|Native to||South Korea|
|Ethnicity||Koreans of Jeju Island|
Jeju (Korean: Jeju-eo, Korean and Jeju: Jeju-mal), often called Jejueo in English-language scholarship, is a Koreanic language traditionally spoken in Jeju Island, South Korea. While often classified as a divergent Jeju dialect (Korean: ? Jeju bang'eon) of the Korean language, the variety is referred to as a language in local government and increasingly in both South Korean and foreign academia.
The consonants of Jeju are similar to those of Seoul Korean, but Jeju has a larger and more conservative vowel inventory. Jeju is a head-final, agglutinative, suffixing language like Korean. Nouns are followed by particles that may function as case markers. Verbs inflect for tense, aspect, mood, evidentiality, relative social status, formality, and other grammatical information. Korean and Jeju differ significantly in their verbal paradigms. For instance, the continuative aspect marker of Jeju and the mood or aspect distinction of many Jeju connective suffixes are absent in Korean. While most of the Jeju lexicon is Koreanic, the language preserves many Middle Korean words now lost in Standard Korean. Jeju is not mutually intelligible with the mainland dialects of South Korea.
Jeju was already divergent from Seoul Korean by the fifteenth century, and was unintelligible to mainland Korean visitors by the sixteenth century. The language was severely undermined by the Jeju uprising of 1948, the Korean War, and the modernization of South Korea. All fluent speakers remaining in Jeju Island are now over seventy years old. Most people in Jeju Island now speak a variety of Korean with a Jeju substratum. The language may be somewhat more vigorous in a diaspora community in Osaka, Japan, but even there, younger members of the community speak Japanese. Since 2010, UNESCO has designated the language as critically endangered, the highest level of language endangerment possible. Revitalization efforts are ongoing.
Jeju is closely related to Korean. It was traditionally considered an unusually divergent dialect of Korean, and is still referred to as such by the National Institute of the Korean Language and the South Korean Ministry of Education. While the term "Jeju language" (Korean: , Jeju-eo) was first used in 1947, it was not until the mid-1990s that the term gained currency in South Korean academia. While "Jeju dialect" was still the preferred usage throughout the decade of the 2000s, the majority of South Korean academic publications had switched to the term "Jeju language" by the early 2010s. Since somewhat earlier, "Jeju language" has also been the term preferred in local law, such as the 2007 Language Act for the Preservation and Promotion of the Jeju Language (Korean: ? Jeju-eo bojeon mit yukseong jorye), and by non-governmental organizations working to preserve the language. The only English-language monograph on Jeju, published in 2019, consistently refers to it as a language as well.
Jeju is not mutually intelligible with even the southernmost dialects of mainland Modern Korean. In a 2014 test for intelligibility, Korean speakers from three different dialect zones (Seoul, Busan, and Yeosu) were exposed to one minute of spoken Jeju, with a control group of native Jeju speakers. On average, Korean native speakers from all three dialect zones answered less than 10% of the basic comprehension questions correctly, while native Jeju speakers answered over 89% of the questions correctly. These results are comparable to the results of an intelligibility test of Norwegian for native Dutch speakers. Diaspora Jeju speakers living in Japan also report that they find it difficult to understand South Korean news media, and resort to Japanese subtitles when watching South Korean TV shows.
Jeju was traditionally spoken throughout Jeju Province except in the Chuja Islands, halfway between Jeju Island and mainland Korea, where a variety of Southwestern Korean is found. The language is also used by some of the first- and second-generation[a] members of the Zainichi Korean community in Ikuno-ku, Osaka, Japan.
Compared to mainland Korean dialect groups, there is little internal variation within Jeju. A distinction between a northern and southern dialect with a geographic divide at Hallasan is sometimes posited, but an eastern-western dialectal divide cutting through Jeju City and Seogwipo may better explain the few dialectal differences that do exist. A 2010 survey of regional variation in 305 word sets suggests that the north-south divide and the east-west divide coexist, resulting in four distinct dialect groups.
An east-west divide is salient in the Jeju words for "lizard".
Eastern Seogwipo uses ? jangkullebi while western Seogwipo uses ? dokdaguri.
The Koreanic languages are likely not native to Jeju Island; it has been proposed that the family has its roots in Manchuria, a historical region in northeastern Asia. It is thought that Koreanic speakers migrated from southern Manchuria between the third and eighth centuries CE. Linguist Alexander Vovin suggests that the ancient kingdom of Tamna, which ruled the island until the twelfth century, may have spoken a Japonic language that left a substrate influence on Jeju. When exactly this putative Japonic language may have been replaced by the Koreanic ancestor of Jeju remains unclear.
Unlike mainland Korea, which was ruled only indirectly by the Mongols, Jeju was placed under direct Yuan administration in the late thirteenth century. Significant numbers of Mongol soldiers migrated to the island during this period, and their language acted as a superstratum that may have accelerated local language change. Linguist Yang Changyong speculates that the formation of Jeju as a language independent of Korean was influenced by Mongol. By the fifteenth century, when the invention of Hangul permits a detailed understanding of Korean phonology for the first time, Seoul Korean and Jeju were already divergent; the Seoul prestige dialect of fifteenth-century Middle Korean disallowed the diphthong /j?/, but Jeju does not.
Sixteenth- and seventeenth-century references to the language of Jeju by mainland Korean literati state that it was already unintelligible to mainland Koreans.Kim Sang-heon (1570-1652), who served as pacification commissioner (Korean: ? anmueosa) on the island from 1601 to 1602, gives six words in the "provincial language" with clear cognates in modern Jeju and also writes:
? ? ? ? ?...
"The exiled man Shin Jangnyeong was originally a government interpreter. He said, 'The language of this island is most like Chinese, and the sounds they make while driving cattle and horses are yet more impossible to tell apart. Is this because the climate is not far from that of China, or because the Yuan dynasty once ruled and appointed officials here and the Chinese mingled with them?'... What is called the provincial language is but high and thin and cannot be understood."
In 1629, the Korean government banned the emigration of Jeju Islanders to the mainland, further restricting linguistic contacts between Jeju and Korean. At the same time, the island was also used throughout the Joseon era (1392-1910) as a place of exile for disgraced scholar-officials. These highly educated speakers of Seoul Korean often tutored the children of their Jeju neighbors during their exile and established a continuous and significant Seoul Korean superstratum in Jeju.
Jeju remained the dominant language of both private and public spheres under Japanese colonial rule (1910-1945), although many Japanese loanwords entered the lexicon, and many speakers were monolingual. Large-scale migration of Jeju people to Japan began in 1911, and 38,000 Jeju Islanders lived in Osaka alone by 1934. Immigration to Japan continued even after Korean independence into the 1980s. Jeju is still spoken by older members of these diaspora communities, although younger individuals speak Japanese as their native language and are not fluent in Jeju.
Severe disruption to the Jeju language community began after the end of Japanese rule in 1945. Popular opposition to the division of Korea and police brutality led to a rebellion against the American military government on April 3, 1948. The Syngman Rhee regime, which succeeded the American administration in August 1948, suppressed the rebellion with mass killings of civilians. As many as sixty thousand Jeju Islanders, or a full fifth of the pre-rebellion population, were killed. Forty thousand more fled to Japan. Out of the four hundred villages of the island, only 170 remained. The devastating impact of the massacres on the Jeju language community was exacerbated by the outbreak of the Korean War in 1950. While Jeju was never occupied by the North Korean army, nearly 150,000 Korean-speaking refugees from the mainland fleeing North Korean invasion arrived in Jeju in the first year of the war. These events shattered the Jeju language's former dominance on the island, and Standard Korean was beginning to displace Jeju in the public sphere by the 1950s.
