J. Noorduyn Some remarks on Javanese chronogram words; A case of localization
In: Bijdragen tot de Taal-, Land- en Volkenkunde 149 (1993), no: 2, Leiden, 298-317
This PDF-file was downloaded from http://www.kitlv-journals.nl
Downloaded from Brill.com09/30/2021 08:11:50AM via free access J. NOORDUYN
SOME REMARKS ON JAVANESE CHRONOGRAM WORDS: A CASE OF LOCALIZATION1
The Javanese tradition of expressing dates in candrasengkalas or chronograms, i.e. by representing the numbers in dates by words having a fixed numerical value, has several times been recorded, and occasionally commented upon. However, so far no one has offered an exhaustive explanation of the origin and development of this tradition. This paper does not attempt to discuss all the aspects of this tradition either. It is hoped, however, that the following observations will show that an overall explanation of this remarkable phenomenon may be possible, at the same time indicating the lines along which this could be undertaken. In order to show how chronograms were used in Javanese historical texts, an example will be quoted from an eighteenth-century chronicle written in metrical form which contains many such dates. This text was edited and translated by M.C. Ricklefs in 1978. On p. 20 of this edition the following verse line is found (with the meaning of each word and the figure denoting its numerical value being added):
1 The author wishes to thank Professor J.J. Ras, whose valuable comments on some details in a first draft of this paper have been gratefully made use of in the revision of this paper. He also acknowledges the help of Mrs. Rosemary Robson, who kindly corrected the English of the text. He feels especially indebted to Professor A. Teeuw, who, because of the author's protracted illness, edited the final draft of this paper. He introduced some valuable corrections and additions. In discussions with the author, Teeuw pointed out that the mnemonic poems quoted by Korn and Ricklefs most probably go back to the text in the Old Javanese kakawin metre Turagagati (Zoetmulder 1974:456); in due time Teeuw hopes to return to this problem of the original version of the poem.
J. NOORDUYN, former Director of the KITLV and lecturer in Bugis, Makasarese and Sundanese with the Department of Languages and Cultures of Southeast Asia and Oceania at the University of Leiden, obtained his Ph.D. at Leiden in 1955. His two principal publications are Een achttiende-eeuwse kroniek van Wadjo', 1955, and 'The Manuscripts of the Makasarese Chronicle of Goa and Talloq: An Evaluation', BKI 147-IV, 1991. Dr. Noorduyn's address is Rijn en Schiekade 119, 2311 AT Leiden.
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 299
guwa jaladri wedang patala2 cave sea hot water (centre of the) earth 9 4 4 1 In his translation, Ricklefs (1978:21) renders this verse line as a coherent sentence, starting with the previous line and adding prepositions between the chronogram words, in the following way: When [they were roughly shaken] in Kadhiri, in 'the cave in the sea was hot water from the centre of the earth' [C1449/March 1527-Febr. 1528]. In this way the text of the chronicle has chronograms embedded in the description of certain events. Here, we shall not enter into the function and use of Javanese chronograms in texts; instead, we shall focus on the question of why and how Javanese chronogram words acquired their numerical value. Why does a word meaning 'cave' stand for the numeral 'nine' and a word meaning 'earth' for 'one'; why, to take another example, has the Javanese word pw-M/i/courageous', the numerical value of 'one'? To begin with, the published records consisted first of all of lists of Javanese words used for each numeral from 'one' to 'nine' plus 'zero'. The first Western investigator to give attention to this phenomenon, as to many other Javanese cultural phenomena, was Raffles (1817:ccii-ccv). He was followed by Gericke (1836), Hageman (1852:391-395), and Van Hien (1895M9336; 1933:382ff). These scholars all added glosses to the chronogram words listed, purporting to indicate their lexical meaning. Korn took a step forward (1922) when he published a Javanese poem of eleven stanzas, each containing four lines of twelve syllables. Each stanza presents the chronogram words for the successive numerals 'one' to 'ten' plus 'zero'. In his comments Korn added translations of some of these words, also mentioning some other words not included in the poem. A similar poem, clearly related to the former but consisting of ten stanzas and including a number of different words from or variants of words occurring in Kom's version, was published by Ricklefs in an appendix to his above- mentioned text edition (1978:243-244). Korn occasionally compared Hageman's data, and Ricklefs those of Van Hien. As an illustration of these published records, Raffles' list of chronogram words for 'three' and his glosses will be given below, along with the third
2 For Modern Javanese words (MJ) the present-day orthography is used, with one exception: e or e are written (hence with the diacritics) where necessary to distinguish the mid-open front vowel from e (pepel). Foir Old Javanese (OJ) the spelling system used in Zoetmulder (1982) has been followed. Sanskrit words (Skt) are presented according to the current transliteration system, though only in such cases where explicit reference is made to words in that language; otherwise such words have been spelt, depending on the context in which they occur, as OJ or MJ words. The term Javanese (also abbreviated as Jav.) is used whenever the latter opposition is not relevant.
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 300 /. Noorduyn stanzas from the chronogram poems in Korn's and Ricklefs' versions (below, these lists will be referred to as K and R respectively).
Raffles: Three Banning: fire (of a furnace), Pawaka: fire (from a volcano), Siking: fire (struck from a flint), Guna: fire (from friction of wood), Dahana: the fire which pervades every thing that has life, extensive fire, fire that has never been extinguished, Tri-ningrana: fire (of the work-room), Uta: a leech, Buja-lana: alligator or crocodile, Huti: worm of the earth, Jata: flame of fire, Weda: fire or heat confined in a vessel, as in a cooking vessel, Anala: the fire which giveth warmth to the heart, the fire of passion, the fire which giveth warmth to passion, Gni: fire (to a match), Utawa: fire, Kea: great fire, Lena: fire of flame of a lamp, Puyika: fire and ashes, mixed, Tiga: three, Uninga: a torch.
