For Part IV, go here.
In the last post in this series, I said we would look at the first line of text in the Old Sundanese narrative poem Bujangga Manik, a lovely and interesting story written in the late fifteenth century. Here is that line (Figure 1):
As I mentioned last time, this says:
saur sang mahapa(n)dita
‘The great sage said’
In this post I’m going to go over this line — let’s call it ‘BM 1’ — in some detail, looking first at the script and then at the grammar of the line, including a light sprinkle of syntax and a dash of phonology. By the end of this post you’ll be able to fully interpret this piece of text from the original manuscript.
The letters of the Roman alphabet — or really any alphabet — come in two basic flavours: vowels and consonants. Alphabets are intended to replicate the phonemes of a language so that there’s a one-to-one correlation between phonemes (the smallest distinguishable units of sound in a language) and graphemes (the smallest distinguishable units of writing in a script). It doesn’t work that well in English, but that’s the idea.
It’s a little more complicated with the scripts used to write Old Sundanese. These scripts aren’t alphabets; they’re alphasyllabaries or abugidas, meaning that the main signs in the script stand for whole syllables rather than individual phonemes. In an alphabet, the syllable [ka] is written with two separate signs, ⟨k⟩ and ⟨a⟩. In alphasyllabaries like the ones used to write Old Sundanese, [ka] is written using only one sign (or aksara), ⟨ka⟩ (Figure 2).
Most of the principal signs in the Old Sundanese script that we’re looking at here are combinations of a consonant and an inherent vowel, a. These signs are called aksara nglegena (Javanese — ‘naked aksaras’) or aksara ngalagena (Sundanese). There are 18 of these signs in total, and they’re the backbone of the script. In the conventional order, they are:
ka ga nga ca ja ña ta da na pa ba ma ya ra la wa sa ha
(ng is pronounced like the ‘ng’ in ‘sing’; c is like the ‘ch’ in ‘church’; ñ is like the ‘ny’ in ‘Tanya’, or as in Spanish mañana; the rest are as in English.)
There are also five aksara swara ‘vowel aksaras’. These represent independent vowels (preceded, incidentally, by an implied glottal stop /ʔ/):
a é i o u
There are other kinds of aksara, but these are the ones we’re interested in today.
Now, here’s BM 1, the line in Bujangga Manik from above, marked up to show the divisions between these aksaras. The line includes both ‘naked’ and ‘vowel’ aksaras (Figure 3):
Notice that some of these aksaras have two or more components, and that the spaces between the aksaras are not obvious if you don’t already know where they should go. The text is written scriptio continua in the sense that there are no spaces or marks between words — but there is nonetheless some punctuation in the manuscript, as we saw last time.
In Figure 4 I’ve added the values of the aksara nglegena and aksara swara to the line. Note that the first and third aksaras are the same:
You’ll notice, of course, that sa u sa ma ha pa da ta does not spell ‘saur sang mahapandita’, which was how I said the line ought to be read. There’s more going on here. Specifically, there are these diacritics above the headline (Figure 5):
These diacritics (Sundanese: aksara rarangken; Javanese: aksara sandhangan ‘clothing letters’) serve to modify the syllable — changing the inherent vowel from a to i, for instance, or adding or removing sounds. There are eight of these in total. We’ll look at just these three this time.
The first diacritic in BM 1 adds an ⟨-r⟩ to the end of the syllable. It’s called panglayar in Sundanese. It floats above the aksara it modifies, and in isolation it looks like this (Figure 6):
In the line above, the panglayar is floating above the aksara swara ⟨u⟩ — so it’s not ⟨u⟩ but ⟨ur⟩.
The second sign is the panyecek (as it is known in modern Sundanese). It floats above the aksara that it modifies and adds ⟨-ng⟩ to the syllable coda, in a similar manner to the panglayar. In MS Jav. b.3. (R) it takes the form of a dot, sometimes with a little trailing line off to the right, much like the octosyllabic line marker we looked at last time (Figure 7):
So the third syllable is not ⟨sa⟩ but ⟨sang⟩.
The third and final diacritic in the line above is the panghulu. It also floats above the aksara, but it changes the inherent vowel of the aksara from ⟨-a⟩ to ⟨-i⟩. It looks like this (Figure 8):
In BM 1, the panghulu rests on the aksara ⟨da⟩, turning it into ⟨di⟩.
