Passive and anti-passive

The most obvious example of valency-changing morphology in English is the passive voice. Example (3) shows a pair of active and passive sentences in English:

In the active sentence, the verb has two arguments, its agent (the one who does the action) and the patient or theme (what gets affected or moved by the action); the agent functions as the subject of the action, and the patient as the object. In the passive sentence, an agent is unnecessary. If it occurs, it appears in a prepositional phrase with the preposition by. The patient is the subject of the passive sentence. In effect there is no longer any object, and the passive form of the verb therefore has one fewer argument than the active form.

Part of what signals the passive voice in English is passive morphology on the verb. English passives are formed with the auxiliary verb be and a past participle, which is signaled for regular verbs by adding -ed to the verb base. Irregular verbs can form the past participle in a number of ways: by adding -en (write ~ written), by internal vowel change (sing ~ sung), or by internal vowel change and addition of -t (keep ~ kept).

Other languages also have morphological means to signal the change in argument structure in passive sentences. Example (4) shows an active and a passive sentence from West Greenlandic, and (5) an active–passive pair from Maori:

In (4b), you can see that the morpheme niqar makes a verb passive in West Greenlandic, and (5b) shows that a verb in Maori can be made passive by adding the suffix -tia. As was the case in English, the addition of these morphemes goes along with passive syntax, that is, making the patient/theme into the subject of the sentence, and making the agent optional. If the agent appears, it is marked with the ablative case in West Greenlandic, and by the preposition e in Maori.

Passive sentences are relatively familiar to speakers of English, but English has nothing like what is called the anti-passive. Like the passive, the anti-passive takes a transitive verb and makes it intransitive by reducing the number of its arguments. What’s different, though, is which argument gets eliminated.

For the passive, it’s the transitive subject that disappears (or is relegated to a prepositional phrase or a case form other than that typical for subject), whereas for the anti-passive, it’s the transitive object that disappears, as the example in (6) from Yidiɲ shows:

In Yidiɲ, the anti-passive is marked on the verb by adding the suffix -:ᶁiŋ. Since Yidiɲ is an ergative case-marking language , the subject of a transitive verb is in the ergative case. The subject of an intransitive verb is in the absolutive case, as you see in example (6). So while ‘eat’ is normally transitive in Yidiɲ, you can see that it has become intransitive here.

As with the passive, it is also possible to express the ‘missing’ argument overtly, but in a case form other than that usually used for the direct object. Whereas in an active sentence, the direct object of a transitive verb is marked with the absolutive case, in an anti-passive sentence, the subject is absolutive and the object, if it appears, is either in the dative or the locative case:

While the translation of (7) makes it look like this sentence means exactly the same thing as an active sentence, this is only because there is no real way of capturing the nuance of this sentence in English, a language that lacks the anti-passive.