Rílin phonology

This phonology is of the Sunuli dialect of Rílin, which is spoken by the group of Ríli who maintained their civilization above ground after the Flight, when the group known as the Lunauli retreated into the denser forests and underground caverns. The Sunuli, by contrast, remained in their original settlements and took up war with the Tosi who had been attacking them for four years. After the first two centuries of cultural and linguistic separation, the Rílin language began to split off into to two mutually intelligible but phonemically and syntactically distinct dialects. The focus of this overview is the Sunuli dialect, but some comparative phonology with Lunauli will be presented.

Phonemic Inventory

Rílin has a fairly large inventory of both consonants (29) and vowels (13). Consonants occur at nine places of articulation and in seven manners of articulation, contrasting voiced vs. voiceless phonemes. Most places and manners of articulation have both a voiced and voiceless version.

Consonants

Labial Labio-dental Alveolar Post-alveolar Retroflex Palatal Velar Uv. Glottal
Stop p b t d k g q ʔ
Nasal m n ŋ
Fricative ɸ β f v s z ʃ ʂ ʐ x ɣ h
Approximate w j
Lateral Approximate l ʟ
Lateral Fricative ɬ
Tap ɾ
The Romanized transcription system for the consonants has two versions for many phonemes—some have diagraphs and a diacritic version of the same sound. For example, /ɬ/ may be written <lh> or <ļ>. Below are the Romanized representations of each phoneme. When two versions exist, the diagraph is listed to the left of the diacritic version. On the far right is the Rílin native orthographic representation. 

/p/ <p> p
/b/ <b> b
/m/ <m> m
/t/ <t> t
/d/ <d> d
/n/ <n> n
/k/ <k> k
/g/ <g> g
/ŋ/ <ng>, <ñ> N
/q/ <q> q
/ʔ/ <’> `
/ɸ/ <ph>, <ŵ> P
/β/ <bh>, <ẃ> B
/f/ <f> f
/v/ <v>v
/w/ <w> w
/l/ <l> l
/ɬ/ <lh>, <ļ> K
/ɾ/ <r> r
/s/ <s> s
/z/ <z> z
/ʃ/ <sh>, <ŝ> S
/ʂ/ <hs>, <r̂> C
/ʐ/ <hz>, <ŕ> Z
/j/ <j> j
/ʟ/ <ll>, <ł> L
/ɣ/ <gh>, <ǵ> G
/x/ <x> x
/h/ <h> h

The two main dialects of Rílin are Lunauli (lunauli) and Sunuli (sunuli). Lunauli does not differentiate between some phones which Sunuli distinguishes as different phonemes.
Lunauli Sunuli
fa ‘breeze’ ɸa ‘breeze’
evaka ‘faction’ eβaka ‘faction’
ewun ‘spy’ eɫun ‘spy’
heʃɛɣɛ ‘corn bread’ heʂɛɣɛ ‘corn bread’
This phonemic merger has caused an inconsistency in the Rílin orthography, however. For example, those words that originally contained a /ɸ/ or /β/ are still written with P or B, despite the fact that they are now pronounced /f/ and /v/. The same applies to words with a former /ʟ/ that is now a /w/. /ewun/ in Lunauli is still spelled as ELun rather that as *Ewun. 

Different words that contrast phonemically in Sunuli often become homonyms in Lunauli.
Lunauli Sunuli
fa ‘breeze’ ɸa ‘breeze’
fa ‘cotton’ fa ‘cotton’

Vowels

Rílin orthography and Romanized orthography for vowels is below:

