Karaam - NP: syntax

Conworlds and conlangs
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Karaam - NP: syntax

Post by Ares Land »

I'm working on the grammar of Simbri; but this required a detour back 3500 years in the past to work out Karaam, the liturgical language.


The Tarande: Karaam in context.
Map of Tarande.png
Map of Tarande.png (96.52 KiB) Viewed 5010 times
The Tarande is both a geographical area and a distinct civilization.

Geographically, it corresponds to the area shaded red on the map; there are no clear dividing border. The best definition is probably ecological; the Tarande corresponds to the areas of Mediterranean or Pontic climate. (Summers are hot and dry; Winters are either severe - in the North or mild - in the South.) It's bordered by mountainous steppe to the East, desert in the south, and the Sea of Serenity and the Sea of Anger to the West. For scale; this is about the size of Western Europe.

One meaning of Tarande is 'calendar.' The Tarandim share a common calendar; this meshes in with religion -- they share a common belief system or more accurately, a family of related belief systems.

Politically we can say that the Tarande is defined by political disunion. There are about 300 independant states, all connected in a complex network of alliances, rivalries and vassal relationships. The natural political unit is the city-state or mansa. There have been imperial periods in the past, but imperial rule in the Tarande has always been difficult (typically city-states would formally accept suzerainty and then do what they pleased) and short-lived anyway.
Currently the largest states are Alwas and Qanan; both about the size of England. (In both cases, most of the area is marginal land, with only theoretical control over the local nomadic population.)

Language in the Tarande

The linguistic situation is one of considerable diversity: there are seven major languages -- each with many distinct dialects, belonging to four distinct language families.

The language with the highest number of speakers is, by far, Simbri. Quite naturally, it serves as the lingua franca for the whole Tarande.

Also relevant is Asuraam . Asuraam was spoken about 1500 years before present; it's the direct ancestor of Simbri and the former lingua franca.
Before Asuraam, the prestige language was Karaam (spoken 3500 years before present.)

Asuraam and Karaam are still widely learned and used. All three languages belong to the Dimoan language family.

In this thread, we will take a plunge 3500 years before present and focus on Karaam.

Karaam is still widely used as the language of priesthood, religion, theology and magic. For that reason it is sometimes called Hieratic Tarandim.

The Dimoan family

Karaam is pretty close to proto-Dimoan. At the time it was spoken, most of central Tarande (in deeper red on the map) spoke a Dimoan dialect; the Dimoan dialects, though diverse, were still mutually intelligible.

Karaam was used for the Mannamis (Karaam Moggagwis), 'the Inquiry', the Tarandim holy books. Further religious commentary used Karaam as well; and Asuraam later on.

Karaam means 'clear' or 'holy'. The other Dimoan dialects are collectively known as Gwemoam or Memoam 'demotic.'

The Tarandim believe that Karaam is the ancestor of Asuraam and Simbri; and indeed of most of the Tarandim language. That is probably not the case; Asuraam derives from another branch of Dimoan. There doesn't seem to be any direct descendants of the Karaam branch.
Asuraam, Simbri and in fact all of the Tarandim languages borrowed extensively from Karaam. In addition they freely calque Karaam word-formation and syntax. So the traditional Tarandim view isn't completely inaccurate either.

The Tarandim believe Karaam is the original tongue, spoken when the world was recreated; which they date back to their year 0 (5192 years before present.)
The Dimoan languages can probably be related to the Newari and Kaysan families; proto-Dimoan-Kaysan probably spread along with agriculture.
Last edited by Ares Land on Thu Jul 28, 2022 8:22 am, edited 1 time in total.
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Writing systems
Karaam was originally written using the Tarandim hieroglyphics; a logographic scripts that was not only used for Karaam, but for other dialects and indeed other languages.

Writing(1).png
Writing(1).png (20.05 KiB) Viewed 5006 times

Many glyphs could be read in several ways; for instance the glyph for 'sacrifice' could be read kwogwar, ɓomīr or pi.

This was, mostly, a feature, not a bug; the script brought together different linguistic communities.
It was a problem, though, for the priestly class that was aiming at codification and standardization of religious practice -- which was a chief factor in writing down tradition in the Mannamis.

Their chief objections were, it seems, to prayers being said and hymns being sung using Demotic pronunciation. One answer to that was to add pronunciation cues to the glyphs; for instance: the second version adds the glyph for 'cloth' hate and 'reed' gwir as a hint to the correct vowel and consonant for -gwar.

The original system only had cues for vowels, the labio-velars and glottalized consonants (the chief areas of dialectal divergence). Scribes seized on the same principle for secular writing, and developped glyphs for the other consonants. This led to the development of the Tarandim alphabet.

Knowledge of the old hieroglyphics was generally lost as the alphabetic system took over. At some point copies of Mannamis switched to full alphabetic.
(A number of hieroglyphs are still known; they're used for esoteric works; as shorthand, as mathematical symbols and for heraldry.)


The priests' punctiliousness has the practical advantage that we can describe Karaam's phonology with some precision. One caveat is that they were most probably very sensitive to outsiders' "incorrect" pronunciation and language mistakes, and probably quite blind to language change when it affected their own speech.

It's very likely the description below doesn't reflect the pronunciations of the original compilers of the Mannamis, but that extant in the priestly class a few centuries later.


Phonology

The sound inventory of Karaam is as follows:



Consonants
Labials Dentals Palatal Velar/Uvular Labio-Velar Uvular Glottal
Stops p ɓ b t ṭ d k g q kw gw qw
Fricatives fs ð š (g) (gw) h
Nasals m (b) n (d) (g) (gw)
Liquids l r
Vowels
Front Central Back
High i u
Close e o
Open ai ə au
Low a
Stops have a three-way disinction: plain, voiced, glottalized.
ɓ is implosive; ṭ, q, qw are ejective /tʼ/, /kʼ/, /kʷʼ/. ð is a voiced dental fricative /ð/; š was post-alveolar /ʃ/
ḥ was most likely a voiceless uvular fricative /χ/; r the voiced counterpart /ʁ/

Nasal harmony:
b, d were realized [m], [n] whenever one of the nasals m, n occured within the same word; they were pronounced [b.], [d] in all other contexts:
ribe 'he claimed' [ʁibe], riben 'he claims' [ʁimen]


g and gw were realized [ŋ], [ŋʷ] wheneve m, n occured in the same word.
In other contexts, they were pronounced [g], [gʷ] word-initially; [ɣ], [ɣʷ] in all other contexts.
keggwis 'he sifted' -- [keɣ.ɣʷis], keggwsin 'he sifts' [keŋ.ŋʷəsin]
gim 'twelve' -- [ŋim]
gaišer 'witness' -- [gaj.ʃer]

Vowels

ai, au are pronounced as diphthongs [aj], [aw] when stressed, as [ɛ], [ɔ] otherwise:
daikwain 'judge' -- ['naj.kʷɛn]
hauhau 'cinnabar' -- ['haw.hɔ]

Implied schwa; phonotactics

As a convention, clusters of three consonants are broken up by an implied schwa between the second and third consonant:
keggwsin > keggwəsin [keŋ.ŋʷəsin]
eṭhqo 'he dictates' > eṭhəqo [et'.hə.k'o]
An initial cluster of two consonant is also broken up by a schwa:
greghom 'mountain lion' > gəreghom [ŋə.ʁeŋ.hom]

More generally, Karaam only allows media clusters of two consonants and no initial clusters.
Syllables are of the form V, VC, CV, and CVC (with no restriction on consonant or vowel); word-finally, Karaam only allows vowels and the following consonants: m, n, g, s, š, ḥ, h, r, l, d, ð

Stress falls on the penultimate syllable; but never on a schwa; if the penultimate syllable has a schwa, stress moves to the preceding syllable:
daikwain ['naj.kʷɛn]
daikwnas 'he shall judge' ['naj.kʷə.nas]

The digraphs ai, au are, as we've seen, diphthongs or simple vowels.

