Shinasthana: Difference between revisions

Jump to navigation Jump to search
Line 239: Line 239:
| 廣韻   ''kwang'-ghwjens'' || 1001 || 丁重  ''teng-drjong'' || Extant || 206 rhymes || With definitions
| 廣韻   ''kwang'-ghwjens'' || 1001 || 丁重  ''teng-drjong'' || Extant || 206 rhymes || With definitions
|-
|-
| 集韻   ''dzjep-ghwjens'' || 1030 || 孫皓青 ''sun-ghu'-ts'ing'' || Extant || 238 rhymes || With definitions
| 集韻   ''dzjep-gwjens'' || 1030 || 孫皓青 ''sun-ghu'-ts'ing'' || Extant || 238 rhymes || With definitions
|-
|-
| 韻鏡   ''ghjwonh-krjangh''|| 1161 || uncertain || Extant || 3,324 syllables || A table-style summary of all known syllables in the language
| 韻鏡   ''gwjenh-krjangs''|| 1161 || uncertain || Extant || 3,324 syllables || A table-style summary of all known syllables in the language
|-
|-
| 簡韻略  ''kren'-ghjwonh-gljak'' || 1210 || 顧平水 ''gah-brjang-hljei’'' || Extant || 106 rhymes || Abridged ver. of ''dzjop-ghjwohn''
| 簡韻略  ''kren'-ghjwonh-gljak'' || 1210 || 顧平水 ''gah-brjang-hljei’'' || Extant || 106 rhymes || Abridged ver. of ''dzjop-ghjwohn''

Revision as of 20:05, 23 April 2019

Shinasthana
Themiclesian, Sinastani
震旦語系
Pronunciationtjən˥˩.tɑn˥˩.ŋjɑ˩˥.ɣis˥˩
Native toThemiclesia
EthnicityThemiclesian
Native speakers
51,240,000 (2010)
Menghic
  • Shinasthana
Early form
Old Menggok language
Standard forms
T'ong-ngja'
DialectsSee "dialects" below
Mengja
Official status
Official language in
Themiclesia
Recognised minority
language in
Language codes
ISO 639-1sh
ISO 639-2tem
ISO 639-3

Shinasthana (震旦語系, tjens-tans-ngja'-ghis) is a group of languages most widely spoken in Themiclesia, with 22 officially-recognized dialect groups. It is a member of the Menghic language family, originally spoken in Menghe and introduced to Themiclesia by Meng merchants and settlers. The variety of Shinasthana that is the official language in Themiclesia is towng-ngja', or the "Common Speech". Language policy in Themiclesia had curtailed the dialectal variety present in the country, with several now critically endangered; modern efforts to preserve them are in effect. Shinasthana has around 41 million native speakers, domestically and abroad, as well as several million second language speakers. The language is widely taught and researched in foreign educational institutions.

Etymology

The word Shinasthana is from an early transliteration of Snjonh (name of a Themiclesian dynasty) into a Maverican Sanskrit, suffixing -sthana, which means "place"; this term was later further transliterated by Casaterran scholars into Shinasthana, Sinastana, Xinastana, Zenaktana, Shenaclana, and Themiclesia. Though technically the same word, Shinasthana is now exclusively used to refer to the language, and Themiclesia the country.

History

The phylogeny of Shinasthana was considered settled in the 19th century, as a branch of the Menghic languages, which is distributed in Menghe, Themiclesia, and communities outside them. However, in 1935, Themiclesian scholar S. N. Kip discovered that the 22 dialects of Shinasthana do not form a monophyletic clade of languages; that is, there is no single language spoken in Menghe that was the ancestor to all Shinasthana language and only them. Rather, the language shows at least four strata of Menghic influence, which may be connected to waves of Menghean immigration to Themiclesia.

Old Meng stratum

The oldest layer corresponds with Menghic during Themiclesia's pre-dynastic period and reflects phonological features of Old Menghean; Kip's student, Mang, places this layer around the 8–7th c. BCE, or the founding of Themiclesia as a trading post. This layer is not represented as any single dialect in Themiclesia, but traces of its structure can be evinced in some of the more archaic ones as irregular readings. Systematic reconstruction of this layer is difficult due to lack of material. The main characteristics of this stratum are:

  • Complex initials and prefixes in *s-, *r-, *l-, *n-, *h-, and *m-;
    • Obstruent prefixes, such as *p-, *t-, and *k- have been proposed but yet to receive broad acceptance;
    • Certain lexical items seems to contain dual or trial prefixes, such as *s-n-glja;
  • Complex codas and suffixes in *-n, *-k, *-ɦ, and *-s;
  • Connections with phono-radicals are more clearly preserved, e.g. 魚 nglja and 魯 gla (the former is the top half of the latter, where it is phonetic).