The decline of Jeju continued into the 1960s and 1970s. The Saemaeul Movement, an ambitious rural modernization program launched by Park Chung-hee, disrupted the traditional village community where Jeju had thrived. The language came to be perceived as an incorrect dialect of Korean, so that students were subject to corporal punishment if they used it in school, and the use of Standard Korean even in the private sphere began to spread from Jeju City outwards. The language attitude of native Jeju speakers in this period was self-disparaging, and even Jeju people regarded the use of Jeju "with contempt." A 1981 survey of language attitudes among high school and university students natively speaking Seoul Korean, Chungcheong Korean, Southwestern Korean, Southeastern Korean, and Jeju showed that Jeju speakers were the most likely among the five groups to ascribe negative traits to their native variety.
A 1992 study of code-switching by native Jeju speakers shows that Jeju was by then in an unfavorable diglossic relationship with Korean, and was largely restricted to informal contexts even between Jeju natives. Within a primarily Jeju-language conversation, speakers might spontaneously switch to Korean to emphasize the rationality or truth value of their statement, while switching to Jeju in a primarily Korean conversation signified that the speaker was making a subjective statement or being less serious.
|Participants||Formality||Intimacy||Social status||Preferred variety|
|Includes mainlander||Standard Korean|
|Only Jeju natives||Formal|
|Informal||Participants are not emotionally intimate||Speaker is socially inferior to addresee|
|Speaker is socially superior to addressee||Jeju / Jeju Korean|
|Participants are emotionally intimate|
The same study notes that by 1992, even this variety restricted to the informal domain was usually a Korean dialect with a Jeju substratum, rather than the traditional Jeju language:
? ? , ? , . ? ' () '?, ? ? ' ' '(?) ?'? ? ?, ? ? ? ? ? .
"As for the Jeju language [lit. 'Jeju speech'] in general use nowadays [as of 1992], the situation is that its differences from Standard Korean are greatly diminishing compared to the past. Its greatest differences with Standard Korean [now] lie especially in the suffix paradigm, and in other areas the differences are being minimized. The Jeju people accordingly understand that Jeju and Standard Korean are in a form of dialect continuum, and refer to the native language formerly in use as "thick (or intense) Jeju language" and the Jeju language currently in use as "light Jeju language" or "mixed (with Korean) language."
Jeju is nearly extinct. As of 2018, fluent speakers in Jeju Island were all over seventy years of age, while passive competence was found in some people in their forties and fifties. Younger Islanders speak Korean with Jeju substrate influence found in residual elements of the Jeju verbal paradigm and in select vocabulary such as kinship terms. The language is more vigorous in Osaka, where there may be fluent speakers born as late as the 1960s. Since 2010,UNESCO has classified Jeju as a critically endangered language, defined as one whose "youngest speakers are grandparents and older... [who] speak the language partially and infrequently."
A 2008 survey of adult residents' knowledge of ninety Jeju cultural words showed that only twenty-one were understood by the majority of those surveyed. Lack of heritage knowledge of Jeju is even more severe among younger people. Four hundred Jeju teenagers were surveyed for their knowledge of 120 basic Jeju vocabulary items in 2010, but only nineteen words were recognized by the majority while forty-five words were understood by less than 10%. A 2018 study suggests that even the verbal paradigm, among the more resilient parts of the substratum, may be in danger; the average middle schooler was more competent in the verb system of English, a language "taught only a few hours a week in school and in private tutoring institutions," than of Jeju.
Revitalization efforts have recently been ongoing. On September 27, 2007, the Jeju provincial government promulgated the Language Act for the Preservation and Promotion of the Jeju Language, which established five-year plans for state-backed language preservation. However, it was not until UNESCO's 2010 designation of Jeju as critically endangered that the provincial government became proactive in Jeju preservation efforts. In 2016, the provincial government allotted ?685,000,000 (US$ 565,592 in 2016) to revitalization programs, and the government-funded Jeju Research Institute has compiled phrasebooks of the language. The provincial Ministry of Education has also published Jeju textbooks for elementary and secondary schools, although some textbooks really teach Standard Korean interspersed with Jeju lexical items. Some public schools offer after-school programs for Jeju, but the short duration of these classes may be insufficient to promote more than "symbolic" use by students. The linguistic competence of many teachers has also been challenged.
Other preservation and revitalization efforts are led by non-state bodies. The Jeju Language Preservation Society (Korean: Jeju-eo bojon-hoe), founded in December 2008, publishes Deongdeureong-makke (), a bimonthly Jeju-language magazine, and holds Jeju teaching programs and speaking contests. Literature in the language has recently been published, including children's books and a 2014 poetry anthology. Local bands and theater troupes have made Jeju-language performances. Regional newspapers such as the Jemin Ilbo and the Halla Ilbo include Jeju-language sections, and local branches of KBS and MBC have launched radio programs and a television series in Jeju. Recent South Korean media with nationwide appeal, including the 2010 television series Life is Beautiful and The Great Merchant, the 2012 drama film Jiseul, and the 2015 television series Warm and Cozy, have also featured spoken Jeju.
Recent surveys show changes favorable towards Jeju in prevailing language attitudes. In a National Institute of the Korean Language survey in 2005, only 9.4% of Jeju Islanders were very proud of the regional variety. When the same survey was reheld in 2015, 36.8% were very proud of the language, and Jeju Islanders had become the most likely among South Korean dialect groups to have "very positive" opinions of the regional variety. In a 2017 study of 240 Jeju Islanders, 82.8% of those sampled considered Jeju to be "nice to listen to," and 74.9% hoped that their children would learn the language. But significant generational cleavages in language attitudes were also found. For instance, only 13.8% of Jeju Islanders between twenty and forty liked Jeju much more than Standard Korean, which 49.1% of those above eighty did.
In a 2013 survey of Jeju natives, 77.9% agreed with the statement that "[the Jeju language] has to be passed down as part of Jeju culture." But a 2015 study of approximately a thousand Jeju Islanders suggests that even though most Jeju Islanders believe the language to be an important part of the island's culture, the vast majority are skeptical of the language's long-term viability, and more people are unwilling than willing to actively participate in language preservation efforts.
Jeju has historically had no written language. Two recently devised standard orthographies are currently in use: a system created in 1991 by scholars of the Jeju Dialect Research Society (Korean: ? Jeju bang'eon yeon'gu-hoe), and a system promulgated by the provincial government in 2014. Both systems use the Korean alphabet Hangul with one additional letter ?, which was used in the Middle and Early Modern Korean scripts but is now defunct in written Korean. Similar to the modern Korean script, Jeju orthographies have morphophonemic tendencies, meaning that transcribing the underlying morphology generally takes precedence over the surface form. The two orthographies differ largely because they are based on different morphological analyses of the language, especially of the verbal paradigm, as seen in the example below.
|Orthography||Underlying morphemes||Jeju word||Necessary analysis|
|Research Society orthography||- nakkeu- "to fish"||- -eomsi CONT||-? -min COND||? nakkeomsimin "if [he] is fishing"||Stem-final vowel -eu is lost before vowel-initial suffix|
|Government orthography||?- nakk- "to fish"||-? -eoms CONT||-? -min COND||? nakkeomsimin "if [he] is fishing"||Conditional suffix -min requires epenthetical vowel -i- after consonant|
This article will use the government's orthography where the two differ.
