Outlook Markers: Dual-Layered Secondary Meaning

@cite{farkas-bruce-2010} @cite{kubota-2026} @cite{potts-2007}

Formalization of @cite{kubota-2026} "Outlook Management: 'Subjective' Meanings of Discourse-Sensitive Adverbs and Particles."

Key Insight

Outlook markers (Japanese nanka, dōse, mushiro, semete, koso, etc.) are discourse markers with two-layered secondary meaning:

1. Presuppositional component: requires a salient counterstance in the discourse
2. Expressive-like component: encodes the speaker's evaluative stance

This dual nature means outlook markers are neither pure CIs nor pure presuppositions, but share properties of both — specifically:

• Like expressives: descriptive ineffability, immediacy
• Unlike expressives: lack independence and nondisplaceability (allow perspective shift)
• Like presuppositions: require discourse antecedent (counterstance)

Three-Way Typology of Secondary Meaning (@cite{kubota-2026}: (14))

Class	Examples	Key Property
Anaphoric presupposition triggers	pronouns, mata 'again'	Discourse-anchored, hard triggers
Lexical preconditions	yameru 'stop', seikō-suru 'succeed'	Soft triggers, overridable defaults
Discourse-sensitive modifiers	nanka, mushiro, dōse, semete	Outlook markers (this file)

Stance Classification

inductive Semantics.Lexical.Expressives.OutlookMarker.StanceType : Type

The type of evaluative stance an outlook marker expresses.

Each stance type characterizes how the speaker situates the prejacent relative to a salient counterstance in the discourse (: §3).

• negative : StanceType

  Negative/pessimistic evaluation: the prejacent is undesirable or implausible. E.g., nanka 'anything like', dōse 'anyway'

• minimum : StanceType

  Minimum standard: the prejacent is the least one could settle for. E.g., semete 'at least', kurai 'at least'

• contrary : StanceType

  Contrary to expectation: the prejacent reverses the expected evaluation. E.g., mushiro 'rather', kaette 'rather', yoppodo 'much more'

• emphasis : StanceType

  Emphatic confirmation: the prejacent is precisely what's expected. E.g., masani 'precisely', koso 'precisely'

instance Semantics.Lexical.Expressives.OutlookMarker.instDecidableEqStanceType : DecidableEq StanceType

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instDecidableEqStanceType x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

def Semantics.Lexical.Expressives.OutlookMarker.instReprStanceType.repr : StanceType → ℕ → Std.Format

Equations

• One or more equations did not get rendered due to their size.

instance Semantics.Lexical.Expressives.OutlookMarker.instReprStanceType : Repr StanceType

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instReprStanceType = { reprPrec := Semantics.Lexical.Expressives.OutlookMarker.instReprStanceType.repr }

instance Semantics.Lexical.Expressives.OutlookMarker.instBEqStanceType : BEq StanceType

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instBEqStanceType = { beq := Semantics.Lexical.Expressives.OutlookMarker.instBEqStanceType.beq }

def Semantics.Lexical.Expressives.OutlookMarker.instBEqStanceType.beq : StanceType → StanceType → Bool

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instBEqStanceType.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)

def Semantics.Lexical.Expressives.OutlookMarker.instInhabitedStanceType.default : StanceType

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instInhabitedStanceType.default = Semantics.Lexical.Expressives.OutlookMarker.StanceType.negative

instance Semantics.Lexical.Expressives.OutlookMarker.instInhabitedStanceType : Inhabited StanceType

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instInhabitedStanceType = { default := Semantics.Lexical.Expressives.OutlookMarker.instInhabitedStanceType.default }

Three-Way Typology of Secondary Meaning

inductive Semantics.Lexical.Expressives.OutlookMarker.SecondaryMeaningClass : Type

Classification of secondary (non-at-issue) meanings following).

This typology cross-cuts the standard CI vs presupposition divide. Each class exhibits different projection behavior, discourse sensitivity, and interaction with attitude predicates.

• anaphoricPresup : SecondaryMeaningClass

  Anaphoric presupposition triggers: pronouns, definites, clefts, additives (too). Presupposition anchored to prior discourse; cannot be overridden. "Hard triggers."

• lexicalPrecondition : SecondaryMeaningClass

  Conditions on lexicalized concepts: aspectuals (stop), factives (know), implicatives (manage, succeed). "Soft triggers" — projectable default assumptions that can be overridden by epistemic uncertainty.