/i/ <í>I Ex.: Ikara íkara ‘cost’ /’ikaɾa/
/ɪ/ <i>i Ex.: zisanu zisanu ‘ninth’ /zɪ’sanu/
/y/ <y>Y Ex.: vAkY väky ‘aim, direction’ /’væky/
/e/ <é>E Ex.: SutrE ŝutré ‘animalistic’ /’ʃutɾe/
/ɛ/ <e>e Ex.: pemYLe pemyłe ‘charmed’ /pɛ’myʟɛ/
/æ/ <ä>A Ex.: nAtEI nätéí ‘source, origin’ /’nætei/
/ø/ <ö>W Ex.: mWn mön ‘goose pinfeather’ /møn/
/u/ <u>u Ex.: ku ku ‘droop’ /ku/
/ɯ/ <ŭ>M Ex.: dMk dŭk ‘same’ /dɯk/
/ʌ/ <û>V Ex.: SamV ŝamû ‘stand up’ /’ʃamʌ/
/o/ <ó>O Ex.: silO ‘follow’ /’sɪlo/
/ɔ/ <o>o Ex.: qo qo ‘king’ /qɔ/
/a/ <a>a Ex.: Ba ẃa ‘spring wind’ /βa/

Historical Rílin Sound Changes in Vowels

In Proto-Rílin, there were two phonemically distinguishable vowel lengths, long and short. The vowels i, e, a, o, and u had both long and short versions. When the ancestral Rílin language transformed into modern Rílin, the vowels shifted. Short vowels became lax in the case of the high- and mid-front vowels (i, e), unrounded in the case of the high- and mid- back vowels (o, u), and the low back vowel became more front. ø was pushed back and became ʌ (causing a merger of ø and o). ə became ɛ. The long vowels simply became short versions of themselves.

i → ɪ
iː → i

ə → ɛ
e → ɛ
eː → e


o → ʌ
ø → ʌ
oː → o

u → ɯ
uː → u

a → æ
aː → a

There were also three diphthongs in the earlier Rílin which became monothongs in modern Rílin. 

au → ɔ
ju → y
eu → ø

In Lunauli, ɔ became merged with a and is no longer a distinguishable phoneme. 

(N.B. In the word “Lunauli”, and other words with the a + u combination, this sequence is not considered a phonemic diphthong but rather a sequence of two separate phonemes, a and u. This applies to all words in either modern Rílin dialect with a + u, j + u, or e + u combinations.)

Histaxa Dialectal Differences

/æ/ vs /a/

Some speaker populations, such as that around the Histaxa fields region, pronounce /æ/ as [a]. Some in these groups pronounce /æ/ as a more back vowel but not as far back as the standard phonemic /a/, so it is possible that the two phonemes /æ/ and /a/ have not split for either group. In other groups around the same area, /a/ has come to be pronounced [ɑ] or [ɒ] to differentiate between /æ/ which has come to be pronounced as [a], created a chain shift. No other vowels are shifted in these dialects, so it is not a systemic change.

Diphthongs

Diphthongs are not very common in Rílin, but they do occur. Of the following, /ai/ and /oi/ are probably the most common. 

ai 
ei
oi
ɔi
ui
yi
ʌi

All other vowel combinations do not make phonemic diphthongs (e.g. the name Lunauli) (though they may do so phonetically). In Rílin orthography (Sér̂a /’seʂa/), diphthongs are written as a sequence of two individual vowels.

Phonotactics

All clusters, codas, etc. refer to possibilities applied to individual syllables, not words. The syllable’s position in the word is irrelevant, except for /ʔ/ <’>and /ʟ/ <ll, ł> which can only occur in an intervocalic position.

Syllable structure

CV(C), VC(C), CCV(V)(C)

Possible codas: nasals, voiceless sibilants, voiceless stops (except ʔ) (n m ŋ s ʂ ʃ p t k q)

Consonant clusters: Complex onsets that are allowed are as follows. 

Voiceless stop + voiceless sibilant: (ps-, pʃ-, pʂ-, ts-, tʃ-, tʂ-, ks-, kʃ-, kʂ)

examples:
psílhut /’psiɬut/ ‘blood’
pŝír̂û /’pʃɪʂʌ/ ‘profession’
pŝa /pʂa/ ‘throw’

tsäu /’tsæu/ ‘shawl’
tŝäní /’tʃæni/ ‘expression’
Tr̂ûlhû /’tʂʌɬʌ/ ‘election’

Ksata /’ksata/ ‘bind’
Kŝet /kʃɛt/ ‘tight’
Kr̂ŭ /kʂɯ/ ‘warmth exuded from light’