If any two other vowels are in hiatus, they belong to different syllables:
guorkau 'enter' → [gu.or.kɔ]

a+i > aj/ɛ and a+u → aw/ɔ without exceptions; ai and au can never be read [a.i], [a.u]. For instance, kara + optative infix ikaira 'may it be clear', pronounced [kaj.ra], not *[ka.i.ra]

Regional and historical variation

From developments in later languages, and contemporary comments, we have some idea of the differences between Karaam and its sister Demotic dialects; this in turn gives us a hint of the earlier phonology.
  • Proto Dimoan had voiceless and voiced approximants: /w/ /ʍ/. In Karaam w ʍ → ∅ f, but most Demotic dialects merged both as /w/
  • There were two dental fricatives: /θ/ and /ð/, or perhaps /ð/ and /ð'/. Karaam merged both. Developements varied in other dialects.
  • There was a distinction of vowel length, which Karaam lost early on.
In other areas Karaam is noticeably conservative. It kept the labio-velar series which most other dialects merged either with the labials or the velars. Karaam has the dentals /t/ /t'/ /d/; later on most dialects underwent the shift /t/ /t'/ /d/ → /d/ /t/ /ɗ/
Last edited by Ares Land on Wed Jul 20, 2022 6:03 am, edited 3 times in total.
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Roots

Karaam, like all Dimoan languages, has root-and-pattern morphology.

Simple and complex roots

Roots fall under two categories:

Simple roots are monosyllabic: bos 'dog', qweg 'king', fer 'grain'

However, most roots are complex roots; mostly (but not always) disyllabic.

Many complex roots can be analysed as compounds of simple roots:
  • ðandam 'trade' ← ðam 'gold' + ṭam 'give'
  • hesqaa 'religion' ← hes 'spirit' + qaa 'know'
  • gaišer 'witness' ← gais 'true' + er 'speak'
  • hešer 'prophet, oracle' ← hes 'spirit' + er 'speak'
  • kauer 'declare' ← kau 'come' + er 'say'
  • ḫaiðos 'anger' ← ḫai 'anger' + ðos 'anger'
  • mores 'human being' ← mor 'man' + es 'human'
  • eṭah 'write' ← e 'write, oracle, ochre' + ṭah 'sign, mark'
Other were formed by adding a derivational suffix to a simple root:
  • šoṭes 'earth' ← šo 'low' + -ṭes
  • šosṭes 'reside, dwell' ← šos + -ṭes
  • sisdah 'come with, be a companion' ← sis- + dah 'go'
  • ṭoṭah 'draw' ← ṭo- + ṭah
  • pehe 'obscure' ← pe- + he 'light'
Other complex roots were formed by adding classifying suffixes (which were, probably, measure words in an earlier stage) to a root:
  • qweṭah 'truthfulness' ← qwe 'true' + -ṭah
  • kaannos 'Hippidion' ← kaan + -ṭos
  • bauṭos 'giant llama' ← bau + -ṭos
  • mirkoh 'beer' ← mir 'drink' + -koh
  • mahug 'oven' ← mah 'hot' + -ug
Many roots are simply unanalyzable:

qenmam 'vizier'
ḫerim 'amaranth'
qaton 'jar, amphora'

Patterns

Karaam morphology makes use of two processes:
  • Vowel change.
  • Vowel deletion, triggered by prefixing, suffixing and compounding.
There are, generally, no purely nominal or verbal roots; verb forms and nouns are formed by implying a pattern to a root that in itself is neither nominal nor verbal.

Many forms are thus derived from a single root; for instance:

From eṭah 'write':
  • eṭah 'he wrote', eṭih 'they both wrote', eṭeh 'they wrote'
  • iṭah 'writer, scribe', iṭeh 'scribes', aṭah 'writing'
  • qahṭah 'unwritten'; qahṭheh 'illiterate'
  • aṭhiš 'book'
  • ṭeṭah 'correspond'; guṭhag 'stylus'; eṭhqo 'dictate'; aṭhqeh 'secretary'
  • qwegṭhug 'royal library'
From kwogwar 'sacrifice, propitiate':
  • kwegwreh 'priest'
  • kwagwir 'they have been propitiated'; kwaggwir 'gods, spirits'
  • takwagwras 'sacrificial animal'; kwagwroam 'sacrificial stone'
  • kwagwriš 'rosary'
  • behkwogwrah 'ceremony involving human sacrifice'
  • šaunkwgwar 'to offer incense'; šaunkwogwroh 'incense'; šaunkwogwriš 'censer'
Last edited by Ares Land on Mon Jul 18, 2022 4:45 am, edited 2 times in total.
User avatar
Rounin Ryuuji
Posts: 2994
Joined: Wed Dec 23, 2020 6:47 pm

Re: Karaam

Post by Rounin Ryuuji »

Little more to say for now than that this looks cool.
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Pronouns

The independant pronouns distinguish:
  • three numbers: singular, dual, plural.
  • 1st person inclusive and inclusive; a 4th person (ie, an obviative).
  • The obviative isn't marked for number.
Singular Dual Plural
1st sai ibi ea
1st inclusive saiti saite
2nd tu ribi rea
3rd ti šabi šea
4th ha ha ha

The independant pronouns are used solely to mark topic and focus:

Sai seaqam.
sai
1incl.sg
se-aqam
1.excl.OBJ-PASS\wash

'As for me, I bathe.'

Sai am ðorau heṭguḫa.
sai
1incl.sg
am
that
ðorau
sun(CL3)
he-ṭ-guḫa
1.excl.SBJ-CL3-make

'I made the suns.'

Pronominal prefixes

Subject Object Possessor
1st he- sa-, s- sau-, s-
1st inclusive a- e- e-
2nd ši- ri, r- ru-, r-
3rd ∅- te-, t-, šau-
4th u- u- u-

The pronominal prefixes do not mark number.

For subjects, the noun or verbs is placed in the appropriate form.
For the first person singular, the 1st exclusive prefix is used:

Heiṭah
he-iṭah
1.EXCL.SBJ-scribe

'I am a scribe.'

Arammam.
a-ramm-am
1.INCL.SBJ-mortal-PL

'We are mortal'

Ušaunkwgwar
u-šaun-kwgwar
4.SBJ-incense-sacrifice\PL

'They offer incense.'

For objects or possessors, number isn't marked at all:
saudaifrag
sau-daifrag
1.EXCL.POSS-knife

'my knife' or 'our knife (not yours.)'

šaukweg
šau-qweg
3.POSS-glory

'His/Her/Their glory'

herkumon
he-s-kumon
1.SBJ-2.OBJ-order

'I gave you (sg/pl) an order.'

As you can see from the sample sentences above:
  • Nominal predicates (eg: Heiṭah.) are formed by adding the appropriate subject prefix to the noun. There is no need for a copula.
  • The 3rd person subject isn't marked. So iṭah 'scribe' may form a complete sentence on its own 'he/she is a scribe.'
  • Both subjects and objects are marked on the verb; the subject prefix comes first.
  • Both subject and possessor can be marked on a noun: šisšorša 'You're my brother.'
Some object and posessor prefixes have long (sa-, ri-, te-, sau, ru-) and short (s-, r-, t-) variants.