Meng Dynasty stratum

The Meng Dynasty stratum corresponds with the lingua franca of the Meng court's scholarly elite in the 2nd and 3rd c. CE. This stratum is reflected predominantly in the modern southern dialects. The main characteristics of this layer are:

  1. Prefixes
    • *n- is lost before obtruents but leave the obtruent voiced;
      • but preserved before continuants;
      • *n-l > d
    • *m- causes obstruents after it to drop;
    • *h- and *s- are retained;
    • *r- metathesizes with following consonants to produce some -rj- clusters in Division III;
      • *r-s > ts
    • *l- metathesizes with following consonants;
      • but *l-s > ts
  2. Nucleus
    • Voiced consonants (b-, d-, and g-) are lost before -r- and -l-, unless prefixed;
  3. Suffixes
    • Suffix -ɦ is lost and gives rise to the rising tone;
      • except before -s, resulting in -hs;
    • Suffix -s > -h after vowels;
      • *-ts > -is; *-ks > -s; *-kws > -ws
      • *-ps, *-rs, *-ls, *-ns, *-ms, *-ngs, and *-ngws are retained.

Middle Meng stratum

The Middle Meng stratum is thought to represent the language of the Chǒllo aristocracy and the migrants to Themiclesia of the 6th century. This stratum sees general phonetic shifts in the syllabic core, unlike in previous strata, where only the onset and coda seem to participate in them. The dialects spoken in northern Themiclesia represent this layer best, and, after such changes shared with most of the Menghean dialects, including what gave rise to Gwanhwa, thus seem much more familiar to speakers of the Menghean language today.

  1. Initials
    • Velars combine with -lj- and merge with palatals.
      • but in Division I and after back vowels in Division III, -l- is lost after unvoiced velars, while voiced velars are lost before -l.
    • h-n- and h-l- > th
    • h-m- and h-ng- > h
    • Dental stops and affricates combine with -r- and -rj- to form retroflexed initials, and stops with -j- to form palatal stops.
      • Prefix s- combines with labials, velars, and alveolars in Division III-a to becomes voiceless or voiced fricatives; the distinction of aspiration is lost.
      • but when s- occurs before larygneals, Division I words are re-analyzed as Division III.
      • when s- occurs before -n, -m, -ng, -ngw, and -l, these consonants are lost.
    • l > j
    • g(w)- > gh(w)-
      • except before -j-, but not before -rj-
  2. Glides
    • Medial -r- centralizes the following vowel and is vocalized as [ɤ] after non-labialized initials and [o] after labialized initials
    • Medial -rj- is pronounced as [ɨ] after non-labialized initials and [ʉ] after labialized initials
  3. Nucleus
    • u merges with e, when the latter follows a labialized consonant
      • Except after velars
    • -in merges with -ian; -en merges with -ian following a dental initial
    • -jan and -jian merge
    • The contrast between -rj- and -j- is lost before back vowels.
  4. Codas and suffixes
    • All remaining -s and -h are lost, generating the departing tone.
    • -kw > -k
    • -r, -l > -n

Modern loans

Modern loans are a blanket term for Menghean influence after the 6th century. As Themiclesia's language has been well-established after then, loans rarely add more than an alternate readings to existing words. In the earlier period, between the 6th and 19th centuries, loans generally agree with the Menghean Gwanhwa; after Gwanhwa was abolished as lingua franca in Menghe, later loans agree with one of two sources, one being the Chǒllo dialects and the other the Menghean Botong-ǒ. A prominent example is the reading meng as in "Menghe"; the standard reading in Themiclesia would be mrangs. The former reading is used exclusively as a abbreviation for "Menghe", while the later is the name of the Meng Dynasty, with which Themiclesia identified for much of history.

