Old Sundanese 101: Part IV — Killing Vowels

This post is the fourth in a series on the Old Sundanese language. For earlier posts in the series, see: · PART I · PART II · PART III ·

You may remember from the last post that the scripts used to write Old Sundanese are based on signs for syllables rather than signs for individual phonemes. These syllables typically end with the inherent vowel -a. With supplemental signs or diacritics, this inherent vowel can be changed and additional sounds can be added to the end of the syllable. Last time we met the diacritics for ⟨-i⟩ (the panghulu), ⟨-r⟩ (the panglayar), and ⟨-ng⟩ (the panyecek), as well as a number of aksara nglegena (‘naked aksaras’) and aksara swara (‘vowel aksaras’).

Here’s a table of the characters and their values to help refresh your memory (Figure 1):

Fig. 1 — A table of all the aksaras in the first line of “Bujangga Manik”.

This time we’re going to look at a different line in the same poem, a few new aksaras, and two ways to ‘kill’ an aksara’s inherent vowel.

Not every Sundanese syllable ends in a vowel (as we have already seen in the words saur ‘to say’ and sang, an honorific). Sometimes the inherent vowel of an aksara has to be done away with entirely — not changed from -a to -i or augmented by consonant diacritics but simply disappeared into the ether. Sometimes we want to go from ⟨na⟩ to ⟨n⟩.

There are two ways of doing this. They’re both represented in the following line from Bujangga Manik (Oxford, Bodleian Library, MS Jav. b.3. (R), f.2v), which I’ll go through in a little detail before touching on the topic of ‘killing’ vowels:

‘the plain of the region of Demak’

This is the 85th line in the poem, so we’ll call it ‘BM 85’ for short. Here it is in the manuscript:

Fig. 2 — “Palataran alas Demak”. Line 85 in “Bujangga Manik”. Oxford, Bodleian Library, MS Jav. b.3. (R), f.2v.

As in the last post, we’ll go over the script and grammar separately.


Note, first, that BM 85 is a complete line of eight syllables, and it is bookended by the octosyllabic line markers I pointed out in Part II, as you can see here (Figure 3):

Fig. 3 — The punctuation in BM 85.

You may also recognise some of the aksaras. We’ve already come across (from left-to-right) ⟨pa⟩, ⟨ta⟩, ⟨-r⟩, ⟨sa⟩, ⟨ma⟩, and ⟨ka⟩ (the latter mentioned only incidentally) (Figure 4):

Fig. 4 — Aksaras you may remember from the last post: ⟨pa⟩, ⟨ta⟩, ⟨-r⟩, ⟨sa⟩, ⟨ma⟩, ⟨ka⟩. You may also recognise the aksara ⟨da⟩ below the ⟨sa⟩. I’ll talk more about that below.

As in the last post, I’ve marked up the line to show the divisions between the aksaras. As before, it is not immediately obvious where one aksara ends and another begins if you don’t already know (Figure 5):

Fig. 5 — BM 85, marked up to show divisions between the characters (not including diacritics).

Below is a table of the aksaras in BM 85 that you won’t recognise (Figure 6). If you combine this with your knowledge of the other aksaras from Part III you ought to be able to make some sense of the line (although not entirely yet):

Fig. 6 — The remaining signs in BM 85.

As before, some are aksara nglegena (‘naked aksaras’ — ⟨la ra na⟩) and one is an aksara swara (⟨a⟩, an independent vowel preceded by an implicit glottal stop). The other two signs are diacritics (or, as they are called in Sundanese, rarangken).

The first one is called the paneleng, and it’s the sign for ⟨-e⟩ (pronounced [ə], like the ‘a’ in ‘about’ or the ‘o’ in ‘memory’). Old Sundanese had a simple system of six vowels, much like modern Malay/Indonesian (and unlike modern Sundanese, which has a seventh vowel, ⟨eu⟩). ⟨e⟩ should be distinguished from ⟨é⟩; we’ll get onto ⟨é⟩ another time, but it was probably pronounced somewhere in the [e~ɛ] range, as in modern Sundanese (like the ‘e’ in ‘egg’ or sort of like the ‘ay’ in ‘day’). Here is the entire vowel system of Old Sundanese as it appears in manuscripts and inscriptions (Figure 7):

Fig. 7 — Old Sundanese vowel chart.

The second tricky sign in BM 85 is called pamaéh ‘killer’ in Sundanese; it ‘kills’ the inherent vowel of the aksara it modifies. (In Javanese it’s called patèn, from pati ‘death’.) Here it looks a bit like a long Z, although it varies a bit within and between manuscripts; stick it immediately to the right of an aksara nglegena and the inherent vowel goes away. We have two examples of this sign in BM 85, at the ends of the words palataran (‘plain’ — but see below) and Demak (a place name). If the pamaéh were not present, these words would be read !palatarana and !demaka instead (Figure 8):

Fig. 8 — The two pamaéhs in BM 85.

There’s something else going on in this line. You may remember from the second post that aksaras are sometimes stacked on top of one another, and that this is done to create consonant clusters. Well, that’s what’s happening here (Figure 9):

Fig. 9 — The aksara ⟨da⟩ hanging down from aksara ⟨sa⟩.

