Split Systems in Linguistics: Definitions & Types | Mini-Lesson

Research reveals recurrent paths of change in tense-aspect systems across languages, suggesting that true universals may be mechanisms of change rather than synchronic patterns. Perfective tenses develop from resultative expressions (finish), then become antepassive, then perfect tenses. Progressive forms develop from locative expressions (be located at) or duplication constructions, then become progressive aspects. Future tenses develop from want-movement or adverbial constructions expressing intention, then future meaning. Bybee argues these recurrent pathways represent true universals, while Anderson notes that similar outcomes can arise from unrelated processes (accidental convergence), distinguishing between genuine universals and coincidental patterns.

Auxiliary systems evolve from proto-language tense-aspect encodings that become mandatory. The copula frequently evolves into an auxiliary verb, encoding either continuous/imperfective aspects or perfective/completed aspects. This evolution can be modeled by tracing how the copula plus different verb forms (participles, infinitives, converbs) came to encode different grammatical meanings over time.

Languages worldwide handle tense and aspect differently across a spectrum. Chinese, Thai, and Vietnamese have no grammatical tense, relying entirely on aspect markers and context. Spanish, Portuguese, and Italian have rich tense systems with aspect built in. English has only two grammatical tenses (past/non-past) with most apparent tenses being tense-aspect combinations. Slavic languages occupy a middle ground with both tense and strict aspect systems. Tagalog joins the tenseless group with aspect doing the heavy lifting. Ancient Greek and Latin already had aspectual distinctions (aorist/imperfect/perfect), but Slavic grammarians in the 18th-19th centuries applied Latin labels purely to aspect. 20th-century linguists used Slavic languages as textbook examples, creating the misconception that aspect concepts originated there. Despite both using perfective/imperfective distinctions, Slavic (lexical) and Tagalog (morphological) systems require separate learning approaches.

In tense-aspect split ergativity, languages display ergative patterns in one tense-aspect paradigm and accusative patterns in another. Past perfect and perfective paradigms almost always show ergative or unaccusative patterns, while non-past and non-perfective paradigms usually display accusative or unergative patterns. For example, in Kashmiri, the present progressive paradigm shows neutral alignment where arguments are not marked for their roles, while the past paradigm shows ergative alignment with subjects and patients unmarked in the absolute case and agents marked in the ergative case. Similarly, Hindi shows neutral alignment in non-perfective paradigms and ergative alignment in perfective paradigms. This system can evolve when speakers develop direct object marking, which may lead to tripartite alignment where subject, agent, and patient are all treated differently.

Ferdinand de Saussure introduced synchrony (synchronic linguistics) as the study of language at a particular point in time, examining how language functions, sounds are formed, words are created, and meanings are conveyed without reference to earlier or later stages; conversely, diachrony (diachronic linguistics) studies language change and evolution across time, comparing different periods to understand how phonological, semantic, and structural changes occur, such as the historical shift in English double negative usage from 'I do not have no problem' to 'I do not have any problem.'

Dyirbal (an Australian language) has a tripartite case system. The speaker explains that in Dyirbal, the subject of intransitive verbs, the subject of transitive verbs, and the object of transitive verbs are all marked with different cases. This system is considered one of the most distinctive case alignment systems in the world.

Government splits occur when governors require different case frames based on criteria beyond canonical expectations. Four types exist: primary lexeme partition (different verb classes take different cases, e.g., Russian quirky case), secondary lexeme partition (different governed items take different cases, e.g., Turkish pronouns vs. nouns), featural split according to secondary lexeme's specification (Belarusian singular/plural distinction), and featural split according to primary lexeme's specification (Georgian tense-aspect-mood changes). Real-world data often shows complexity, with splits evolving gradually over time (Belarusian locative invasion).

Tense and aspect influence case marking choices globally. Present tense and imperfective aspect use accusative systems; past tense and perfect aspect use ergative systems. Perfect constructions emphasize result states (patients), making ergative marking appropriate. Georgian shows this: present uses accusative, past uses ergative, perfect uses neutral. Embedded clauses often show accusative patterns like perfective constructions. This temporal-aspectual sensitivity explains why languages choose different marking systems based on temporal properties.

The relationship between information type and sequential ordering shows that old information (theme) appears at sentence beginnings while new information (rheme) appears at ends. This ordering is not arbitrary but serves communicative purposes like contrast. Markedness occurs in morphology and syntax. Unmarked forms are more frequent, have broader contexts of occurrence, are simpler, and are acquired earlier by children. For example, singular is unmarked (no 's' ending) and acquired before plural. In syntax, 'Eu uso esta roupa' is less marked than 'Esta roupa eu uso' (which inverts the canonical order). Markedness is relative and context-dependent. The principle of economicity manifests in three subprinciples: quantity of information (greater information quantity leads to greater form complexity), integration (cognitively closer content appears syntactically closer), and sequential ordering (element order reflects temporal sequence of events).

Differential Object Marking is a phenomenon where some objects with patient function receive case marking while others do not. Studies recognize a strong correlation between case marking occurrence and the semantic and pragmatic properties of the object. Two key properties are animacy hierarchy (human > animate > inanimate) and definiteness hierarchy (pronouns > proper nouns > definite noun phrases > indefinite specific > indefinite non-specific).

A split-S case system (also called split-S alignment) is a morphosyntactic alignment where arguments are marked differently based on their semantic roles. Agents (doers of actions) take the nominative case, while patients (receivers of actions) take a different case. This system differs from nominative-accusative alignment because the patient marking is not the same as the object marking of transitive verbs. The split-S system can arise from the grammaticalization of passive constructions.

Sora exhibits differential subject marking where agentive subjects are encoded with subject markers while undergoer-type subjects are encoded with object agreement markers. For example, in the sentence 'I see you,' the subject uses subject markers, but in 'You are scared of her,' the undergoer-type subject uses object markers. This creates a transitive split system where the same syntactic function (subject) can be marked differently based on semantic role.

Differential Object Marking (DOM) is fundamentally a thematic and semantic phenomenon involving the transfer of argument structure from an original lexical category to the main verb, as demonstrated by comparing Romance 'a' (reanalyzed from a directional preposition to a nominal case marker) and Chinese 'ba' (reanalyzed from a verb to a verbal voice marker), where both constructions share universal properties of requiring animate/referential objects and strongly effective verbs but differ in their grammatical categories and analogical distributions.