AMA on Indonesian

[Creyeditor]

Edit: Table of Contents
Phonology
Segments: Plosives, Fricatives, Nasals, Liquids and Glides,
Suprasegemntals: Stress/Accent (I, II)
Morphosyntax
Nominalizations: phrasal ke--an
nominal syntax: adnominal possession
Voices: passives (I, II, III), applicatives and causatives
Questions: gap questions
Lexicon
Nouns: kinship terms, personal pronouns amd negation

So, I thought about making a thread about Indonesian varieties and I always feared that I was not competent enough, but I figured I won't get any more competent soon, so I could very well just start a thread now.

I am pretty confident about my mesolectal/acrolectal Papua Indonesian as spoken in the Province of Papua. I also know some colloquial Jakarta Indonesian, Standard Indonesian and Standard Indonesian the way it is actually used in urban areas of Java. This is a thread, where you can ask me anything about the Indonesian language and I will figure out, if I know the answer. Still, I also plan on regularily posting some smallish bits of Information. Let's start with the consonant inventory, because that's the way it's done

The next post will probably be on the consonant inventory, if noone has any other questions.

[elemtilas]

So, I thought about making a thread about Indonesian varieties and I always feared that I was not competent enough, but I figured I won't get any more competent soon, so I could very well just start a thread now.

I am pretty confident about my mesolectal/acrolectal Papua Indonesian as spoken in the Province of Papua. I also know some colloquial Jakarta Indonesian, Standard Indonesian and Standard Indonesian the way it is actually used in urban areas of Java. This is a thread, where you can ask me anything about the Indonesian language and I will figure out, if I know the answer. Still, I also plan on regularily posting some smallish bits of Information. Let's start with the consonant inventory, because that's the way it's done

The next post will probably be on the consonant inventory, if noone has any other questions.

Any good comparative works that deal with Indonesian and Philippine languages? For preference delving back into historical & ancient forms of the languages.

Any good grammar of Proto-Austronesian, for that matter?

[Creyeditor]

Indonesian and Philippine languages are both part of the Malayo-Polynesian languages. The closest I would get to a contrastive, historical description is probably the collection "the austronesian languages of asia and madagascar" by ‎Alexander Adelaar and ‎Nikolaus Himmelmann. It has a bit of everything and if your library can get their hand on it, you should definitely give it a read. It discusses all Austronesian languages, except the Oceanic languages. Most (older) literature on Austronesian in general is on Malayo-Polynesian, often with a heavy focus on Indonesian, Phillipine and Oceanic languages, so you might find something by searching for either term.

I don't know of a good reference grammar of Proto-Austronesian. As for vocabulary, I usually check the online database here. If you have any specific grammar topic, Robert Blust has probably published something on it. A lot of the older papers are found on Project MUSE now.

[elemtilas]

Indonesian and Philippine languages are both part of the Malayo-Polynesian languages. The closest I would get to a contrastive, historical description is probably the collection "the austronesian languages of asia and madagascar" by ‎Alexander Adelaar and ‎Nikolaus Himmelmann. It has a bit of everything and if your library can get their hand on it, you should definitely give it a read. It discusses all Austronesian languages, except the Oceanic languages. Most (older) literature on Austronesian in general is on Malayo-Polynesian, often with a heavy focus on Indonesian, Phillipine and Oceanic languages, so you might find something by searching for either term.

I don't know of a good reference grammar of Proto-Austronesian. As for vocabulary, I usually check the online database here. If you have any specific grammar topic, Robert Blust has probably published something on it. A lot of the older papers are found on Project MUSE now.

Cool, thanks! Salamat ha ALAM! Looks interesting.

May be a little outside your manor, but do you know any sources for old level Philippine languages? Again, would be looking mostly for a reference grammar type thing.

[Creyeditor]

I am afraid this is a bit outside of my expertise. I checked the glottolog and they do not seem to have any entry for old Philippine languages, and not a lot of literature on low level Proto-languages inside the Phillipine branches.