The transliteration scheme generally used in Korean linguistics, including when transcribing Jeju, is the Yale Romanization system. Yang C., Yang S., and O'Grady 2019 instead uses a variant of the Revised Romanization system with the addition of the sequence aw for ? /?/. This article also uses Revised Romanization with the addition of aw, but without Yang C., Yang S., and O'Grady 2019's one-to-one correspondence between Hangul glyphs and the Latin alphabet.
The non-approximant consonants of Jeju correspond to the nineteen non-approximant consonants of Standard Korean, and Jeju displays the three-way contrast between stops and affricates characteristic of Modern Korean. Whether the voiced glottal fricative /?/, absent in Standard Korean, exists as a phoneme in Jeju or merely as an allophone of /h/ remains disputed. A 2000 acoustic and aerodynamic study of eight native Jeju speakers concludes that "the consonants of the two languages seem to be the same in every respect... the phonetic realization of all [Jeju] consonants are the same as those found in [Seoul] Korean."
|Nasal||m ?||n ?||? ?[b]|
|lax||p ?||t ?||t? ?||k ?|
|tense||p? ?||t? ?||t ?||k? ?|
|aspirated||p? ?||t? ?||t ?||k? ?|
|Fricative||lax/aspirated||s ?||h ? |
Jeju allophony involves a number of phonological processes also found in Seoul Korean. As in Korean, /l/ surfaces as [?] intervocally. Also as in Korean, lax stops and affricates have fully voiced allophones in medial position, all obstruents have unreleased allophones in final position, and syllable-final sibilants surface as [t?]. Whether non-lax stops and affricates[c] can appear in final position is controversial. The morphological analysis necessary for the government's orthography permits them, while the analysis behind the Jeju Language Research Society's orthography forbids them.
Most non-morphophonological consonant assimilation rules of Standard Korean are also found in Jeju. /s/ and /s?/ are regularly palatalized to [?] before /i/ or /j/. Lax obstruents are tensed following another obstruent. /h/ aspirates both the preceding and the subsequent lax obstruent. A nasal consonant nasalizes a preceding obstruent or /h/. /l/ becomes [n] following all consonants except itself or /n/, and this [n] can itself nasalize the preceding obstruent so that the underlying sequence /pl/ is realized as [mn]. On the other hand, underlying /ln/ and /nl/ both produce [ll].
|Jeju word||Underlying phonemes||Realization|
|? "strength"||/sim/||[?im]||Palatalization before high vowel /i/|
|"on purpose"||/j?kpul/||[j?k?p?ul]||Lax obstruent tensed after another obstruent|
|"bloodsucker"||/h?phelkwi/||[h?p?elgwi]||/h/ aspirates surrounding obstruents|
|"pig's tooth"||/tosn(?)i/||[tonni]||Nasals nasalize preceding obstruent|
|"cleverness"||/me?la?/||[me?na?]||/l/ is realized as [n] after most consonants; underlying /l/ will nasalize preceding obstruent|
|? "blade"||/k?aln?l/||[k?all?l]||/l/ assimilates both preceding and subsequent /n/|
Jeju also has consonant allophones that appear only at morpheme boundaries. Some of these are found in Standard Korean, such as the insertion of [n] before i- or j- at most word-internal morpheme boundaries; the palatalization of /t/ to [d?] before an affixal -i; and the tensing of obstruents following certain morpheme-final nasals. Other rules are absent in Standard Korean. For instance, a sonorant-final word or morpheme can trigger aspiration (for older speakers) or tensing (for younger speakers) in a subsequent lax consonant. In some cases this is due to an underlying consonant cluster, but not all cases can be explained in this way. Other Jeju-specific processes include the doubling of a word-final consonant when followed by a vowel, glide, or /h/, and the lenition of /p/ to [w] at some word boundaries.
|Jeju word (morphemes hyphenated)||Underlying phonemes||Realization|
|ssog-ip "inner leaf"||/s?okip/||[s?o?nip?]||n-insertion before /i/|
|? mawd-i "eldest child"||/m?ti/||[m?d?i]||/t/ palatalized before /i/, devoiced in medial position|
|keom-su-da "to be black" [honoring addressee]||/k?msuta/||[k?ms?uda]||Obstruent tensed after verb stem-final nasal|
|sul-beng "alcohol bottle"||/sulpe?/||[sulp?e?]||Aspiration after sonorant (for older speakers)|
|il-weol "January"||/ilw?l/||[illw?l]||Consonant doubling|
|? jijib-ai "girl"||/t?it?ipai/||[t?id?ip?p?ai]|
|dae-bat "bamboo field"||/tæpat/||[tæwat?]||Lenition of /p/|
Verbal conjugation can also lead to consonantal changes. Verb stem-final /l/ and /h/ are lost before /n/. In the case of verb stems ending in -d, -p. -s, and -k, the final consonants are always preserved in so-called regular verbs, but in irregular verbs, -d and -p are lenited to [?] and [u~w] respectively while -s and -k are lost when followed by a vowel.
|Underlying morphemes||Surface realization||Regular verb|
|- dawd- "to run"||-? -gok CONN||? dawt-gok "runs, and"||bat-gok "receives, and"|
|-? -a SE||? dawr-a "runs"||bad-a "receives"|
|?- bib- "to pour"||-? -gok CONN||bip-gok "pours, and"||ip-gok "wears, and"|
|-? -eo SE||biw-eo "pours"||ib-eo "wears"|
|?- jis- "to compose writing"||-? -gok CONN||jit-gok "composes writing, and"||jit-gok "builds, and"|
|-? -eo SE||ji-eo "composes writing"||jis-eo "builds"|
|?- nug- "to lie down"[d]||-? -gok CONN||nuk-gok "lies down, and"||meok-gok "eats, and"|
|-? -eo SE||nu-eo "lies down"||meog-eo "eats"|
While not permitted in the surface representation of Jeju, morpheme-final consonant clusters can exist in the underlying form. Many cases of post-sonorant aspiration involve morphemes whose Middle Korean cognates feature a final -h, suggesting that an underlying final -h after the sonorant should be posited in Jeju as well. Besides these h-final clusters, Jeju permits a number of other final consonant clusters, including -lk, -lm, -mk~?k, -sk, and (in the analysis of the government's orthography) -ms. These clusters surface as a single consonant in isolation or before a consonant, but are fully realized when followed by a vowel.
|Underlying form||Realization in isolation/before consonant||Realization before vowel|
|- /p?lk/ "to be bright"||? [p?k?k?o]||? [p?lg?n]|
|? /salm/ "life"||? [sam]||[salmi]|
|?~?~? /namk~na?k~na?/ "tree"||?~? [nam~na?]||~~ [namgi~na?gi~na?i]|
|-? /?ms/ CONT||[tek?j?mga]||[tek?j?ms?]|
|? /pask/ "outside"||? [pat?]||[pat?k?i]|
|Close||? i||/i/||? eu||/?/||? u||/u/|
|Mid||? e||/e/||? eo||/?/||? o||/o/|
|Open||? ae||/æ/||? a||/a/||? aw||/?/|
Among younger and less fluent speakers, /æ/ and /?/ have both raised to /e/ and /o/ or /?/[f] respectively, resulting in a seven-vowel system identical to the vowel inventory of Seoul Korean. The raising of Jeju /æ/ occurred before the raising of /?/, and may have predated Standard Korean's ongoing merger of /æ/ and /e/. The subsequent loss of /?/ may have been motivated by a language-internal desire for symmetry in the vowel system. On the other hand, the vowel mergers are accelerated among Jeju speakers living in coastal communities more exposed to Standard Korean.
