Active–Stative Alignment¶

A morphosyntactic alignment pattern in which some intransitive arguments receive the coding used for transitive A arguments and others receive the coding used for transitive P arguments, conditioned by predicate class or event semantics.

Version: v1 · 2026-08-30 · History
Domain-specific #: 1235
Origin domain: linguistic typology
Subdomain: morphosyntactic alignment
Aliases: Active-stative alignment, Active-inactive alignment, Agent-patient alignment

Core Idea¶

Active–stative alignment is a morphosyntactic alignment pattern in which the sole argument of an intransitive predicate is not coded uniformly. Some intransitive arguments are treated like the more agent-like argument of a transitive clause; others are treated like the more patient-like argument. In standard typological notation, let A be the transitive agent-like argument, P (or Dixon's O) the transitive patient-like argument, and S the sole argument of an intransitive clause. Active–stative alignment contains both

\[ S_A=A \qquad\text{and}\qquad S_P=P \]

within the relevant case, agreement, or indexing subsystem.^[1]^[2]

The split is semantically motivated but need not transparently mirror a single feature. In a split-S system, lexical predicate classes conventionally determine which coding an S takes. The classes may correlate with volition, control, instigation, affectedness, change of state, stativity, or inherent aspect, yet lexicalization and historical change can leave exceptions. In a fluid-S system, at least some predicates permit both patterns, and the coding varies with a constrained construal of the event or participant—for example, greater control versus accidental involvement—not with unconstrained stylistic whim.^[1]^[3]

The identity is construction-specific. A language can display active–stative alignment in verbal person indexing but another alignment in nominal case, or only in a particular tense-aspect construction. Calling an entire language “active” is convenient shorthand, not a claim that every grammatical subsystem has one pure alignment. Likewise, A and P are comparison positions defined by the language's transitive coding; they need not be literal intentional agents and affected patients in every clause.

The candidate survives as a mature domain-specific abstraction. Major typological monographs and a dedicated Oxford volume use it to organize recurrent morphosyntactic systems across unrelated language families.^[2] Existing primes cover mapping, classification, roles, markedness, and conventional form, but none supplies the A/P/S comparison, the two intransitive coding classes, or the split-S/fluid-S diagnostic. The construct is not a prime because its indispensable variables and evidence remain those of linguistic argument structure and grammatical coding.

Structural Signature¶

An active–stative alignment analysis requires all of the following roles:

A transitive comparison frame. The language or construction must distinguish the coding of its A and P arguments in at least one observable subsystem.
An intransitive domain. The relevant predicates have one core argument, S, under the analysis being tested.
Two S patterns. Some S arguments share the relevant coding with A (\(S_A\)); others share it with P (\(S_P\)). A third unique S coding would instead support tripartite alignment.
A common coding dimension. The comparison is made within case marking, agreement/indexing, clitic series, constituent order, or another demonstrable morphosyntactic subsystem. One may not equate case in one construction with agreement in another without an argument.
A conditioning rule. Lexical predicate membership, participant properties, event construal, inherent aspect, or a combination decides which S pattern appears.
A semantic tendency. Agent-like properties such as volition or control commonly favor A-like coding; patient-like properties such as affectedness or undergoing change commonly favor P-like coding. The exact feature bundle is language-specific.^[4]
Conventional grammatical realization. The contrast must be encoded by the grammar, not merely inferred from the real-world meaning of the predicate.
Constructional scope. The analyst specifies where the pattern holds and where another alignment, derivation, or paradigm applies.
A subtype diagnosis. Split-S assigns predicates to relatively fixed lexical classes; fluid-S permits a controlled alternation for some predicate under different event construals. A language can contain both.

Formally, let \(C(z,k)\) be the coding of argument token \(z\) in construction \(k\). An alignment subsystem is active–stative when there are admissible intransitive tokens \(s_1,s_2\) such that

\[ C(s_1,k)=C(A,k), \qquad C(s_2,k)=C(P,k), \]

and the contrast is governed by a recurrent lexical or semantic rule rather than accidental allomorphy. For split-S, the main selector can be modeled as a lexical function \(\kappa(V)\in\{A,P\}\). For fluid-S, the input must include the event construal \(e\): \(\kappa(V,e)\in\{A,P\}\). Treating the predicate alone as input would falsely make fluid-S “inconsistent.”

