- This article may contain information that is already published in another article on the same subject or may be repeating information already explained.
Split ergativity is shown by languages that have a partly ergative behaviour, but employ another syntax or morphology—usually accusative—in some contexts. In fact, most of the so-called ergative languages are not pure but split-ergative.
Split conditions 
The split is usually conditioned by one of these:
- The presence of a discourse participant (a first or second person) in the proposition. The Australian language Dyirbal behaves ergatively in all morphosyntactic contexts, except when one of these is involved. When a first or second person pronoun appears, however, it is marked according to a nominative–accusative pattern (with the least marked case when it is the agent or intransitive, and with the most marked case when it is the patient). This can be explained in terms of the high animacy of a first or second person speaker in the animacy hierarchy.
- The use of certain tenses and/or aspects in the verb. The Indo-Iranian family, for example, shows a split between the perfective and the imperfective aspect. A verb in the perfective aspect causes its arguments to be marked using an ergative pattern, while the imperfective aspect triggers accusative marking. (Related languages and others always tend to associate past tense and/or perfect aspect with ergativity.)
- The type of marking involved. Some languages (including various Austronesian languages of New Guinea, such as Sinaugoro) exhibit an ergative–absolutive pattern with respect to case marking, but a nominative–accusative pattern with respect to agreement.
- The agentivity of the intransitive subject. In languages like Dakota, arguments of active verbs such as to run are marked like transitive agents, as in accusative languages, while arguments of inactive verbs such as to stand are marked like transitive objects, as in ergative languages. Languages with this kind of marking are known as split-S languages, and are formally a subtype of active languages.
An example of split ergativity conditioned by tense and aspect is found in the Hindustani language, which has an ergative case on subjects in the perfective aspect for transitive verbs in the active voice, while in other aspects (habitual, progressive) subjects appear in the nominative case. Example sentences for Hindustani are:
लड़का किताब ख़रीदता है laṛkā kitāb xarīdtā hai boy-Msg.NOM book.Fsg.NOM buy-IMPERFECTIVE-Msg be.PRES.3sg 
- "The boy buys a book."
लड़के-ने किताब ख़रीदी laṛke-ne kitāb xarīdī boy.Msg-ERG book.Fsg.NOM buy-PERFECTIVE.Fsg
- "The boy bought a book."
In the first sentence, the verb in present tense has the masculine ending -ā, agreeing with laṛkā (boy). But in the second sentence, the verb in past tense has the feminine ending -ī, agreeing with the feminine noun kitāb (book). "Boy" here appears as laṛke-ne, literally "by the boy".
- The morph-by-morph analysis has been simplified to show the features relevant to the topic of split ergativity.
- Dixon, R. M. W. (1994). Ergativity. Cambridge University Press. ISBN 0-521-44898-0.