Repilian languages/Owl

From FrathWiki
Jump to: navigation, search

The Owl language was spoken in 0 AD in the area that later became the Moonshine Crown. It was the proto-language of many others that survived for a few thousand more years.


The phonology of Owl, like most related languages, was small, but its grammar was such that one coudl talk about two phonologies: the surface level and the grammatical one. In grammar, MR languages icluding Owl treat only a few hard consonants as proper consonants, and these are what the noun rotos are made of.


Surface consonants are

HARD:    t  k  s  p  ṇ
SOFT:    r  ŋ     v  n  m  l  y  ʔ  ʕ  h

The soft and hard consonants behave differently in the grammar; the soft consonants never appear in roots, only in affixes. The soft consonants are inherited from the MRCA of Owl and Tapilula, and this is why Tapilula's infixes preferentially use the consonants /l h g k ḳ ŋ/. Even though the set of consonants used is very different, the concept is the same and survived 14,000 years of evolution in both branches.[1]

There may also be /x g/ in the soft row, but these would be restricted to certain positions.

The nasals m n (and ŋ?) only appear in the coda, whereas behaves as a hard consonant and can appear in all positions, even in clusters like /tṇ/.

Note: it is possible that the position restrictions on /m n/ mean that they would never be distinct from each other. Even though Owl allows disharmonious clusters like /tp/, /kn/, etc, it is possible that the nasals would not be allowed to do this since they are "weak" and considered part of the vowel nucleus. If so, then /n/ is a hard consonant and /m/ a soft one, with no overlap.

Resolution of initial consonant clusters

In most words, a vowel intervenes between the first two consonants. No word has a vowel before the first consonant, and no word has a soft consonant before the first hard consonant. Thus, any word which acoustically begins with anything other than /t k s p n/ can be recognized as undergoing elision of one of those five sounds. For example, the clusters /tl kl sl pl/ are pronounced as spelled in initial position, but /nl/ is reduced to a simple /l/, and word-initial /l/ never surfaces in any other context.

     R    Ŋ    V    M    L   Y   ʔ   ʕ   H       T   K   S   P   N
T    tr        tw        tl  tš          th      t*  tk  ts  p*  n* 
K    kr   ŋ*   kw        kl  kš          kh      kt  k*  ks  kp  kn
S    sr        sw   sm   sl  š           sh      st  sk  s*  sp  sn
P    pr        pw   m*   pl  pš          ph      pt  pk  ps  p*  pn 
N    r    ŋ*   nw   m*   l   ň           h       d   ġ   z   b   n*

It is possible that consonants like /ʔ/ and /ʕ/ simply never occur in this slot. Alternatively, they serve to voice the initial consonant, which may or may not merge with the nC clusters.

*Indicates a surface phoneme that conflicts with another, either single or double. IF THERE ARE NO MERGERS, these will probably need to change.

More likely, though, mergers are allowed, and there may even be dissimilatory pairs such as /t...p/ which occur in preference to /p...p/ but then merge into /pp/ when no vowel intervenes.

It is possible that /nt/, /ŋk/, should be allowed initially after all, BUT only as part of a sandhi process involving clusters like /tm/.

Since the nasals /m ŋ/ are weak and mostly occur adjacent to a consonant, there may in fact be only one of them, meaning that there is no contrast between /m/ and /ŋ/ even intervocalically. Alternatively, they contrast *only* intervocalically, meaning that many forms merge when the vowel padding is absent on one side.

Resolution of final consonant clusters

Since no vowel or soft consonant can occur after the final hard consonant in a word (unless there are single-vowel clitics), there may be similar processes of consonant cluster reduction at the end of each word. But because words have final stress, these will likely be less intense than the initial reductions.


a  i  u  ə

The schwa is spelled as /e/. The IPA /e/ vowel appears in some MR languages, and must be spelled with a different glyph, because even these languages also have the schwa.

Allophones of vowel-glide sequences

These are not sequences of two vowels, because one must always be a glide. None of this is reflected in the spelling. Note that three-member sequences also exist.


    ʕ   Y   V
A   ā   ɛ̄   ɔ̄ 
I       ī   ȳ
U       ɨ̄   ū
E   ə̄   ē   ō



Stress is probably word-final unconditionally at least on nouns. Verbs might have dynamic stress but final is the most common there as well.


All nouns are consonant shells, and they take auxiliary verbs as infixes.


There are only about 360 roots in the language. All noun roots consist of four consonants, and there is no other information to set them apart. Any root whose surface form appears to have less than four consonants only appears so because the surface form contains a cluster such as /ttt/ which is pronounced as /t:/ rather than /t::/, and even these nouns will still separate the consonants in most of their inflectional forms.