Jeju has two or three glides: /j/, /w/, and possibly /?/. /j/ can occur with all vowels except /i/ and /?/. /jæ/ and /je/ have merged even among speakers who distinguish the monophthongs, and many speakers who retain /?/ also merge /j?/ with /j?/./w/ cannot occur with the three back vowels or with /?/. /?/ occurs only with /i/, and the resulting diphthong /?i/ is generally realized as [?] word-initially and [i] otherwise.
|/je/||?, ?||yesukjekkillak||"quiz; riddle"|
|/jo/||?||yore||"here; this place"|
|/we/||?, ?, ?||wenchak||"left side"|
|/wa/||?||warida||"to be in a hurry"|
Several phonological processes affect the surface realization of Jeju vowels. In one process shared with Standard Korean, a bisyllabic vowel sequence may be contracted to a monosyllabic polyphthong.
|English||Uncontracted Jeju form||Contracted Jeju form|
|"it was caught"||? jepieotjeo||jepyeotjeo|
Vowel-affecting processes are particularly numerous in the verbal paradigm. Verb stem-final -eu is lost before a vowel-initial suffix. Similar to Standard Korean, a stem-final -i diphthongizes a subsequent vowel by inserting the onglide [j]. Unlike in its sister language, Jeju j-insertion may occur even with an intervening consonant, and between a verb stem ending in -e, -ae, or -aw and a suffix with initial eo-.
Many of Jeju's consonant-initial verbal suffixes take an initial epenthetic vowel if the previous morpheme ends with a consonant.[g] The default epenthetic vowel is -?- -eu- /?/, but the vowel surfaces as -?- -i- [i] following a sibilant and as -?- -u- [u] following an underlying labial.
|English||Underlying morphemes||Surface realization|
|"is sad, and"||- seulpeu- "to be sad"||-? -eong CONN||seulp-eong|
|"rests, and"||?- swi- "to rest"||-? -eong CONN||swi-yeong|
|"was fast"||?- je- "to be fast"||-? -eos PFV||-? -eo SE||je-yeos-eo|
|"mixed with water"||?- gae- "to mix with water"||-? -eos PFV||-? -eo SE||gae-yeos-eo|
|"did"||haw- "to do"||-? -eos PFV||-? -eo SE||? haw-yeos-eo|
|"if [it] burns"||?- ka- "to burn"||-? -min COND||ka-min|
|"if [he] believes"||?- mid- "to believe"||mid-eumin|
|"if [it] is bad"||?- jus- "to pick"||jus-imin|
|"if [he] puts in the soup"||- jawm- "to put into soup"||jawm-umin|
Like Standard Korean but unlike Middle Korean, Koreanic vowel harmony is no longer generally applicable in all native morphemes but remains productive in sound symbolism and certain verbal suffixes. Jeju has two harmonic classes, yin and yang. The neutral vowel /i/ can occur with either class.
|Harmonic class||Vowel correspondences||Sound symbolism|
|Yin||?||u||e||?||Dark; heavy; dull; negative|
|Yang||a||o||æ||?||Bright; light; sharp; positive|
|Yin-class allomorph||Yang-class allomorph|
|meog-eos-eo "ate"||gar-as-eo "plowed"|
|guj-eos-eo "was bad"||gob-as-eo "hid"|
|geus-eos-eo "drew a line"||dawr-as-eo "ran"|
In certain cases, suffix allomorphs do not match the harmonic class of the previous vowel. Verb stems with final vowel /u/ or /?/ take the yang allomorph if their Middle Korean forms were /?/, thus conserving their original harmonic class while violating their current one. Disyllabic stems that end in -u also take the yang allomorph, but monosyllabic -u stems or disyllabic -uC stems do not.
Jeju syllable structure is (C)(G)V(C) with G being a glide.
|V||? i "this"||[i]|
|CV||? tta "earth"||[t?a]|
|GV (or VV)||? we "cause; principle"||[we]|
|CGV (or CVV)||? swi "filling" [for dumplings, etc.]||[swi]|
|VC||? al "egg"||[al]|
|GVC (or VVC)||? yut "neighbor"||[jut?]|
|CGVC (or CVVC)||? kwang "lunatic"||[kwa?]|
As in Standard Korean, ng- /?/ cannot occur syllable-initially, and l- /l/ does not occur word-initially in native words.
Jeju does not have phonemic vowel length, stress, or tone. Its phonological hierarchy is characterized by accentual phrases similar to those of Standard Korean, with a basic Low-High-Low-High tonal pattern varying according to sentence type, but there are also important differences in the two languages' prosody. Jeju has a weaker tonal distinction within the first half of the accentual phrase than Seoul Korean does, while its aspirate consonants do not produce as significant a high pitch as their Seoul equivalents. Jeju uses more contour tones, where the pitch shifts within a single syllable, than Seoul Korean. Unlike in Seoul Korean, older and fluent speakers of Jeju will also lengthen the final vowel of both clauses in alternative questions.
Jeju is typologically similar to Korean, both being head-final agglutinative languages. However, the two languages show significant differences in the verbal paradigm, such as Jeju's use of a dedicated conditional suffix.
Jeju nouns may be a single morpheme, a compound of multiple nouns, or a base noun with a merged attributive verb, or form through derivational affixes attached to nouns or verb stems. In compound nouns that include a native morpheme, the phoneme -s- may intervene between the two elements. Because this "in-between s" appears only after a vowel and before a consonant, it is never realized as [s] but almost always surfaces as [t?].
Some Jeju nouns are bound nouns, meaning that they cannot appear independently without a noun phrase. The example below features the bound noun ? chim "worth" accompanied by the obligatory attributive verb.
? ? bo-l chim eus-jeo
see-FUT.ATTR worth not.be-DEC
"hardly worth seeing"
Jeju has two suffixing plural markers, which are obligatory for plural nouns accompanied by determiners and optional otherwise. The plural marker -? -deol can occur with all nouns and pronouns. The marker -? -ne is restricted for humans and pronouns, and can also have an associative meaning: e.g. Mansu-ne "Mansu and his family" (lit. 'Mansu and his associates'). The combined sequence - -ne-deol is sometimes also used.
Nouns accompanied by numerals usually take a variety of classifiers, such as ? jae for counting trees and gokji for counting songs. Classifiers for cardinals are unmarked, but those for ordinals are followed by the ordinal-marking ? che.
? ? nang hawn jae
tree one CLF
? ? ? ? nang du jae che
tree two CLF ORD
"the second tree"
Jeju marks noun case and other semantic relations through suffixing noun particles. Particles that mark the nominative, accusative, and genitive cases are very frequently omitted. The table below is not exhaustive and lists only some of the most significant particles.
|Function||Particle[h]||Allomorphy and variants||Example||Usage notes|
|After vowel: ? -ga||?