[Creyeditor]

Let's start small, since I won't be able to write up a post on the whole consonant inventory anytime soon: Plosives. Indonesian has 6 to 8, depending on the analysis:

/p b t d c~tʃ ɟ~dʒ k g/

Starting from the lips, [p] is in free variation with [f] for some speakers, since native Indonesian words do not have contrastive /f/. However, due to loanwords from English, Dutch and Arabic, /f/ has become pretty stable for most educated younger speakers. Nevertheless, I met both older educated and younger less educated speakers that still had the free variation. /b/ does not show such a variation.

/t/ and /d/ are coronal for all speakers, but the exact place of articulation varies. From my experience in Papua, it is basically free variation between apical denti-alveolar and apical alveolar. For Indonesian speakers on Java, the difference is sometimes more extreme, with the voiced plosive being further back than the voiceless plosive. This is part of what is called 'medok', having a thick Javanese accent.

The palatal plosives/postalveolar stops are a bit hard to analyse. To my mind, they sound like very ordinary SAE postalveolar affricates (minus the lip rounding), but I talked to a speaker of Hungarian, which has /c/ and /ɟ/ and he was convinced that they were palatal stops. Go figure. Also, these sounds cannot occur at the end of a root, an affix, or a word.

/k/ is realized as a glottal stop word finally. There is some variation though. Sometimes, this is more of a glottalized [k] and some other speakers even have an unreleased plosive in that position. There is also some weird interaction with reduplication, that linguists are not quite sure what to make of. /g/ does not show such a variation.

Voiced plosives cannot occur root, word or affix-finally. In general, plosives are unreleased word finally, and by unreleased I mean really, really unreleased. Sometimes they are almost inaudible. In Papua Indonesian they are also sometimes aspirated in word initial position, but this seems to be largely optional.

[Salmoneus]

I guess my main question with Indonesian is: what the hell is actually going on with passives? It seems as though every account I've seen has differed, and none have seemed complete.

What's the difference between the symmetrical and asymmetrical passives, and how does this interact with hierarchy and transitivity!?

[Creyeditor]

I guess I am going to split my answer up into to posts. So here is part 1: Why do accounts on Indonesian passives differ so much?

Standard Indonesian has a complicated voice system, that I will talk about in the next post. I suspect some spoken variety of Indonesian has the exact same system, but I have encountered very few speakers who come close to it. I suspect that the Standard Indonesian voice system 'leaks* into many spoken varieties of Indonesian to some extent. This might be a reason why actual descriptions that are based on speaker intuitions differ so wildly.

Anyway, just to show you that not all Indonesian varieties are strange, here is something about passives in Papua Indonesian and generic colloquial Indonesian of Java.

In Papua Indonesian, passive is mostly formed periphrastically with 'dapat' which means 'get', a structure similar to English 'get'-passive. It can be used with all persons and the agent is deleted alltogether. It can be reintroduced with the preposition 'sama' or 'deng' both meaning 'with'. This construction is related to a more marginal construction in Standard Indonesian which is formed with 'dapat di-V', where V is the verbal root. This constructions has a meaning similar to the English suffix -able/ible as in edible.

Sa dapat pukul sama mereka.
I get hit with they
I got beaten up by them.

Ini dapat di-maken.
this get PASS-eat
This is edible.

Spoken Indonesian on Java uses the di-passive that is also used in Standard Indonesian. In the spoken variety, however, there are no constraints on the person value of either agent or patient. The agent can only be reintroduced if it is preceded by the word 'sama' again meaning 'with'. More acrolectal varieties of Papua Indonesian basically use the same construction.

Aku di-pukul sama mereka.
I PASS-hit with they
I got beaten up by them.

Most Indonesian varieties have an additional adversative passive ke- -an, used when bad things happen to someone against ones will, and aninvoluntary passive with a prefix ter-, used when things happen on their own without any agent or against someones will. Both constructions are usually highly lexicalized and can also apply to intransitive verbs. Neither construction allows us to reintroduce the agent in the usual contexts.

Kita ke-hujan-an
we ADVER-rain-ADVER
We got rained on./We got surprised by the rain.

Pintu ter-tutup.
door INVOL-close
The door (is) closed.

The next part will be about the strange passive voice in Standard Indonesian.