Alternatively, there could be three consonant lengths, but with the weaker grades turning voiced allophonically. This requires that there be a true contrast between voiced /k/ and /g/.

Even the very simplest nouns, like "water", consist of four consonants. However, as above, some noun shells have the same consonant two, three, or even four times in a row, and the surface forms of these words merge the resulting clusters to make them sound as if they have no more than two stuck togther.

Because of the 4-consonant minimum, all words for nouns are quite long. Thus, many concepts are not expressed by nouns. For example, there are no pronouns; all of this is marked on the verb. Inalienable possessions are often not expressed by nouns either but by affixes of manner such as "by hand", "on foot", etc. To form a word for hand, foot, etc in isolation one would then need to put a further affix on this word, and this word would be as long as a 4-consonant noun would, but is not used often.

Nouns for simple concepts are often reduplicative in origin. For example, the root KTSK means owl, but SKSK means bird.

Consonants shells

Only hard consonants can be part of the shell. In Owl, these constonants were t k s p n. Any other consonant is treated by Owl's grammar as part of the vowel nucleus of the syllable; as well, /n/ has a second form that also occurs as part of the vowel nucleus.

Helper verbs

Nouns must agree with the verb they are the agent of, if there is one. This is done by sticking an infix into the noun; this also can happen when there is no verb in the sentence, and in this case it takes the meaning odf a simple tenseless verb. THis can be called an auxiliary verb. The infixes have very broad meanings mostly, but some are more specific such as "sleep".


All infixes go in one of three specific slots.

Slot A

  1. C-a-CCC null (used on most animates)

Noun case inflections might go here.

Slot B

  1. CC-ev-CC male
  2. CC-vi-CC female

Slot B did not exist in the early days of the language. It only emerged when the four-consonant minimum emerged and primordial two-consonant roots got stuck to other ones. This is probably the most nounlike slot, but may also be the one that contains the auxiliary verbs. It may also contain body parts.

Slot C

  1. CCC-am-C progressive aspect (auxiliary verb)

Slot C is probably the most verblike slot. If the auxiliary ("helper") verbs are not in B, then they will be here.

Words for emotions are here. The emotions are those of the agent, not the speaker, and like Poswa, things like "hungry" and "sleepy" are considered emotions.


Thus for example, from the root KTSK "owl", one can say katevsk "male owl" and katvisk "female owl", using the animate infix and one of the gender infixes. The surface pronunciations of these are [katōskə] and [katʷiskə].

Noun and verb gender

Owl, like other Repilian languages, was spoken in a hyper-feminist society in which men had lower social status than pest animals such as rats. Grammatical traits found in Owl and its descendants contributed to the grammaticalized gender bias later found in the Icecap Moonshine language, though it is likely that the Owl languages were never as extreme as IMS, and that IMS innovated most of its traits on its own.

Relationship with nearby languages

Repilian languages suffered at the hands of the male warriors of Sakhi, who believed that they could clean up the world by genociding all of the dirty peoples to their north. But then the Sakhi women stood in front of the men and told them the war was over. The soldiers tried to fight back but their wives overpowered them and then signed a surrender treaty handing over all Sakhi land to the Repilians. The Repilian women and the Sakhi women teamed up to keep their men in line and soon the Moonshine women saw what was happening and joined in as well. Thus Moonshine became the language of the ruling class, while Sakhi was spoken by slaves, but Repilian traits spread in both directions even as the Repilians themselves began to blend into the Moonshines.


Brief timeline of history:

  • 11000 BC: Repilian-Zenith splits from Paleo-Andanese. These are tribes, not languages; not one of the three groups still speaks a language directly inherited from the primordial split by 4000 AD.
  • 8000 BC: Repilian splits from Zenith. This is where the people start to be physically distinct from each other. The language by this time has evolved into something like Australian, with probably just three vowels, no tones, and no fricatives. Those three vowels are probably /a i u/, meaning that the vertical inventory grew as the labialized consonants were lost and then shrank again. The language never evolved a voice distinction either, and although the prenasals are probably still around, they no longer contrast with sequences.
  • The "semivowels" /ʕ j w/ may be still around, and can occur in the coda to form long vowels and diphthongs, so long as no other consonant is in the coda.



This page has a NOTES section. See the notes below for additional information on these languages.




  • Notes
  1. Notes:


Note:This is the beginning of the NOTES section.
  1. Tapilula also did /r/ > /y/, and has one common infix with a /y/, but this sound otherwise came to pattern with the vowels and appeared in no other common infixes.