What It Is Not¶

Not nominative–accusative alignment. A consistently accusative subsystem codes every S like A and P differently. Active–stative has both \(S_A=A\) and \(S_P=P\).
Not ergative–absolutive alignment. A consistently ergative subsystem codes every S like P and A differently. Active–stative does not put all S into the P-like class.
Not split ergativity as such. Split ergativity alternates accusative and ergative patterns under conditions such as tense, aspect, person, or clause type. Active–stative divides or alternates intransitive S coding by predicate or participant/event semantics. A language can combine both phenomena, so surface labels must be assigned construction by construction.^[1]
Not merely an unaccusative/unergative semantic distinction. Many theories divide intransitive predicates by argument structure even where overt case or agreement codes all S alike. Active–stative requires a demonstrated A-like versus P-like morphosyntactic realization.
Not lexical aspect alone. Dynamic versus stative meaning may condition the system, but the semantic opposition must correlate with transitive-reference coding. A list of activity and state verbs without an A/P coding contrast is not alignment.
Not case marking only. Head-marking languages can realize the pattern through distinct agreement or pronominal series. Calling the two series “agentive case” and “patientive case” can misdescribe their formal locus.
Not semantic roles themselves. Volition, control, affectedness, and change of state motivate class membership. The alignment is the grammatical distribution built around those properties.
Not free optionality. Fluid-S choice is constrained by what a construction means or how an event is construed. It is not arbitrary alternation between interchangeable affixes.
Not a claim that a whole language is uniformly active–stative. Alignment can differ between case and agreement, nouns and pronouns, aspects, or clause types.
Not voice or valency alternation by default. Passive, antipassive, causative, applicative, and derivational operations can change argument structure. A core active–stative diagnostic compares underived intransitive coding to transitive A and P coding unless a particular analysis justifies a broader domain.^[4]

Scope of Application¶

The abstraction belongs to linguistic typology, descriptive grammar, morphosyntax, lexical semantics, and historical linguistics. It is used to compare language-specific systems, to decide whether an apparent “subject” category is uniform, to document pronominal or agreement paradigms, and to reconstruct how semantic motivations become lexicalized or grammaticized. The Oxford volume The Typology of Semantic Alignment includes general chapters and case studies from Eurasia, eastern Indonesia, and the Americas, including Tundra Nenets, Basque, Amis, Lakota, Otomi, Guaraní, Arawak languages, and Pilagá.^[2] This breadth establishes recurrence without implying that all those systems are structurally identical.

Klamer's eastern Indonesian survey illustrates how the same recognition framework accommodates genuine variation. It compares S with A and P in nine Austronesian and Papuan languages and finds systems conditioned by volition, change of state, default coding, participant type, predicate aspect, or combinations.^[4] The abstraction is therefore not “use volition to choose a case.” Its scope is the higher-order alignment relation; each grammar supplies its own selector and formal realization.

Historical work also lies within scope. Mithun shows that apparently opaque active/agentive case systems can be products of successive, individually motivated developments followed by grammaticalization and lexicalization.^[3] A synchronic exception does not by itself refute the typology, but neither may a proposed ancient active–stative system be accepted from etymological speculation alone. It needs converging evidence for two S coding patterns and their relation to transitive A and P.

Clarity¶

A reliable analysis begins with four tables rather than with English translations. First, establish how A and P are coded in the target subsystem. Second, inventory the coding of S across underived intransitive predicates. Third, test whether the S forms match the A and P series. Fourth, model the distribution using lexical class and contextual semantic variables. Only then assign the alignment label.

The split-S/fluid-S diagnostic is especially clarifying. If each predicate selects one series across contexts, even when a particular occurrence is involuntary, the pattern is split-S. If the same predicate licenses contrasting series with a systematic meaning difference, the relevant portion is fluid-S. Fluidity can coexist with a large fixed lexical core, so a grammar need not be placed in exactly one box. Dixon's contrast between lexically assigned split-S and context-sensitive fluid-S remains foundational, while Mithun cautions that semantic motivations can become obscured over time.^[1]^[3]

Manages Complexity¶

The abstraction compresses a potentially disorderly inventory of agreement prefixes, clitics, cases, and verb classes into a comparison with three grammatical positions and a conditioning rule. Instead of memorizing that dozens of intransitive predicates take one of two series, an analyst asks which share A coding, which share P coding, and which semantic or lexical features predict the distribution. Exceptions then become diagnosable as lexicalization, competing features, historical residue, or misanalysis rather than undifferentiated irregularity.

