. Tīmū ‘man’ > tīmūm

Some inanimate nouns ending in d, t, k, g will mutate this last consonant to z, s, h, j. Kud ‘water’ > kuz

Finally, some words are just plain irregular: bə ‘this one’ > biə

There is no real way to obtain plural forms. Sometimes, the augmentative bā- will denote a big quantity (bāmut ‘pine forest’), the diminutive gu- will denote a paucative or singulative (gugugə ‘little ant/few ants/only one ant’), and there is a collective in nā-. But these are lexical derivations, to the same level as the honorific miə- (miəwad ‘crone’), not regular inflexions.

Possession is marked through a suffixed -ā/-a to the stem of the possessed noun: dākā nap ‘the thumb of the mother’, literally ‘her thumb the mother’. Same thing with the “pronouns”: ŋin duta ‘my gander’, literally ‘me my gander’.

Verbs

In my current conception, verbs inflect only for evidentiality: -∅ “seen it happen”, -tî/-tə “it seems like it happened”, -kī/-ki “they told me it happened” and -nīh/-nih “it didn’t happen but let’s pretend it did”.

Negation is with the auxiliary mwōd, which takes the inflexion if any: jumən məŋ mwōddî ‘seems like you haven’t fallen’.

Modal particles include mî “prohibitive” and wā “imperative”.

What is more interesting with verbs is that they each have a preferred argument and a preferred gender attached to that argument. So, for example, kān ‘to eat’ wants its subject to be animate, and its object to be inanimate. So both nouns will stay in the plain stem: kān mimin bīk ‘the goat (bīk) eats the moss’. But if both arguments are animates, the object will features in the augmented stem: bīk kān gugəm ‘the goat eats ants’.

It’s not really an accusative system however, with the marking of the less animate patient. That’s more clearly seen with the verb pîbūm ‘to kill’, that wants a Ʌ main argument. So, it would be pîbūm tīmū ŋwitənta ‘the man kills the puppy’ (sorry), where the subject ‘man’ is Ʌ and the object ‘puppy’ is V; both are in the plain stem. However, if the bitch wants her revenge and kills the man, it would be tīmūm ŋwit pîbūm with ‘man’ in the augmented stem, cause it’s been demoted to patient despite being a Ʌ. And if a sister of the fallen puppy decide to join in, it would be pîbūm tīmūm ŋwitəntam with both arguments in the augmented stem. Were one unmarked and the other not, they would both have got the same role: “the man and the puppy kill (something)” or “(one) kills the man and the puppy”.

Verbs of movement work with the dispreferred argument as the destination: with mjud ‘to go up’ a Ʌ verb, the sentence ‘the man goes to the summit of the oak tree’ is tīmū mjud bīmîn. ‘Oak’ is in the augmented stem because it is a Ʌ noun too, but is the destination nevertheless.

Of course, it only applies to active verbs. Stative verbs (corresponding to adjectives in IE languages) take one argument only, e.g. bai ‘be red’.

There is also a copula, ŋat, used with nouns: biə wad ŋat ‘this one is a woman’.

Syntax

As my example sentences may have clued you in, there is no set order between S, V and O, or between possessor and possessee. They move freely along purely pragmatic parameters. Eat that, Chomsky!

Vocabulary

Rather poor. In fact, I must have used 80% of the lexicon with this post.

[vilikemorgenthal]

I’m bored in class #1

At the university, some classes are less interesting than other. So what I do usually is scribble in my margins or deface the handouts given to us. I am no visual artist, so instead of doodles there is conlanging. It may build upon a previously existing work of mine (sample sentences, new words, etc), but more often than not, I go about jetting down a random paradigm and try to expand on that. So today, in the English class...

... I wrote heili on the page. What does it mean? I did not know, neither do I now. I came to think “It’s a good shape for a plural... how could its singular differ from it?”. I’m kinda obsessed with Finnish these days, and spent some time researching the origin of its root alternations. My “goal” with this draft was to imitate it with regular sound changes.

I decided on heigo. Then, the idea was that the stem heig- saw its final consonant disappear before the plural ending -li, and added an epenthetic -o in the singular as to avoid ending a word with a voiced stop. That -o was to be derived, by regular a sound-change, from a -ə. Problem: it seemed to be possible only with some sort of vowel harmony going on; and how could a front unrounded diphthong -ei- influence a schwa to a back rounded vowel -o? I had to dive deeper in the history of the language (all while my teacher read the corrections for our tests).

Here is what I came up for the singular (numbers indicate a fixed point in time; macrons denote a long vowel, ogoneks a mid-open vowel):

0 hog >0.5 hǫgə >1 hǫgo >2 huogo >3 hø̄go >4 heigo

And the plural:

0 hogli >1 hǫʎi >2 huoʎi >3 hø̄ʎi >4 heili

Precisions: vowel openness became phonemic at 0.5 (for the singular) and 1 (for the plural), when formerly closed syllables got opened. It gave rise to a seven-vowels system, which introduced diphthongs at 2, that later monophthongized and re-diphthongized. The change in 2>3 is inspired by Old French. I went then on creating more sg/pl pairs, to explore other vocalic and consonantal alternations:

0 hud >1 hudo >2 hydo >3 hyro >4 hiro

0 hudli >1 huʎi >2 hyʎi >3 hyʎi >4 hili

Vocalic changes still inspired by Vulgar Latin to Old French (except for 3>4).