[Creyeditor]

Okay, so now something on Passive in Standard Indonesian. Like I said, I guess in some Indonesian variety (maybe even some kind of Riau Malay?) this must be real, but I have not met many people who speak like this. I will try to remember everything as correctly as possible. The basics are simple: In Standard Indonesian, you cannot have a passive with an implied first or second person agent. Instead, something akin to topicalization is used.

So, Standard Indonesian has the di-Passive. You put a prefix on the verb, you can delete the Agent and reintroduce it with a special preposition 'oleh'. However, the use of the preposition is sometimes described as optional. This means, it's hard to tell if this is symetric or not. In general, I would say that the version without the preposition sounds a bit more likely to get interpreted as an attributive struture, but this is just my non-native intuition.

Anak di-gigit (oleh) anjing.
child PASS-bite by dog
The child (was) biten by the dog.

A note on word order. Even though, officially, the word order is described as SPOK, subject - predicate - object - adverbial (keterangan), it is pretty free in general in the spoken varieties. There is one context, where even the Standard Indonesian, as taught in schools allows for a variation on the verb order and this is when you try to reintroduce a first or second person agent (in some descriptions also a third person pronoun agent) into a di-passive sentence. Instead of SPO you get the reintroduced agent before the verb. Also, note that the verb cannot be prefixed with the 'active' voice marker 'meN-' in this context.

Mobil aku cuci.
car I wash
The car was washed by me.

I am really not so sure, if this can really be considered a passive. The 'active' voice marker is known to be absent in sentences with deviating word order anyway in Standard Indonesian, so you could just call this 'object fronting' or 'topicalization' if you want to. It probably makes sense to call this a voice for historical reasons, but synchronically it's not. I hope this clears it up a bit and does not create more confusion.

[Salmoneus]

Thank you!

So the ter- and ke-an passives aren't productive, but are lexicalised? Or is it just that the default choice is largely lexicalised, but the constructions are productive if you really want to defy the defaults?


Regarding the di-passive and the null-passive: you only mention first and second person with the null-passive; I've also read that the two constructions are both permissable with the third person provided that the argument is a pronoun?

I gather that the null-passive is considered a voice because it changes the subject: the subject (patient) of either passive can be relativised, while the agent cannot be? [and this is part of confusion whether the di-passive is symmetrical or not; the fact that using the di-passive at all is a form of agreement with the (superficially demoted) agent is also confusing!]

It's weird that historically it's the di-passive that was the symmetrical object voice, yet this has transferred to the null-passive in modern speech. And I seem to recall that in some Indonesian languages (not Malay), both passives are still marked on the verb - but maybe one of them is actually the equivalent of the ter-passive instead? I don't know.

[Creyeditor]

Thank you!

So the ter- and ke-an passives aren't productive, but are lexicalised? Or is it just that the default choice is largely lexicalised, but the constructions are productive if you really want to defy the defaults?

Ke- -an is more lexicalized than ter-, so the last question is a yes for ter- and a no for ke--an, I guess. The problem is that there is a homophonous nominalizer ke- -an and words derived by any of the two ke- -an's have sometimes rather opaque meanings.

Regarding the di-passive and the null-passive: you only mention first and second person with the null-passive; I've also read that the two constructions are both permissable with the third person provided that the argument is a pronoun?

I actually mention this in brackets, in some descriptions third person singular and plural pronouns are possible in both constructions. I have no further experiences here.

I gather that the null-passive is considered a voice because it changes the subject: the subject (patient) of either passive can be relativised, while the agent cannot be? [and this is part of confusion whether the di-passive is symmetrical or not; the fact that using the di-passive at all is a form of agreement with the (superficially demoted) agent is also confusing!]

Right, the so-called pivot vs. non-pivot idea. This might be true for Standard Indonesian, but it could equally well be a property of topicalization. Actually, some linguists used to equate promotion to pivot with topicalization. In spoken acrolectal Papua Indonesian at least the agent and patient of a di-passive can be relativized. I heard in the typology of Austronesian voice systems this is sometimes called post-Indonesian.

Anak di-pukul sama orang yang jahat sekali.
child PASS-beat with person REL evil very
The child got beaten up by a very evil person.

It's weird that historically it's the di-passive that was the symmetrical object voice, yet this has transferred to the null-passive in modern speech. And I seem to recall that in some Indonesian languages (not Malay), both passives are still marked on the verb - but maybe one of them is actually the equivalent of the ter-passive instead? I don't know.