It also prevents the traditional category “subject” from hiding significant structure. Two sole arguments can share clause count and syntactic position while receiving different indexing because one is construed as controlling and another as undergoing. Conversely, superficially distinct affixes can belong to a coherent system because each repeats a transitive coding series. The alignment model links paradigmatic form, argument role, event semantics, and diachrony without reducing any one of them to the others.

Abstract Reasoning¶

Several inferences follow from the signature. If an alleged active marker never occurs with transitive A, its analysis as \(S_A=A\) requires additional justification. If P-like S marking appears only on derived passives, voice may explain it better than active–stative alignment. If the same intransitive predicate alternates with a stable contrast in control, affectedness, or event viewpoint, fluid-S is supported; if it does not alternate, a split-S lexical analysis is safer. If semantic features predict the distribution imperfectly, historical lexicalization is a live explanation, not permission to invent ad hoc roles for every item.

The mapping formalism also identifies hidden variables. A “same verb, different marker” pair does not violate grammatical regularity if event construal is part of the input. Conversely, adding unconstrained context until every form is predictable makes the analysis unfalsifiable. A sound account states which semantic factors are licensed, how they are diagnosed independently, and which residual verbs are lexically listed.

Knowledge Transfer¶

The abstraction transfers literally among case-marking studies, agreement and indexing analyses, lexicon–grammar interface research, field grammar, and diachronic typology. In each practice the analyst identifies A, P, and S; verifies two S coding patterns; and determines whether lexical class or event semantics conditions the selection. This common procedure supports comparison even when one language uses suffixal case and another uses verbal pronominal prefixes.

Transfer outside linguistics is analogical. A software classifier that routes events by agentivity does not thereby have active–stative alignment, because it lacks grammatical arguments and an A/P/S coding system. The portable residues—conditional mapping, classification, role assignment, and markedness—already have prime nodes. The linguistic construct should therefore remain domain-specific.

Examples¶

Schematic split-S system. Suppose transitive A takes series ka- and P takes series ma-. The intransitive predicate run always indexes its S with ka-, while fall always indexes its S with ma-. Even an accidental running episode remains in the lexical ka- class. The grammar exhibits \(S_A=A\) for run and \(S_P=P\) for fall. The invented forms show the diagnostic without pretending that English nominative/object pronouns reproduce another language's morphology.

Schematic fluid-S system. In the same comparison frame, a motion predicate permits ka- when the participant deliberately controls the motion and ma- when the participant slips involuntarily. The predicate is constant, but event construal changes. This is fluid-S only if the alternation is grammatically licensed and the meaning contrast recurs; random form variation or performance error is not.

Eastern Indonesian recurrence. Klamer's survey finds languages where a volitional S is A-coded and a nonvolitional S is P-coded, others where undergoing a change of state motivates P coding, and systems with A-like defaults plus semantically restricted P-like departures.^[4] These cases demonstrate why “active–stative” names the alignment architecture rather than one universal semantic switch.

Otomi complexity. Palancar reports both agent–patient and active–stative semantic-alignment patterns among Otomi intransitive verbs and investigates their diachronic emergence.^[5] The example shows that a single language can contain more than one organization within the broader family; the active–stative label does not erase the distinction.

Non-example: uniform accusative coding. If both run and fall take the same nominative or subject agreement as transitive A, semantic differences in volition do not create active–stative alignment. The system remains \(S=A\) in that coding dimension unless another subsystem displays the split.

Structural Tensions¶

Semantic motivation versus lexical conventionalization. Transparent agentivity makes the system learnable, while historical lexicalization stabilizes forms but creates exceptions. Analysis must distinguish productive selection from inherited verb class.
Whole-language labels versus construction-specific evidence. “Active language” is concise, but it can hide different alignments in case, agreement, aspect, or person paradigms. The smallest evidenced subsystem should control the claim.
Split-S stability versus fluid-S expressivity. Fixed verb classes make coding predictable; controlled alternation lets speakers construe participation differently. Mixed systems balance both and resist a binary language-level classification.
Cross-linguistic comparison versus language-specific semantics. A/P/S notation enables comparison, but forcing every grammar into volitional agent versus affected patient can erase aspect, viewpoint, change-of-state, and lexical factors.
Surface coding versus deeper syntax. Matching affixes establish morphological alignment, not automatically syntactic pivot behavior. Case, agreement, control, coordination, and extraction may organize arguments differently.

Structural–Framed Character¶

Assessment: predominantly structural with a mild analytical frame. Once a construction's A, P, S, coding series, and conditioning distribution are established, the alignment relation is a formal comparison. It supports reproducible paradigms and counterexamples. The framed component enters through analytical choices: how to delimit a construction, which argument counts as A or P in low-transitivity clauses, and which semantic feature best generalizes over a verb class.

