Modality (Semantics)¶
A natural-language construction qualifies a claim as necessary or possible relative to a contextually supplied background, with force and reading determined for that language.
Core Idea¶
In natural-language semantics, modality concerns how an expression qualifies a claim rather than simply stating it as fact. A modal word, affix, or construction relates a prejacent proposition to possibilities or necessities evaluated against a relevant background. Which possibilities matter, and whether the reading is strong or weak, must be established for the language and utterance being studied.[1][2]
English must can express a necessity in view of obligations or in view of available evidence. Washo verbal modal -eʔ occurs with necessity and possibility readings across epistemic and deontic contexts. These examples share modal qualification of a prejacent, but they do not license one universal lexical rule for how all languages divide force from context.[1][2]
Structural Signature¶
- Linguistic carrier — constitutive. Identify an attested word, affix, or construction. A bare formal necessity operator in a calculus is a related model, not by itself a natural-language instance.[1][2]
- Prejacent — constitutive target. State the proposition or event description whose status is being qualified. Without that target, a claim that something is “modal” is too indeterminate to test.[1]
- Modal contribution and force — constitutive operation. Determine how the expression relates the prejacent to necessity, possibility, or a context-sensitive range of force. Do not stipulate that Washo -eʔ always means the same English must or may.[1][2]
- Contextual background or flavor — interpretive restriction. Identify which facts, obligations, dispositions, or preferences constrain the relevant alternatives. English must can keep a relative modal core while the selected background changes; how another language composes force and flavor requires its own evidence.[1][2]
- Licensed reading and limit — evidential boundary. Record the reading supported in the particular construction, including uncertainty in its analysis. An irrealis label or subjunctive form alone does not establish existential possibility.[2]
What It Is Not¶
Modality here is not every use of the letters must: the noun for grape juice lacks the auxiliary's modal meaning. Kratzer uses that contrast to separate accidental word-form overlap from the related contextual readings of an English modal verb.[1]
Nor is this entry identical to classical modal logic. Possible-worlds models help analyze linguistic modals, but a formal operator alone does not tell us which actual construction bears a reading in English or Washo. The distribution of -eʔ in conditionals and generics, and the contexts in which it is interpreted strongly or weakly, are linguistic facts needing field evidence.[1][2]
Scope of Application¶
Apply this entry when a natural-language device qualifies a prejacent and the context supplies relevant alternatives. Modal flavor may be epistemic, deontic, dispositional, or preferential; the cases cited here establish particular English and Washo patterns, not an exhaustive taxonomy for every language.[1][2]
Do not classify an unexamined irrealis, evidential, subjunctive, tense, or aspect marker as a simple possibility modal from its grammatical name. Washo -eʔ also appears in conditionals and generics, but Bochnak's proposed analysis of how context selects the right partition is explicitly incomplete.[2]
Clarity¶
Separate two questions that the phrase “modal meaning” can hide: what does the expression contribute, and what does the context supply? For Kratzer's English must, a duty context and an evidence context differ in background while retaining a relative necessity analysis. Treating each reading as a wholly unrelated homonym obscures the common contribution; treating all readings as the same unconditional rule ignores their background.[1]
Also separate force from flavor. Necessity versus possibility concerns strength; epistemic versus deontic concerns the kind of alternatives under consideration. Washo shows why even that analytic separation cannot simply be read off an English translation.[2]
Manages Complexity¶
A single modal form can appear in many conversations and grammatical environments. A first-pass analysis records five items: the form, the prejacent, the apparent force, the relevant background, and the construction's limits. That record reduces a proliferation of apparent word senses without pretending to settle every language's compositional semantics.[1][2]
The reduction must retain uncertainty. Bochnak models Washo variable force using a necessity modal that may be weakened by context, including a proposed partition of accessible worlds. He states that exactly how context determines the partition remains a challenge, so the data should not be converted into a fully predictive algorithm.[2]
Abstract Reasoning¶
Given a modal utterance, identify the prejacent and ask what alternatives the speaker's context makes relevant. Then compare how the same form behaves under different backgrounds or grammatical environments. If the form appears to permit both strong and weak readings, test whether the difference follows from the lexical contribution, context, scope, or a combination, rather than assigning an English word as its definition.[1][2]
For English must, Kratzer's duty and knowledge examples support a relative necessity analysis. For Washo, the primary data show strong and weak readings of -eʔ in epistemic and deontic settings; Bochnak argues for an underlying necessity modal with contextual weakening, while leaving the exact partition mechanism unresolved. The reasoning pattern is comparative, but its empirical premises are language-specific.[1][2]
Knowledge Transfer¶
The analysis transfers between unlike linguistic carriers: an English auxiliary and a Washo verbal modal. Both require a prejacent and a context-sensitive relation to alternatives; the Washo form's variable-force distribution prevents simply copying English lexical meanings into it.[1][2]
A portable formal skeleton—qualifying a proposition relative to possible states—also appears in models outside linguistics. That skeleton is related to live Prime Modal Reasoning, but that Prime is an inferential act by a reasoner. An attested modal construction is not necessarily a reasoning act, so the shared machinery does not by itself establish a typed parent edge.