The short variants are used whenever their use wouldn't result in an initial cluster of two consonants, or a medial cluster of three consonants:

heskumon / rikumon
he-r-kumon
1.SBJ-2.OBJ-order
/
ri-kumon
2.OBJ-order

'I gave you an order / he gave you an order.'

sineš
s-ineš
1.POSS-cria\PL

'my crias (young llamas)'

saubauṭeš
sau-bauṭeš
1.POSS-llama\PL

'my llamas'

Genitive construction use the possessor prefixes:
sauqaatas tešosdaur
sau-qaatas
1.EXCL.POSS-father
te-šosdaur
3.POSS-house

'My father's house.'

qweg tegammo
qweg
king
te-gammo
3.POSS-palace


The fourth person

The fourth person prefix is used to resolve potential ambiguity between different referents:
šaušosdaur dauqen sauqaatas
te-šosdaur
3.POSS-house
dau-qen
CL.8-repair
sauqaatas
1.EXCL.POSS-father

'My father repaired his (own) house'

ušosdaur dauqen sauqaatas
u-ušosdaur
4-house
dau-qen
CL.8-repair
sauqaatas
1.EXCL.POSS-father

'My father repaired his (someone else's) house'

Usage of the 4th person is somehow intricate; I'll try to give a brief overview:

The topic is the default subject and possessor:

Regghas šaušosdaur dauqen sauqaatas
Regghas
Regghas
te-šosdaur
3.POSS-house
dau-qen
CL.8-repair
sauqaatas
1.EXCL.POSS-father

'My father repaired Regghas' house'; ušosdaur would not be acceptable here.

Regghasa teqor sauqaatas
Regghas-a
Regghas-TOP
te-qor
3.OBJ-kill
sauqaatas
1.EXCL.POSS-father

'Regghas killed my father.'

Regghas a uqor sauqaatas
Regghas
Regghas
a
TOP
u-qor
4-kill
sauqaatas
1.EXCL.POSS-father

'As for Regghas, my father killed him.'

Constituents can be freely reordered. The most common orderings are SOV, VSO and confusingly OVS. So all of these use the 3rd person prefixes:

Sauqaatas Regghas teqor
Regghas teqor sauqaatas
Teqor sauqaatas Regghas
'My father killed Regghas.'


SVO word order requires the 4th person prefix:
Sauqaatas uqor Regghas.
'My father killed Regghas.'

The use of the 4th person can be determined semantically; for instance:

Uranin teqor qweg.
uranin
panther
te-qor
3.OBJ-kill
qweg
king

'The king killed a panther.'

Uranin teqor figsšes.
uranin
panther
te-qor
3.OBJ-kill
figsšes
old.woman

'The panther killed an old lady.'

E he ðonnahal udaikwnis bauṭeš ṭasfiqeam
and
e
ðonnah-al
time-at
u-daikwn-is
4-judge-FUT.PL
bauṭeš
llama\PL(CL2)
ṭas-fiqe-am
CL.2-<HAB>eat<PL>-REL

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

Inanimates.
The pronouns listed above refer to animates only. For inanimates you use a noun class marker instead:

Hedaubaša he šosdaur.
hedaubaša
1.EXCL.SBJ-CL.8-build
he
that
šosdaur
house(CL8)

'I built that house.'

šosdaur dauqe
šosdaur
house(CL8)
dau-qe
CL.8-entry

'The house's entrance.'
Last edited by Ares Land on Mon Jul 18, 2022 10:58 am, edited 1 time in total.
User avatar
Raphael
Posts: 4401
Joined: Sun Jul 22, 2018 6:36 am

Re: Karaam

Post by Raphael »

Ares Land wrote: Mon Jul 18, 2022 6:16 am


The fourth person

The fourth person prefix is used to resolve potential ambiguity between different referents:
šaušosdaur dauqen sauqaatas
te-šosdaur
3.POSS-house
dau-qen
CL.8-repair
sauqaatas
1.EXCL.POSS-father

'My father repaired his (own) house'

ušosdaur dauqen sauqaatas
u-ušosdaur
4-house
CL.8-repair]dau-qen[/gloss]
sauqaatas
1.EXCL.POSS-father

'My father repaired his (someone else's) house'


YES!
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Nouns

Karaam nouns belong to one of nine noun classes. The native names are hesðis '~ of soul', ðedonnasðis '~of living things', amhoðis 'of thought', fangðis 'of food', mnauðis 'of hand', daifruðis 'of cutting', qorðis 'of heaviness', basðis 'of buildings', gonððis 'of mountains'. The names are cumbersome and often misleading; I use numbers instead.

They're marked for two numbers dual (buð) and plural (baguð). Both are generally marked by vowel change.

Class 1 (hesðis)
Class 1 nouns are animates. This class is defined by pretty clear semantics: they cover human beings, animals, and supernatural entities.
The dual of class 1 nouns is generally formed by changing the last vowel to -i; the plural by changing the last vowel to -e
  • qemmam 'merchant' → dual qemmim, plural qemmem.
However, many class 1 nouns end in -eh; in those cases, the dual is -ih (as expected) but the plural is -am:
  • heseh 'spirit' → dual hesih, plural hesam
Class 2 (ðedonnasðis)

Class 2 nouns are defined morphologically; all nouns ending in -Vs (any vowel followed by -s) belong to class 2.
Morphology trumps semantics; nouns in -Vs will belong to class 2, even if animate and this class, indeed, includes many animates; the class namer is ðedonnas 'angel, numinous being, vision', most of class 2 nouns are, however, names of domestic animals: kaannos 'Hippidion'; bauṭos 'llama'

The dual is in -is but the plural is in -eš
  • ðedonnas → Dual ðedonnis, Plural ðedonneš,
  • bauṭos → Dual bauṭis, Plural bauṭeš,
  • ines 'cria' → Dual inis, Plural ineš,
Class 3 (amhoðis)

Class 3 is the largest class in numbers; it covers all inanimates that don't fit in the classes listed below. Many of these are abstractions or infinitive, hence the native name.
The dual and plural are formed by vowel change; the last vowel changes to -i- in the dual, -e- in the plural.
  • hesṭah 'spiritual force' → Dual hesṭih, plural hesṭeh
  • dagwum 'good deed' → Dual dagwim, plural dagwem
Class 4 fangðis

Class 4 regroups all nouns ending in -o, -uh, -oh.
The dual and plural are again formed by vowel change: Dual -i-, plural -e-:

Generally this covers substances and food: moho 'linen', ḫerimuh 'amaranth', nakoh 'milk' but there are many exceptions: gammo 'palace'.
  • moho → Dual mohi, plural mohe
  • ḫerimuh → Dual ḫerimih, plural ḫerimeh
  • gammo → Dual gammi, plural gamme
Class 5 mnauðis

Class 5 regroups all nouns ending in -iš. These tend to be small, hand-held objects:
The dual is in -is and the plural is in -eš
  • lilliš 'fruit' → Dual lillis, plural lilleš
Class 6 daifruðis

Class 6 regroups all nouns ending in -ag. These cover objects with a blade, but also insects:
The dual and plural are formed by vowel change: Dual -i-, plural -e-:
  • daifrag 'knife' → Dual daifrig, plural daifreg
  • kalhag 'mosquito' → Dual kalhig, plural kalheg
Class 7 qorðis

Class 7 regroups all nouns ending in -oam. They're generally bulky objects: ṭassoam 'column', danhoam 'table', though not necessarily so: hemhoam 'kite'.
The dual and plural are formed by vowel change: Dual -i-, plural -e-:
  • ṭassoam → Dual ṭassoim, plural ṭassoem
  • danhoam → Dual danhoim, plural danhoem
  • hemhoam → Dual hemhoim, plural hemhoem
Class 8 basðis

Class 8 covers all nouns ending in -Vr (vowel followed by -r) or in -ug. They're generally buildings or placenames. This class also regroups most trees.
The dual and plural are formed by vowel change: Dual -i-, plural -e-:
  • šosdaur 'house' → Dual šosdir, plural šosder
  • ṭugwmnar 'olive tree' → Dual ṭugwmnir, plural ṭugwmner
  • ḫerimug 'granary' → Dual ḫerimig, plural ḫerimeg
Class 9 gonððis
Class 9 covers nouns ending in -du, -do, -no
The dual and plural are formed by vowel change: Dual -i-, plural -e-:
  • qehdu 'hill' → Dual qehdi, plural qehde
  • henno 'plain' → Dual hendi, plural hende
Determining class
Noun class can be determined in the following manner -- starting from a singular noun:
  • Is it animate? If so, does it end in -Vs? → Class 2
  • Is it animate with any other ending? → Class 1
  • If inanimate, does it end in -Vs? → Class 2
  • In -oh, -o, -uh but not in -do, -no? → Class 4
  • In -iš? → Class 5
  • In -ag? → Class 6
  • In -oam? → Class 7
  • In -Vr? → Class 8
  • In -ug? → Class 8
  • In -du, -do, -no? → Class 9
  • None of the above → Class 3
The classes have strong semantic connotations, but generally relying on morphology is safer. (The semantics of noun classes are actually fairly reliable in Karaam; in other Dimoan language class alignment is trickier.)