"Did grandmother say that?"
|Does not appear in the complement, unlike in Standard Korean. Cannot be topicalized.|
|Rare, formal post-vowel form: ? -reul|| .
siri-re gawreu-l dam-eura
"Put the flour into the steamer."
|Unlike in Korean, can be followed by other particles, e.g. neu-l-gwang 2SG-ACC-COM "with you".|
jib-i bas-eun eodui is-eo
"Where is your family's field?"
|-i is rare, but required when the subsequent noun phrase begins with an adnominal clause.|
|? ? ?
san ui-t nang
"the trees on the mountain"
|Called "pseudo-genitive" in Yang C., Yang S., and O'Grady 2019. Appears in certain compounds, as mentioned above. May also follow a locative marker to attribute a noun.|
gaui-gawra gongbui-haw-ren haw-b-seo
"Please tell him to study."
|Restricted to human addressees of verbs of speaking.|
geu sareum-sindi gawr-eup-di-ga
"Did you talk to that person?"
|May be used with the verb sitda "to exist" to form possessive constructions, with the dative marking the possessor.|
|Due to interference from Korean cognate -hante: -hanti, -ante||? .|
"I will give [it] to you."
"To whom did you give [it]?"
|Rarely, after vowel: ? -neun||? .
"Today is Sunday."
|Either introduces a new topic or establishes a contrast. Must have a contrastive meaning sentence-internally.|
hareubang-irang jeo-re aj-ib-seo
"Grandfather [and not anyone else], sit there."
|Contrastive meaning only.|
|Also used: (?) -(i)rageune||? .|
neu-rageun jib-i ga-ra.
"You [and not anyone else], go home."
|After -i and possibly -l: ? -e
Occasionally after any vowel: ? -ye
seong badang-i sy-eo
"Is my older sibling at the sea?"
|Refers to location for stative verbs and direction for dynamic verbs; may also refer to time.
According to Kim Jee-hong 2015, -(y)e is not an allomorph but a different locative morpheme used for clearly bounded spaces, such as tables or containers.
|Post-vowel form -seo sometimes occurs after consonant.||?? .
"Let's study at school."
|Refers to location for action verbs.|
"In the field"
|Variants of -i, -iseo used to emphasize the boundedness of the referent.
Analyzed by Yang S., Yang C., and O'Grady 2019 not as a separate morpheme but as a bound noun meaning "place," juxtaposed with the locational noun.
|After liquid consonant -l and sometimes after vowel: / -leure/leore
Initial syllable deu- also found as deo-, teu-, ti-, di-, de-, or ri-
"Pour [it] to this side."
|Denotes direction of movement, like English "to; into; toward."
Moon S. and Kim W. 2017 analyzes ? -re and -deure as distinct particles, with -re having a solely directional meaning while -deure simultaneously emphasizes both the direction and the location of the direction's destination. Most sources treat the two as allomorphs, especially when appended to nouns.
|Comitative and conjunctive||
|N/A|| ? .
gaui-n eomeong-iyeong sawdab-haw-yeoms-u-ge
"S/he is doing the laundry with his/her mom."
|Kim Jee-hong 2015 notes:
Like in Middle Korean but unlike in Modern Seoul Korean, comitative markers may occur on the final element being linked and also take other case markers.
"potatoes and eggs"
|Also used: -hawgo||? .|
nang-hawgok gojang hawsseul singgeu-ra
"Please plant some trees and flowers."
The Jeju verb consists of a root that is followed by suffixes that provide grammatical information such as voice, tense, aspect, mood, evidentiality, relative social status, and the formality of the utterance. Jeju verbs include not only action verbs familiar to English speakers such as meokda "to eat" or berida "to see," but also adjectival verbs such as beochida "to be heavy" or hultta "to be thick." Verbs can take derivational suffixes to form adverbs and nouns.
nal-do yeong daws-im-gwang
day-even like.this warm-NMLZ-COM
"What a warm day!" (lit. 'With the warmth of even the day like this!')
"Vegetables that s/he will pluck"
Jeju has a number of pre-final verbal suffixes: tense-aspect-mood markers which follow the verb stem but cannot appear at the end of the inflected verb. The exact number of these suffixes is unclear because scholars disagree on the correct morphological segmentation. One analysis of the suffix paradigm, as presented in Yang C., Yang S., and O'Grady 2019, is given below.
There is relatively widespread agreement on the existence of the following four discrete TAM morphemes, presented in the order they co-occur: the continuative aspect marker -eoms, the perfective aspect marker ? -eos, the prospective mood marker ? -(eu)k, and the realis mood marker (?)? -(eu)neu. Depending on the analysis of the aforementioned epenthetical vowels that precede many verbal suffixes, the base forms of the three morphemes may alternately be analyzed as -eomsi, -eosi, ? -keu, and ? -neu.
-eoms(i) is an imperfective or continuative aspect particle, referring to a process perceived as ongoing and similar to the English construction "be VERB-ing." With an adjectival verb, it has an inchoative ("beginning to; become") meaning. A verb with -eoms(i) is interpreted as either present or future by default, and some analyses interpret the particle as also conveying the present tense for specific events and states. The suffix has a vowel-harmonic variant -ams(i), as well as allomorphs -yeoms(i), -yams(i), and -ms(i) when following certain vowels.
Often characterized as a perfective aspect marker,-eos(i) has also been described as a present perfect marker and as behaving as a perfective marker with some verbs and as a past tense marker with others.-eos(i) can express non-past events in certain constructions that call for verbs "conceptualized in their entirety," such as a hypothetical future event. In adjectival verbs, it may also refer to a current state that contrasts with a past situation.-eos(i) can also be doubled for a habitual or a past perfect interpretation. Also like -eoms, this suffix takes the vowel-harmonic variant -as(i) and has allomorphs -yeos(i), -yas(i), and -s(i) after certain vowels.
The prospective mood marker ?/ -(eu)k/(eu)keu[l] marks the subject's intention in first-person-subject declarative sentences or second-person-subject interrogative sentences, and the speaker's conjecture otherwise. -(eu)k may also have a future-tense interpretation.
-(eu)k can only be followed by a small number of suffixes in Yang C., Yang S., and O'Grady 2019's analysis. Some analyses treat the initial vowel of the following suffix as part of an allomorph or nuanced variant of -(eu)k, so that gakeora "[I] will go" may be segmented as ga-k-eora or ga-keo-ra.
The realis or indicative mood marker (?)? -(eu)neu[l] indicates "a fact or habitual action in the nonpast" which the speaker perceives to be true in general, permanently, or over a longer duration of time, as demonstrated in the contrast below. The putative non-past-tense marker -(eu)n may also be analyzed as an allomorph of -(eu)neu. In this context, the morpheme -(eu)n(eu) has also been interpreted as a perfect marker (not to be confused with the perfective marker).
The existence of the Korean subject-honorific marker (?)? -(eu)si is controversial for Jeju, with some scholars arguing that it was entirely absent and others that it was restricted to higher registers. Ko J. 2011b notes that it was used only "by officials while referring to people of very high status and by the seonbi of the educated classes."
The segmentation of verb-final elements is controversial. The two recent extensive treatments of the topic, Yang C., Yang S., and O'Grady 2019 and Kim Jee-hong 2015, give incompatible analyses of the suffix paradigm.
They further divide verb-final suffixes into three categories: Type 1, which cannot occur with tense markers; Type 2, which must occur with either a tense marker or the aspect marker -eoms, which loses its underlying -s before a Type 2 suffix; and a mixed type, which can occur with the non-past marker but not with the other two tense markers. The vast majority of suffixes are categorized as Tense 1 and thus cannot follow a tense marker. Uniquely among pre-final suffixes, the past tense marker -eon can also appear without a final suffix.
