Grammatical category

Organize a language's grammar around a recurrent feature dimension—such as number, case, tense, aspect, mood, person, or gender—and the constrained values that expressions may realize within it.

Version

v3 · 2026-09-19 · History

Domain-specific #

1951

Domain group

Humanities

Origin domain

Linguistics & Semiotics

Subdomain

Morphosyntax → Linguistics & Semiotics

Aliases

Grammatical feature, Morphosyntactic category

Core Idea

A grammatical category is a dimension of contrast that a language makes available to its grammar, together with an organized inventory of values on that dimension. Number may contrast singular, dual, trial, paucal, and plural; case may contrast nominative, accusative, ergative, absolutive, locative, and other values; tense, aspect, mood, person, gender, evidentiality, polarity, and definiteness provide further familiar dimensions. The point is not that every language has the same list. It is that a grammatical analysis can identify a recurrent choice-set whose values participate systematically in word formation, agreement, government, clause structure, or interpretation. Bybee treats such categories through the semantic content and patterns of grammatical expression that become bound into morphology, while typological work emphasizes that inventories and realization strategies vary substantially across languages.

How would you explain it like I'm…

Grammar Choice Menus

Languages have built-in choices you have to make when you talk, like 'one dog' or 'many dogs'. The choice 'how many' is like a menu, and 'one' and 'many' are the things on the menu. Different languages have different menus — some even have a special pick for exactly two.

Built-In Grammar Choices

A grammatical category is a kind of choice a language builds into its grammar, together with the options you can choose. Number is one example: English has "one" and "many," but some languages also have a special form for exactly two. Tense, which tells when something happens, is another category. The options get shown in different ways, like an ending on a word, a little extra word, or sometimes nothing at all. Not every language has the same categories or the same options.

Category, Value, and Exponent

A grammatical category is a dimension of contrast that a language's grammar uses, plus the set of values on that dimension. Linguists keep three levels apart. The category is the dimension, such as number. A value is one choice within it, such as dual. An exponent is the actual material that expresses the value, such as a suffix, a particle, a stem change, word order or even nothing at all. One ending can express several values at once, like number and case together. A single value can also have different exponents in different words or contexts. A contrast counts as grammatical only when it is built into the language's regular system. Merely being describable with ordinary words, like saying 'a pair of,' does not make a contrast grammatical.

 

A grammatical category is a dimension of contrast integrated into a language's grammar, together with an organized inventory of values. Number may include singular, dual, trial, paucal and plural. Case may include nominative, accusative, ergative, absolutive, locative and others. Tense, aspect, mood, person, gender, evidentiality, polarity and definiteness are further examples. Inventories and realization strategies vary substantially across languages. The analysis separates three levels: the category (the dimension), the value or grammeme (a position in it) and the exponent (the formal realization). Exponence is not one-to-one. A single exponent can realize several values, one value can have several exponents, and agreement can reveal a category on a controller that carries no overt marker. Categories participate systematically in word formation, agreement, government, clause structure and interpretation. The line between grammatical and lexical expression is gradient over time: lexical material can grammaticalize, and old inflections can become optional or restricted. Corbett's canonical typology compares real systems against idealized criteria without forcing them into one prototype.

Scope of Application

The construct is literal wherever a language or construction conventionalizes a structured feature dimension. Its comparative use is legitimate only when language-specific evidence and possible mismatches are kept visible.

• Descriptive grammar. Stating the categories and values realized by nouns, verbs, modifiers, pronouns, phrases, and clauses.
• Morphology. Analyzing paradigms, syncretism, exponence, allomorphy, and feature bundles.
• Syntax. Modeling agreement, concord, government, selection, and feature checking or unification.
• Semantics and pragmatics. Relating grammatical values to time, event structure, modality, information source, participant properties, and discourse status.
• Typology. Comparing how languages partition a broadly similar semantic or functional space without assuming identical systems.
• Historical linguistics. Tracing grammaticalization, reanalysis, category loss, renewal, and shifts in obligatoriness.
• Language technology. Defining annotation features and tagsets while documenting where computational labels simplify the language.
• Language learning and documentation. Replacing isolated ending lists with an explicit account of choices, contexts, and interactions.

Clarity

A clear description names the category, lists its values, specifies which expressions can bear them, and distinguishes inherent values from contextual or agreement values. It records whether marking is obligatory and whether the contrast is neutralized in particular environments. It also separates semantic comparison from language-internal distribution: two forms may express similar meanings but belong to different grammatical systems, while one formal exponent may fuse several categories.

Manages Complexity

Grammar distributes information across forms, positions, controllers, targets, and interpretive contexts. Treating each surface expression as unrelated produces an unwieldy inventory. Grammatical categories compress that inventory by grouping recurring alternatives along dimensions and by separating a value from its many possible realizations. A number system can explain suffixes on nouns, agreement on verbs and modifiers, pronoun choices, and restrictions on numerals within one coordinated account.

Abstract Reasoning

1. Identify a recurrent grammatical contrast rather than beginning from a translated label.
2. Establish the expressions and constructions for which the contrast changes form, distribution, or interpretation.
3. Propose the smallest value inventory that explains the contrasts.
4. Distinguish inherent, assigned, and contextually valued features where the grammar requires it.
5. Map values to exponents, including cumulative, discontinuous, suppletive, and zero realization.
6. Test whether agreement, government, or selection transmits or restricts the values.

Knowledge Transfer

The portable move is Classification: replace a mass of observations with a dimension and a controlled set of alternatives. The linguistic node adds constitutive constraints. Values must be conventional parts of a grammar, have identifiable loci, and participate in realization or combinatorial rules. Outside linguistics, calling a database column a 'grammatical category' is metaphor unless expressions, paradigms, agreement, and grammatical interpretation are genuinely involved. Within linguistics the framework transfers across morphology, syntax, semantics, typology, diachrony, and computation because the same category–value–exponent distinction prevents recurrent errors in each.

Relationships to Other Abstractions

Current abstraction Grammatical category Domain-specific

Parents (1) — outgoing typed relations

• Grammatical category is a kind of Classification Prime

  Classification is the strict parent because a grammatical category partitions a recurrent grammatical dimension into organized values and rules for assigning or realizing them.

Hierarchy path (1) — routes to 1 parentless root

• Grammatical category → is a kind of Classification

Neighborhood in Abstraction Space

Grammatical category sits in a sparse region of the domain-specific corpus (95^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2769 abstractions)

Nearest neighbors

• Dimensionless Quantity — 0.79
• Syllabary — 0.79
• Distinctive feature — 0.77
• Retrieval Facet — 0.77
• Noun — 0.77

Computed from structural-signature embeddings · 2026-10-25