Here’s a table summarising the signs/graphemes found in BM 1 (Figure 9):
And here’s the line again (Figure 10):
saur sang mahapa(n)dita
You might notice that we don’t have an ⟨n⟩ in the word mahapa(n)dita; if you follow the guide above then BM 1 would read saur sang mahapadita. That’s because nasal consonants like ⟨n⟩, ⟨ng⟩, and ⟨m⟩ are sometimes absent from words in which they certainly belong (and an ⟨n⟩ certainly belongs in this word). They’re usually missing before other consonants; there are a couple of ways to make consonant clusters, but it seems scribes often preferred not to. This is a quirk of Old Sundanese orthography. Interpolated nasals are usually transliterated with brackets (n).
In the next post I’ll introduce some more aksaras. In fact, it might be a good idea to put up a table of all the characters in the script and go through a few sample lines so that you can start to transliterate and interpret the text on your own. Now, though, it’s time to look at the grammar of BM 1.
Old Sundanese is a predicate-initial language. For the most part. What that means is that, like in Irish and Classical Arabic and other such noble languages, the verb comes first in the sentence. This is not the case with modern Sundanese, and it isn’t always the case with Old Sundanese — but it is to some extent the standard word order. Lines in Old Sundanese poetry are typically complete (if simple) sentences comprising a subject and a verb or other predicate, although they can also be noun phrases or serial verb constructions.
In the case of BM 1, the line is a sentence:
saur sang mahapandita
‘the great sage said’
The verb here is saur ‘word; to say’. As you can see, it’s the first word in the line/sentence.
Old Sundanese verbs are not conjugated for tense or gender, and they are only occasionally conjugated for number (using an optional infix -ar-, which goes after the first consonant of the word when the subject of the verb is plural — we can look at this another time). There are plenty of affixes that can be applied to verbs to change their valency and to perform other functions, but this verb is good enough on its own.
There are two syllables in saur, incidentally; it’s [sa.ʔur], not [saʊr] or something like that. As I mentioned above, independent vowels/aksara swara are preceded by an implicit glottal stop [ʔ] — the middle consonant in stereotypical Cockney pronunciations of ‘button’ (‘bu’n’) — so in most cases they don’t form diphthongs.
The subject of this line is sang mahapandita ‘the great sage’.
Sang is a sort of honorific-slash-definite article; it precedes the names or titles of honoured people (or deities). It can’t be translated directly into English: ‘the’ is necessary for the syntax of the line in English translation anyway, and we’d translate it as ‘the great sage said’ even if the sang were not present. Old Sundanese also has an anti-honorific, by the way: si, the inverse of sang.
Mahapandita ‘great sage’ is a Sanskrit word. It probably entered Old Sundanese by way of Old Javanese, and in both Sanskrit and Old Javanese it has the form mahāpaṇḍita, with a long vowel in mahā ‘great’ and retroflex consonants in paṇḍita ‘scholar, pundit, learned sage, (etc.)’. (Retroflexes are the ones transliterated in the Roman alphabet with dots underneath — ṇ, ḍ, ṭ, etc. There are special aksaras for retroflexes in all the scripts used for writing Sanskrit and Old Javanese.)
Well, Old Sundanese doesn’t have any long vowels or retroflex consonants. They’re sometimes written in some Old Sundanese texts when the word in question comes from Sanskrit or Old Javanese, but they do not appear anywhere in MS Jav. b.3. (R), and were almost certainly absent from the phonology of spoken Old Sundanese.
So, to sum up:
Old Sundanese scripts are based on signs (‘aksaras’) that represent whole syllables. These aksaras may be modified by diacritics (‘rarangken’ or ‘sandhangan’). Here we’ve met the aksaras ⟨sa u ma ha pa da ta⟩ (and also incidentally ⟨ka⟩) and the rarangken ⟨-r⟩, ⟨-ng⟩, and ⟨-i⟩. Together, this allows us to read the line saur sang mahapandita, the very first line of Bujangga Manik — which, like many octosyllabic lines of Old Sundanese poetry, is a grammatical clause with a verb (saur ‘say’ — at the beginning) and a subject (sang mahapandita ‘the great sage’ — after the verb).
Next time we’ll look at some more aksaras and some more lines of poetry. Now that the basics are out of the way it should be a little simpler and a bit more fun. I hope you enjoyed reading this post and working through the line anyway.
A. J. West — Leiden, September 2021