Korn: III Ricklefs: 3 Bahni pawaka siki guna dahana, bahni pawaka siking guna dahana trini rama uta ujwala nauli, trinira mauta bujal anauti jata weda nala gni ulawaha, jatha weda nala agni utawaka kaya lena puyika tiga uningan. kaya lena puyika tiga uninga
The poems have no poetical value or pretensions; they are obviously a kind of mnemonic verses. Comparison of Raffles' text with the other two shows that the former goes back to a similar poem which Raffles or his informant (possibly the aged Adipati Sura Adimenggala of Semarang, Ras 1992:305) must have had at his disposal. The sequence in which Raffles' words are listed per numeral is almost exactly the same as that displayed in the relevant stanzas of the poems, as becomes evident from a comparison of the examples given above. Raffles' list for 'one' is similar to the first stanza of R except for sasadara, which is absent, and ron or godong, which occurs instead of rani. There are more variants in Raffles' lists for the other numerals, but in the main the order is the same. Therefore, Raffles' list may actually be taken as another version of the two poems published by Kom and RickJefs. The other published lists are arranged differently: in Gericke they occur in an arbitrary order; Hageman gives them in the order of the Javanese alphabet; while Van Hien uses that of the European alphabet. In one case it can be shown that Gericke's data also must go back to a similar poem to those published by K and R. Gericke's list for 'six' (written in a column) starts with the words rasa, sadrasa, winayang, gana, and retu, which also happens to be the first line of the sixth stanza of K and, allowing for wisaya instead of winayang, of R. These are also the first words in Raffles' list of words for 'six', with the exception that in the latter, rasa is not the first
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 301 word but the very last; the first word here is marga, which does not occur in K and R. Apparently, Gericke did not take his list for 'six' from Raffles' list, but he, or his informant, also had a chronogram poem at his disposal; in most cases, however, he changed the sequence of the words. It is obvious from the above remarks that these words should not be studied in isolation, as they are presented in the lists by Van Hien and Hageman, but in their contexts in the poems, including Raffles' list. Ricklefs rightly remarks in his comments that the origin of these poems is probably to be sought in some OJ manual for poets (Ricklefs 1978:241), but does not go into this matter in more detail. A closer look at the two verses clearly reveals their Old Javanese origin. Out of a number of cases, only one particularly convincing example will be discussed here. One of the words for 'three', which we find in the lists of Raffles, Gericke, Hageman and Van Hien, and also in the two poems, is puyika. This is a word of unknown meaning in MJ, although it is glossed in a similar way in all the lists, namely as 'fire and ashes mixed' (Raffles), 'a fire smouldering in the ashes' (Gericke), and 'smouldering ash fire' (Hageman, Van Hien). In Van Hien's list, this word is written slightly differently, as pucika or puciyika. In the third stanza of both poems the word puyika occurs in the fourth line, and is preceded by the word Una, which means either 'careless' or 'dead' in Javanese. As is well known, the Javanese script is continuous, with no spaces between the words. This means that the text gives the reader no clue as to the correct division of words. If an earlier attempt to read and interpret a text proves unsatisfactory, another solution may be sought by dividing the words differently. In the present case, the alternative reading len apuy ika is grammatically and semantically completely clear, though only if we read this as an OJ phrase; it means 'and fire that'. The words apuy for 'fire' and ika for 'that, it' occur in OJ but not in MJ, whereas len for 'other, and' is found in both. It will be clear that the reading puyika, found in all the published instances, is in fact a non-word, which has only been thus transcribed out of ignorance of OJ. The glosses given for this word, originating with Raffles' and Gericke's informants, are mere fantasy, while the variants given by Van Hien are yet more non-words invented by himself. In principle, the glosses presented in the published lists should not as a rule be accepted without further proof. In many cases they were clearly provided by someone who did not know the real meaning of the words concerned. It must be assumed, however, that originally all the words used in the poems had a meaning which explained their use for a particular numeral. If we now look at the fourth line of other stanzas, it becomes clear that some of the words here are not meant as chronogram words in the same way as the others. The fourth line of Korn's first stanza reads: liyan . ulan yata (for ya ta) tunggal ika kabeh and also moon then that one that all
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 302 J. Noorduyn
In these words the (presumably OJ) author summarizes the purpose of the first stanza by saying 'and also moon, all these (words) may be used in the sense of one'. This means that the last words of the lines and of the stanzas are not meant to be chronogram words themselves, in contradiction with, for instance, Ricklefs' remark concerning the word kabeh being 'also given for 1' in K, but 'not normally used in modern Javanese chronograms' (Ricklefs 1978:243). In fact, this same word also occurs at the end of the fourth line in the second stanza of K. This line reads: liyan tangan suku elar ma ika kabeh and also hands feet wings two that all i.e., 'and hand, foot, wing, all these mean two'. Similarly, the final words of the third stanza are: ika ('that') tiga ('three' in OJ) uninga(n) ('know' or 'be known'), meaning 'know that these [words] mean three'. The last word of the fourth stanza is simply pat ('four'), that of the fifth stanza lima ('five'). The seventh stanza again has the explanation ya ('then') pitu ('seven') ika ('that'), 'these are [the words for] seven' in K, and ya pitu in R. The final words in the eighth stanza in K are: uluka (owl) ika kabeh, which probably should be read as wolu ('eight') ika ('those') kabeh ('all'). In R the final words of the ninth stanza are nawa sanga, both meaning 'nine', while the tenth stanza ends with sapuluh, 'ten'. The last word of the ninth stanza in K is uninga ('know'), and of the tenth stanza kengetakna ('to be remembered'). The second line of the eleventh stanza in K ends with the words yata (for ya ta 'then that') kengetana (kerjit as a secondary base for derivatives such as keqetakin and kinegitan is already found in early OJ, see Zoetmulder 1982 s.v. j'g^t), while the third and fourth lines of this stanza contain no chronogram words at all. These lines are rather an explanation and an exhortation to remember all the words that have been given earlier, viz. kawruhana ta denira sangengaji I yan sira arep wruha pasasakala, meaning 'you should know all the words if you wish to know how to use chronograms'. Such numerals occurring at the end of a stanza can, of course, be used in chronograms for the numbers they denote (as is, in fact, often done). They themselves are not chronogram words proper, however.