Examples¶
English must across duty and evidence contexts¶
Kratzer's example of children required to learn ancestral names uses a duty background; her example of ancestors arriving from Tahiti uses what is known as the background. Each utterance has a different prejacent. The auxiliary must contributes a relative necessity rather than two completely unrelated lexical items.[1]
Mapped back: English must is the linguistic carrier; learning names or arriving from Tahiti is the prejacent; the modal contribution is necessity; tribal duties or evidence provide the background; the licensed readings are deontic and epistemic in their respective contexts. The contrast does not show that the same utterance changes over time.[1]
Washo -eʔ across epistemic and deontic contexts¶
Bochnak reports primary Washo fieldwork. His examples include a snow-covered mountain road interpreted as epistemic necessity, a possible visitor as epistemic possibility, a dog required to stay outside as deontic necessity, and a question about permission to play as deontic possibility. He argues that the form is a necessity modal whose interpretation can be weakened; that analysis should be distinguished from the attested readings.[2]
Mapped back: verbal -eʔ is the carrier; the road, visitor, dog, or play proposition is the prejacent in each clause; the attested readings vary in force; evidence or permissions supply flavor; the limit is that the exact contextual partition deriving a given strong or weak reading remains unsettled. Washo is an unlike setting because it is a verbal modal with variable force, not an English auxiliary with a fixed necessity contribution.[2]
Structural Tensions¶
Stable form meaning versus contextual variation. If every use of a modal form is given one rigid reading, the analysis misses English flavor variation or Washo's strong and weak readings. If every contextual reading is declared a different lexical item, the form's repeated distribution ceases to explain anything. Kratzer seeks a stable relative core for English must; Bochnak preserves a necessity-based analysis of Washo while acknowledging unresolved contextual partitioning.[1][2]
Diagnostic: which part of a reading stays with the form, and which part is supplied or weakened by the surrounding construction and context? The answer determines whether to revise the lexical analysis, the context model, or both.
Structural–Framed Character¶
Evaluative weight: a modal reading is a semantic classification, not a claim that the utterance is good or bad. Human-practice dependence: language use and speaker contexts supply the data; an analyst's possible-worlds model is a theory of those data, not the utterance itself. Institutional origin: no one institution defines modal meaning for all languages. Vocabulary travel: “possible” and “necessary” appear in logic and planning as well as language, but this entry concerns attested linguistic devices. Import versus recognition: it is recognized when a construction qualifies a prejacent in the documented contexts; importing it to any uncertainty talk without a linguistic carrier would change the subject.[1][2]
The portable sign-to-content skeleton is live Prime Meaning: an expression conveys a contextually licensed content to a competent interpreter. The approved parent relation is strict subsumption: this is Meaning plus modal force in an attested linguistic construction, while ordinary meaningful nonmodal utterances show the parent without the child. Prime Modal Reasoning contains related possibility/necessity machinery but requires an inferential act; a modal construction need not be such an act. A broader Prime for modal qualification independent of language and an agent's reasoning is a future-Prime question and receives no edge here. Its character: structural within a linguistic frame, because the form, language, and contextual distribution remain constitutive even where the formal skeleton travels.