Irregular nouns

Monosyllabic animate nouns form their plural in -e, -i:
  • bos 'dog' → Dual bosi, plural bose
  • qweg 'king' → Dual qwegi, plural qwege
These are assigned to class 1.

Other monosyllabic nouns are trickier; you have to look at the dual and plural forms.
  • šah 'sea' → Dual šahdi, plural šahde-- Class 9
  • fer 'grain' → Dual ferkih, plural ferkeh -- Class 3
  • haan 'river' → Dual haandi, plural haande -- Class 9
Of note is the archaic singular bau 'llama' (instead of bauṭos) → bauṭis, bauṭes -- Class 2.
We may also note the irrgular gonðe 'mountain' (class 9) → Dual gonðdi, plural gonðde.

As you can see, the noun class of irregular/monosyllabic mouns are pretty predictable from their semantics.

Another irregular case: hai 'eye' has the dual and plural u. It falls under class 3.

Agreement
The following prefixes are used as both object and possessive prefixes:
Prefix
Class 1 te-, t-, šau- / u-
Class 2 ṭas- / te,- t-, šau- / u-
Class 3 ṭah-, ṭ-
Class 4 ho-, k-
Class 5 ši-, š-
Class 6 ga-, g-
Class 7 hoa-
Class 8 dau-
Class 9 do-


The prefixes for Class 1 are simply the third/fourth person pronouns; they are used as described above.

For Class 2 the situation is more complex.

In the Mannamis you will find both a ṭas- prefix:

ðedonnas bago ṭasu
ðedonnas
angel
bago
many
ṭas-u
CL.2-eye.PL

'The angel many eyes'

Or the 3rd (or 4th, as the case may be) person prefixes:

E ðedonnas tesišer.
e
and
ðedonnas
angel
te-sišer
3.OBJ-talk.with

'And he talked with the angel.'

This was rationalized by later grammarians; you use ṭas- with animals and more generally non-human animates; te- with human animates (or non-humans who act like humans). Eg.:

ines ṭastamom
ines
cria
ṭas-tamom
CL.2-leg

'the cria's leg'

figsšes šautamom
figsšes
old.woman
šau-tamom
3.POSS-leg

'the old woman's leg.'

This is one of these grammatical that seems to have little basis in fact. In fact the Mannamis' usage is inconsistent (this could be attributed to different writers) and so is later usage. FWIW, Simbri merges classes 1 and 2 and Simbri speakers, when writing in Karaam, tend to use 3rd/4th person pronouns in all cases or hypercorrect and use ṭas- with all animals, even those in class 1.

The rest of the classes are unremarkable; there are, again, short forms for some classes that are used when they do not form a cluster:

lilliš šiferqau
'he ate the fruit' / lilliš hešferqau 'I ate the fruit.'

Dual and plural agreement

The same prefixes are used for singular, dual or plural nouns:

Hedaubaša he šosdaur. Hedaubaša he šosdir. Hedaubaša he šosder.
hedaubaša
1.EXCL.SBJ-CL.8-build
he
that
šosdaur
house(CL8)
/
hedaubaša
1.EXCL.SBJ-CL.8-build
he
that
šosdir
house(CL8)\DU
/
hedaubaša
1.EXCL.SBJ-CL.8-build
he
that
šosder
house(CL8)\PL

'I built that house. I built those two houses. I built those houses.'
Last edited by Ares Land on Fri Jul 29, 2022 8:27 am, edited 1 time in total.
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Demonstratives.

Karaam has three demonstrative particles:
  • proximal i
  • medial he
  • distal ši
i šosdaur
i
PROX
šosdaur
house

'This house'

he ṭugwmnar
he
MED
ṭugwmnar
olive.tree

'That olive tree'

ši mader
ši
DIST
ma-der
pyramid-PL

'Yonder pyramids.'

Demonstrative particles are invariable.

Demonstrative pronouns

There are three series of demonstrative pronouns:
Proximal Medial Distal
Class 1 ibeh hebeh šibeh
Class 2 ? iṭos heṭos šiṭos
Class 3 iṭah heṭah šiṭah
Class 4 ikoh hekoh ? šikoh
Class 5 hiš šiš
Class 6 ? ihoam hehoam ? šihoam
Class 7 ? igahega ? šiga
Class 8 ? idaur hedaur šidaur
Class 9 ? ito heto šito

Only the medial series (in he-) sees any consistent use. A number of the forms listed above -- indicated with a question mark -- are not attested in the Mannamis.
(It's likely the spoken language made good use of all three series.)

The demonstrative pronouns are not inflected for numbers.

Heṭah raṭam am šiṭqikumnis
heṭah
that.CL3
r-aṭam
2.OBJ-PASS\give
am
REL
ši-ṭ-qikumn-is
2.SBJ-CL.3-OPT\obey-FUT

'You have been given these [laws] that you might obey them.'

E qag hesih terumi e hatmen ṭahberqim hebeh.
e
and
qaq
exist
hesih
god\DU
te-rum-i
3.OBJ-see-DU
e
and
hatmen
puppet\PL(CL3)
ṭah-ber-qim
CL3-dance-CAUS.PRS.DU
hebeh
that.CL1

'And she saw two gods, and those ones were making puppets dance.'

Interrogative pronouns

The interrogative pronouns are formed in a similar fashion:
]
Interrogative
Class 1 labeh
Class 2 laṭos
Class 3 laṭah
Class 4 lakoh
Class 5 liš
Class 6 lahoam
Class 7 lag
Class 8 ladaur
Class 9 lato

They can be translated as 'what', 'who' or 'which.'

When expecting an answer of unknown class, laṭahis used.

-- Laṭah am im fonim hohšais?
-- Ḫuggoh.

laṭah
what.CL3
am
that
small
im
fonim
white
hohšai-s
feather-with
/
ḫuggoh
onion

'What is a small white thing with feathers?
-- An onion.'

Other interrogatives

laṭah is also used in the sense of 'how much/how many/how many times':

Laṭah qere heṭuṭas?
laṭah
what.CL3
qer-e
road-PL
he-ṭuṭ-as
1.EXCL.SBJ-walk-FUT

'How many roads shall I walk?'

Laṭah qag ekon?
laṭah
what.CL3
qag
exist
wine
ekon

'How much wine is there?'

We can also list the following interrogative pronouns:
]
Interrogative
Where? larai
How? lasam
When? laṭe
Why? lagau
And here are their declarative equivalents:

Proximal Medial Distal
how many ; this much / that much iṭah heṭah šiṭah
here / there / yonder i / irai herai ši / širai
this way / that wayhisam šisam
now / then iṭe heṭe
for this reason, therefore igau

As you can see, not all of these maintain the three way distinction.

The demonstrative particles i and ši can also be used for 'here' and 'there/yonder', along with their equivalents irai, širai.
Last edited by Ares Land on Tue Jul 19, 2022 4:57 am, edited 1 time in total.
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Raphael wrote: Mon Jul 18, 2022 8:52 am YES!
Rounin Ryuuji wrote: Sun Jul 17, 2022 8:43 pm Little more to say for now than that this looks cool.
Thanks!
User avatar
foxcatdog
Posts: 1652
Joined: Fri Nov 15, 2019 7:49 pm

Re: Karaam

Post by foxcatdog »

I see no *f in the phonology table but it is mentioned in the historical section.