Voiceless stop: (not q or ʔ) + ɾ: (kɾ-, tɾ-, pɾ-)

examples
Kraíña /’kɾaiŋa/ ‘bite’
Trófó /’tɾofo/ ‘build’
Pre /pɾɛ/ ‘attention, care’

p + t; p + k
 
Pta /pta/ ‘should, must’
Pkó /pko/ ‘tail’

Complex codas:

Complex codas can consist of a nasal plus any voiceless sibilant, and any voiceless stop (except q and ʔ) plus any voiceless sibilant. Often these complex codas will not appear on syllables that are not word-final. Most occurrences of these clusters exist in monosyllable, monomorphemic words. 

n, m, ŋ+ ʂ, s, ʃ | p, t, k + ʂ, s, ʃ 

méms /mems/ ‘year, cycle’
mûts /mʌts/ ‘poor, broke’
níns /nins/ ‘back (of)’
täñs /tæŋs/ ‘worry’
xuns /xuns/ ‘wide’


A very small number of archaic words have retained pf-. In most cases, Old Rílin pf- became Modern Rílin ɸ- 

pfó /pfo/ ‘white’ 

pfaí /pfai/ ‘jab’

pfókala /pfo’kala/ ‘pale stripe of skin down the back of all Ríli’ 


All vowels may appear in sequence with each other, but it is uncommon rare to have a sequence of more than two within the same morpheme especially if they are not separated by a semi-vowel. Across morpheme boundaries, it is more common, such as in words like daó-ö ‘to my maternal grandfather’ or sóu-es ‘guiltiness (erg.)’. 

In the concatenation of different morphemes, phonotactics also prevent the sequence of certain phonemes that otherwise may occur in a monomorphemic word. For instance, /-tɾ-/ can occur across or within syllable boundaries in a monomorphemic word, such as /’ʃut.ɾe/ ‘animal’ or /’win.tɾa/ ‘elbow’. However, with the attachment of any bound morphemes, this sequence does not exist. Consider the bound derivational suffix -ɾa, which can act as both a gerund suffix for verbs or as a nominal suffix indicating an abstract quality related to X, with X being the free morpheme to which it is attached. When the free morpheme ends in a voiceless alveolar stop (/t/), the /ɾ/ of the -ɾa suffix is deleted.

/ɛlisɛt/ ‘goddess’ 

/ɛlisɛt - ɾa/ ‘goddesshood’ 

[ɛlisɛt-a] 

This deletion does not occur after other final stops -p, -k, -q, nor does it occur after final -s. After -ʃ and -ʂ, however, there is deletion. 

natʃæk ‘stubborn’ → natʃæk-ɾa ‘stubbornness’ 

geis ‘brown’ → geis-ɾa ‘brownness’ 

salɪs ‘content’ → salɪs-ɾa ‘contentedness’ 

zɛʂ ‘dry’ → zɛʂ-a ‘dryness’

oʃ ‘young’ → oʃ-a ‘youth’

Stress Parameters

Stress placement follows nine rules that determine whether the stress will be onn the first, second from last, or third from last syllable. 

Two-syllable words have stress on the first syllable. Words of more than two syllables follow the eight rules below:
  1. If one of the following vowels is on penultimate (2nd from last) syllable: i, y, e, ø, a, æ, u, o, then the stress is on the penultimate syllable.
  2. If there is a double vowel or diphthong on penultimate, then stress is on the penultimate.
  3. If the onset of last the syllable is nasal, then the penultimate is stressed.
  4. If the onset of the last syllable is a lateral, then the penultimate is stressed.
  5. If the onset of the last syllable is fricative, then the penultimate is stressed.
  6. If the penultimate has coda and last syllable has a consonantal onset, then the penultimate is stressed.
  7. If the onset of last syllable is h or ʔ, then the penultimate is stressed.
  8. If the onset of last syllable is an approximant, then the penultimate is stressed.
If none of the above eight rules apply, the word of more than two syllables will have an antepenultimate stress. That is, the third from last syllable is stressed.

da'lana 'elation' – parameter 1
o'ɾaipa 'relatives' – parameter 2
miɛ'lʌmu 'marzipan' – parameter 3

nɛ'tɛle 'tranquility' – parameter 4
da'zʌxɛ 'specific' – parameter 5
na'xɛnti 'study' – parameter 6
ɛ'laʔas 'treacherous' – parameter 7
βɛ'ɣɛwa 'viscera' – parameter 8
da.'la.na = onset of last syllable is nasal, therefore penultimate stress o.'ɾai.pa = diphthong in penultimate, therefore penultimate stress, etc.