• discourseSensitive : SecondaryMeaningClass

  Discourse-sensitive modifiers and connectives: outlook markers, scalar adverbs (almost, barely), discourse particles (doch, ja), evidentials. Sensitive to discourse state; dual-layered meaning.

instance Semantics.Lexical.Expressives.OutlookMarker.instDecidableEqSecondaryMeaningClass : DecidableEq SecondaryMeaningClass

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instDecidableEqSecondaryMeaningClass x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

instance Semantics.Lexical.Expressives.OutlookMarker.instReprSecondaryMeaningClass : Repr SecondaryMeaningClass

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instReprSecondaryMeaningClass = { reprPrec := Semantics.Lexical.Expressives.OutlookMarker.instReprSecondaryMeaningClass.repr }

def Semantics.Lexical.Expressives.OutlookMarker.instReprSecondaryMeaningClass.repr : SecondaryMeaningClass → ℕ → Std.Format

Equations

• One or more equations did not get rendered due to their size.

def Semantics.Lexical.Expressives.OutlookMarker.instBEqSecondaryMeaningClass.beq : SecondaryMeaningClass → SecondaryMeaningClass → Bool

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instBEqSecondaryMeaningClass.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)

instance Semantics.Lexical.Expressives.OutlookMarker.instBEqSecondaryMeaningClass : BEq SecondaryMeaningClass

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instBEqSecondaryMeaningClass = { beq := Semantics.Lexical.Expressives.OutlookMarker.instBEqSecondaryMeaningClass.beq }

Presupposition Trigger Strength

inductive Semantics.Lexical.Expressives.OutlookMarker.TriggerStrength : Type

Hard vs soft presupposition triggers.

Hard triggers project robustly in all embedding environments. Soft triggers allow non-projective readings under epistemic uncertainty.

shows that Japanese mata 'again' (hard) projects more robustly than yameru 'stop' or seikō-suru 'succeed' (soft).

• hard : TriggerStrength

  Projects robustly under questions, modals, conditionals. E.g., mata 'again', definites, clefts.

• soft : TriggerStrength

  Allows non-projective readings when speaker is epistemically uncertain. E.g., yameru 'stop', seikō-suru 'succeed', factives.

instance Semantics.Lexical.Expressives.OutlookMarker.instDecidableEqTriggerStrength : DecidableEq TriggerStrength

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instDecidableEqTriggerStrength x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

instance Semantics.Lexical.Expressives.OutlookMarker.instReprTriggerStrength : Repr TriggerStrength

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instReprTriggerStrength = { reprPrec := Semantics.Lexical.Expressives.OutlookMarker.instReprTriggerStrength.repr }

def Semantics.Lexical.Expressives.OutlookMarker.instReprTriggerStrength.repr : TriggerStrength → ℕ → Std.Format

Equations

• One or more equations did not get rendered due to their size.

def Semantics.Lexical.Expressives.OutlookMarker.instBEqTriggerStrength.beq : TriggerStrength → TriggerStrength → Bool

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instBEqTriggerStrength.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)

instance Semantics.Lexical.Expressives.OutlookMarker.instBEqTriggerStrength : BEq TriggerStrength

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instBEqTriggerStrength = { beq := Semantics.Lexical.Expressives.OutlookMarker.instBEqTriggerStrength.beq }

Outlook Marker Semantics

structure Semantics.Lexical.Expressives.OutlookMarker.OutlookMeaning (W : Type u_1) : Type u_1

An outlook meaning combines a presuppositional and an expressive component.

The key innovation of: these two layers are not independently stipulated but fall out from the marker's basic function as a counterstance marker. The marker presupposes a salient counterstance in discourse, and the expressive-like evaluation arises from the relationship between the prejacent and this counterstance.

For an outlook marker OM applied to prejacent p:

• prejacent: p (at-issue content, unchanged by the marker)
• counterstance: there exists a salient issue Q in the discourse to which p responds
• stance: the speaker's evaluative orientation toward Q vis-à-vis p

• prejacent : W → Bool

  At-issue content: same as the prejacent without the marker.

• counterstance : W → Bool

  Presuppositional component: a salient counterstance exists in the discourse.

• stance : StanceType

  The type of evaluative stance expressed.

def Semantics.Lexical.Expressives.OutlookMarker.OutlookMeaning.toPrProp {W : Type u_1} (om : OutlookMeaning W) : Core.Presupposition.PrProp W

Extract the presuppositional layer as a PrProp.

The presupposition is the counterstance requirement; the assertion is the prejacent. This connects outlook markers to the standard presupposition projection machinery.

Equations

• om.toPrProp = { presup := om.counterstance, assertion := om.prejacent }

def Semantics.Lexical.Expressives.OutlookMarker.OutlookMeaning.toTwoDimProp {W : Type u_1} (om : OutlookMeaning W) : TwoDimProp W

Extract the expressive layer as a TwoDimProp.

The at-issue content is the prejacent; the CI content is the counterstance (that it is salient). Unlike pure CIs, this "CI" component is discourse-anchored rather than being about the utterance situation per se.

Equations

• om.toTwoDimProp = { atIssue := om.prejacent, ci := om.counterstance }

theorem Semantics.Lexical.Expressives.OutlookMarker.OutlookMeaning.atIssue_eq_prejacent {W : Type u_1} (om : OutlookMeaning W) : om.toTwoDimProp.atIssue = om.prejacent

Outlook markers do not change the prejacent's truth conditions.

