Old Javanese or Kawi is an Austronesian language and the oldest attested phase of the Javanese language. It was natively spoken in the central and eastern part of Java Island, what is now Central Java, Yogyakarta and East Java Provinces, Indonesia. As a literary language, Kawi was used across Java and on the islands of Madura, Bali, and Lombok.
History The oldest example written entirely in Ancient Javanese, called the Sukabumi inscription, is dated 25 March 804 CE. This inscription, located in the district of Kepung in the Kediri Regency of East Java, is a copy of the original, dated some 120 years earlier (only this copy has been preserved). Its contents concern the construction of a dam for an irrigation canal near the river Śrī Hariñjing (now shortened to Srinjing). This inscription is the last of its kind to be written using Pallava script; all consequent examples of Old Javanese are written using Kawi script.
Development Old Javanese was not static, and its usage covered approximately 800 years – from the Kalingga kingdom until the founding of the Majapahit empire in 1292. The Javanese language which was spoken and written in the Majapahit era already underwent some changes and is therefore already closer to the Modern Javanese language.
Austronesian origins The most important shaping force on Old Javanese was its Austronesian heritage in vocabulary, sentence structure, and grammar that it shared with its sister languages in Southeast Asia.
Sanskrit influence The Indian linguistic influence in the Old Javanese language was almost exclusively Sanskrit influences (approximately 25–40%). There is no evidence of Indian linguistic elements in Old Javanese other than Sanskrit. This is different from, for example, the influence of Indian linguistics in the (Old) Malay language.
Sanskrit has had a deep and lasting impact on the vocabulary of the Javanese language. The Old Javanese–English Dictionary, written by Professor P.J. Zoetmulder in 1982, contains approximately 25,500 entries, no fewer than 12,500 of which are borrowed from Sanskrit. This large number is not an indication of usage, but it is an indication that the Ancient Javanese knew and employed these Sanskrit words in their literary works. In any given Old Javanese literary work, approximately 25% of the vocabulary is derived from Sanskrit.
Influences Medieval poems written in Old Javanese using the Kawi script continued to be circulated within the courts of Kartasura, Surakarta, and Yogyakarta. The poems were called layang kawi (Kawi books) or kakawin and were held in high regard. Starting in the 18th century, literature inspired by Old Javanese was written using the modern Javanese language and verse.
Phonology Sanskrit has also influenced both the phonology and the vocabulary of Old Javanese. Old Javanese also contains retroflex consonants, which might have been derived from Sanskrit. That is disputed by several linguists, who hold the view that it is also possible that the occurrence of these retroflex consonants was an independent development within the Austronesian language family.
Vowels Old Javanese has six vowels. Those vowels are ⟨a⟩, ⟨ĕ⟩ /ə/, ⟨e⟩ /e/, i, u, and o in Latin transliteration. Little can be said about the pronunciation of Old Javanese. It is believed that it has not been much different from the pronunciation of modern Javanese. However, the major difference is the pronunciation of ⟨a⟩ in open syllables: now ⟨å⟩ /ɔ/, then /a/, such as in (forest). Although, Old Javanese made a distinction between those "short vowels" and "long vowels" in writing such as ā, ö, e, ī, ū, and o, however, these "long vowels" have no distinction in phonology with those "short vowels". This distinction is generally found with unadapted loanwords from Sanskrit which differentiates the short and long vowels.
Consonants There are twenty consonants in Old Javanese which are written as b, c, d, ḍ, g, h, j, k, l, m, n, ñ, ŋ, p, r, s, t, ṭ, w, and y in Latin transliteration. The consonant ⟨ñ⟩ sometimes is written as the digraph ⟨ny⟩ and IPA ɲ, while the consonant ŋ sometimes is written as the digraph ⟨ng⟩.