Some Allophonic Rules

Lateral Fricativization
/l/ → [ɬ] / [t] __ A lateral approximate becomes a voiceless fricative after /t/ Ex.: déntsutla ‘amber /den’tsutla/ → [den’tsutɬa]
Palatal Fricativization
/j/ → [ʝ] / V__V A palatal approximate becomes a voiced palatal fricative between two vowels. Ex.: äja ‘song’ /’æja/ → [‘æʝa]
Fronting of glottal fricative
/h/ → [ç] / __ [+high front vocoid] A voiceless glottal fricative becomes a voiceless palatal fricative when before a high front vocoid. Ex.: bíha ‘rotate, change’ /’biha/ → [‘biça]
Final stop aspiration
/+voiceless stop/ → [+ aspirated] / __ # A voiceless stop becomes aspirated at the end of a word. Ex.: zuk ‘weak’ /zuk/ → [zukʰ]
Vowel lowering and centralization
/+ high front V/ → [+ centralized; lowered] q __ A high front vowel becomes centralized and lowered when following /q/. Ex.: bíqi ‘ugly’ /’biqɪ/ → [‘biqï̝]
Glottal fricative voicing
/h/ → [ɦ] / [+voiced] __ [+voiced] A voiceless glottal fricative becomes voiced between two vowels. Ex.: buhí ‘word’ /’buhi/ → [‘buɦi]
Lengthening of tense vowels
/+ tense V/ → [+ long] / __ # A tense vowel becomes long at the end of a word. Ex.: bí’é ‘problem’ /’biʔe/ → [‘biʔeː]
Nasal lateralization
N → [l] / __ /l/ A nasal consonant becomes a lateral approximate when preceding another lateral approximate. Ex.: jóekunla ‘alveolar ridge’ /joɛ’kunla/ → [joɛ’kulla]
Vowel nasalization
V → [+nasalized] / __ N A vowel becomes nasalized when preceding a nasal consonant. Ex.: jentóí ‘easy’ /’jɛntoi/ → [‘jɛ̃ntoi]
Nasal place assimilation
N → [α. + velar β. + alveolar γ. + bilabial] / __ [-cont α. + velar β. + alveolar γ. + bilabial] A nasal consonant assimilates in place of articulation to a following stop. Ex.: ŕysézyñsé ‘sick’ /ʐyse’zyŋse/ → [ʐyse’zynse]
Final vowel centralization
/ʌ/ → [ə] / __ # A mid-low back unrounded vowel becomes centralized to a schwa at the end of a word. Ex.: bíqû ‘bad’ /’biqʌ/ → [‘biqɘ]
Alveolar tap assimilation
/ɾ/ → [+retroflex +cont] / [+retroflex +cont] __ An alveolar tap becomes a retroflex fricative when after another retroflex fricative. Ex.: pir̂raí ‘sp. of small freshwater fish’ /’pɪʂɾai/ → [‘pɪʂʂai]
Velar approximation
/ɣ/ → [ɰ] / /ɯ/ __ A voiced velar fricative becomes a velar approximate when following a high back unrounded vowel. Ex.: leǵŭa ‘ghastly’ /ʽlɛɣɯa/ → [‘lɛɰɯa]
Alveolar stop affricativization
/t/ → /tʃ] / __ /j/ A voiceless alveolar stop becomes a post-alveolar affricate when preceding palatal approximate. Ex.: tjaẃe ‘understandably’ /ʽtjaβɛ/ → [‘tʃaβɛ]
Stop voicing assimilation
/+ voiced C/ → [+voiceless] / [+voiceless C] __ Ex.: bésga ‘digusting person’ /’besga/ → [‘beska]