The published records of chronogram words are merely word lists, printed either in columns or in the stanzas of mnemonic poems. Such lists only explain that specific words should be used for particular numerals, but do not usually give the reasons why these particular words have been selected. Occasionally, however, such reasons have been stated. In other cases they can be inferred by analysing the semantic spectrum which the relevant words display. A straightforward example is provided by the first stanza of the above poems. When the 18 words contained in lines 1-3 of this stanza are
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 303 considered, the following semantic pattern emerges: 1. 'form' (once: rupa); 2. 'moon' (three times: candra, sasih, sasadara); 3. 'navel' (once: nabi); 4. 'earth' (six times: bumi, bu, darani, dara, siti, awani); 5. 'arse' (once or twice: medi, guda, the latter only in K); 6. 'tail' (once: iku); 7. 'man' (once: jalma); 8. 'one' (once: eka); 9. 'voice' (once: wak); 10. 'child' (once: suta, only in R). This makes a total of 17, not 18, because the words sasa dara have been taken as one word representing Skt s'as'adhara (literally 'bearer of hare marks' = 'moon'). It is not difficult to see that a number of these 9 (in K) or 10 (in R) semantic domains have in common the fact that they refer to an item which is unique in reality, or one which as a rule occurs only singly: so 'moon', 'navel', 'earth', 'arse', and 'tail' refer to 'one'. Together they cover 14 of the 17 words involved. If and why the other three or four meanings, viz. 'form', 'man', ('child', only in R), and 'voice', also belong to this group is not immediately clear; we shall return to some of them below. Two crucial aspects of the words discussed may be noted. One is that all of them but two, i.e. medi and iku, originate from Skt. The second is that the use of some of them in Javanese chronograms is at variance with what their Skt meanings suggest. A good example of this is Skt awani 'earth', which is used in the sense of 'courageous' or 'courage' in chronograms because of an OJ homonym meaning 'courageous'. The latter is an adjectival form derived from wani 'courage', which in MJ is the nominal, adjectival and verbal form. Subsequently, more Javanese synonyms of this were used, such as purun 'courageous' (for instance, Ricklefs 1978:88). In this way a new semantic domain was added to the chronogram words for 'one', on the basis of the fact that the Skt word happened to have a Javanese homonym. In such cases the original meaning of the Skt word was disregarded. Here one innovational route along which chronogram words might be semantically expanded is distinguishable, namely the switch to a homonym, which may be styled 'the homonymy shift'. In the cited example of wani > purun, a further innovational route was pursued through what may be styled 'synonymy extension'; this is in fact the most common route towards expansion and innovation of the system. Another example of a homonymy shift in the first stanza is provided by the word dara, which is Skt dhara, lit. 'bearing' (female), i.e. 'the earth'. In chronograms it is used in the Javanese sense of 'dove' (Ricklefs 1978:18-19; Gericke-Roorda 1901 1:568). According to Korn's informants (Korn 1922:669), this word means 'woman', cf. MJ rara or lara. An example of this in the second stanza is paksa (Skt paksa 'wing') for 'two', which in chronograms is used in its (Javanese) meaning 'force, compulsion' (e.g., Ricklefs 1978:125). Another example is the word retu in the sixth stanza, which is used in chronograms in its Javanese meaning 'uproar' (Ricklefs 1978:125), but originates from the Sanskrit word rtu 'season', on account of the six Indian seasons, Skr. fadrtu. A third innovational route may be styled 'the emended homonymy shift'.