Other than that its cool. I don't think i would have ever made something like this since it doesn't fit my particular aesthetic taste in languages but still cool.
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Case suffixes

Nouns can take the following case suffixes:
  • Dative -de / -ed
    Locative -al / -la
    Ablative -qe, -eq
    Comitative -s
    Instrumental -n
Dative.

For monosyllabic nouns, simply add the suffix -ed after a vowel, or de after a consonant:
  • qweged 'to the king'
  • rumoade 'to your mother'
For all other nouns, the last vowel is deleted; then the suffix -ed is added:
  • qenmam 'vizier' → qenmmde 'to the vizier'
  • raisa 'cousin' → raised 'to a cousin'
Locative.

Monosyllables

For monosyllabic nouns, simply add the suffix -al after a vowel, or la after a consonant:
  • ma 'temple, pyramid' → mala 'at the temple'
  • dan 'swamp' → danal 'in a swamp'
haan, river has an irregular locative: haan → harnal 'at the river.'
Some nouns are inherently locative; ie. they don't take any suffix: qer 'a road, on the road'; sa 'city, in the city'

Other nouns
For all other nouns, the last vowel is deleted; then the suffix -al is added:
  • šosdaur 'house' → šosdral 'in the house'
  • gonðe 'mountain' → gonðal 'in the mountains'
Ablative

Similarly, for monosyllabic nouns, the ablative is marked with the suffix -qe, -e after a consonant:
  • ma 'temple, pyramid' → maqe 'from, out of the temple'
  • dan 'swamp' → dane 'from the swamp'
The ablative of haan is harne. Inherent locatives form the ablative regularly:
qere 'from the road'; saqe 'city, in the city'

For polysyllabic nouns ending in a vowel, it is deleted and the suffix -qe is added after three consonants, -e otherwise:
  • šosdaur 'house' → šosdre 'from the house'
  • gonðe 'mountain' → gonðqe 'from the mountains'
Comitative

The comitative is formed with the suffix -s:
  • mamas 'with a temple'
  • dandans 'with a swamp'
  • haanhaans 'with a river'
  • qerqers 'with a road'
  • šosdauršosdaurs 'with a house'
  • gonðegonðes 'with a mountain'
There are no notable irregularities; however if -s follows a final consonant, it is realized with an implied schwa; that schwa isn't counted for determining stress.
  • haans ['ha.an.əs]
  • šosdaurs ['ʃos.dɔ.rəs]
Instrumental

The instrumental is formed with the suffix -n:
  • maman 'by the temple'
  • dandann 'by the swamp'
  • haanhaann 'by the river'
  • qerqern 'through the road'
  • šosdauršosdaurn 'by the house'
  • gonðegonðen 'by the temple'
Similarly, -n is realized with an implied schwa after a final consonant:
  • haann ['ha.an.ən]
  • šosdaurn ['ʃos.dɔ.rən]
Case and number

Case suffixes are applied to the singular form only; nouns with case suffix do not mark number in any way.
  • haann 'by the river, by the rivers'; maqe 'from the temple; from the temples.'
qer and sa take normal plural markings, even when used as locatives:
qere 'roads, on the roads'; sae 'cities, on the cities'

With pronouns or class markers

The object pronoun, or the class markers, can be added to the following case markers:
  • Dative -de
  • Locative -la
  • Ablative -qe
  • Comitative -sis
  • Instrumental -nu
A few examples:
  • ude 'to him/her' (4th person)
  • sade 'to me' (1st person exclusive
  • hoala 'on it' (class 7)
  • ganu 'with it, using it' (class 6)
  • sala 'near me, by me' (1st peson exclusive)
  • tesis 'with him'
  • rinu 'through you, thanks to you.'
Relational nouns

Relationals in Karaam are the equivalent of pre- or postpositions.
They agree in person or in class with the noun phrase they refer to. For instance:

šosdaur daugor
šosdaur
house(CL8)
dau-gor
CL8-inside

'Inside the house.'

gammo hogor
šosdaur
palace(CL4)
ho-gor
CL4-inside

'Inside the palace.

sauhqo
sau-hqo
1.POSS-side

'At my side'

qweg šauhqo
qweg
king
šau-hqo
3.POSS-side

'At the king's side.'

Relationals are often placed after the noun phrase, so they are not unlike inflected postpositions... But they can be, in fact, moved freely and separated from their noun phrase:

Šosdaur mah daugor.
šosdaur
house(CL8)
mah
warm
dau-gor
CL8-inside

'It's warm inside the house.'

When used with inanimates, relationals, as you can see above, take the appropriate class prefix.
With animates, some relationals will require possessive prefixes:

šauhqo
šau-hqo
3.POSS-side

'At his side.'

But others require object prefixes instead:

qweg tepsis
qweg
king
te-psis
3.OBJ-without

'Without a king.'

Note that neither the relational, nor the noun take any case markers.

Here's a list of the most common relationals:

Relational Meaning
OBJ-sna'like, similar to'
OBJ-qeh'over'
OBJ-dah'towards, until'
OBJ-kha'from'
OBJ-sha'under'
POSS-mur'behind, after'
POSS-qamo'before, in front of'
OBJ-dem'among, during'
POSS-gor'inside'
POSS-hqo'besides, at the side of, by'
POSS-pušah'across, beyond'
POSS-qem'near, around'
OBJ-qi'at the entrance of, near, close to'
OBJ-psis'without'

There is no real equivalent of 'left' and 'right'; POSS-hqo can be translated 'right' and POSS-pan 'left', but judging both from textual evidence and present-day Dimoan language, it seems POSS-hqo really refers to the dominant hand and POSS-pan for the other. So when a left-handed person says sauhqo, he really means his left hand.

For directions, Karaam offers several alternatives:

The cardinal points:
  • POSS-ro 'north'
  • POSS-mdam 'south'
  • POSS-tur 'east'
  • POSS-ig 'west'
Towards the sun-facing side of the mountain: POSS-ðoraus or the shady side: POSS-bian.
'Uphill' POSS-gonðe or 'downhill' POSS-qog.

(The modern Tarandim often use the cardinal points; they have an uncanny talent for orienteering.)

This list of relationals is not exhaustive; in fact the process of forming relationals was still productive; you could find, for instance, nonce constructions:

Regghas tegammo
Regghas
Regghas
tegammo
3.POSS-palace

'Under the rule of Regghas' (lit. 'Reggas' palace.'

Kobgonðe doðuboa
Kobgonðe
Kobgonðe
do-ðuboa
CL9-foot

'In the foothills near Mt. Kobgonðe.'

Again, no case markers are used.

Absolute directions

Nominal forming absolute directions can be formed from relationals; the derivation is often unpredictable. A few example:
  • -qeh 'above' → qean 'what is above, heaven'
  • sha 'under' → šotis 'what is below, earth'
  • POSS-gor 'inside' → guorṭes 'the inside'
  • POSS-ro 'north of ', POSS-mdam 'south of', POSS-tur 'east of' POSS-ig 'west of' → rol, maam, tur, ig
Last edited by Ares Land on Wed Jul 20, 2022 6:05 am, edited 2 times in total.
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

foxcatdog wrote: Wed Jul 20, 2022 5:31 am I see no *f in the phonology table but it is mentioned in the historical section.

Other than that its cool. I don't think i would have ever made something like this since it doesn't fit my particular aesthetic taste in languages but still cool.
Thanks. (And that's fixed, there is indeed a /f/)
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Numbers and quantifiers

Numbers

The Tarandim count in base twelve; this derives from a method of finger-counting (you count the finger bones of one hand, using the thumb as a pointer.)
I use b12 for base twelve numbers, also a for 10 and b for 11.

As far as we can tell, they developped a positional system, including zero, before writing.
One difficulty is that Karaam speakers always used the numeral system and rarely spelled out numbers. So we have the numbers from zero to 12, the word for 144 (100b12 ), the word for 1728(1000b12)

The numbers are as follows:

pezero
ho1
bi2
qwen3
kerka4
ligeh5
kešo6
nihðe7
eme8
barto9
main10 = ab12
dau11 = bb12
gim12 = 10b12
ṭemah144 = 100b12
kwuggwos1728 = 1000b12


The Simbri and Asuraam practice, for higher numbers is to simply spell out the numeral, and the same convention is used for Karaam.