Writing Systems

Traditional Characters

Inherited from the Menggok language, each character is monosyllabic and morphemic. This writing system is common to all Shinasthana languages and is the standard script. Under traditional reckoning, there are six methods that govern the formation of characters:

  1. Pictogrammatic (象形), where the character depicts the physical appearance of an object; however, the actual meaning of the character needs not be constrained by the object depicted
  2. Pictophonetic (形聲), where a pictogrammatic character used determinatively is combined with another, unrelated grapheme that represents the character's pronunciation
  3. Ideogramamatic (指事), similar to pictogrammatic characters, but the depiction (or at least part thereof) does not correspond to a physical attribute or object
  4. Ideosynthetic (會意), where multiple pictogrammata are combined and meaning deduced from the relationship between them
  5. Transideographic (轉注), where the definition of an existing character is given to a new character
  6. Substitution (假借), where a completely unrelated character acquires the definition of a word that otherwise cannot be formed according to the principles above

Sylvanization

Various schemes, based on the Sylvan Alphabet, have been used by Casaterrans to notate Shinasthana since the 14th century, when the first Casaterran merchants landed in Themiclesia and began to study its language. Few Casaterran languages have the same phonemic or phonetic inventory as Shinasthana, resulting in the use of several alphabets to represent a single phoneme or the agglomeration of several phonemes into one Casaterran alphabet; the former leads to cumbersome writing and reading, and the latter, ambiguity. Between the 14th and 19th centuries, Sylvan, Sieuxerrian, Rajian, Tyrannian, and Columbian scholars each Sylvanized Shinasthana according to idiosyncratic rules. That they often draw on different dialects or even mix them further reduces intelligibility. As printing Shinasthana characters was nigh-impossible in Casaterra, academic work was hampered by this confusion.

In 1807, the first Sylvanization reference with government backing appeared, devised by Columbian scholar, George Figley, who worked for over three decades on the Shinasthana. Keenly aware of the difficulties that Casaterran scholars encountered in writing about Themiclesia, not only about its tongues, but also whenever proper terms are mentioned, they presented to and received sanction for their proposal from the Magistrate of Tonning. As he lived in Tonning, the Sylvanization plan was based on the dialect there, albeit aware of other dialects. While it found a somewhat sympathetic audience in Casaterra, other scholars, based elsewhere in Themiclesia, found their choice of dialect to generalize for Shinasthana, as a whole, unacceptable. Figley continued to defend their system and improve upon it, publishing a final revision in 1829, then presented it to the Imperial Court. Ratification was granted, but its use was not made mandatory. This standard eventually gained widespread acceptance and today is known as the Figley Transcription System, or F for short.

Revision

As phonemic theory gained prominence late in the 19th Century and crested in the early 20th, there was significant impetus to rework Figley according to phonemic lines (it was originally phonetic) and suggestions from the government to adopt a more generalist system that worked for all dialects in some way. In theory, by applying phonemic analysis to the most archaic dialect of Shinasthana then known, one can describe a set of syllables that encompass all syllable varieties in any dialect, since it was assumed that the oldest stage of the language must be the most complex, and subsequently phonemes could not be added; this was later convincingly disproven. Nevertheless, the revision was announced in 1930, known as Revised Figley or R. Though widely acclaimed by lingiusts as a successful and accurate practice of phonemic theory, phoneticians and lay persons criticized it for being distant from the actual, spoken language. The government maintained that the revision better represented all the major dialectal groups unambiguously. To demark this scheme from the previous, a drop capital R was printed ahead of the sylvanized phrase.

Phonology

Consonants

Shinasthana dialects generally distinguish between four places (labial, alveolar, velar, and larygneal) and three manners (lenis, asper, and voiced) of articulation. Labialization is present in velars and larygneals. This structure is inherited from the shared Old Menghean ancestral language. Not every variation within this gamut is fully preserved in each dialect, but vestiges of such a structure are more or less visible in each.