Consonant articulationPlace of articulationPancawalimukhaSemivowelSibilantFricativeUnvoicedVoicedNasal Unaspirated Aspirated Unaspirated Aspirated Velar ka (kha) ga (gha) ṅa (h)aPalatal ca (cha) ja (jha) ña ya (śa)Retroflex ṭa (ṭha) ḍa (ḍha) (ṇa) ra (ṣa)Dental ta (tha) da (dha) na la saLabial pa (pha) ba (bha) ma wa Notes Aspirated consonants are pronounced as their unaspirated counterparts. The retroflex nasal consonant is pronounced as the dental counterpart. The sibilants are pronounced as the dental counterpart. The presence of such aspirated consonants, retroflex nasal, palatal sibilant, and retroflex sibilant are used for unadapted loanwords from Indo-Aryan languages (specifically Sanskrit).
Sandhi Sandhi is a cover term for a wide variety of sound changes that occur at morpheme or word boundaries.
If a word ends in a vowel and the next word in the same sentence begins with a vowel, both words may merge into one, with one vowel instead of two vowels, such as dewatādi instead of dewata + adi. A merger of ĕ with the preceding vowel results in assimilated ĕ to the preceding vowel, such as wawan (load, cargo; vessel, carrier, container, setting) from wawa (to carry, to bring) + ĕn. Similar vowels without short-long vowels consideration are assimilated as "long vowels". For example, rĕngön (hearing) is constructed from rĕngö (hearing, listening) + ĕn. The open vowel /a/ followed by close-front vowels of /e/ or /i/ are assimilated as /e/, such as bhinna ika become bhinneka (those are different). Meanwhile, the open vowel /a/ followed by close-back vowels of /o/ or /u/ is assimilated as /o/, such as mantra oṣadha becomes mantroṣadha. The semi-vowel y or w will replace the corresponding vowel i, u, or ö, when followed by a dissimilar vowel. For example, kadi amṛta become kadyamṛta (i + a → ya), ri ubhaya become ryubhaya (i + u → yu), milu āśā become milwāśā (u + a → wa), māsku ibu become māskwibu (u + i → wi), and angangsö agawe become angangswagawe (ö + a → wa).
Grammar
Verb Old Javanese verbs are morphologically complex and are conjugated by taking on a variety of affixes reflecting focus/trigger, aspect, voice, and other categories.
Voice/Focus/Trigger The active voice is derived through either prefix (m)aN- or infix -um-. The prefix (m)aN-, which is realized as maN- and aN-, is the prefix to make transitive verbs, for example, amati (to kill) from pati (death) and mangan (to eat) from pangan (food), if the base word is a verb. However, if the base word is a noun, the derivation can result in both transitive and intransitive verbs, such as angjanma (to be born, incarnate) from janma (man), whether the result will be either transitive or intransitive can not be predicted. In the base word is an adjective, the derived verbs are causative, such as anghilang (to cause something to disappear) from hilang (disappeared). The derivation can undergo denasalization in particular situations such as pamangan instead of mamangan (to eat) and panginum instead of manginum (to drink). The infix -um- is the prefix to show active verb which generally shows no difference in meaning with the derivation with prefix (m)aN-. Sometimes, there is a difference in meaning between the prefixed (m)aN- and infixed -um-, such as anahur (to repay) and sumahur (to answer) from sahur (answer, return). Passive voices are derived through either the prefix ka- or infix -in-. It is not necessary to express the actor in a passive sentence. If the actor is explicitly mentioned, the actor is introduced by de and put after the subject, such as "Katon pwa ta de sang Śrutasena". The prefix ka- refers to passive voice. If it is put before the consonant of the stem, it shows no change. However, if it is put before vowels, the sandhi is