0 med >1 męde >2 miede 3> mīre 4> maire

Old French again, for ę>ie. I did not write down the sound changes for the plural, as by that point it became clear that, at least for voiced-stop stems, the plural was just to replace the last syllable with -li: so maili from medli.

0 lok >1 lok 2> loʔ >3 loʔ >4 lo

0 lokli >1 loʎi 2> loʎi >3 loʎi >4 loli

No opening for that /o/! It strikes me as odd now. Could I say that a syllable closed by a voiceless stop is somehow “less” closed than with a voiced one? Mmh...

0 elot >1 elot 2> eloʔ >3 elʔo >4 ello

0 elotli >1 eloʎi 2> eloʎi >3 eloʎi >4 eloli

Now there is interesting alternations again! The metathesis of ʔ in final syllable is inspired by some Finnish dialects that did it with /h/.

0 mar >1 mar 2> mar >3 ma >4 ma

0 marli >1 marri 2> marri >3 marri >4 marri

And, at least, the plural suffix has a variant!

And now?

I usually grow disinterested very quickly with these tiny conlanging bits, and move to something else entirely the following day. But here, the end result has the phonotactics of what I want to be the language of Bai De Ho, an important country in my conworlding project (of which I touch a bit there). So I think I will save these sound changes for a future point in my designing of Proto-Da.

[vilikemorgenthal]

Some thing I learned in my attempt to create a proto-language

Some months ago, I decided to throw away my then-main conworld, which began as a rip-off of Zompist’s Almea (if not in shape, at least in spirit) and whose conlangs I never really involved myself in (actually, I was way more involved in Old Greedian, whose setting I didn’t create, and Ɣu, which is a true personal language ; only Worbon, spoken by isolated islanders, has any substantial work attached to it). I started again from scratch, with a different concept in mind (dieselpunk), and with the ultimate goal of telling the sort of stories I like. So, I drew a partial map, political boundaries and all, and began my writing. Quickly, it became apparent that at this early stage, I had the opportunity to create a realistic linguistic landscape (that, and I couldn’t wait anymore to play with language families).

I started to create the proto-languages of the two I knew shall be more useful to me (the first because its languages encompass the country I draw emphasis to, the second because it’s the cultural equivalent of Indo-European in that world). The starting point would have been 4000 years before the “current year”, when the second ones’ exiled speakers would have come to the shore of the continent.

It first ran smoothly. I devised a simple phoneme inventory for Proto-Da, a simple but culturally indebted gender system, a simple case system with the potential to birth monstrous irregularities in the daughter languages ; I wrote example sentences and filled the blank in the vocabulary. Some of the sentences were already complex, thanks to the approach permitted by the syntax:

kiənnih ŋin bājūpiə ŋig wibə mapən: go-HYP I big-river be_real see father-ACC “If I go to the sea I’ll see my father”

I had real fun devising that proto-lang, that I envisioned less like a matrix and more like a conlang I could use in texts (barring the limitations inherent to the Late Neolithic setting, of course). Owing to the simplicity, there was no need for me to imagine an complex earlier stage from which the rare irregularity (like the stem-changing suffix) could have arisen.

But then there was Proto-Suner, the IE-y conlang. In that world, contrarily to Proto-Da, its speakers would have an already complex culture, with cities, ships, social stratification, and writing. In order to convey such a feel, I created a complex language... that is, I mixed implosives and prenasalized consonants with five points of articulations (labial, dental, palatal, velar, labio-velar), nasal vowels and a syllabic structure allowing words like ɓliso, śporʄeⁿba, ⁿdrife, fkʷara. The verbe was inflected for agent and patient, its root had two epenthetic vowels whose distribution was arbitrary, and the case system was following an older ergative-absolutive pattern changed to nominative-accusative (like some scientists posit for Proto-IE). Moods were of desiderative, assertive, negative, and timitive. As for the the gender system, it had a loose semantic basis and was coded by the alternation of vowels between singular and plural... (oh, and there was a dual). I aimed to wreak havoc with that in the daughter languages.

The result was not a pretty language. I did not like the sound of the sentences I came up with, like:

ɓliso ʄejaⁿdjok kʷavol: fish eat-PAST.PART-1S see-3S<2S “you see the fish that I’m eating”

The complexity of the morphology required also that I imagined a former stage of the language without creating it, something I routinely did with my former conlangs but was more hard work and turned me from the main goal of the project: looking forward, not backward.

As a result, I was less interested in the project overall and let it sleep, waiting for a new spur of creativity. The task seemed to me too much for the moment.

A week ago, I applied for the creation of four naming languages for The Chained Adept, by Karen Myers, and while I didn’t get the job, I worked on some example words and morphology along the guidelines of the author. One of those four languages was to be “vaguely Afghan”, and I devised a phonetic inventory and syllable structure aiming to that goal. I grew fond of the words I created and liked the shape of the gender system ; I found it a pity not to let that sketch live on its own. So I added a verb conjugation, expanded the case system... and then it struck me: that conlang would be perfect as a proto-language, even if the consonants weren’t exotic and the morphology was rather stable. I realized that to simply toss instable structures around and waiting them to collapse was not what I was conlanging for, and that every language is a potential proto-language. I already saw it with Proto-Da, but it took that job appliance (and the episode on Bad Conlangs by @dedalvs) to get the courage to throw away the first version of Proto-Suner and replace it with “vaguely Afghan”.

Vikhas bis griturmanos lagyos ristaz: you-NOM I-INST eat-PASS.PRES.PART-GEN fish-GEN see-PRES

It sounds way better.