When you put one aksara below another, the inherent vowel of the aksara on top (i.e. the first one in the combo) is ‘killed’. Any diacritics that surround this combination apply to the aksara on the bottom — which is referred to as an aksara pasangan ‘paired aksara’. In this case we have ⟨sa⟩ on top, ⟨da⟩ below, and a paneleng ⟨-e⟩ above the pair. This thus reads ⟨sde⟩.

Not all aksaras can be put in such a position, and some of those that can change their forms considerably. ⟨da⟩ is barely altered in this role — you may recognise it from the previous post. This consonant cluster does not have to be within a single word; in this example the s and the de are parts of different words, alas and Demak respectively, connected in both writing and reading aloud. (I should probably point out at this juncture that this text was almost certainly read aloud or intoned, probably with some sort of musical accompaniment.)

This means that there are two ways to kill the inherent vowel:

  1. attach the pamaéh to the right-hand side of the aksara, or
  2. stack (certain) aksaras on top of one another.

Anyway, here is a literal transliteration of BM 85 so you can see what’s going on overall:

palataran alas Demak
‘the plain of the region of Demak’

You might note that the r in palataran appears both as a panglayar on the ⟨ta⟩ and as a separate aksara ⟨ra⟩ (Figure 10). This happens fairly often in Old Sundanese texts — with ⟨-r⟩, ⟨-ng⟩, and ⟨-h⟩. Gemination/consonant lengthening like this was not (as far as we know) a feature of Old Sundanese phonology, and in normalised transliteration we don’t usually include both geminated consonants (!palatarran).

Fig. 10 — Two syllables: ⟨tar ra⟩, but read “tara”.

Grammar & Vocabulary

Last time I mentioned that octosyllabic lines in Old Sundanese poems tend to comprise whole sentences with subjects and predicates, and while that is often the case, a line can also be made up of a noun phrase, a string of onomatopoeias, serialised verbs, place names, personal names, epithets, or titles. In this case we have a noun phrase consisting of three nouns in a row.

There is no genitive in Old Sundanese — in fact there are no case endings at all. Adjectives follow the nouns they modify (not e.g. ‘red car’ but ‘car red’), and the names of things, people, or places come after other nouns to indicate possession. So our head noun here is palataran ‘plain(s)’, and this palataran is possessed by the alas ‘region’ of Demak (a place). Not so tricky. The referents of these words are more mysterious than the translation above might suggest, though.

There is currently no complete Old Sundanese dictionary. There have been some older attempts at glossaries, and even an Old Sundanese-Indonesian dictionary, but a lot of texts have been published and our understanding of other texts improved since these were put together. We can’t turn to a single authoritative source for the interpretation of Old Sundanese vocabulary (although one day I expect that will come about). There are therefore lots of problems in the interpretation of certain lexical items, and we depend to a large extent on dictionaries of modern Sundanese, Old Javanese, Malay, and other related languages to figure out the meanings of words in Old Sundanese texts.

In the case of the word alas, we infer its meaning (‘region, area’) from its use in Bujangga Manik. In Old Javanese the primary meaning of alas is ‘forest’. It doesn’t seem to mean ‘forest’ in Old Sundanese, though — at least not consistently. A couple of lines earlier there’s a reference to alas Jawa (BM 82), which is much more likely to mean ‘region of Java’ than ‘Javanese forest’. Alas probably meant ‘area’ or ‘region’, apparently of greatly varying extent — the alas of Demak is notably within the alas of Java.

There are similar problems with palataran. It’s probably related to the Old Javanese word natar (variant latar, ‘yard, grounds (in front or around a building, house, temple, etc)’) and to Malay pelataran ‘platform’. ‘Plain’ is a reasonable translation in this context, but the Dutch scholars Jacobus Noorduyn and Andries Teeuw, who worked on Bujangga Manik and published the first transliterated text of it in 2006 (after Noorduyn had died, incidentally), also connect latar to words in other Old Sundanese texts with the apparent meanings of ‘rampart’ or ‘fortification’. This is a significant difference. (‘Plain’ is nonetheless the word they opted for in their English translation of the poem.)

Fig. 11 — Demak on a map of modern Java. The town is less important now than it used to be, and it’s quite a bit further inland (due to progradation of the coastline, a common phenomenon in Java).

Demak (Figure 11) was in any case a pretty important place when Bujangga Manik was written in the late fifteenth century. In the 1470s an Islamic polity was established there (not yet using the title of ‘Sultan’ for its ruler — that seems to have come later), one that eventually separated itself entirely from Majapahit, the Javanese kingdom based near what is now Mojokerto in East Java. It is, in fact, this reference to Demak, alongside mentions of Majapahit and Melaka, that allows us to definitively date the text of Bujangga Manik — and you may remember from Part I of this series that it was an army from Demak that started the piecemeal conquest of Sunda in the early sixteenth century. A consequential place indeed.

In the next post we’ll go over the last remaining weird bit of the Old Sundanese script — a character not used in Old Javanese but quite important in the text we’re going through in this series.

A. J. West — Leiden, October 2021

Part I, introducing this series, can be found here.

Part II, on punctuation and the page, can be found here.

Part III, on reading a simple line of poetry, can be found here.

Posting about ancient and medieval Indonesia, up to ~1500 CE. Mainly into 14th & 15th century stuff, but earlier is fine too.