The morphosyntactic alignment of a language is determined by the morphosyntactic treatment[1] of the arguments of intransitive and monotransitive verbs. Simple indicative sentences that meet the following criteria should be considered:

• either S (the sole argument of an intransitive verb) or A (the agentive[2] argument of a transitive verb) is the topic
• the verb and, if applicable, P (see definition below) constitute the comment or rheme (new information in the sentence)
• the functions S and A are fulfilled by nouns, while P is fulfilled by a pronoun

The arguments in question are the following:

Sa  – the argument of an intransitive verb expressing conscious, willful action (intransitive agent)[3]

Snona – the argument of an intransitive verb expressing non-action, such as a state or event (intransitive non-agent)[4]

Aa  – the agentive argument of a transitive verb expressing conscious, willful action (transitive agent)[5]

Anona – the non-patientive argument of a transitive verb expressing non-action, such as a state or event (transitive non-agent)[6]

P – the patientive argument of a transitive verb; it expresses the result or affected entity (imagined or actual) of the action or event expressed by the verb[7]

Types:

Tri: Three categories are distinguished from a morphosyntactic perspective: 1. [Sa and Snona]; 2. [Aa and Anona]; 3. [P]. This type is known as tripartite or triadic.

Nom: Two categories are distinguished from a morphosyntactic perspective: 1. [Sa, Snona, Aa and Anona];  2. [P]. This type is known as nominative or nominative–accusative.

Erg: Two categories are distinguished from a morphosyntactic perspective: 1. [Sa, Snona and P]; 2. [Aa and Anona]. This type is known as ergative or ergative–absolutive.

SemiAct: Two categories are distinguished from a morphosyntactic perspective: 1. [Sa, Aa and Anona]; 2. [Snona and P]. This type is known as semi-active or semi-agentive.[8]

Act: Two categories are distinguished from a morphosyntactic perspective:1. [Sa and Aa]; 2. [Snona, Anona and P]. This type is known as active or agentive.

Else: The morphosyntactic alignment of the language does not fit any of types listed above.[9]

When a language displays more than one type, two values can be listed. If one type appears only infrequently, a slash (/) can separate the two values, with the dominant value appearing first; if their distribution is determined by syntactic or semantic considerations, they should be listed with an ampersand (&) separating the two. The use of parentheses around a non-dominant type indicates that the strategy is not obligatory.

[1] Morphosyntactic treatment includes case marking, adposition use, and fixed word order.

[2] The agentive argument is determined in contrastive to the patientive argument (P), which is affected by or the result of an action.

[3] For example, consider the boy in the sentence The boy walks.

[4] For example, consider the boy in the sentence The boy faints.

[5] For example, consider the boy in the sentence The boy builds it.

[6] For example, consider the boy in the sentence The boy senses it.

[7] For example, consider it in the sentences The boy will build it, The boy will paint it, The boy will inherit it, and The boy imagines it.

[8] It is doubtful whether such languages exist, although active languages are often misrepresented as belonging to this type. The treatment of Aa and Anona should be examined closely to confirm that the two roles are not distinguished. If they are, the language generally belongs to the Act type.

[9] This type includes thematic (focus-oriented or Austronesian-type languages), duplicative (double-oblique) languages, and others.