Examples of Yang C., Yang S., and O'Grady 2019's segmentation are given below.
In Kim Jee-hong's analysis, verb-final single morphemes are termed "canonical endings." Canonical endings are contrasted with a wide variety of "non-canonical endings," formed by the fusion of various grammatical elements such as multiple canonical endings, truncated conjuctive and embedded sentences, and bound nouns[o] connected to the verb stem or a canonical ending via an attributive or a nominalizer. The most common canonical component of these non-canonical endings is the suffix ? -eo (vowel-harmonic allomorph ? -a), which Kim calls the unmarked "default ending."
Since Yang C., Yang S., and O'Grady 2019's tenses align with the aforementioned attributive suffixes, sentences they analyze as "Tense-Type 2 Suffix" sequences are often analyzed as non-canonical endings with a "Canonical ending-Attributive-Bound noun" composition by Kim Jee-hong. Many of Yang C., Yang S., and O'Grady 2019's Type 1 suffixes are also interpreted as polymorphemic non-canonical endings. Kim Jee-hong also segments some of Yang C., Yang S., and O'Grady 2019's mixed-type suffixes so that the base form of the suffix includes the -n of the latter's non-past tense marker.
Examples from Kim Jee-hong 2015's analysis, directly corresponding to the examples above of Yang C., Yang S., and O'Grady 2019, are given below. The "default ender" -eo is bolded.
Jeju has a number of clause-final suffixes, called "sentence enders" in Yang C., Yang S., and O'Grady 2019 and "terminal suffixes" (Korean: ? jonggyeol eomi) in Korean, that provide information such as degree of formality, social status, evidentality, and modality. Sentence enders may consist of one or multiple morphemes. Kim Jee-hong argues for four speech levels in Jeju, defined by the degree of formality and deference their sentence enders connote: informal and plain (non-honorific); formal and plain; informal and honorific, marked by the morpheme ? -u-, and formal and honorific, featuring the morpheme ? -eup. An archaic speech level showing extreme deference is attested from shamanic chants.
As different segmentation hypotheses produce different sentence enders. the chart below will list only a small, illustrative sample of the dozens of suffixes that appear in Yang C., Yang S., and O'Grady 2019 and Kim Jee-h. 2015. The classification is based on Kim Jee-hong 2017, which differs from Kim Jee-hong 2015.[p]
"Is s/he pretty?"
|Kim Jee-hong considers -eo the unmarked sentence ender. Depending on suprasegmentals, the suffix may be used in a plain statement, a question, a command, an exclamation, or a construction in which the speaker informs the addressee of information that the latter did not know and expects a confirmatory response. The suffix is also found in Standard Korean with a similar degree of versatility and widespread use.|||
|(?)?||-(eu)ju|| ? ? ?.
Mansu mal jal gawd-ju
"Mansu talks well."
|According to Kim Jee-hong, -ju conveys a statement of presumption or assumption without direct supporting experience, and invites the addressee to confirm the statement's veracity. Kim also states that -ju may end a confirmatory question with the implication that the addressee should agree with the speaker.
According to Yang C., Yang S., and O'Grady, -ju expresses a statement of intention or strong assertion with a first-person subject and a statement of judgement or assumption with a third-person subject, and may also convey regret or advice.
"[I see that] the day was cloudy."
|(-eun)-ge generally conveys a statement of fact that the speaker has directly observed, or has inferred from a direct observation.
In Yang C., Yang S., and O'Grady 2019, the suffix is given as -ge, a special Type 2 suffix which can only combine with the two tense markers ending in -n. Kim Jee-hong classifies it as a non-canonical ending composed of a fused attributive -eun and bound noun.
neu musigeo sichi-men
"What are you washing?"
|(-eu)men is used for both statements and questions, but only when the speaker and addressee are emotionally intimate. When the verb is inflected for aspect, (-eu)men is used to refer to a past event that was observed or inferred from observation. If uninflected, the suffix denotes an ongoing event.|||
|?||-da|| ? .
dosaegi geot meog-eot-da
"The pig ate the fodder."
|Unmarked formal statement ender. Dictionary citation form.|||
|(?)?||-(eu)jeo|| ? .
hareubang mom gawm-at-jeo
"Grandfather took a bath."
|-jeo expresses a factual statement with the premise that the addressee is unaware of the fact, and may implicitly either urge the addressee to accept this new information or rebuke the addressee for not having known it. With a first-person subject, -jeo conveys the speaker's intention to do something. Whether these two uses of -jeo are connected uses of the same morpheme, or whether they are two different homophonous morphemes, is disputed.|||
|?||-na||? ? ? .
i pul sareum-deor-i meong-na
"This plant is edible."
(Gloss: this plant person-PL-NOM eat-na)
|Expresses a statement of fact with the implication that it is an intrinsic or permanent quality or state; commonly found with proverbs and aphorisms. -na is also used to ask questions about facts (including non-permanent facts), where it has a "somewhat authoritative tone." As with -jeo, whether these two uses reflect the same morpheme or two homophonous ones is disputed.|||
|-gona|| ? !
i-di mul saer-a-na-t-gona
"The water had leaked here!"
|Expresses a statement of surprise or excitement.
Yang C., Yang S., and O'Grady 2019 gives the suffix as -guna instead. Kim Jee-hong reports that the suffix can be shortened to a single syllable ? -go.
geu-di musa daws-as-i(-)n-ga
"Why is that place warm?"
|Conveys a question directed to the addressee.
Yang C., Yang S., and O'Grady 2019 analyzes the suffixes as Type 2 ender -ga, with -n-ga not a genuine ending but -ga following a tense marker ending in -n. Kim Jee-hong distinguishes the canonical ending -ga with the non-canonical -eun-ga, which is analyzed as having a fused attributive.
seonsing musigeo tew-as-i(-)n-go
"I wonder what the teacher distributed."
|Has a conjectural connotation. Often used in questions addressed to oneself, and is less direct than -eun-ga when asked to the addressee.
Yang C., Yang S., and O'Grady 2019 analyzes the suffix as mixed type ender -go, with -eun-go being -go preceded by the non-past tense marker (and gaining this conjectural meaning only in the presence of the non-past tense marker). Kim Jee-hong analyzes it as a non-canonical ending with a fused attributive.
neu-n musa eor-eo(-)n-dia
"Why are you cold?"
|Used to ask a question about which the addressee has direct relevant experience. In most cases the addressee is the subject of the verb, although third-person subjects have been attested. Appears in Kim Jee-hong's work in the contracted form ? -dya.|||
"Let's get finished eating!"
|Used in propositions (not commands as in Standard Korean).|||
|Used to command immediate action.|||
|Used by older women when talking to younger adults not old enough for honorifics and not young or emotionally close enough for informal speech. May convey statements, questions, requests, and proposals.|||
The honorific verbs, which show deference to the addressee, are formed by a special suffix that can be followed only by a small number of sentence enders.
The informal honorific forms are marked by ? -u- or (?)? -(eu)u-.[l] The former is used with the copula verb ida and with all inflected verbs, and the latter is used with uninflected adjectival verbs. -u- and -(eu)u- may take the alternative form ? -su- after a verb inflected for aspect and a non-liquid consonant, respectively. The informal honorific form cannot occur with uninflected action verbs. The two suffixes may only be followed by the sentence enders in the table below. Informal honorific requests cannot be formed morphologically.
|(?)? -(eu)u-||? -da||? ?.