Initials Bilabial Alveolar Velar Velar
labialized
Glottal Glottal
labialized
Stop Lenis p t k ʔ ʔʷ
Asper kʰʷ
Voiced b d g
Nasal m n ŋ ŋʷ
Affricate Lenis t͡s
Asper t͡sʰ
Voiced d͡z
Fricative Unvoiced s h
Voiced ɦ ɦʷ
Trill r
Approximant w l j

Vowels

Front Central Back
Close i u
Mid e ə o
Open a
Diphthongs aɪ   eɪ   iɪ   oɪ   əɪ   uɪ
eu   au   iu

Syllable structure

The syllable structure Shinasthana varies considerably due to its paraphyletic composition. While the Old Menghean language is agreed to have a syllabic structure at least similar to CCCCCVCCC, most dialects have undergone some sort of secondary simplification. For example, the Common dialect permits five onsets, with much more restrictive phonotactical options, and tolerates only two codas, of which the second must be -s. The Nom-t'ur dialect preserves four onsets but permits two or more obstruents to co-occur, while similar latitude is found in the coda sequences, in which -kʔs (realized as [ks], as opposed to [kʰs]) is a maximal example. While such long onset sequences are typologically uncommon, it is noted that, with rare exceptions, there could only be one obstruent in each onset sequence; linguists working with Shinasthana and, in broader terms, the Menghic family, term this obstruent the onset nucleus, and elements aside from it often correspond to morphological affixes. The Menghic syllable also permits a number of glides between the onset and syllable nucleus. Nom-thur and the Common dialect both permit two glides, the first of which has a retroflective value (typically /r/ and /l/, but not simultaneously), and the second has a palatalizing effect. Though most dialects descended from the Old Menghean language have phonetic tonal contrasts, all can be attributed to lost but attested suffixes; most scholarly work on Old Menghean phonology therefore do not recognize phonemic tonal contrasts. Synchronic studies may consider such suffixes and tonal contrasts allophones of each other, in an unusual way.

Those dialects that descend directly from the Old Menghean language generally allow more complex syllable structure, as exemplified by the two above. In contrast, those that are descended from Middle Menghean are relatively simple. Their syllabic structure can be observed in the Menghean Gwanhwa and dialects spoken in the Themiclesian North. Scholars have debated what the maximal syllabic structure of Middle Menghean was, without much consensus; the most-commonly cited form is CCCVCT. The first C, representing the onset nucleus, can either be a consonant inherited from Old Menghean or an affricate or fricative resulting from a prefixed obstruent. The second C represents the pre-vocalic glide, which could have one of three contrastive values (/ɯ/ /ɨ/ /i/); the third C represents a rounding element /w/. This rounding element is thought to result from Old Menghean labialized consonants and diphthongized, rounded vowels.

Comparison

Menghean

Within the Menghic family, Shinasthana is more closely related to the Menggok-descended group of dialects. Having diverged around 2,700 years ago at the oldest stratum, differences between modern Menggok dialects and Shinasthana are considerable. An obvious example is the absence of phonemic tones in Standard Menghean, though they remain in conservative dialects of the south and west of Menghe. These dialects are more closely related to Gwanhwa, the language of the Menghean court; however, scholars do not classify them as descendants of Gwanhwa, since Gwanhwa itself lost tones between the 9th and 11th centuries. Not all these dialects retain the three tones presumed to have existed in the Old Menggok period. Moreover, under the influence of modern Standard Menghean, some of the younger generation have dispensed with tonal distinction when speaking a dialect that canonically retains it.

The collapse of a three-way distinction between between voiceless, aspirated, and voiced consonants has occurred in most of Menghe. In Themiclesia, a similar but less pervasive change occasioned the phonemization of relative pitch in the tonal system (previously only distinguishing contour), while in Menghe it seems to have simply devoiced formerly voiced consonants, merging them with their unvoiced and unaspirated counterparts.

Dialectal variation

Common dialect

Comparative phonology

Mengja Menghean Sinmun Menghean Sinmun
Sylvanized
Middle Meng Stratum Old Meng Stratum
國之語音
異乎中國
與文字不相流通
故愚民有所欲言
而終不得伸其情者多矣
予爲此憫然
新制二十八字
欲使人人易習便於日用耳
국지어음
이호중국
여문자불상유통
고우민유소욕언
이종불득신기정자다의
여위차민연
신제이십팔자
욕사인인이습편어일용이
guk ji eo eum
i ho jung guk
yeo mun ja bul sang yu tong
go u min yu so yo geon
i jong bul deuk sin gi jeong ja da ui
yeo wi cha min yeon
sin je i sip pal ja
yok sa in in yeo seup pyeon eo il lyong i
kwek chi ngjo' iim
jik gho tiung kwek
ngjo' mjun dzji' pjot sjang lju t'ung
koh ngjo miin ghwju' sho' yoh ngjan
nje tiung pjot tek zjin gji dzjing ta ghwji'
yo wji ts'ji' miin njen
sjin chiai njih zip peat dzji'
yok sri' njin njin yik zjek pianh yo njit yungh ni
kwek tje ngja' 'rjem
lis ga trjung kwek
la' mjen dzjeh pje smljang rju hlung
kah ngrjo mrjing gwje' sk'rja' gljuk ngjan
nje trjung pje tek stjin gje dzjing tjaw' tai ghlje'
gla gwrjarh sdi' mrjen' njen
sjing tjiais njih gljop priat dzje'
gluk srjo' njin njin lias gljap prjans gwja njit lungs nje'