Structural Core vs. Domain Accent¶
The skeletal relation is linguistic expression → qualified prejacent → contextually relevant alternatives and force. English must and Washo -eʔ are accents of that relation, but their lexical contribution and distribution must be established separately. English duty/evidence variation does not prove that Washo shares its exact force mechanism; Washo's variable force does not prove that all modals weaken in context.[1][2]
The portable expression-to-content relation belongs to live Prime Meaning and is the approved strict subsumption parent: this modal linguistic contribution is a kind of meaning, while Meaning also covers nonmodal content. Evaluating propositions against possible states may support a broader future Prime. Live Prime Modal Reasoning is related but requires reasoning by an agent, so it cannot be made an all-instance parent for every linguistic modal expression. This entry does not clear the Prime bar: without an attested linguistic carrier and its construction-specific readings, what remains is a formal model or a broad idea of possibility, not this named natural-language semantic relation.
Instantiates / Related Primes¶
This entry is a kind of Meaning.
Modality (semantics) is, in every case, a kind of Meaning. It is the language-specific modal content of an expression, not an act of interpretation; every English must or Washo -eʔ case requires the form to convey a licensed meaning, while nonmodal expressions show Meaning without modality. Modal Reasoning is nearby because it handles possible and necessary conclusions, but it is a reasoning act and does not cover every modal morpheme. Classical Modal Logic describes one formal system and does not cover all natural-language modality. Representational Modality concerns choice of representational medium, a different sense of “modality.” These are related neighbors, not three broader abstractions that modality automatically inherits from.
Relationships to Other Abstractions¶
Current abstraction Modality (Semantics) Domain-specific
Parents (1) — more general patterns this builds on
-
Modality (Semantics) is a kind of Meaning Prime
A linguistic modal contribution is a context-bound subtype of the Meaning relation.The modal construction qualifies a prejacent by conveying necessity or possibility content to a competent interpreter within context. This is a linguistic species of the live Meaning genus; nonmodal expression content is Meaning without this child. The modal construction and context are stable differentiae, making strict subsumption the approved typed edge after independent Gate 4 review.
Hierarchy path (1) — routes to 1 parentless root
- Modality (Semantics) → Meaning → Representation → Abstraction
Neighborhood in Abstraction Space¶
Modality (Semantics) sits in a sparse region of the domain-specific corpus (77th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Syntax, Semantics & Grammar Theory (24 abstractions)
Nearest neighbors
- DisCoCat — 0.84
- Phonocentrism — 0.83
- Telicity — 0.82
- Resultative — 0.82
- Case grammar — 0.82
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
Modal logic supplies formal operators and semantics but not an attested natural-language expression. Evidentiality marks information source and can interact with modality without being identical to it. Mood/irrealis labels grammatical categories whose modal readings need separate analysis. The noun must is accidental homography with the English auxiliary. Deontic versus epistemic distinguishes contextual flavor, while necessity versus possibility distinguishes force; neither distinction alone classifies the entire lexical contribution of Washo -eʔ.[1][2]
References¶
[1] Angelika Kratzer, “What ‘must’ and ‘can’ must and can mean,” Linguistics and Philosophy 1 (1977): 337–355, especially pp. 338–344 and examples (2)–(5), (2)–(5). Original full text; supports the relative modal core and contextual background analysis for English. https://semantics.uchicago.edu/kennedy/classes/s08/semantics2/kratzer77.pdf registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t
[2] M. Ryan Bochnak, “Variable force modality in Washo,” Proceedings of NELS 45 1 (2015): 105–114, author-hosted full paper, §§1–4, 6–7, examples (1)–(4). Original primary fieldwork and author analysis; strong/weak readings are attested while the exact context-partition mechanism remains unsettled. https://semanticsarchive.net/Archive/DVkNzVjZ/Bochnak_NELS45.pdf registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u