"S/he drew water."
|Used for statements.|
"I have made side dishes."
|Used to report new information; restricted to inflected verbs.|
"It is deep."
|Used to report new information or an opinion; restricted to uninflected verbs.|
"Will you throw away the eggs?"
|Used in questions. -gwa(ng) is generally restricted to inflected verbs. Due to sound symbolism, the tense endings are considered emphatic.|
|? -(eu)p-||?/? -ne/ni||? -da||? ? ? ?.
geu sin Mansu-sindi jog-eup-ne-da
"[I know] those shoes must be small for Mansu."
|A formal statement founded on prior knowledge, e.g. of Mansu's foot size. Implies that the rationale for the statement continues in the present and may be shared or experienced by the addressee.|
|?/? -kka(ng)|| ?
na-ga mawncheo aj-ip-ne-kka
"Should I sit first?"
|A formal question that the addressee is expected to be able to answer without direct observation.|
|?/? -de/di||? -da||? ? ? ?.
geu sin Mansu-sindi jog-eup-de-da
"[I saw] those shoes were small for Mansu."
|A formal statement motivated by a direct, external past observation that cannot be experienced firsthand by the addressee. As the observation must be external, the first-person singular subject is prohibited except in highly atypical situations such as dissociation.|
"[Based on your observation,] will the day be cold?"
|A formal question that the addressee is expected to answer based on a past observation relating to a third party.|
|N/A||? -seo|| .
"Please eat this."
|A formal request.|
|? -ju|| .
"Let's gather the potatoes."
|Expresses speaker's intention with a first-person subject and advice or judgment otherwise; widely used for suggestions and propositions.|
Some Jeju connectives, such as the suffixes ?/? -eon/eong "and", occur in pairs with one variant ending in -n and the other in -ng. Hong Chong-rim and Song Sang-jo both note that the choice between -n and -ng is often determined by the inflections of the subsequent clause; certain pre-final suffixes and sentence enders require a n-connective in the previous clause, while others require a ng-connective. Hong suggests that -n is used for specific and objective events and states, while -ng implies a general and subjective event or state. Song argues that -n is used for completed or achieved verbs, and -ng for incomplete or unachieved verbs. The nuances below are thus possible.
?? meog-eon ga-min
"Having eaten, if [I] go"
?? meog-eong ga-min
"If [I] eat and [then] go"
? ?? ? ? ? oneul bi o-ran il mot haw-n-da
today rain come-CONN work cannot do-PRES-SE
"It's raining, so [we] can't work"
? ?? ? ? ? oneul bi o-rang il mot haw-n-da
today rain come-CONN work cannot do-PRES-SE
"[I presume that] it's raining/going to rain, so [we] can't work"
The distinction between -n and -ng does not exist in mainland Korean varieties. Yang C., Yang S., and O'Grady 2019 reports that "the contrast between -eong and -eon appears to be disappearing, and the distinctions that remain are subtle and variable."
An important class of connectives, used for reporting speech and thoughts, is formed by the suffix ?/? -en/-eng, which fuses with sentence enders as in the example of -da below.
"'I am going.'"
? .[t] ga-ms-den gawr-at-su-da
"[S/he] said that [s/he] was going."
Similarly, informal honorific conjectural k-u-da becomes ? -ken; plain forms -ju and -jeo become ? -jen; question enders -ga and -go become ? -gen and ? -gon; honorific imperative -eup-seo becomes -eup-sen; and so forth. These fused suffixes may be used for both quotative and reportive purposes. In Standard Korean, indirect speech is strictly distinguished from the quotative by the removal of addressee honorifics and the switching of pronouns. In Jeju, the lines between direct and indirect speech are more blurred. All four forms below--given in order of increasing indirectness--are in use, and have the same meaning, "He said [to a superior] that he was going home."
"? "? . na-n jib-i ga-ms-u-da-en gawr-at-su-da [direct quote]
1SG-TOP house-LOC go-CONT-HON-SE-QUOT say-PFV-HON-SE
? ? . na-n jib-i ga-ms-u-den gawr-at-su-da [quotative fused]
1SG-TOP house-LOC go-CONT-HON-REP say-PFV-HON-SE
? . ji-n jib-i ga-ms-u-den gawr-at-su-da [pronouns changed]
3SG-TOP house-LOC go-CONT-HON-REP say-PFV-HON-SE
? . ji-n jib-i ga-ms-den gawr-at-su-da [honorific neutralized]
3SG-TOP house-LOC go-CONT-REP say-PFV-HON-SE
Jeju has many auxiliary verbs that are linked to the preceding main verb by the morpheme ?/? -eo/a.[s] These include anneda "to give," for an action that benefits a superior; bulda "to throw away," for an action yielding a complete result; and jida "to become," for a change of state. Jida is also used to indicate ability.
? ??. il hayeong he-nonan jug-eo-ji-k-yeo
work a.lot do-because die-eo-become-PROSP-SE
"I will become close to dying because I work a lot."
Jeju also uses light verbs, which have little semantic meaning but combine with nouns to form verbs. The most common light verb is hawda "to do," e.g. bureumssi "errand" -> bureumssi-hawda "to run an errand". There is also a large inventory of periphrastic phrases that convey modality.
Jeju has a small group of particles that commonly occur at the very end of phrases or sentences, many of which play important roles as discourse markers. The four principal ones are the formality marker -masseum and the emphatic markers ? -ge, ? -i, and ? -yang.
-masseum (variants -massim, -massi) may occur after subsentential phrases such as a bare or case-inflected noun, or attach to a small number of mostly plain sentence enders. The particle shows the speaker's deference towards the addressee, but is considered more emotionally intimate than the verbally inflected honorifics. In certain contexts, -masseum may be used with an intention to snub the addressee.
-ge is a discourse marker that attaches to adverbs, nouns and noun particles, and both sentence enders and connectives. It adds emphasis to the utterance and is often used to agree with or confirm something the addressee has just said.-i is used similarly to -ge, but is weaker in its emphasis. Both cannot be used while addressing a social superior, and -i also cannot appear in formal speech. Both particles can also appear in isolation: ge as a strong affirmation to a question, i as an indication that the speaker has not heard or does not believe what has been said.
-yang shows deference, but is considered more informal than -masseum. At the end of a sentence, it emphasizes the speaker's beliefs or attitudes. For example, a question becomes a rhetorical one when -yang is attached: is-ik-a "Could there be?" -> ? is-ik-a-yang "How could there be?" The particle is also commonly used for sarcastic mock deference, such as by parents while scolding children. Sentence-initially or internally, the suffix may establish the preceding element as the topic of discourse.Yang is also used in isolation as an interjection to get the attention of unfamiliar individuals, such as a shopkeeper, or to request the addressee to repeat what they have just said.
In the example below from Yang C. 2009, three of the four particles discussed above are used.
halmani, idi-ga mal-loman deureo-nan Moseulpo-ukkwa?
"Grandmother, is this place Moseulpo, which I've only heard of?"
Grandmother: ?, ?.
aw matda-ge, idi-ga Moseulpo-yeo.
"Yes, you're indeed correct, this place is Moseulpo."
Granddaughter: ?... ? ? ? .
Moseulpo-yang... gemin eotteong-heoyeon Moseulpo-yen hawyeosinsingo-massim? hawkgyo-eseo sukje naeyeon-massim.