applicable, such as in kālap (taken) from ka- + alap. Other than ā from a + a, the other sandhi is ā from a + ĕ /ə/, e from a + either i or e, and o from a + u (there is no example with a + o exist). The other passive voice derivation is through infix -in-, such as inalap (taken) from alap. +Nasalization rule for prefix (m)aN-Initial of base wordSandhiHarmonized prefixExamplesnasal (m-, n-, ng-)(m)aN- + N- → (m)a-(m)a-maga → amaga (to disappoint)k(m)aN + k- → (m)ang-(m)ang-kĕmit → angĕmit (to guard)p, w(m)aN- + p-,w- → (m)am-(m)am-pahat → amahat (to tap)s, t(m)aN- + s-,t- → (m)an-(m)an-sambut → anambut (to seize)c(m)aN- + c- → (m)any-(m)any-cangking → anyangking (to carry)vowels(m)aN- + V- → (m)ang- + V-(m)ang-abĕn → angabĕn (to attack)d, g, h(m)aN- + d-,g-,h- → (m)ang- + d-,g-,h-(m)ang-haḍang → anghaḍang (to stand off)j(m)aN- + j- → (m)ang- + j-(m)ang-jajah → angjajah (to explore)semivowel (r, l, w)(m)aN- + H- → (m)ang- + H-(m)ang-liput → angliput (to envelop)b(m)aN- + b- → (m)am- + b-(m)am-bawa → ambawa (to bring) +Nasalization rule for infix -um-Initial of base wordSandhiHarmonized infixExamplesvowels-um- + V- → umV- umV-alap → umalap (to take)labials (b-,p-,m-,w-)-um- + C- → um-um-wawa → umawa (to carry)othersno changeno changejawil → jumawil (to touch)
Case The beneficiary-orientedness or plurality can be indicated with the suffix -i and -an. The suffix -i is used for active transitive verbs (with prefix (m)aN- or infix -um-) which harmonized into either -i (amatī, to slay, from pati) and -ani (amatyani, to slay, from pati) after a vowel. However, passive transitive verbs use the suffix -an (with the prefix ka- or infix -in-). In case of a final vowel -a, -an is attached, not -anan, for example, kapaḍan. Causative can be indicated by the suffix -akĕn from verbal and nominal bases (either prefix (m)aN- or infix -um-). The verb with -akĕn is object-oriented. There is no combination between the passive ka- and with suffix -akĕn. Applicative can be indicated by prefix maka- and pinaka- with sandhi rules applied. Prefix maka- is used for active voice (with (m)aN- and -um- integrated), while passive voice uses the prefix pinaka- (with -in- and ka- integrated). The denasalisation phenomenon can happen.
Mood Irrealis mood can be indicated by suffix -a to the verbs. The active irrealis verb can be constructed as active affixes (prefix (m)aN- or infix -um-) with suffix -a (such as manghuripa from manghurip). However, passive irrealis can be constructed with suppression of infix -in- and addition of the suffix -ĕn (such as huripĕn), while the suffix -a is added to passive verbs prefixed with ka-. The presence of sandhi sometimes makes it impossible to see whether -a of irrealis mood. In the case of pronominal suffix presence, the irrealis suffix is prioritized. There are three ways to put a verb in the imperative mood. ...by the bare indicative form without any changes, such as mijil (please come out) and anunggangi (please mount), which is the polite form. This form is just recognized by the context. ...drop the verbal prefix, such as wijil and tunggangi. ...place either t(a) or p(a) before the unchanged or reduced form, such as ta mijil, ta wijil, pamijil or pawijil and tānunggangi, ta tunggangi, pānunggangi or patunggangi. Prohibition is expressed by the word haywa (do not), such as "haywa ta kita malara!" (Don't be sad!).
+Construction of verbs in irrealis moodIrrealis moodBenefactive case -iCausative case -akĕnActive voice prefix (m)aN- or infix -um- the prefix (m)aN- or infix -um- present the suffix -anathe prefix (m)aN- or infix -um- present the suffix -akna or -akĕnPassive voice infix -in- the infix -in- absent the suffix -anathe infix -in- absent the suffix -akna or -akĕn
Noun and pronoun
Particle