Let's see how it works, using Simbri and Karaam.
  • 24 → 20b12 → S. me piyo K. bi pe 'two zero'
  • 238→ 17ab12 → S. poma nitli man, K. ho nihðe main 'one seven a'
If there's more than one zero, the number of zeros is given, followed by po
  • 288 → 200b12 → S. me mepo K. bi bipo 'two two-zero'
  • 10368 → 6000b12 → S. čiso binpo K. kešo qwenpo 'two three-zero'
  • 2985984 → 1000000b12 → S. poma čisopo K. ho kešopo 'one six-zero'
For 10000, 100000, 1000000, the initial S. poma / K. ho 'one' can be omitted → S. keččapo/ K. kerkapo, S. linipo / K. ligehpo, S. čisopo K. kešopo

100 and 1000 (base 12 → 144, 1728 base 10) have specific names: toma/ṭemah, doma/kwuggwos. These can't be used for the hundreds or the thousands: 200 (base 12): *me toma but S. me mepo, K. bi bipo.

For large numbers and a precise count, the -po number can be given to indicate how many digits are expected:
  • 2234567b12 → S. me cisopo me bin kečča lini ciso nitli (two six-zero two three four five six seven), K. bi kešo bi qwen kerka ligeh kešo nihðe
Again, if the first digit is 'one', it may be omitted if followed by a -po number:
  • 13456b12 → S. keččapo bin kečča lini čiso (four-zero three four five six) K. kerkapo qwen kerka ligeh kešo
Modern speakers use this system when souding out Karaam numbers; it was not the system used by native speakers when Karaam was a living language. The suffix -po is not Karaam to begin with. (It's an Asuraam/Simbri cognate of pe)

A few (very few) inscriptions provide vowel markings for numbers; for instance 55b12: I-E / I-AA-E, 132b12: I / E-E

That would be consistent with:

55
*
ligeh
*five
ligaagm-e
fifty-ABL

Assuming 50b12 is *ligaagVm

132
*
(ṭemah)
*(hundred)
bi
two
qweggm-e
thirty-ABL

Assuming 30b12 is *qwenVgm


Numbers are invariable particles. bi 'two' precedes noun in the dual, naturally enough. pe 'zero' requires the plural, as do, of course, all other numbers except ho 'one'

pe gonðde
pe
zero
gonð-de
mountain-PL

'No mountains'

ho bos
ho
one
bos
dog

'one dog'

bi hesih
bi
two
hesih
spirit\DU

'two gods'

kerka hatmen
kerka
four
hatmen
puppet\PL

'four puppets'

Fractions

The following fractions are attested:

bibun '1/2'
qwemun '1/3'
kerkun '1/4'
gimun '1/12'

The other fractions are attested as numerals only; we do not know their pronunciation either.
By analogy, it is assumed they were formed by adding -un main 'ten' → mainun 'one tenth' or removing the last vowel and suffixing -un: emeemun 'one eighth'

The fractions were invariable as well: bibun habmug 'half a field'

Ordinals

Ordinal numbers are formed with a suffix -an:
  • pean 'zero-th, number zero'
  • hoan 'first'
  • bian 'second'
  • qwenan 'third'
Higher ordinals are not attested.

We can add uroan 'last.'

Ordinals are again, invariable, when used as modifiers.

hoan → Du. hoin → Pl. hoen

As for agreement, they're treated as animates when referring to animates, as class 3 otherwise.

The Tarandim count starting as zero. This explains why Karaam has an ordinal for zero.
The first year in the Calendar is really pean fig 'year zero'; the first man is Pean

Nouns

Numerals are turned into nouns by adding the suffix -hes :
  • pehes 'zero, absent'
    hohes 'one, unique'
    bihes 'two, dual, twofold' (note the vowel change; this necessarily implies a dual!)
    qwenhes 'three, threefold'
Quantifiers

The quantifiers are invariable; the most common are:
  • pe 'zero, no'
  • no 'some, a few, a little'
  • noṭah 'a few, several'
  • bago 'many, much'
  • bagṭah 'many, a lot'
  • dopoa 'full, entire'
  • qan 'all'

The following nouns are derivations:

  • pebeh 'no one'
  • pehes 'nothing, no one'
  • qameh 'everyone'
  • qannah 'everything'
  • pebgal (locative) 'nowhere' , ablative pebge 'out of nowhere'
  • pesam 'never'
  • qansam 'always'
  • qamgal (locative) 'everywhere', ablative qamge 'from everywhere'
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Transitive verbs

The active infinitive.

The citation form is the active infinitive. It translates as an infinitive or a gerund. A close translation could be 'the act of X'

eṭah 'to write, writing'
ɓennes 'to put, (the act of) putting'
ðeri 'to like, liking'
kumon 'to command'

The active infinitive is a verbal noun; as such it has a noun class -- determined morphologically. Mostly infinitives end up in class 3, eṭah (class 3), ðeri (class 3), kumon (class 3) but that's not a hard rule: ɓendes belongs to class 2, as it ends in -Vs.

Monosyllabic verbs are class 3:

qen 'to fix'
gwaug 'to govern'

The active infinitive triggers agreement:
eṭah ṭahṭam Irukah
eṭah
write
ṭahṭam
CL.3-give
Irukah
Irukah

Irukah gave (us) writing.

It isn't marked for plural, or for case.

The passive infinitive

The passive is formed by changing the first vowel to a. It translates as 'the result of X' or 'the undergoer of X':
aṭah 'writing, a text, written word'
ɓannes 'what has been put down.'
ðari 'what one likes, things one likes.'
kamon 'orders' or 'subordinate'

The passive is assigned a class; it triggers agreement and is marked for case and number:
aṭeh 'writings'; ðare 'things that are liked.'

(The dual and plural, however, are consistently marked by vowel change; the dual and plural may merge, as in the case of ðari).

For monosyllabic verbs, the passive is formed with a prefix a-

aqen 'a repair'
agwaug 'what is governed; the state'

The antipassive infinitive

The antipassive is usually formed by changing the first vowel to e; if the first vowel is i, it changes to i instead.
Essentially the antipassive infinitive is an agent noun:

eṭah 'to write' → iṭih 'writer, scribe'
kumon 'command' → kemon 'commander, leader.'

For monosyllabic verbs, the antipassive is generally unpredictable. (In fact it is formed by reduplication, but the process is obscured by sound changes.)

kqen 'repairman, one who fixes'. (There's an implied schwa here: kqen [k'ək'en]
mgwaug 'governor'

The antipassive is often an agent noun, but there is often an unpredictable shift in meaning:

ṭam 'give' → tṭam 'a giver' but also 'to hold a potlatch'

The purpose of voices.

Starting with the active infinitive, adding subject and object prefixes gives us a finite verb form (the past tense, in fact; but more on tense later).

eṭah
ṭeṭah, heṭeṭah
∅-ṭ-eṭah
3.SBJ-CL.3-write
he-ṭ-eṭah
1.SBJ-CL.3-write

'he wrote it, I wrote it'

(remember that there is no 3rd person subject prefix; or, more precisely, it's ∅-)

Voice in Karaam has a dimension of telicity. The active voice is telic; there is an implied endpoint.

How do we get from here to an intransitive?

Passive
When the agent is unimportant, or not in focus, we switch to the passive voice:
aṭah
∅-aṭah
3.SBJ-PASS\write

'It is written.'

aqen šosdaur
∅-aqen
3.SBJ-PASS-repair
šosdaur
house
'the house is repaired.'

The former object becomes the subject; here the subject ("marked" with ∅-) is of coure the house.

There's not far from this to a patient noun: aṭah 'it is written/ what is written'

Note that the distinction of tense shifts to aspect in the passive; the active past tense corresponds to a passive perfective.