"Moseulpo, right... So why do they call it Moseulpo, please? They gave us an assignment at school, please."
Note the granddaughter's use of the verbally inflected honorific -u- and the deference-marking massim and yang while addressing the grandmother.
Jeju has the following basic personal pronouns.
|1st person||?/? na/nae "I; me"||(?) uri(-deol) "we; us"|
|2nd person||?/? neu/ni "you (s.)"||(?) neu-ne-(deol) "you (pl.)"||For younger, emotionally intimate, or socially inferior individuals|
|? ji "you (s.)"||(?) ji-ne(-deol) "you (pl.)"||For younger individuals, but more respectful than neu/ni|
|No overt pronoun||For older individuals|
|3rd person||None per se.
Informally, demonstratives used before ai: gai "him/her" (lit. 'that person'), etc.
According to Yang C., Yang S., and O'Grady 2019, there are four basic deictic demonstratives in Jeju. Most other sources mention three, which are identical to those of Standard Korean.
Most of the Jeju lexicon is Koreanic, and "a sizeable number" of words are identical with Korean. There are false friends between the languages, such as Korean kamda "to wash hair" and Jeju ? kawmda "to wash the body." Jeju also preserves many Middle Korean terms now lost in Korean, such as ? kas "wife; woman" and eosi "parent." Like Korean, Jeju uses many Sino-Korean words based on local readings of Classical Chinese.
Jeju Island was ruled by the Mongols in the late thirteenth century and some Middle Mongol terms still survive in the language, though the extent of Mongol influence is disputed. Popular claims of hundreds of Mongol loans in Jeju are linguistically unsound. Uncontroversial Mongol loans are most common in terms relating to animal husbandry.
|two-year-old cattle||dagan||da?a?an "two-year-old horse"|
|classifier for houses||geori||ger "house"|
Jeju may have loans from an ancient Japonic substratum. As the last fluent generation of Jeju speakers were born under or shortly after Japanese rule, remaining speakers also use many loans from Modern Japanese.
|noisy||? ureusai||? urusai|
Jeju sound symbolism operates with both consonants and vowels. The intensity of a Jeju word may be strengthened by using tense and especially aspirate obstruents. The sound symbolism may also be emphasized through the addition of consonants, by adding the sequence -? -rak to both reduplicated segments, and with fortition or lenition. The yang harmonic class of vowels has a bright, small connotation, and the yin vowel class gives a dark, large connotation. Ko Jae-hwan also gives examples of three or four layers of vowel sound symbolism.
Multiple sound-symbolic strategies may combine in a single word. Kang S. 2008 gives eight sound symbolic variants of the ideophone mawltak-mawltak "the shape of many objects being blunt," each more intense than the other:
mawltak-mawltak -> ? mawlteurak-mawlteurak -> mawlchak-mawlchak -> ? mawlchirak-mawlchirak -> ? moltak-moltak -> ? molchak-molchak -> molteurak-molteurak -> molchirak-molchirak
The kinship terminology of Jeju has been the focus of particular attention. Jeju has a complex kinship system that distinguishes the gender of both the speaker and the relative. Gender distinctions are particularly noticeable in sibling terminology. The words ? seong and asi refer to "older same-gender sibling" and "younger same-gender sibling" respectively, while orabang and nui refer specifically to "brother of a female" and "sister of a male" respectively. Female speakers also tend to refer to relatives with native compounds, whereas male speakers prefer Sino-Korean terms. For instance, the same cousin may be referred to by a man as sawchun "cousin" but by a woman as gomo-nim ttawl "paternal aunt's daughter." A major distinction between Jeju and Korean kinship terms is that women do not use honorifics to refer to her in-laws, reflecting weaker historical influence from Confucian patriarchal norms.
Jeju also uses supplementary prefixes to clarify the type of kinship, equivalent to "step-" or "maternal" in English. These include ?- chin-, ?- seong-, and ?- dang- for paternal relations, ?- we- for maternal relations, - daseum- for step-relations,?- cheo- and - gasi- for a male's in-laws, and ?- si- for a woman's in-laws. Five other prefixes, which may be combined, mark relative age: ?- chet- or ?- keun- "eldest," ?- set- "second eldest of three or more," - maljet- "third eldest of four or more," and - jogeun- "youngest." These are used to distinguish relatives of the same generation.
Other prefixes include ?- wang-, used in ? wang-hareubang "great-grandfather", and ?- neot-, used to refer to a sibling of one's grandparent generally.
The following is an excerpt from a version of the Menggam bonpuri, one of the epic chants recited by Jeju shamans. In this myth, the poacher Song Saman discovers an abandoned skull in the hills and cares for it as if it where his own ancestor. The skull reciprocates by warning Song Saman of his early death and advising him on how to avoid the chasa, the three gods of death.
This version was transcribed between 1956 and 1963 from the recitation of the shaman Byeon Sin-saeng, born c. 1904. The transcription predates both standardized orthographies of Jeju. The transcriber openly notes that the orthography is inconsistent. No attempt was made in this article to standardize or update the orthography.
|Jeju original||Korean translation[v]|
"? , ? ? ?, ? ? ... ? ? ? ? ? ? ..."
? ? , ? ...
? " . ? ... ? "
"? , ? ? ? ? ... ? ? ? ? ?[ ]? ..."
? ? ? ? ? ? , ? ...
" ? ?"
, " ? . ? ? ? ? ?... ? ? ? ? "
|neu Song Sawman-i jeonmaeng-i gawt seoreun-i maeg-inan, seoreun naneun hae-e amu dawl amu nar-eun maeng-i maeg-ini neu-ga bal sarang omong-hawyeo-jil ttae, na-reul nang-gos-euro gawjyeoda dora... si mawseul gang simbang si gae geoduugok madang-i keun dae sewang du ilhwe yeol-naeul gus-eul hawra...
maetttak chawllyeo-nowan baekbo baekgyeot-dillo gan jeor-eul hawyeon, gawmani kkulyeon eopdeojyeo-duseo bonan samchesaw-ga nawryeo-omeong...
"Song Sawman-i-ne jib-iseo jeongseong-eul ani deuryeomsin'ga?"
Maljai oneun chesaw-n "Song Sawman-i-ne jib-i baengnyeon-daegang'i-reul mosamtta. geu baengnyeon-daegang'i-ga Song Sawman-i simeure oramsen gawra-bun saeng-iyeo... badeum-eun badat-jumaneun simeong oraen hawn sigan-i siyeo-bunan eotteong-heulko?"
|"You, Song Saman, your life will end at only thirty, and the year you turn thirty, your life will end at any day of any month, so take me to the wooded forest while your feet are still alive and you can move... Go to three villages and gather three shamans and raise a great flagstaff in the household hall, and hold the Great Gut [lit. 'the gut of two weeks and fourteen days']..."
Once they laid out everything, [Song Sawman and his wife] went back a hundred steps and prostrated themselves. Quietly kneeling and lying prone, they saw the three chasa descend...
"Are they not doing devotional acts at Song Saman's household?"
The chasa coming in last [responded], "At Song Saman's household, they are worshipping a hundred-year-old skull. It seems that the hundred-year-old skull told [them] that we were coming to capture Song Sawman... We have partaken of the offerings [lit. 'received what is received'], but there is a date that they told us to capture him by, so what should we do?"
gai jire-ga keu-keu-ra
3SG height-NOM grow-PROSP-SE
See section Tense-marking.