Scholarship

Dedicated references

The canonical anthology of ancient phonology consists of the following twelve works.

Title Year Compiler(s) Status Categorization Notes
聲類   hling-rjed 2nd c. 李登  ’rje’-teng Fragmentary 322 rhymes extant Compiled in Menghe
韻集   ghjwonh-dzjep 275 呂靜  glja'-dzjingh Extant 539 rhymes
釋韻   st'jak-ghjwonh 334 李軌  ’rje’-krju’、徐邈 sdja-mrawk Extant 421 rhymes
四聲   sjed-hleng 472 沈約  st'jem-'jawk Extant 252 rhymes First work to describe tones concretely
玉篇   ngjowk-p'rjan 531 梁惠帝 Emperor ghwid of Rjang Fragmentary 260 rhymes Reflects speech in Kien-k'ang
經典音義 king-ten'-'rjem-ngjars 560 陸明  rjuk-mrjang Extant 212 rhymes With definitions
切韻   ts'it-ghwjens 601 陸法延 rjuk-pjap-lan Mostly extant 193 rhymes Reflects speech in Glak-lang
廣音錄  kwang-'rjem-rjuk 732 張常拱 trjang-djang-kjung’ Mostly extant 2,742 dialectal peculiarities
廣韻   kwang'-ghwjens 1001 丁重  teng-drjong Extant 206 rhymes With definitions
集韻   dzjep-gwjens 1030 孫皓青 sun-ghu'-ts'ing Extant 238 rhymes With definitions
韻鏡   gwjenh-krjangs 1161 uncertain Extant 3,324 syllables A table-style summary of all known syllables in the language
簡韻略  kren'-ghjwonh-gljak 1210 顧平水 gah-brjang-hljei’ Extant 106 rhymes Abridged ver. of dzjop-ghjwohn

Epexegesis references

To support the study of classical literature that often possessed much philosophical and political esteem in imperial-period Themiclesia, many authors have written epexegeses (a work that explains another) that offer analyses of the canonical texts, including definitions of terms, notable contrasts with other works and other epexegeses, and correct pronunciation of words. The corpus of such explanatory literature has since accumulated, over many centuries, to be dozens, if not hundreds, of times longer than canon. Authors provide their opinion on how individual characters should be read, often reflecting the speech of their own environment and age.

Poetry analysis

In 972, Sungh Dynasty scholar Mak Mjon-ghwed developed a method of the phonological relationship between rhyming syllables in verse. This is known as the Rhyme Identification Method (歸韻法). It stipulates:

  1. If a poem has lines XXXA / XXXB / XXXC / XXXD, and if each line is known to rhyme with each other, then syllables A, B, C, and D must rhyme with each other.
  2. If a second poem has lines XXXA / XXXE / XXXF / XXXG, then syllables E, F, and G must rhyme with each other and B, C, and D, since they are "connected" by syllable A.

These two principles proved incredibly potent in reconstructing the phonological structure of ancient languages based on a logical and consistent analyses. Though Mak himself did not pursue this subject much further, his students proceeded to apply this method to hundreds of poems dating to the archaic and classical eras of Menghean history. In the 1100s, scholars concluded that ancient Menghean poetry had nine "rhyme departments"; for many generations afterwards, this conclusion was refined by exhaustive studies to remove connections made by incorrect rhyme identification (i.e. connecting two words that do not actually rhyme) and glyph conflation (i.e. assuming each glyph corresponds to only one syllable). This heralded the birth of Ancient Phonology (古音學) in Themiclesia.