The label also carries a history of competing terms—active, agentive, split intransitive, split-S, fluid-S, and semantic alignment—with partly overlapping extensions. That terminological framing warrants caution but does not make the object merely interpretive. A reasonable aggregate framed score is about 0.25, below a boundary-case threshold.

Structural Core vs. Domain Accent¶

The portable core is a conditional mapping: items in one domain are assigned to one of two reference classes by a lexical or contextual rule. Classification and Function Mapping already cover that skeleton. The domain accent is constitutive here: A, P, and S roles; transitive and intransitive valency; case, agreement, and indexing paradigms; predicate classes; event semantics; and language-specific construction boundaries.

Removing those linguistic roles produces generic binary classification. Retaining them creates a diagnostic that linguists can apply across unrelated languages while respecting different formal exponents and semantic selectors. That combination justifies a domain-specific node rather than either a prime or a loose topic page on grammatical case.

Active–Stative Alignment presupposes prime:function_mapping: a grammar maps a fully specified predicate-and-context input to an A-like or P-like coding output. In split-S, lexical predicate identity supplies the main input; in fluid-S, event construal must be included. prime:classification is related because predicates or argument tokens fall into \(S_A\) and \(S_P\) classes. prime:role helps distinguish grammatical positions from their occupants and from prototypical semantic agents or patients. prime:markedness can analyze default versus overt series in particular languages, but no universal direction of morphological marking is part of the node. prime:arbitrariness_of_symbolic_conventions explains why semantic motivation can coexist with conventionalized, historically opaque lexical membership.

Relationships to Other Abstractions¶

Current abstraction Active–Stative Alignment Domain-specific

Parents (1) — more general patterns this builds on

Active–Stative Alignment presupposes Function (Mapping) Prime

Active–Stative Alignment presupposes prime:function_mapping: a grammar maps a fully specified predicate-and-context input to an A-like or P-like coding output.

Hierarchy path (1) — routes to 1 parentless root

Active–Stative Alignment → Function (Mapping)

Neighborhood in Abstraction Space¶

Active–Stative Alignment sits in a sparse region of the domain-specific corpus (88^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

The frozen semantic neighbors prime:coordination, prime:threshold_triggered_rule_activation, and prime:paradigmatic_vs_syntagmatic_relations match generic “alignment,” selection, and paradigm vocabulary but do not cover morphosyntactic argument alignment. prime:markedness describes default/specified asymmetry, not the \(S_A=A\) and \(S_P=P\) structure. prime:code_switching concerns alternation between linguistic codes or varieties, not alternate coding of intransitive arguments inside a grammar. prime:linguistic_universals concerns cross-language generalizations; active–stative alignment is one typological pattern whose distribution and realization vary.

“Semantic alignment” is often used for this family, especially in the Donohue–Wichmann volume, but its extension can be broader than the narrow active–stative contrast. “Split intransitivity” and “unaccusative/unergative” also have theory-specific uses that do not always require overt A-like/P-like coding. These terms belong in a context-qualified vocabulary review rather than an unexamined unrestricted alias list.

References¶

[1] R. M. W. Dixon, Ergativity, chapter 4, “Types of Split System,” Cambridge University Press, 1994, pp. 70–110. https://doi.org/10.1017/CBO9780511611896.006 registry ↩a ↩b ↩c ↩d

[2] Mark Donohue and Søren Wichmann, eds., The Typology of Semantic Alignment, Oxford University Press, 2008. https://doi.org/10.1093/acprof:oso/9780199238385.001.0001 registry ↩a ↩b ↩c

[3] Marianne Mithun, “Active/Agentive Case Marking and Its Motivations,” Language 67(3), 1991, pp. 510–546. https://doi.org/10.2307/415036 registry ↩a ↩b ↩c

[4] Marian Klamer, “The Semantics of Semantic Alignment in Eastern Indonesia,” in Donohue and Wichmann, eds., The Typology of Semantic Alignment, 2008, pp. 221–251. https://doi.org/10.1093/acprof:oso/9780199238385.003.0009 registry ↩a ↩b ↩c ↩d

[5] Enrique L. Palancar, “The Emergence of Active/Stative Alignment in Otomi,” in Donohue and Wichmann, eds., The Typology of Semantic Alignment, 2008, pp. 357–379. https://doi.org/10.1093/acprof:oso/9780199238385.003.0014 registry ↩