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 304 J. Noorduyn
An example of this is provided by the word winayang in the first line of Ricklefs' sixth stanza. This word is an 'emendation' of winaya, which itself is a curious example of how Skt words could acquire a chronogram value in (Old) Javanese. The meaning of Skt windyaka is 'remover (of obstacles)', but the chronogram value of 'six' derives from $a4windya(ka), 'remover of the six obstacles', a name for Gana or Ganes"a in OJ (Zoetmulder 1982:1590). Through its 'emendation' into winayang it acquired the meaning of 'played upon the screen' (lit. 'used in the wayang', the shadow play theatre using leather puppets, although in MJ, -in- forms often occur as near-synonyms of monomorphemic words). Other examples of this innovational route are provided by buda roni for 'one', from Skt bhu and dharani, both meaning 'earth', but used in chronograms for 'naked' (cf. Jav. wuda, possessing the same meaning), and ron 'leaf respectively; and warna 'form, colour' for 'four', from warnawa for Skr. arnawa 'ocean' or, more probably, from Skr. warna '(the four) castes'. As is apparent from some of the above examples, the reason for the selection of a word or concept in several cases is furnished by Indian mythology. A good example is provided by the words for 'elephant' for the numeral 'eight'. These words occur in the eighth stanza, and are: kunjara, gajah, dipa (Skt dvlpa, lit. 'drinking twice'), asti (Skt hastin, lit. 'having a trunk'), kari (Skt karin, lit. the same meaning), liman, and dirada (Skt dvirada, lit. 'two-toothed'). These seven words all of them except liman derive from Sanskrit. The reason why they are used for the numeral 'eight' is that, in Indian belief, the earth is supported by the eight mythical elephants in the eight quarters of the sky, the astadiggaja. In this case, too, we find a homonymy shift. For another Skt word for 'elephant' is ndga, which, however, also means 'snake' in Skt. The word also came into use in this latter sense in Javanese chronograms, adding the semantic domain of 'snake, serpent, dragon' for the number 'eight'. Subsequently other words for 'snake' were introduced, such as Javanese ula and Skt sarpa (lit. 'creeping'), which are found in R and K respectively. A second homonymy shift took place when another Skt word for 'snake', bujaga (lit. 'moving crookedly'), came to be used for 'eight', which word in its alternative form bujangga in Java had acquired the meaning 'brahman', thus introducing another new semantic domain for 'eight', including.the word brdhmana itself. Similarly the use of words for 'body', such as sarira, tanu and murti, all of them from Skt, as chronogram words for 'eight' rests on the Indian concept of the eight parts of the body in prostration, when making an obeisance, namely the asfdrjga, which comprises the hands, breast, forehead, eyes, throat and the small of the back (see Zoetmulder 1982:145). The reason why words for 'fire' have the numerical value 'three' is provided by the Indian idea that there are three kinds of sacrificial fires, the tryagni. Words for 'horse' stand for the numeral 'seven' because in Indian mythology the sun is drawn by seven horses, the saptdSwa. Words for
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 305
'gate', 'door', 'hole', and 'cave' are used for 'nine', lastly, because there are nine 'orifices of the body (six in the head and three, including the navel, in the abdomen)', the navadvara. On the basis of the approaches and conclusions outlined above, it is possible to explain a large number of the words found in mnemonic poems. The stanza for 'five' in the above poems may be taken as an example. Here the first word, buta, is included because in Indian thought there are five natural elements, the pahcabhata (earth, water, fire, air, and ether). Skt bhata also means 'demon', however, as does Jav. buta, which word has acquired the sense of 'five' through a homonymy shift. The inclusion of the second word, Pandawa, is self-evident as in Indian and Javanese mythology it denotes the five sons of Pandii- The words indriya ('senses') and wisaya (Skt visaya 'fields of the senses') are included because there are five senses, pahcendriya, in the human body. Then there are two words meaning 'arrow', sara (Skt fara) and OJ warayang, since the Indian god of love is represented as having five arrows, pahcabdna. There are six words for 'wind' because of the pahcaprdna or pancabdyu, the five vital breaths (prdqa), or 'winds' (bdyu), in the human body. These words for wind are, firstly, maruta, Skt Manila, one of the names for the god of the wind; then pavana, samlrana, and bdyu, or rather pancabdyu (the five winds) itself, all three of them Sanskrit, and the Javanese word wisikan 'secret whispering' (for this latter see below, however). The sixth word is ambaramdrga, a Skt compound literally meaning 'moving through the air', from which K has preserved ambarana for ambara, and R bana 'arrow' and rnarga 'road'. The word buna has apparently been included by someone who knew that the semantic domain of 'arrow' was appropriate to the numeral 'five*. Here we have an example of an innovational route which may be styled 'emended synonymy shift'. Tata (often used in its Javanese sense of 'arrange', 'order') and gad 'course, action, movement', taken together, call to mind Skt tathdgata, the term pancatathdgata referring to the five Buddhas, a well-known concept in Javanese Buddhism (Zoetmulder 1982:1268). This again would be a case of emended homonymy shift. However, see also Jacquet's interpretation of gati below. There remain a few cases where the texts may be corrupt. One of these is yaksa 'demon' in R and (ya)ta in K, in the stanza containing words for 'five'. Here Raffles provides a clue, giving the word astra, which means 'arrow', in the corresponding place. From this word K has apparently preserved the syllable la and R the syllable ksa. In both versions the syllable ya belongs to the preceding word indriya. Thus yaksa is another example of emended homonymy shift. Another Indian numeral symbol which has not yet been commented upon (in the stanza containing words for 'two') is ama in R, which may be a 'hypercorrection' from yama influenced by the preceding i of dresihi; cf. yama[lo] in K. These words go back to Skt yama 'pair'. Finally, there is gulingan, translated as 'air circulating in a room' (Raffles), or 'wind that is in a bedroom' (Gericke), or 'wind in a closed
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 306 J. Noorduyn