Minimal Pairs

Vowels

mí /mi/ ‘girl’
mi /mɪ/ ‘jealous’
my /my/ ‘silk’
mé /me/ ‘row’
me /mɛ/ ‘thin’
ma /ma/ ‘mother’
mä /mæ/ ‘snarl’ (v.)
mó /mo/ ‘oh really?’
mo /mɔ/ ‘pan’
mö /mø/ ‘device’
mŭ /mɯ/ ‘indistinct’
mu /mu/ ‘core’
mû /mʌ/ ‘duckling’
maí /mai/ ‘free’
méí /mei/ ‘bright yellow’
móí /moi/ ‘deceit’
moí /mɔi/ ‘weaving’
mûí /mʌi/ ‘drip’ (v.)
myí /myi/ ‘smooth’

Consonant Minimal Pairs

awu /awu/ ‘upright’ 
aǵu /aɣu/ 'blind'
ału /aʟu/ 'machete'

uga /uga/ 'douɡh'
uxa /uxa/ 'axe'
uła /uʟa/ ‘old woman’
uka /uka/ ‘companion’
uqa /uqa/ ‘piglet’

da ‘rock’ 
ɾa ‘cunning’
sa ‘deep’
ʂa ‘supply’
za ‘prostitute’
ʃa ‘petal’
ʐa ‘blizzard’
ta ‘face’
na ‘shame’
la ‘sorrow’
ɬa ‘swing’


pa ‘wrap’
ma ‘mother’
fa ‘cotton’
ɸa ‘breeze’
va ‘corset’
βa ‘energy’
ba ‘father’
wa ‘stretch’


iha ‘plan’
iʔa ‘insect’

imo 'eyelash'
ino 'color'
iŋo 'sugar'

ke ‘hand’
kɛ ‘lettuce’

ko ‘intend’
kɔ ‘tilt the head’

ɾe ‘hot’ 
ɾɛ ‘write’

ʂy ‘sorrow’
ʃy ‘be.NPRES

Orthography

The writing system is called Sér̂a (meaning also ‘be made real, manifest’). It exists in three modes:  Ture ('soft'), which is a cursive script, R̂ek ('hard'), which is a runic system, and the default, which is standard. 

It is a phonemic alphabet with 42 symbols, one for each phoneme in the language.

There is also a standardized Romanization for the Rílin alphabet, as seen below in brackets. In some cases, there are options between using a diacritic to modify a Latin character and using a digraph, such in the case with /β/, which may be written ẃ or bh.

Consonants:

Labial Labio-dental Alveolar Post-alveolar Retroflex Palatal Velar Uv. Glottal
Stop p:p b:b t:t d:d k:k g:g q:q ʔ:`
Nasal m:m n:n ŋ:N
Fricative ɸ:P β:B f:f v:v s:s z:z ʃ:S ʂ:C ʐ:Z x:x ɣ:G h:h
Approximate w:w j:j
Lateral Approximate l:l ʟ:L
Lateral Fricative ɬ:K
Tap ɾ:r

Vowels:

/i/ <í> I
/ɪ/ <i> i
/y/ <y> Y
/e/ <é> E
/ɛ/ <e> e
/æ/ <ä> A
/ø/ <ö> W
/u/ <u> u
/ɯ/ <ŭ> M
/ʌ/ <û> V
/o/ <ó> O
/ɔ/ <o> o
/a/ <a> a

Romanised Rílin Orthography

Some Romanizations have two options (or more), one that is a diagraph and one that is a diacritic.

aghûbhadógha ‘stare into someone’s eyes’ 
aǵûẃadóǵa  /ɑɣʌβɑdoɣɑ/

bíhzara ‘custom’
bíŕara  /biʐɑɾɑ/ 

öhsa ‘soft
öŕa /øʂɑ/ 

tshimlŭ ‘mush, goo’
tŝimlŭ  /tʃɪmlʌ/