The di-passive could be viewed as optionally symmetric, if you want to describe it that way, since the preposition introducing the agent can be dropped. Does that make sense? Still, I guess you could argue that the dropping of agents shows that this is not symmetric.

Well, it is marked compared to the active voice, because the prefix "meN-" is missing. In Balinese for example, (more symmetric) patient voice is similarly zero marked, whereas agent voice is marked with a cognate of "meN-", IINM. Maybe you were thinking of some Indonesian language that uses nasal mutation alone to mark the active voice and has lost the prefix part?

[Creyeditor]

Just another phonology post, I thought I'd continue with fricatives. Indonesian has two, or maybe three, or four, or five fricatives. It definitely has /s/ and /h/. The maximal inventory, including loan phonemes is the following.

/f v s z ʃ x h/

As mentioned before /f/ for many speakers is neutralized completely with /p/. Phonologically, it behaves very similar to /p/ mostly, e.g. in standard Indonesian you can sometimes see /f/ becoming /m/ in the morphophonological change triggered by the active voice/verbalizer prefix me-, e.g. foto /foto/ -> memoto /məmoto/, which is similar to pukul /pukul/ -> memukul /məmukul/. /f/ does not occur in native words.

/v/ is really rare, even in loanwords, and only used by the most educated speaker. Other speakers mostly neutralize it with /f/ (maybe due to Dutch influence) and some speakers therefore with /p/. It does not usually become /b/ though. It became /w/ in loanwords from Sanskrit, IINM. An example of a recent loanwords, where it becomes /f/ is TV /tifi/.

/s/ is common and has a lot of different realization. The exact pronunciation varies from speaker to speaker and even the same speaker might not have a fixed pronunciation. The most common realization is a slightly post-alveolar, laminal or apical sibilant, which I usually transcribe as [s̠]. There is some evidence that it is palatal/post-alveolar at least for morphophonological considerations. In the morphophonological change mentioned above it becomes a palatal nasal, e.g. senang /sənaŋ/ 'happy' becomes menyenangkan /məɲənaŋkan/ 'be happy'. On the other hand, in contrast to other palatals, it can occur at the end of roots in native vocabular, e.g. bagus /bagus/ 'very good'.

/ʃ/ is only found in loanwords and varies freely with /sj/ for most speakers. Educated speakers will use /ʃ/, especially when using religious terms. It does not occur in native words, but is very frequent anyways, e.g. bersyukur /bərsjukur/~/bərʃukur/ 'to be grateful'

/x/ has a similar distribution to /ʃ/ in that it occur mostly in loanwords, in this case especially from Arabic and Dutch. The Arabic cultural and religious loanwords are again the context in which you will find the 'proper' pronunciation most often. It is neutralized to /k/ or /h/ in other contexts, depending on the speaker and the word it occurs in. Example akhir /axir/~/ahir/ 'end, last' vs. khotbah /kotbah/~/xotbah/ `to preach, sermon'

/h/ is a very normal fricative. It can occur in the coda of a syllable, where I feel it partially devoices the preceding vowel. After front vowels and in some interjections, it is often as pronounced as a kind of weak [ç], e.g. bersih /bərsih/ [bərsiç] 'clean'. In Papua Indonesian it is sometimes deleted when it occurs as a coda consonant, e.g. Standard Indonesian basah /basah/ wet'' vs. Papua Indonesian /basa/.

[Creyeditor]

Nasals are maybe a good topic for a short post. They can occur underlyingly or derived from plosives if prefixed by the active verb marker. Nasals occur at four places of articulation.

/m n ɲ ŋ/ m n ny ng

/m/ is pretty boring. Nothing special about it. It can be derived from /f/ and /p/ in active verbs. It can occur in any position. It is very frequent because it occurs in several affixes.

I am not sure about the specific place of articulation of/n/ , but I would guess its mostly apical denti-alveolar or something similar. /n/ can also be derived from /t/ in active verbs. It is also the result of word-final neutralization between /ŋ/ and /n/ in some eastern dialects, e.g. kurang /kuraŋ/ [kuran] 'less'.