compartment' (Hageman, Van Hien), which in Javanese is a verb meaning 'rolling', which remains unclear. One wonders whether this may be an early mistake for uningan, 'one should know', which often occurs in OJ explanatory notes; see also stanzas 3,6 (R) and 9 (K), which have uninga(n) with the same meaning in the final line. It is clear from the above that the chronogram words with their numerical value in many cases originally came from Sanskrit, and were expanded and innovated along several routes in Javanese practice and theory. The obvious question is: did the system itself have an Indian background?3 The answer to this question appears to have already been given long ago in an enlightening article published by the French orientalist E. Jacquet in the Journal Asiatique of 1835. After mentioning some earlier publications, by A.W. von Schlegel, W. Jones and others, on the Indian symbolic notation of numbers, which was also employed in Tibet, Jacquet discusses the system itself and gives a list of Skt words used for the numerals from 0 to 22, 24- 27, 32, 33, and 49, as well as of those used by the Tibetans and, in the second part of the article, those used by the Javanese according to the list published by Raffles. As becomes apparent from Jacquet's list, the semantic domains in general are similar to those found above in the Javanese chronogram verses. They are, briefly: 'zero' - 'emptiness', 'sky' (gagana, etc.); 1 - 'earth' (bhami, etc.), 'moon' (candra, etc.), 'form' (rupa, etc.); 2 - 'wing' (pak$a, etc.), 'eye' (netra, etc.), 'arm' (bhuja, etc.), 'pair, twin' (yama); 3 - 'fire' (yahni, etc.), 'Rdma' (because there are three mythological heroes called Rama), guna (the three qualities); 4 - veda (the four Indian scriptures entitled ' Vedas'), 'ocean' (qbdhi, etc.), 'world age' (yuga); 5 - 'arrow' (vana, etc.), 'breath' (prdna); 6 - 'part of the body' (arjga), 'flavour' (rasa), 'musical mode' (rdga), 'season' (rtu); 7 - 'sage' (muni, etc.), 'vowel' (svara), 'mountain' (parvata, etc.), 'horse' (aSva); 8 - 'elephant' (gaja, etc.), 'serpent' (n<3ga, etc.), 'good fortune' (maggala); 9 - 'numeral' (arjka), 'opening' (cidra), 'planet' (graha); 10 - 'region of the sky' (di$a)\ 11 - the eleven names of the god Rudra or Shiwa; 12 - 'sun' (surya, because of the twelve solar months), 'circle' (cakra, because of the zodiac). In his introduction, Jacquet discusses the reason why this system came into being. It served not only as a mnemonic device, as W. Jones believed, but first and foremost as a means of preventing all-too-easy mistakes in the writing of numbers and of ensuring the immutability of the figures through their fixed positions in the metrical schemes of the verses in which the
3 A more recent publication on the use of chronogram words in India and a list of such words is to be found in D.C. Sircar, Indian Epigraphy (1965), the section 'Numbers expressed by words', pp. 228 ff. The use of chronograms is also widespread in the Islamic world. They are different from those used in Java, however, as they are based not on words but on letters of the alphabet having a numerical value. For Persian poetry see, e.g., De Bruijn 1991:550-1 (personal communication Professor De Bruijn).
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 307 chronograms were used (Jacquet 1835:8). Jacquet refers to several Skt astronomy texts and includes a number of quotations from one of these, the Suryasiddhanta, containing words with numeral values (Jacquet 1835:34-40). He also adds a supplementary list of such Skt words (Jacquet 1835:40-42). In the second part of his article, Jacquet lists and discusses all the chronogram words mentioned by Raffles. In many cases he corrects Raffles' glosses by demonstrating the Skt (or occasionally Jav.) origin of the relevant word. For each number he also lists the words which he is unable to explain. The following notes on the successive numerals outline the most important of Jacquet's additional findings. 1. Jacquet explains medi as coming from Skt mediril 'earth'; suta 'child' from Skt silti 'birth', in the same way as janma 'birth' > Jav. 'human being', while disregarding Skt suta 'son' and not recognizing Skt bhu 'earth' and dhararfl, same meaning, in buda ron. 2. Jacquet explains hama [= yama (K), ama (R)] from Skt yama 'couple'; and sikara from Skt tikara 'armpit' (?), but does not recognize anamba [= anembah (R)] 'to pay respect with both hands'. 3. Pointing out that initial
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 308 J. Noorduyn
as a whole runs: trini rama hutabuj jwalana huti ('three Rama oblation-eater fire sacrifice'). Other words not recognized by Jacquet, because he did not know that Raffles' list came from the continuously written lines of a poem, are, for instance, jataweda, from Skt jatavedas 'fire god' (lit. 'knower of all beings'), which is given by Raffles as two words, jata and wida. 4. Jacquet explains wedang from Skt veda '(the four) Vedas'; warna (Raffles, K) from Skt varna 'caste'; and suci (Raffles) [supti (K), sutri (R)] as perhaps from Skt saci 'pyramid'. He does not, however, recognize wahana, from Skt vahana 'flowing, ship*. 5. Jacquet identifies bana and margana (Raffles; marga R) as Skt bana/vdna and margana 'arrow'; gati 'road' as Skt gati 'the directions and issue of the five vital vapours in the human body (prdna, apdna, uddna, samdna, and wydna)'; and wisikan (Raffles, R; Jav. 'whisper') as Skt vtfikha 'arrow' (see wisika in K); and perhaps tata Jav. 'arrange' from Skt tattva in the sense of bhata '(the five) elements'. He does not identify OJ warayang 'arrow' because the OJ language was still practically unknown in his lime. Line 3 of this stanza then must have originally read bana margana samirana warayang, comprising three words for 'arrow' and one (Skt satriirana) for 'wind', together coming to 12 syllables. 6. Jacquet explains gana as deriving from Skt gana in the sense of 'the six philosophical schools according to the Jainas'; and anggas from Skt arjga 'member', perhaps from sadagga, the six auxiliary works supplementing the Veda (Jacquet 1835:17). 7. Jacquet explains the word mandala in Raffles' list, instead of the second turangga in K and R, as coming from Skt mandate 'circle' in the sense of 'the circle of each of the seven dvlpa or continents'. Jacquet has failed to recognize aksa as standing for Skt aiva 'horse', or wiku for Skt bhik$u 'religious mendicant'. 8. Jacquet explains panagan as deriving from Skt pannaga 'serpent' (lit. 'moving by creeping'); samadya from Skt samdja (for sdmaja) 'elephant' (as Indra's mount); and manggala from Skt marjgala 'good luck', of which elephants are considered to be the special bearers in Indian belief; cf. asfamarjgala 'collection of eight lucky things'. 9. The first word, rudra (Raffles, R; indra in K), is explained by Jacquet as coming from Skt randhra 'opening', a word that is also mentioned in his supplementary list (Jacquet 1835:41). He is unable to explain several other words, among them gatra and wadana. Gatra presumbly comes from Skt gatra 'limb', but does not seem to have any connection with the numeral 'nine'. For wadana one could think of Skt vadana 'face', as a synonym of mukha, which derives its numerical value through the meaning 'opening' (Skt navamukha = navadvdra, 'the nine orifices of the body'). 10. Jacquet explains bwna (Raffles; boma in R) as standing for buwa, Skt bhuvas 'air, atmosphere', but does not recognize widik as deriving from Skt vidiS 'intermediate region of the sky'.
