Morphosyntactic Alignment¶

Compare how a grammatical subsystem treats intransitive S versus transitive A and P arguments.

Version: v1 · 2026-10-03 · History
Domain-specific #: 13447
Domain group: Humanities
Origin domain: Linguistics & Semiotics
Subdomains: Linguistic Typology, Morphosyntax → Linguistics & Semiotics
Aliases: Grammatical alignment, Argument alignment

Core Idea¶

Morphosyntactic alignment compares how a language treats three core argument positions in specified constructions: S, the sole argument of an intransitive predicate; A, the more agent-like argument of a transitive predicate; and P (also called O), its more patient-like argument. The analyst chooses a grammatical channel—such as nominal case, pronoun case, verbal person marking, or a syntactic behavior—and asks which positions are treated alike and which differently. The World Atlas of Language Structures (WALS) uses this frame to compare case and verbal person marking across languages.^[1]^[2]

For instance, nominative–accusative alignment groups S with A against P; ergative–absolutive alignment groups S with P against A. Neutral systems group all three in the channel being studied, while tripartite systems distinguish all three. Active systems can code some intransitive arguments like A and others like P; additional systems and splits require their own conditioning rules. These are patterns of grammatical treatment, not universal claims that every A acts intentionally or every P is passive.^[1]^[2]

Structural Signature¶

Sig role-phrases:

Comparable positions — Identify S, A, and P from intransitive and transitive clauses under a consistent analysis.^[1]
Coding channel — Specify the observable dimension: full-noun-phrase case, pronoun case, verbal person marking, or a declared syntactic diagnostic. Never combine distinct channels without saying so.^[1]^[2]
Grouping relation — Record which positions share a marker or behavior in that channel and which receive distinct treatment.
Constructional scope — State the person, nominal class, tense-aspect, clause type, or voice-like alternant in which the pattern holds.^[2]
Subtype diagnosis — Map the scoped grouping to a pattern such as accusative, ergative, active, tripartite, neutral, or a documented split, without making the subtype the umbrella itself.

The comparison is relational. One marker's form alone is not an alignment type; the analyst needs observations across the three positions and must decide what counts as sameness in the chosen channel.

What It Is Not¶

Not a single universal alignment for a language. English full noun phrases have neutral case marking while some personal pronouns show nominative–accusative case; verbal marking is another channel.^[1]
Not equivalent to ergativity. Ergative–absolutive is one S/P-versus-A grouping within the broader comparison.
Not a purely semantic agent/patient classification. The labels A and P identify comparison positions by prototypical orientation, while actual thematic roles can vary.^[1]
Not necessarily case marking. Verbal person markers and certain syntactic behaviors can instantiate alignment questions, but their evidence must not be silently mixed.
Not every difference in word order. Word order counts only where a specified S/A/P comparison makes the grammatical treatment diagnosable; a free permutation without role-conditioned behavior is not enough.

Scope of Application¶

For nominal case, WALS separately compares full noun phrases and pronouns. Its English example says full noun phrases are neutral for case, while pronouns such as we in S or A position contrast with us in P position. The pronoun subsystem therefore shows an accusative pattern without licensing the claim that all English nominals carry overt nominative–accusative case.^[1]

For verbal person marking, WALS gives Konjo as an ergative example: a third-person form on the verb is the same for S and P and different for A in the cited clauses. Other WALS examples show active and tripartite patterns. Person, aspect, mood, polarity, word order, and main-versus-subordinate clause status can condition splits; a typology must say which of these contexts its label covers.^[2]

Clarity¶

Alignment clarifies the difference between the grammatical relation of one argument and the cross-construction pattern among arguments. Calling an expression “subject” does not yet state whether its case or verbal index patterns with S, A, or P elsewhere. The S/A/P frame makes that comparison explicit without presuming that the English subject/object distinction fits all languages unchanged.

It also prevents a misleading whole-language label. “English is accusative” becomes the testable statement “some English personal-pronoun case forms group S and A against P,” with the separate fact that full noun phrases are neutral in that case dimension.^[1]

Manages Complexity¶

Languages contain many nominal and verbal forms. The S/A/P comparison compresses these into a few grouping patterns that can be compared across systems. It supports a typological map because the same three positions are evaluated in each language or construction.^[1]^[2]

The compression is lossy. It can hide split paradigms, optional marking, alignment differences between case and agreement, or disputed choices of the “basic” construction. A careful account retains its diagnostic forms and conditioning factors rather than letting a subtype label replace the data.

Abstract Reasoning¶

To diagnose an alignment claim, gather comparable clauses with S, A, and P; choose one channel; and compare the relevant coding or syntactic behavior. If S and A match while P differs, the channel is accusative in that scope. If S and P match while A differs, it is ergative. If forms vary by person or aspect, partition the evidence by that conditioning factor before assigning a pattern.^[1]^[2]

This procedure also catches false generalization. If pronouns support one partition and full noun phrases another, a claim of one language-wide case alignment fails unless qualified. If only semantic agency differs but the grammar provides no relevant contrast, the alignment inference is unsupported.

