Austronesian alignment, commonly known as the Philippine- or Austronesian-type voice system, is a typologically unusual morphosyntactic alignment that combines features of ergative and accusative languages. It is best known from the languages of the Philippines, but is also found in Taiwan's Formosan languages, as well as in Borneo, Northern Sulawesi, Madagascar, and Guam, and has been reconstructed for the ancestral Proto-Austronesian language. (Only traces of this system remain in other Austronesian languages, such as Malay and Old Javanese.)
Whereas most languages have two voices which are used to track referents in discourse, a transitive 'active' voice and an intransitive 'passive' or 'antipassive' voice, prototypical Philippine languages have two voices which are both transitive. One of the two Philippine voices is similar in form to the active voice of ergative–absolutive languages, while the other is similar to the active voice of nominative–accusative languages. These perform functions similar to the active and passive/antipassive voices, respectively, in those languages.
The ergative-like Philippine voice has in the past often been called the "passive", and the accusative-like voice has often been called the "active". However, this terminology is misleading and is now disfavored—not least because the "passive" is the default voice in Austronesian languages whereas a true passive is a secondary voice—though no substitute terms have been widely accepted. Among the more common terms that have been proposed for these voices are patient trigger (the ergative-like voice) and agent trigger (the accusative-like voice), which will be used here. These phrases are taken from the terms 'agent' and 'patient', used in semantics for the acting and acted-upon participants in a transitive clause.
The three types of voice system and the grammatical cases of their core arguments can be contrasted as follows:
|Morphological alignment||Case of basic intransitive clause||Cases of basic transitive clause||Cases of the secondary voice|
(as most European languages)
(same case as Agent)
|Active voice||Passive voice|
|nominative (Agent)||nominative (Patient)|
(as most Australian languages)
(same case as Patient)
|Active voice||Antipassive voice|
|absolutive (Patient)||absolutive (Agent)|
(as most Philippine languages)
(the case common to the two transitive voices)
|Patient trigger||Agent trigger|
|"direct" (Patient)||"direct" (Agent)|
|ergative (Agent)||accusative (Patient)|
The Philippine cases are only approximately equivalent to their namesakes in other languages, and are therefore placed in quotes. ("Direct" as used here is commonly called "nominative" or "absolutive", for example.) The "ergative" case is identical in form to the Philippine genitive case, but it is common in ergative languages for the ergative case to have the form of an oblique case such as a genitive or locative.
Lynch et al. 2002 (p. 59) illustrate the Philippine system with reconstructed Proto-Malayo-Polynesian examples. (The asterisks indicate a reconstruction.) The unmarked clause order was to have the verb first and the "direct" phrase last. The voice was indicated by an affix to the verb (suffix -ən for patient trigger and infix ⟨um⟩ for agent trigger). In modern Philippine languages, the practical effect of this voice distinction is rather like the difference between the use of a and the in English, and it is assumed that it played a similar role in the protolanguage.
*ka’ən-ən na manuk a wai eat-(patient trigger) (ergative) chicken (direct) mango
- 'The chicken is eating the mango', or 'The mango is being eaten by the chicken'
*k⟨um⟩a’ən ta wai a manuk ⟨(agent trigger)⟩eat (accusative) mango (direct) chicken
- 'The chicken is eating a mango.'
Some scholars maintain that Philippine-type languages have four voices, rather than two. Beside the ones shown above, there were also locative and benefactive voices. However, these are not as central as the other two. The locative is illustrated here; the suffix on the verb indicates that the noun marked by the direct case is the location of the action rather than a participant:
*ka’ən-an na manuk a kahiw eat-(location trigger) (ergative) chicken (direct) tree
- 'The chicken is eating in the tree', or 'The tree is being eaten in by the chicken'
A broadly similar system is found in Tagalog, the most thoroughly documented language of this type. In Tagalog, the ergative and accusative have been conflated into an "indirect" case, in contrast to the direct case. (Ng is an abbreviation of the indirect-case particle for common nouns, nang.) Note that the root of the Tagalog verb is basa "to read."
b⟨in⟩asa ng tao ang aklat. ⟨(past:patient trigger)⟩read (indirect) person (direct) book
- The book was read by a person.
b⟨um⟩asa ng aklat ang tao. ⟨(past:agent trigger)⟩read (indirect) book (direct) person
- The person read a book.
- One is that Tagalog focus is voice. The following voices are then posited for Tagalog:
- Active voice
- Passive voice (AKA direct passive)
- Local voice
- Instrumental/benefactive voice
- Another is that Tagalog focus is case-marking. For example, ang is used when the prepositional phrase is in focus, while sa is used when it is not in focus. In the example given below, note that the root of the Tagalog verb is bilí, which means "to buy."
b⟨in⟩il-hán ng tao ng aklat ang tindahan. ⟨(past:patient trigger)⟩buy-(locative suffix) (indirect) person (indirect) book (direct) store
- The book was bought by the person at the store. (Store is the focus.)
b⟨um⟩ilí ang tao ng aklat sa tindahan. ⟨(past:agent trigger)⟩buy (direct) person (indirect) book (preposition) store
- The person bought the book at the store. (Person is the focus.)
- "Austronesian languages." Encyclopædia Britannica. Encyclopædia Britannica Online. Encyclopædia Britannica, 2010. Web. Oct. 2010. <http://original.search.eb.com/eb/article-75212>.
- Lynch et al. 2002. In Fay Wouk & Malcolm Ross, eds., The history and typology of western Austronesian voice systems. The Australian National University. ISBN 0-85883-477-4, ISBN 978-0-85883-477-4
- Lynch, John, Malcolm Ross and Terry Crowley. The Oceanic languages. Richmond, Surrey: Curzon Press, 2002.