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 309
The conclusion seems justified that the development from the Indian to the Javanese system is marked in two ways. Firstly, while initially in Sanskrit each word was motivated as a chronogram word either by certain natural phenomena (such as the fact that a person has two eyes and two hands, or that there is only one moon) or by particular (Indian) cultural phenomena or beliefs (such as that the sun is drawn by seven horses), later on the system in Javanese was increasingly invaded, as a result of several homonymy shifts, by words belonging to semantic domains which bear no relation whatsoever to the numerical value acquired by these words. There has been a development from motivated to arbitrary relationship, in the sense which these words have acquired in linguistics and semiology since De Saussure's Cours de linguistique generate. Secondly, this development was increasingly marked by what has been termed 'localization'. One of the first scholars to apply this concept to Indian culture pervading Southeast Asia was the eminent historian O.W. Wolters, in his publication of 1982. Briefly put, he stated: 'I believe .... that Indian materials tended to be fractured and restated and therefore drained of their original significance by a process which I shall refer to as "localization" ' (Wolters 1982:52). The chronogram word system is a good example of such localization. This system was adapted to the local language, Javanese, from the beginning. To the Sanskrit words used in the system, new words from Javanese were added, and not only synonyms but also new semantic domains came to be introduced. Finally, the chronogram words largely lost their motivation and the system came to consist of a number of fixed Javanese words arbitrarily denoting some numeral.
This localization process can be illustrated in greater detail by comparing the chronogram words used in two Old Javanese historical texts, the Nagarakrtagama and the Pararaton. Both these texts have been repeatedly edited and translated and frequently studied, but so far the chronogram words occurring in them have never been discussed separately, let alone compared. The chronogram words in the Pararaton have been published according to their numerical value in a separate list (Brandes and Krom 1920:340-342). Those used in the Nagarakrtagama are not even available in such a list. Therefore a comparative list of Nagarakrtagama and Pararaton chronogram words is appended to this article (see below). The age difference of over 125 years between the Nagarakrtagama (composed in 1365; from here on referred to as Nag.) and the Pararaton (written after 1489; hereafter abbreviated as Par.) is reflected in several respects in the chronogram words found in the two texts. In this connection the following points deserve to be mentioned. 1. Though the great majority of chronogram words found in both texts are Sanskrit loanwords, both also include Javanese words. The number of the latter is, however, much smaller in the Nag. than in the Par., namely 4 out of 66 vs. 20 out of 75.
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 310 J.Noorduyn
2. The Nag. chronogram word system still reveals its Indian origin by the frequent use of words for numerals higher than 9, namely 10, 11, 12 and 14 (in 27 dates), whereas these are completely absent in the Par., as in later Javanese texts. It may be regarded as a reflection of the earlier value of 12 of the word for 'sun' that this word is never used for 'one' in later chronograms, as might have been expected. In reliefs, however, the numerical value of 'sun' does change from 'twelve' to 'one' in successive Saka centuries; so it denotes 'twelve' in a scene from Saka 1272 (Krom 1926:41) and 'one' in a relief from Saka 1372 (Van Erp 1939:44). 3. The two texts clearly reflect different stages in the development of the Javanese system in another respect as well. The system used in the Nag. is still much closer to the original Indian system than that used in the Par., 125 years later, as is apparent from the following facts. a. The few innovations found in the Nag. are cases of synonymy extension, the only clear instance of this being provided by the Javanese word angin 'wind' (5). A second example, sa ('one'), though not dubious, is exceptional because it is a Javanese numeral prefix. A third example, san, must possess the value 'nine' but is otherwise unknown as a Javanese word (Kern 1919:115). Kern's supposition that this word is an abbreviation of Javanese sanga 'nine' (1919:115) is not very plausible in view of the different nasal consonant. Another monosyllabic word, na, apparently stands for 'one', according to Kern's translation (1919:114), but no explanation of this is given by either Kern or Pigeaud (Pigeaud 1960 11:60). Perhaps it is an abbreviation of Skt ndsa 'nose'. However, this word is not found anywhere else as a chronogram word, even though the numerical value of 'one' for 'nose' is obvious. The case of masa for 'six' is more complicated, as this word represents Skt mdsa 'moon'. It is used here as Javanese masa 'season', however, with its numerical value going back to the six Indian seasons (Kern 