Knowledge Transfer¶

The S/A/P procedure transfers literally from nominal case to pronominal case and verbal person marking when each channel is analyzed separately. Comparison with syntactic alignment requires its own behavioral diagnostic rather than a direct substitution of marker forms. Beyond language, people may analogize to grouping inputs and outputs, but without grammatical argument roles and morphosyntactic treatment that is not this abstraction.

Examples¶

English personal-pronoun case¶

WALS contrasts we used for S and A with us used for P. That establishes an S=A≠P pattern for the relevant pronoun case forms. The same chapter reports neutral case for full English noun phrases, demonstrating why the subsystem qualifier is part of the analysis.^[1]

Mapped back: positions = pronouns occupying S, A, and P; channel = personal-pronoun case; grouping = S=A≠P, nominative–accusative; scope = selected pronoun paradigm, not all English noun phrases.

Konjo verbal person marking¶

In WALS's Konjo examples, a third-person verbal marker treats the intransitive S and transitive P alike, while A receives distinct marking. This is S=P≠A in the verbal-person channel, a different pattern and different observable medium from the English pronoun example.^[2]

Mapped back: positions = third-person S, A, and P in the cited clauses; channel = verbal person marking; grouping = S=P≠A, ergative–absolutive; scope = the illustrated Konjo verbal paradigm.

Structural Tensions¶

Comparability versus subsystem specificity. A compact type makes languages comparable, but imposing it on an entire language can conceal that noun phrases, pronouns, and verbs align differently. Preserve the channel and context, even if that makes the summary less terse. Diagnostic: which exact paradigm and construction demonstrate the claimed S/A/P grouping?^[1]

An evidence boundary, rather than a second opposed-cost tension, concerns semantic overreading. S/A/P positions allow systematic comparison across languages, but calling A “agent-like” does not mean it is volitional in every clause. A semantic story can suggest a hypothesis; grammatical evidence must establish the alignment pattern. Diagnostic: is the asserted likeness a case, agreement or syntactic fact rather than only an account of participant meaning?^[1]

Structural–Framed Character¶

Morphosyntactic Alignment is mixed-structural and conventionally framed. Its evaluative weight is low: an accusative or ergative pattern is not inherently better. Human analysis selects comparable constructions and coding channels, while attested paradigms constrain the diagnosis. Its institutional history lies in linguistic typology, but the grouping relation is formally inspectable once the observations and scope are stated. The words alignment and Role travel widely, yet outside grammar their use imports analogy rather than recognizes literal S/A/P systems. The portable skeleton is classification by shared treatment under a declared relation, already addressed by Classification; the named linguistic construct remains domain-specific. Its character: an empirical-structural comparison whose validity depends on language- and construction-specific evidence.

Structural Core vs. Domain Accent¶

The skeletal operation groups comparable positions by an observed relation. The indispensable domain content is the intransitive/transitive S/A/P frame and the morphosyntactic coding or behavioral channel. Remove those, and one has a generic classification or partition, not Morphosyntactic Alignment. Conversely, adding a new case or agreement example does not turn the umbrella into a prime; its literal transfer remains within grammatical systems. Broader portability belongs to Classification, not to this particular linguistic identity.

This entry presupposes Classification.

The live Classification is a strict analytic prerequisite under composition/presupposes: diagnosing an alignment groups observed S/A/P positions by one declared grammatical channel. This does not claim a language's grammatical facts depend on an analyst classifying them. Grammatical argument positions and coding evidence remain the domain residual. Existing Ergative–absolutive alignment, Active–Stative Alignment, and Transitive alignment may warrant future child-to-umbrella edges; this staged package does not change them. Grammatical Relation concerns an individual function, not the comparison across functions.

Relationships to Other Abstractions¶

Current abstraction Morphosyntactic Alignment Domain-specific

Parents (1) — more general patterns this builds on

Morphosyntactic Alignment presupposes Classification Prime

The analytic alignment diagnosis requires criterion-bound grouping of S, A and P grammatical positions.

Hierarchy path (1) — routes to 1 parentless root

Morphosyntactic Alignment → Classification

Neighborhood in Abstraction Space¶

Morphosyntactic Alignment sits in a sparse region of the domain-specific corpus (66^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Language, Mind & Meaning-Making (57 abstractions)

Nearest neighbors

Multivariate Glyph — 0.86
Tautophrase — 0.84
Rhyme — 0.84
Folk Etymology — 0.84
Verbal Reasoning — 0.83

Computed from structural-signature embeddings · 2026-10-08

Not to Be Confused With¶

Ergative–absolutive alignment: the S=P≠A subtype, not the whole umbrella.
Active–stative alignment: an S split into A-like and P-like patterns under stated conditions.
Grammatical relation: the function of one expression in a clause, not the comparative S/A/P coding relation.
Semantic role: a participant's interpretation, which cannot substitute for observed case, agreement, or syntactic behavior.

References¶

[1] Bernard Comrie, “Alignment of Case Marking of Full Noun Phrases”, World Atlas of Language Structures Online, especially §§1 and 3 on S/A/P, alignment types, and English pronouns. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m

[2] Anna Siewierska, “Alignment of Verbal Person Marking”, World Atlas of Language Structures Online, especially §1 on Konjo and split conditioning. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h