Antipassive
Finally, we can omit the patient; in which case we use the antipassive:

eṭah 'he writes'
kqen 'he repairs'

Here the sense is atelic; there is no clear endpoint; that endpoint (marked with the object prefix) has been removed.
Again tense switches to aspect; this corresponds to an habitual:

Active past = Stative perfective = Antipassive habitual.

Again the shift to a verbal noun is not great eṭah 'he writes habitually' = kqen 'he is a scribe'

Further 'voices'

There are three further verbal forms triggered by a change in the first vowel; for convenience they're classified as voices though they have a modal aspect as well.

The optative

The optative is formed by changing the first vowel to -i-
kumonkimon 'to be willing to command.'

The optative patterns with the active; that is, it's transitive and requires an object.

For verbs that have -e- as a first vowel, the optative infinitive is identical to the antipassive. In finite forms, you can count the prefixes :)
ṭeṭah 'he wrote it' → ṭiṭah 'he wants to write it' (again, the indication of tense is lost in the optative.)

For monosyllabic verbs, the optative is marked by a prefix i-

qeniqen 'to be willing to fix'

The deontic
The deontic is marked by changing the first vowel to a, then adding -n, or -m before labials:
kumonkammon 'orders that must be given'
eṭahanṭah 'what must be written'

Essentially this is a passive version of the optative; there's a (fairly natural) semantic shift to 'what must be Verbed'

Monosyllabes use a prefix na-

qennaqen 'what must be fixed.'

The deontic patterns with the passive.

The potential
The potential is warked by changing the first vowel to -o:

kumonkomon 'capable of giving orders'
eṭahoṭah 'literate'

Monosyllables have a prefix bo- instead:

gwaugbogwaug 'capable of ruling'
qenboqen 'capable of fixing'


The potential patterns with the antipassive -- ie. it's intransitive.


A summary of possible forms is given below:

Active Optative Passive Deontic Antipassive Potential
'to order' kumon kimon kamon kammon kemmon kommon
'to write' eṭah iṭah aṭah anṭah iṭah oṭah
'to govern' gwaug igwaug agwaug nagwaug mgwaug bogwaug
'to put' ɓennes ɓinnes ɓannes ɓannnes ɓinnes ɓonnes
'to give' ṭam iṭam aṭam naṭam tṭam ɓoṭam
'to like' ðeri ðiri ðari ðanri ðiri ðori
Last edited by Ares Land on Thu Jul 21, 2022 3:13 am, edited 1 time in total.
User avatar
Rounin Ryuuji
Posts: 2994
Joined: Wed Dec 23, 2020 6:47 pm

Re: Karaam

Post by Rounin Ryuuji »

Do you have an idea of how these vowel alternations came about, or are you sticking to what's actually internally "attested"?
User avatar
Man in Space
Posts: 1651
Joined: Sat Jul 21, 2018 1:05 am

Re: Karaam

Post by Man in Space »

I don't have much constructive other than to say this looks really well-done. I especially am fond of the noun classes and the potential.
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Man in Space wrote: Wed Jul 20, 2022 9:55 pm I don't have much constructive other than to say this looks really well-done. I especially am fond of the noun classes and the potential.
Thanks!
Rounin Ryuuji wrote: Wed Jul 20, 2022 4:43 pm Do you have an idea of how these vowel alternations came about, or are you sticking to what's actually internally "attested"?
In the proto-language, verbs were monosyllabic:

*t'ah 'count, tally, make a mark'
*poh 'hit'
[*]*tʰwes[/i] 'get, do'

or sesquisyllabic:

*kumən 'order, command'
*dəkwain 'judge'


And the passive was formed by a prefix ʕa-:
*ʕat'ah 'be counted'
*ʕapoh 'be hit'
*ʕatʰwes 'gotten, done'
*ʕakumən 'be ordered'
*ʕadəkwain 'be judged'

Disyllabic verbs first arose as compounds:

*we-t'ah 'ochre-tally' → 'write'
*ð'wah-poh 'go-hit' → 'fight'
*pāin-tʰwes 'place-do' → 'put'

By analogy, a full vowel was restored in some sesquisyllables, giving another class of disyllables:
*kumən*kumon 'order, command'
*dəkwain*daikwain 'judge'

The passive prefix stayed in its initial place, so it was either an infix or a prefix:

*weʕat'ah 'be counted'
*ð'wahʕapoh 'be hit'
*pāinʕatʰwes 'be hit'
*ʕakumon 'be ordered'
*ʕadaikwain 'be judged'

Stress fell on the penultimate; and preceding vowels were lost; was (mostly) lost as well:

*wat'ah 'be counted'
*ð'waapoh 'be hit' (*a + *hʕa → *aʕaaa)
*pnatʰwes 'be put'
*kumon 'be ordered'
*daikwain 'be judged'

Difficult clusters were resolved by metathesis: *pnatʰwes*pantʰwes.

Hence the Karaam forms, with vowel change: eṭahaṭah; dahoh → daahoh, bennesbannes; the earlier sesquisyllables had lost the passive marker -- but started using vowel change by analogy: kumonkamon, daikaindakain.

The antipassive derives from an earlier inanimate plural prefix e-: *we-tah → *we-e'tah → *wey'tahiṭah; *ð'wah-epoh → *ð'whepoh → *ð'wehpohdehoh.
(An habitual was formed by reduplication, and that construction was used instead for monosyllables.)

Likewise, the optative is from an earlier prefix *yi-, and the deontic from *na- (*we-tah → *we-natah → *wnatah → *wantahanṭah)
The potential derives from a prefix *pwo-; disyllabic verbs have vowel change instead of an infix by analogy with the other forms.

(I think in-universe there wouldn't be enough data to reconstruct the original system, but hey.)
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Some difficulties.

Vowel mergers
Karaam has seven vowels, five constructions with vowel change, and almost no constraint on what the original first vowel is. Naturally there are mergers.

The first vowel is e
That's the most common case, so common in fact I couldn't really ignore it in the preceding post. As we've seen, both the antipassive and the optative are in -i-
eṭah 'to write' → optative and antipassive iṭah

To distinguish the two, the optative requires subject and object marking; the antipassive only the subject.

The first vowel is a
There's no merger in that case; the passive is in -aa-. Additionally, the optative is in -ai- instead of the expected -i-

dahoh 'to fight' → passive daahoh, optative daihoh.
The other forms are as expected: antipassive dehoh, deontic danhoh, potential dohoh

The first vowel is i
There's no optative. riqaa 'to seek'. Those verbs tend to have an optative meaning anyway.

The first vowel is o
The potential is identical to the active: dopoa 'to pour' → potential dopoa. Both voices are distinguished by the presence (active) or absence (potential) of an object prefix.

The first vowel is u, ai, au
This class of verbs can run through all patterns of vowel change, with no mergers.

qauṭes 'to cook' → optative qiṭes, passive qaṭes, deontic qanṭes, antipassive qeṭes, potential qoṭes

Polysyllables

Verbs may consist of more than two syllables. Most are the result of noun incorporation and are generally intransitives; they'll be discussed below.

There are a few anomalous, unanalyzable polysyllables, notably qokumon 'to obey'. Essentially the middle vowel is ignored; only the first vowel changes:

qokumon → passive qakumon, antipassive qekumon, deontic qagkumon (and so on.)

Medial consonant

The nasal infix in the deontic interacts with the middle consonant in various ways.