1919:118). Another example of synonymy extension, this time in a Skt word, is provided by Skt tilaka ('zero'), which has this value because it is synonymous with windu 'nought'; it is used in the sense of 'sect mark on the forehead', however (this explanation is taken from Kern 1919:102). This particular example is interesting because the shift from one meaning of windu to another approaches a homonymy shift, and the result is a chronogram word with an unmotivated numerical value. An exceptional case of synonymy extension is represented by mukti 'release' or 'eight', which is explained by Kern (1919:124) as deriving from Skt vimoksa, possessing the same meaning, since Buddhists recognize eight vimoksa. Nandana, with the meaning 'nine', because there were nine Nanda princes in the mythological history of India (Hinzler and Schoterman 1979:483), is a case apart, because there is no connection with any other chronogram word.4
4 Teeuw points out that it is not necessary to assume that Nandana is an expanded form of Nanda, 'with the nominal suffix -na, which is not unusual in Archipelago Sanskrit' (Hinzler and Schotcrman 1979:483). A more likely word
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 311
b. The Par. chronogram words, in contrast to those in the Nag., include several cases of homonymy shift. These include examples that have already been cited above, such as Skt bhata 'element* > Jav. buta 'demon'; Skt janma 'birth' > idM.janma 'human being' > (via the synonymy route) Jav. wong 'person'; Jav. iku 'tail' > Jav. iku 'that', and others such as Skt Rama > Jav. rama 'father'. Skt awani 'earth' > Jav. wani 'courageous' is a case of the above-mentioned emended homonymy shift. The complicated case of emended homonymy shift represented by 'nahut 'bite' for 'three' also is already conspicuously present in several Par. chronograms, in the frequent combinations anahut wulan 'biting the moon' and anahut wong 'biting a man'. The emendation developed from bhuj j[w]alanahuti 'eating fire sacrifice', through bujal anauti or ujwala nauti > bujal anaut > anahut or nahut 'bite, grasp with mouth'. This emendation must have occurred in the pre-modem stage of development of the Javanese language, as MJ has saut : nyaut rather than sahut: nahut as in the Par. This shows that a text similar to the mnemonic poems of Ricklefs and Kom must have been in existence at the time of and used (and misinterpreted!) by the author of the Par. This in turn implies an unequivocal (and rather early!) terminus ante quern for the existence of mnemonic verse in Java! The case of lara (2) 'ill' in Par. (31,15) represents an error of representation, namely lara nahut instead of lar anahut. Similarly, in the oldest Par. colophon, one should read lar-apaksa (22) instead of lara-paksa (22), which is a misreading by Brandes of two • chronogram dates in his Par. edition, not corrected by Krom in the second edition (Brandes and Krom 1920:40, 341). In both these cases we have the OJ word lar for 'wing' as a chronogram word for 'two', followed by a verbal form with the prefix a, namely anahut, 'biting', and apaksa, 'resolved*. Of two other unmotivated cases, one, daging (1) 'flesh, prey', seems inexplicable. The other, kaya (3) 'as if, may derive from the fourth line of a chronogram stanza for 'three', containing a statement something to the effect of 'words such as [kaya] .... have the value three'. Another, and perhaps better assumption is that it reflects Skt. kaya, cf. trikaya, 'the three bodies, i.e. the three bodily actions whereby one practises the dharma' (Zoetmulder 1982:2038). Thus, the chronogram words used in the Nag. and Par. demonstrably reflect a progressive development from the earlier Skt/Indian system to subsequent stages, from the 14th century (Nag.), through the 15/16th century (Par.) and the 18th century (Ricklefs' poem), and the later system. Generally speaking, the direction of this development was towards an increase in the number of arbitrary, unmotivated chronogram words, resulting not only from synonymy extension but also, more particularly, from changes along the homonymy and emended homonymy routes. To what extent the same development is also discernible in the most recent,
separation at the beginning of the relevant verse is lhat resulting in gdni fdka yugddrinanda n agadeg, etc.
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 312 J. Noorduyn
20th-century, Javanese publication of chronogram words, in which at any rate the homonymy shift is an accepted part of the system (Bratakesawa 1956), must remain a subject for further study.
Comparative Table of Chronogram Words in the Nagarakrtagama and Pararalon
English Nagarakrtagama Pararalon 1. moon indu 5x = indu sasi sitangsu 7x Sasangka wulan (J) 7x ma[sa] 2x earth awani = awani bhumi = bhumi [a]ga dara ksaya mcdini ksiti naga lemah (J) birth jana (> jana) janma 2x wong (J) 16x form rapa = rupa4x navel nabhi god dewa nose na[sa] tail iku (J) > 'that' 5x beginning wit (J) flesh daging (J) concentration samadhi king ratu(J) one eka tunggal (J) 7x sa(J) words 16 18 meanings 8 10 Jav. 1 7
2. wing paksa 2x = paksa 14x
eye alcsi netra2x mata (J) sight drsti anon (J)
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 313