/ɲ/ can be derived from /tʃ/ and /s/. The latter might be a bit surprising, but as I said earlier, /s/ is actually slightly retracted, so it might make sense. /ɲ/ cannot occur root-finally or affix-finally or word-finally or syllable-finally. Example: cuci /tʃutʃi/ 'wash, clean' menyuci /məɲutʃi/ ACTIVE VOICE.

/ŋ/ can be derived from /k/ in active verbs. It can occur in any position. In some dialects word-final neutralization results in /ŋ/ instead of the above mentioned /n/, e.g. ukuran /ukuran/ [ukuraŋ] 'size'

I will probably have a whole post about what happens in active verb forms. One additional important tidbit. People have said that Indonesian has nasal place assimilation, but this is not true inside roots, i.e. there are some frequent exceptions such as tanpa /tanpa/ 'without'

[All4Ɇn]

Really interesting thread so far. Don't have any questions yet but I'm looking forward to seeing more

[Creyeditor]

So, adnominal possession is interesting, because there is a lot of dialectal variation. In Papua Indonesian, possessors occur before the head noun and are followed by the particle pu. pu is derived from the verb mempunyai meaning to have. I guess this was a reanalysis with the prefix men- being the verb marker and the possor marking clitic =nya. Anyway, this kind of structure is really common in Papua Indonesian.

anak pu mainan
child POSS toy
a/the toy of a/the child

In other Indonesian varieties, possessor follow their head noun. Often, the two nouns are simply juxtaposed.

mainan anak
toy child
childrens' toys

In these varieties, the possessed noun can also be marked with the third person possessor clitic =nya. Some people have suggested that this is actually a definiteness marker, but I remain somewhat skeptical, since the marker cannot just occur on any definite noun. Nevertheless, the possessed noun in these structures is usually understood to be definite.

mainannya anak
mainan=nya anak
toy child
the toy of a/the child

[Tsugar]

In Indonesian, there are the mem- and di- prefixes that are related to Austronesian alignment. However, for the purposes of nominative-accusative-language-speakers, they are often referred to as active and passive in language-learning material. How do they differ in practice?

[Creyeditor]

Sorry, I somehow missed your post. As explained above di- acts very much like a passive voice in many Indonesian varieties in that (a) there is a special marker on the verb (b) the former object is promoted to subject position (c) the former subject becomes optional and can be reintroduced by a preposition. You can almost translate the morphemes 1-to-1 from English.

Apel di-makan oleh kudah.
apple PASS-eat by horse
Apples are eaten by horses.

di- differs from a more prototypical passive in three ways (a) in some varieties (including Standard Imdonesian) the passive voice is restricted to third person or full nouns (b) in some varieties the preposition can be dropped and the agent will thus merely shifts its linear position but might stay a full argument (c) the active form of the verb is not morphologically unmarked.

The actual use of meN- is much more complicated. It shows up in some active voice contexts, but is absent in others. Sometimes it acts as a verbalizer, even for loanwords. Some varieties (including Papua Indonesian) only use it as a verbalizer and in some lexicalized forms. For Standard Indonesian I have seen several proposals. One stated that you only use meN- in main clauses, but there are some exceptions. Another one assumed syntactic movement to delineate the contexts in which meN is unexpectedly blocked.

[Creyeditor]

I guess some sounds are still missing from our consonant inventory: non-nasal sonorants:

/r l j w/

This is a pretty standard inventory in terms of crosslinguistic tendencies. Phonetically, these sounds are nothing special, but phonologically they are a mixed bag.

/r/ is usually articulated as an apical alveolar trill or (more frequently) tap. This sound does not undergo nasal mutation at all, e.g. me-racuni 'to poison'. In many varieties it can also be syllabic. i.e. a sequence of schwa+/r/+another conosonant is realized as syllabic /r/+another consonant, e.g. <berbagi> /bər.ba.gi/ [br̩.ba.gi] 'to share'. Some varieties (including some basolectal Papua Indonesian ones) realize this syllabic version as a full vowel [a], e.g. <terlingkar> ' [tərliŋkar]~[taliŋkar]
encircled, surrounded, entangled'

/l/ is similar to the alveolar rhotic in that it does not undergo nasal mutation, e.g. me-lihat 'to see'. It differs in that it cannot be syllabic. It is phonetically realized as a dental or alveolar lateral. In some varieties it can become velarized in certain positions.