Nasal infix -am-

The nasal infix is -am- before a labial: p, b, ɓ, f, m:
kumonkammon
daifar 'to cut' → damfar

It is -am as well when there is no medial consonant:

ṭoer 'to tell, to say' → ṭamer

Before velars

Before velars or labio velars: k, g, q, kw, g, qw, the form of the infix is -an:

,kwogwar 'sacrifice, propitiate' → kwangwar

Except if m or n occurs anywhere in the same word, in which case it's -ag

daikwaindagkwain

That's most likely a writing convention. Native speakers probably had */kʷaŋŋʷaʁ/, */naŋkʷɛn/, ie. the infix was */aŋ/ in all cases.
There's no velar nasal in Simbri; Simbri speakers use a spelling pronunciation and thus have two realizations for the infix: /an/ in /kwanʁar/ - and /aʁ/ in /naʁkwɛn/. (The pronunciation given here is indicative; Simbri speakers approximate Karaam pronunciation as best they can; realizations vary.)

In all other cases, the form is -an- . baša 'build' → banša.
-n- triggers nasal harmony, so baša is [ba.ʃa] while banša is [man.ʃa]

Medial clusters

A number of verbs have a medial consonant cluster:
  • ɓennes 'to put'
  • ðopfer 'to announce'
  • kuggug 'to free'
  • qahɓo 'to flay'
  • ferqau 'to eat'
How do these form the deontic?

h and r are deleted:

qahɓoqambo
ferqaufanqau

In theory, medial geminates are kept. The cluster is broken up with a schwa:
ɓennesɓannnesɓann(ə)nes
kuggugkanggugkang(ə)gug

That is, at least, what native grammars state.

In practice there's a lot of variation; and it seems native speakers simplified the geminates.
The Mannamis for instance seems to have ɓannnes in free variation with ɓannes and you'll also find kangug instead of kanggug.
This may be scribal error; or maybe medial geminates were only kept in careful speech.

All other clusters are kept, though broken up with (an unwritten) schwa:
ðopferðamp(ə)fer
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Intransitive verbs

The citation form is the base infinitive:
  • šog 'to sleep, asleep'
  • reṭa 'to run, running'
  • hauhau 'cinnabar, to be cinnabar red'
  • ea 'liquid, to flow, to be liquid, to run'
  • ara '(to be) supernatural'
As you can see, the class of intransitives covers words that could be translated as intransitive verbs or adjectives.

In fact, the Karaam equivalent of adjectives all fall into this class.

There's a fine to nonexistent line between nouns, the base infinitive of intransitives, and nouns:

ara can be translated as 'to be supernatural', 'supernatural' or 'it was supernatural event, a supernatural occurence'
ea: 'it's running water', 'running water', 'stream', 'brook.'

Indeed a lot of zero-derivation goes on, from noun to intranstitive: rumeh n. 'dead person' → rumeh, vi, 'to die.'; ðam 'gold' → ðam, vi, 'to be made of or plated with gold'

Vowel change

Voice isn't really applicable to intransitive. Several patterns of vowel change apply anyway; we'll call them A-Form (equivalent to a passive) E-Form, AN-form and O-form.

These forms are less used than their active equivalent. Unlike their active equivalents, it's also hard to point out a specific meaning for each; mostly they trigger unpredictable semantic shifts. It can be said that these forms belong to derivational morphology rather than to inflectional morphology.

A-form
This construction is analogous to the passive. The meaning here is that of a causative + a passive, ie. it described a state that is the result of a previous action. There's often a semantic shift as well.
Monosyllables use a prefix a-
  • šog → ašog 'to be asleep, to have fallen asleep'
  • reṭaraṭa 'chased' but also 'tired'
  • hahau 'to be painted with cinnabar'
  • aa 'poured down, pumped, irrigated'
  • aara 'a miracle'
E-form
This form is analogous to the anti-passive. It's formed in the same way: vowel change to i or e for disyllabic verbs, reduplication for monosyllables.
Often the E-form softens the meaning of the active form with an added semantic shift.
  • šog → sšog 'sleepy, lazy, dozing off'
  • reṭariṭa 'run a little, run slowy, amble'
  • hahauhehau 'vaguely/partly the color of cinnabar'
  • ia 'to trickle, a trickle'
  • ira 'nice, talented, unexpected, surprising'
AN-Form
The I form expresses a wish, a need, or a tendancy; but it may also mark possibility. It's formed with a prefix na- (for monosyllables), or by replacing the first vowel with -an-:
  • šog → našog 'needing sleep, tired'
  • reṭaranṭa 'to want to run'
  • hanhau 'to have to be cinnabar red'
  • ana 'dry, a drought'
  • anra 'in need of a miracle.'
O-Form
This form marks possibility, uncertainty or potential. It's formed with bo- or a vowel change to -o-:
  • šog → bošog 'possibly sleeping'
  • reṭaroṭa 'capable of running / possibly running'
  • hohau 'maybe red'
  • oa 'maybe flowing'
  • ora 'possibly supernatural'
There's no I-Form (equivalent to an optative) for intransitive verbs. However, intransitives have an additional form, the elative.

The elative
The elative is formed by prefixing a- to the noun and:
  • Doubling the initial consonant for monosyllables.
  • Removing the first vowels for disyllables.
The elative forms a comparative or a superlative; occasionally an agent noun.
  • šog → aššog 'more asleep'
  • reṭaarṭa 'running more' or 'a runner'
  • ahhau 'redder' but also 'royal'
  • aa has no distinct elative.
  • ara has no distinct elative.
Morphological difficulties.

The rules given above for transitive verbs apply.
  • reṭa 'to run' → 'rita 'run a little, amble'
  • aqwe 'certain' → anqwe 'must be certain'
In addition verbs begin with a vowel have an elative either identical to the A-form: aa or to the base form: ara.

The different forms of intransitives can be summarized in the following table:

Base form Elative A-Form E-Form AN-Form O-Form
'to sleep' šog aššog ašog sšog našog bošog
'to run' reṭa arṭaraṭa riṭa ranṭa roṭa
'(cinnabar) red' hauhau ahhauhahau hehau hanhau hohau
'to flow' ea aaaa ia ana oa
'supernatural' ara araaara era anra ora
Ares Land
Posts: 2949
Joined: Sun Jul 08, 2018 12:35 pm

Re: Karaam

Post by Ares Land »

Noun incorporation

Another way to reduce valence is noun incorporation; a noun is combined with a verb to form a new verb.

Eg.:
  • šaun 'incense' + kwogwar 'propitiate' → šaunkwogwar 'to offer incense'
  • kog 'hair' + qam 'wash' → kogqam 'to wash one's hair'
  • šeḫḫa 'deer' + uran 'hunt' → šeḫḫuran 'to hunt deer'
  • dogweh 'tribute' + gwaug 'to govern' → dogwhgwaug 'be a tributary lord'
  • bauṭos 'llama' + kwogwar 'propitiate' → bauṭskwogwar 'to sacrifice a llama'
In Karaam, noun incorporation is used to narrow the scope of the verb semantically, thus creating new verbs for unitary concepts. (Mithun, 1984.)

The process is as follows: the noun comes before the verb in the compound; if the noun is more than two syllables long, the last vowel is deleted:
  • dogwehdogwh-
  • bauṭosbauṭs-
Resulting clusters are resolved by (an unwrittent) schwa: dogwh(ə)gwaug, bauṭs(ə)kwogwar.

A verb with incorporated nominal can be taken to be equivalent to an antipassive. Such verbs are thus always intransitive, and form no distinct antipassive.

However, all other voices can be formed; the verbal component is modified alone, with no change to the incorporated nominal:
  • šaunkwogwar 'to offer incense' → šaunkwagwar 'offered incense', šaunkwangwar 'incense that must be offered', šaunkwigwar 'willing to offer incense',šaunkwogwar (identical to the base form, but also potential: 'capable/allowed of offering incense')
  • kogqam 'to wash one's hair' → kogaqam, kognaqam, kogiqam, kogboqam (ie, inflects like qam)
  • dogwhgwaug 'be a tributary lord' → dogwhagwaug, dogwhnagwaug, dogwhigwaug, dogwhbogwaug
One exception though; when both nominal and verb are monosyllabic, there's a definite tendancy to treat it as a unitary, regular ditransitive verb.
So the forms kagqam, kaggqam, kigqam, kogqam are also allowed alongside kogaqam etc.
Post Reply