English Nagarakrtagama Pararaton arm/hand bhuja tangan (J) 2x ear karna pair yama foot pada two dwi 2x = dwi total words 7 8 meanings 7 6 Jav. 0 3
3. fire agni 2x = agni 9x anala bahni ba(J) api (J) quality guna = guna8x walking ngambah (J) Rama rama > 'fatto as if kaya (J) 12x weda weda bite anahut eel welut three tri4x tclu tnni antelu 'egg' total words 5 13 meanings 3 9 Jav. 0 4
4. ocean samudra sagara abdhi5x four on dice krta weda weda period yuga 3x four pat total words 3 4 meanings 2 4 Jav. 0 0
5. wind anila = anila
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 314 J. Noorduyn
English Nagaralcrtagama Pararaton agin (J) pawana bayu wihat arrow bana panah (J) Sara3x isu dice aksa senses artha wisaya 4x element bhuta>'demon' 3x five panca total words 6 9 meanings 3 5 Jav. 1 1
6. taste rasa3x = rasa3x crowd gana = gana3x season rtu masa (J) 2x member agga total words 5 2 meanings 4 2 Jav. 1 0
7. mountain parwata = parwata adri 3x naga 2x [a]ga date sage rsi 2x wiku horse kuda(J) turangga seven ssapt; a = sapta total words 6 6 meanings 3 4 Jav. 0 1
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 315
English Nagarakrtagama Pararaton 8. serpent naga = naga8x ula (J) 2x elephant gaja 2x matangga liman (J) brahman brahmana brahmana 3x body tanu 2x release . mukti eight asta2x total words 6 5 meanings 6 3 Jav. 0 2
9. mouth, face asya 2x muka 3x opening dwara leng (J) cattle go2x Nanda princes nanda(na, see note 4 above) nine nawa nawa san(J)=(?) sanga (J) 4x manawa (?) total words 6 5 meanings 5 3 Jav. 1 2
0. void sunya = Sunya 4x kha destroyed bhasmi sky wiyat ambara byoma gagana zero windu nora (J) sign lilaka total words 6 5 meanings 4 4 Jav. 0 1
10. division of space dcsa
Downloaded from Brill.com09/30/2021 08:11:50AM via free access 316 J. Noorduyn English Nagarakrtagama Pararaton total words 1 meanings 1 Jav. 0
11. Syiwa Rudra Sangkara 2x Sambhu Bhawa total words 4 meanings 1 Jav. 0
12. sun rawi 5x surya4x arkas4x ina4x aryama 3x aruna 2x [h]ari total words 7 meanings 1 Jav. 0
14. Manu Manu (the fourteen creators of the world, see Jacquet 1835:20) total words 1 meanings 1 Jav. 0 sum total words 66+13 (numbers >9) = 78 75 meanings 41 + 4 (numbers >9) = 44 46 Jav. 4 20
February - September 1993
Downloaded from Brill.com09/30/2021 08:11:50AM via free access Some Remarks on Javanese Chronogram Words 317
REFERENCES
Brandes, J.L.A., and N.J. Krom, 1920, Pararaton (Ken Arok) of net boek der koningen van Tumapel en van Majapahit, 's-Gravenhage: Nijhoff / Batavia: Albrecht. [Verhandelingen Bataviaasch Genootschap van Kunsten en Wetenschappen 62.] Bratakesawa, 1956^, Katrangan tjandrasengkala, Djakarta: Balai Pustaka. Bruijn, J.T.P. de, 1991, 'Chronograms', in: Encyclopaedia Iranica, edited by Ehsan Yarshater, V, pp. 550-1, Costa Mesa, California: Mazda. Erp, Th. van, 1939, 'Een merkwaardig Hindoejavaansch chronogram in beeld in het Rijksmuseum voor Volkenkunde te Leiden', Cultureel Indie 1:40-47. Gericke, J.F.C., 1836, 'lets over de Javaansche tijdrekening', Verhandelingen Bataviaasch Genootschap van Kunsten en Wetenschappen 16:65-80. [Signed in 1833.] Gericke, J.F.C., and T. Roorda, 1901, Javaansch-Nederlandsch hand- woordenboek, edited by A.C. Vreede and J.G.H. Gunning, Amsterdam: Miiller / Leiden: Brill. Hageman, J., 1852, Handleiding tot de kennis der geschiedenis, aardrijkskunde, fabelleer en tijdrekenkunde van Java, 2 delen, Batavia: Lange. Hien, H.A. van, [1933, 34, 35], Javaansche geestenwereld en de betrekking, die tusschen de geesten en de zinnelijke wereld bestaat, 3 delen, Batavia: Kolff. Hinzler, H.I.R., and J.A. Schoterman, 1979, 'A Preliminary Note on Two Recently Discovered MSS of the Nagarakrtagama', Bijdragen tot de Taal-, Land- en Volkenkunde 135:481-484. Jacquet, E., 1835, 'Mode d'expression symbolique des nombres employes par les Indiens, les Tibetains et les Javanais', Nouveau Journal Asiatique 16:5-41. Kern, H., 1919, Het Oud-Javaansche lofdichl Nagarakrtagama van Prapanca (1365 A.D.), VGravenhage: Nijhoff. Kom, V.E., 1922, 'Balische chronogrammen', De Indiscne Gids 44:667-674. Krom, N.J., 1926, 'Nota betreffende den staanden, gedateerden GaneSa in het Museum te Rotterdam', Jaarverslag 1926 van het Museum voor Land- en Volkenkunde en van het Maritiem Museum 'Prins Hendrik' te Rotterdam, pp. 19, 40-41. Pigeaud, Th.G.Th., 1960-1963, Java in the Nth Century, 5 vols. The Hague: Nijhoff. Raffles, T.S., 1817, The History of Java, London: Black, Parbury and Allen. Ras, J.J., 1992, The Shadow of the Ivory Tree; Language, Literature and History in Nusantara, Leiden: Vakgroep Talen en Culturen van Zuidoost-Azifi en Oceanie. Ricklefs, M.C., 1978, Modern Javanese Historical Tradition, London: School of Oriental and African Studies. Saussure, Ferdinand de, 1916, Cours de Linguislique Generale ... , Paris: Payot. Sircar, D.C., 1965, Indian Epigraphy. Delhi, etc.: Motilal Banarsidass. Wolters, O.W., 1982, History, Culture, and Region in Southeast Asian Perspectives, Singapore: Institute of Southeast Asian Studies. Zoetmulder, P.J., 1974, Kalangwan; A Survey of Old Javanese Literature, The Hague: Nijhoff. -, 1982, Old Javanese-English Dictionary, 's-Gravenhage: Nijhoff.
Downloaded from Brill.com09/30/2021 08:11:50AM via free access