In linguistics, actant is the general term for one of the participants in a sentence, such as the subject or object. All languages grammatically distinguish subject from object, and many do so by marking them with a different case.

In the English sentence The servant killed the master, there are two actants, the servant and the master. We can tell who killed who by the syntax, the order of the participants. The subject comes first and the object follows the verb. Reversing them gives The master killed the servant, with different meaning. In Latin, case marking of the two actants is obligatory, and the word order is less important: the two sentences are Servus dominum necavit and Dominus servum necavit.

The object in the above sentences is also called the direct object. The indirect object or recipient is another actant. In English there are two ways of showing that the master is the recipient of the action: The servant gave the apple to the master, or The servant gave the master the apple. In the first way a word ('to') is used as a marker, in the second way it is purely positional. In Latin a different case ending is used, the dative: Servus domino malum dedit.

Other actants include goal, instrument, and beneficiary. Languages don't necessarily distinguish all of these. The Latin dative case covers the recipient and the beneficiary. The English preposition 'to' covers both recipient and goal.

Actants are roles or slots. Whatever fits in that slot in the syntax is the same actant. So nouns and pronouns are equivalent: in He killed the master, 'he' is the subject. Grammatically, a pronoun is often treated differently from a noun, even when they are in the same actant roles. So in English, we say The servant killed him, where 'he' is this time the object, but has a case-marked form him. In English all the personal pronouns except you and it have different subject and object forms, but no noun has.

Consider these two sentences, meaning essentially the same thing, one being the transformation of the other into the passive voice:

The servant  was killing  the cats.
The cats     were killed  by the servant.
The subject of the first is the servant, the subject of the second is the cats. They have the same actant role. This is a syntactic matter: they are at the head of the sentence, and govern the verb, which is singular or plural (was/were) according to whether the subject is.

The first sentence also has a direct object the cats. The second sentence has none, but its other actant is now a different role (marked by the preposition by). The transformation of active to passive involves a change in actants, with the old object being promoted to subject, and the old subject demoted to a minor role. It is technically called a chômeur (unemployed person), and can be omitted: The cats were killed.

The role marked by by is conventionally called the agent, but I don't like this. The terms agent and patient are semantic terms, relating to the meaning of the sentence and what actually happens, as contrasted with the syntactic terms describing actants. If The servant killed the master then the servant is agent and the master is patient regardless of whether you say it that way or The master was killed by the servant. A grammatical description of the language will show how semantic roles are mapped to syntactic roles.

Typology is the classification of languages according to grammatical features, and actant marking is now one of the principal features of interest in typology. Languages typically use a mixture of syntax (word order), case marking (endings on nouns and pronouns), and verb marking (concord or agreement).

In English verb concord consists of the -s after a third person singular subject ('she walks'), and more forms of the verb 'to be': I am, you are, she is. Many languages, like Latin, have a fully battery of endings for each person and number, and often therefore do not have an overt subject in front of the verb: neco 'I kill', necas 'you kill'. In such cases it may be argued that the ending -o, -as is itself the actant, and not merely a marker of agreement. In Philippine languages, there are four grammatically distinct actant roles, subject, object, recipient, and location, and one of those four must be chosen as topic: then the verb has one of four different forms depending on which actant is topic.

Both English and Latin are examples of accusative languages, where a one-actant sentence has that actant marked the same way as the subject of a two-actant sentence. Servus ridet 'the servant laughs'. There is no space here to go into ergative languages (such as Basque), but they mark the single actant like an object.

I have mentioned that there may be a difference in treatment between pronouns and nouns, where they fill the same actant roles. Other differences may occur. In Georgian there is a difference between tenses: some tenses mark actants according to an accusative system, other tenses in an ergative way. Some languages mark actants differently according to how animate they are, and there tends to be a so-called Silverstein hierarchy from you and me, down to her and him, down to it, down to animals and finally inanimate nouns.

Terms with much the same meaning as actant are valent and argument. The latter is used particularly in grammatical analyses that treat the verb as a head word governing its actants: sleep takes one argument, see takes two, and give three.

Log in or register to write something here or to contact authors.