In 1594, it was known that ancient poetry rhymed according to 33 departments; upon this conclusion, scholar Drjon Teh (陳第) asserted that the pronunciation of some words have changed since ancient times, but a number have not. He (using methods not rigorous) identified some such syllables as "anchor rhymes" and consulted well-established analyses to assert that words, in his time not rhyming, must have been pronounced in a way that did rhyme with his so-called "anchor rhymes". The majority of modern scholars believe that Drjon's theories, though factually flawed, pioneered the assignment of phonetic values to abstract phonological structures and the principle that sound changes occurred under specific phonological conditions (i.e. not randomly). His successors focused on the criteria under which such changes occurred, though until the modern period and the arrival of Casaterran phonology in 1757, their results have been limited. Of the types of criteria they could assert as a conditioning factor for a phonetic change, place of articulation (i.e. glottal, velar, alveolar, and bilabial), manner of articulation (i.e. plosive, nasal, approximant), voicedness, aspiration, tone, and vowel quality have all been used to explain the causes of sound change. This in turn allowed more recent scholars to describe accurately the phonological and phonetic characteristics of languages lost to time, e.g.

  • What has been analyzed as two separate groups of consonants under Modern Phonology (今音學), the /t/ and /tr/ groups, were in complimentary distribution and should be the same group in Ancient Phonology, and the /tr/ group diverged under the criteria of second-division and third-division-type-B factors.
  • The /pf/ group in some modern dialects diverged from /pj/ in others, and /pj/ is asserted to be the "ancient pronunciation".

Neologisms

Term Themi. reading Themi. meaning Alternate reading Alternate meaning
社會 sghljas-gweps society, social Sah-hwoi socialist
皇帝 gwang-tih Emperor of Themiclesia hwang-tsrjiah emperor (generally)
mrangs first; ethnonym of Themiclesia meng Menghe (abbreviation)
東京 Tongw-krjang Glak-lang Tung-kjung Donggyŏng
dzjian a unit of mass; 1/100 of one mjon tsuen Jŏn, 1/100 of a Menghean Won

In popular media

Being a major branch of the Menghic family of languages, Shinasthana has often been compared with Menghean by scholars, students, and laymen alike. With surprising uniformity, most students consider Menghean both easier to learn and more pleasant to the ear; prominent Internet personages have described some forms of Shinasthana as "creeping", "tinny", and "slimey". A rather large portion words contain /j/ or /i/, which tend to front the following vowel, leading to a sensation of constriction. The presence of /r/ and /l/ in combination with /j/ also can occlude a speaker's voice, making vowels less resonant. In response, several linguists believe that these features are necessary to produce as many distinct syllables as possible, highly motivated in a language in which most words are monosyllabic.

In the early to mid-20th century, linguists attribute this difference to sound changes in Yang-influenced dialects that occurred in the Middle Menghean period. In one famous (and extreme) case, Dr. Mran Tsod, research fellow at the Academia Themiensis, produced a long list different syllables [1] corresponding with only one syllable in Botong-ǒ Menghean; since these syllables may themselves represent more than one word, Mran questioned the aptitude of the Sinmun writing system devised in the 13th century. Other researchers rejected Mran's "emotionally-motivated" assessment, showing that the Yang languages were responsible for deleting via assimilation some distinctions that were present in Old Menggok. The dialects spoken in the Haedong region, under Yang influence, thus did not maintain the syllable structure that characterized the Old Menggok language, from which Shinasthana and most Menggok-derived forms of Menghean have descended.

These scholars also assert that the contribution of Old Menggok language has been considerably over-estimated in the formation of Modern Standard Menghean, particularly on a phonetic and structural level. They state further that the straightforward syllable structure of Yang-based dialects (and by extention the historical Yang proto-language) was not a "shortcoming" as Mran has claimed, since native Yang vocabulary was predominantly polysyllabic, and borrowings from Old Menggok represented a small and specialized subset of words used by Yang-dialect speakers. With fewer words to distinguish, there was no motivation to retain some contrasts that originally existed in Old Menggok and still now exist in Shinasthana.

Notes

  1. k'jot, k'joh, k'jo, k'rje', k'rjeh, grje, grje', kljeh, hjo', 'jar', 'jarh, kjoi, gjo, krjod, gjoi, kjar', gljed, k'rjod, grjwed, gjar, kjar, grjeh, kja, kjoi', gjod, k'jar, gjoh, k'rjo, k'je', klje', k'jod, ngjod, kjod, gje, k'joi', totalling 36.

See also