/j/ is written as <y> in the standard orthography. It is not very common as a simple onset, but it sometimes occurs to resolve vowel hiatus, e.g. <bayi> /ba.i/ [ba.ji] 'baby'. It does not undergo nasal mutation, which is unexpected from an orthographic perspective because the palatal nasal is written <ny>, e.g. me-yakin-kan 'to ensure'.

/w/ is similar to /j/ in that it does not undergo nasal mutation (e.g. me-wajib-kan 'to make obligatory') and occurs more frequently between vowels. It is slightly more common root-initially, mainly due to loan words, e.g. waktu 'time' from Arabic.

[Creyeditor]

I recently thought about kinship terms. IINM, it is very dependent on the variety. Therefore, I will focus on kinship terms in Papua Indonesian. Papua Indonesian kinship terms are like if you boarded a flight to Hawaii and met a couple of Iroquois people on the plane and you all had a party on arrival.

First, the Hawaiian part. All your cousins and siblings are either kakak older sibling/cousin or adek younger sibling/cousin. If you want to specify that you only talk about your parents' childrens you can soecify this as kakak kandung or adek kandung. Note that kakak/adek is also used in compounds in a wider sense that includes friends and the like. It is also a common form of address. Sometimes the Sanskrit loans saudara 'brother, male cousin' is also used to refer to distant relatives lime cousins. The female form saudari is less common, women are sometimes also refered to as saudara. Note that the Sanskrit forms are also used as a forn of address in some political contexts.

Similarly, your mothers sister and your mother are all referred to as mama. And your father and his brother are all referred to as bapak. But here's the catch and this is where things start getting Iroquois. Your mother's brother is your om and your father's sister is your tante. Both terms are derived from Dutch and cannot be specified any further. Mama and bapak can however indicate relative age. Your mother's younger sister is your mama adek or madek for short. Similarly, your mother's older sister is your mama tua. Bapak adek ~ padek and bapak tua are used in the same way for your father's brother. Note that om and tante are also used by children and young adults to adress people from their parents generation that they don't know very closely. Madek and the like cannot be used in that function. Mama and bapak are common terms for adult men and women respectively and also used as a term of adress similar to Emglish Madam/Sir/Mister/Miss(es). For women, the Standard Indonesian form ibu is sometimes used, for men the forms are identical anyway.

Children are generally referred to as anak, independent of gender and age. Your siblings' children (i.e. nieces and nephews) are also referred to as anak. Again, biological parenthood can be indicated by adding kandung, i.e. anak kandung. In many regions, additional terns are loaned from local languages that specify more information, e.g. on gender and birth order. I also heard the term keponakan used for more distantly related nieces and nephews. However, this term is also used for distantly related cousins.

Your grandparents are your nenek 'grandmother' and your kakek 'grandfather'. More basilectal forms also use the form tete 'grandfather', which is also used by children as a term of address for the christian god. Since traditional polygamy was still more common a few generations ago, I also learned the terms for your grandfather's wives that are not your grandmother. They are dependent on age/order if marriage: nenek mudah 'younger wive of your grandfather' and nenek tua 'older wive of your grandfather'. Literally they can be translated as 'young grandmother' and 'old grandmother'. Your grandchildren are referred to as cucu, again independent of age and gender.

Step-parents are generally stigmatized. The terminology can be derived by adding tiri after the terms for parents, children and siblings. They are not usually added to more complex kinship terms. Adoption is common and not usually indicated in speech, except by contrasting it against kandung 'biological'. There is also a looser form of adoption, where children still keep their biological parents but go and live with some other family who pays for their education. This is common, especially inside larger 'clan'-like structures, and children are often terminologically integrated into their host family as well as their biological family.

Something I just learned recently puzzled me a bit. Your cousin's grandchildren are your cucu and you are their nenek/kakek. I guess it makes sense from some perspective, but it was surprising for me at least.

[eldin raigmore]

That “looser form of adoption” sounds like it could be fostering. Nicht wahr?