This is Kubota's observation that omitting an outlook marker like nanka from (13a) does not change the truth-conditional content of the sentence.

theorem Semantics.Lexical.Expressives.OutlookMarker.OutlookMeaning.presup_eq_counterstance {W : Type u_1} (om : OutlookMeaning W) : om.toPrProp.presup = om.counterstance

The presupposition of an outlook marker is the counterstance requirement.

Diagnostic Properties

@cite{potts-2007} identifies six properties of expressives. shows that outlook markers share some but not all of these, which is what makes them a distinct class of secondary meaning.

structure Semantics.Lexical.Expressives.OutlookMarker.SecondaryMeaningProperties : Type

Properties of a secondary meaning expression, extending @cite{potts-2007}).

These diagnostics distinguish outlook markers from both pure expressives and pure presuppositions.

• independent : Bool

  CI contributes to a dimension separate from at-issue content.

• nondisplaceable : Bool

  Predicates something of the utterance situation (not the described situation).

• perspectiveDependent : Bool

  Evaluated from a particular perspective (usually the speaker's).

• descriptivelyIneffable : Bool

  Cannot be fully paraphrased by descriptive, non-expressive terms.

• immediate : Bool

  Achieves its effect simply by being uttered (like a performative).

• repeatable : Bool

  Repetition strengthens rather than creating redundancy.

• allowsPerspectiveShift : Bool

  Allows perspective shift to a non-speaker attitude holder under embedding.

• requiresDiscourseAntecedent : Bool

  Requires a salient issue/counterstance in prior discourse.

instance Semantics.Lexical.Expressives.OutlookMarker.instReprSecondaryMeaningProperties : Repr SecondaryMeaningProperties

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instReprSecondaryMeaningProperties = { reprPrec := Semantics.Lexical.Expressives.OutlookMarker.instReprSecondaryMeaningProperties.repr }

def Semantics.Lexical.Expressives.OutlookMarker.instReprSecondaryMeaningProperties.repr : SecondaryMeaningProperties → ℕ → Std.Format

Equations

• One or more equations did not get rendered due to their size.

def Semantics.Lexical.Expressives.OutlookMarker.expressiveProfile : SecondaryMeaningProperties

Canonical properties of expressives.

Expressives like epithets and honorifics satisfy all six Potts properties and do NOT typically allow perspective shift or require discourse antecedents.

Equations

• One or more equations did not get rendered due to their size.

def Semantics.Lexical.Expressives.OutlookMarker.outlookMarkerProfile : SecondaryMeaningProperties

Canonical properties of outlook markers (: §3).

Outlook markers share descriptive ineffability and immediacy with expressives, but crucially lack independence and nondisplaceability. They allow perspective shift under attitude predicates (42) and require a discourse antecedent (37–38).

Equations

• One or more equations did not get rendered due to their size.

def Semantics.Lexical.Expressives.OutlookMarker.hardPresupProfile : SecondaryMeaningProperties

Canonical properties of hard presupposition triggers (e.g., mata 'again').

Equations

• One or more equations did not get rendered due to their size.

Key Theorems: What Outlook Markers Share With and How They Differ From Expressives

theorem Semantics.Lexical.Expressives.OutlookMarker.outlook_shares_descriptiveIneffability : outlookMarkerProfile.descriptivelyIneffable = expressiveProfile.descriptivelyIneffable

Outlook markers share descriptive ineffability with expressives.

–(41)): when B denies A's outlook-marked utterance, the denial targets the propositional content, not the evaluative component. The negative evaluation expressed by nanka or dōse is not what "no" targets.

theorem Semantics.Lexical.Expressives.OutlookMarker.outlook_shares_immediacy : outlookMarkerProfile.immediate = expressiveProfile.immediate

Outlook markers share immediacy with expressives.

Both are meta-level effects tied to specific linguistic expressions that cannot be paraphrased by descriptive, non-expressive terms.

theorem Semantics.Lexical.Expressives.OutlookMarker.outlook_lacks_independence : outlookMarkerProfile.independent = false ∧ expressiveProfile.independent = true

Outlook markers LACK independence (unlike expressives).

–(43)): the evaluative meaning of outlook markers interacts with the propositional content and can be attributed to a non-speaker attitude holder, violating the independence criterion.

theorem Semantics.Lexical.Expressives.OutlookMarker.outlook_lacks_nondisplaceability : outlookMarkerProfile.nondisplaceable = false ∧ expressiveProfile.nondisplaceable = true

Outlook markers LACK nondisplaceability (unlike expressives).

): under attitude embedding, nanka/dōse evaluate from the attitude holder's (not the speaker's) perspective.

theorem Semantics.Lexical.Expressives.OutlookMarker.outlook_requires_discourse_antecedent : outlookMarkerProfile.requiresDiscourseAntecedent = true ∧ expressiveProfile.requiresDiscourseAntecedent = false

Outlook markers REQUIRE discourse antecedent (unlike expressives).

–(38)): nanka is felicitous only when a counterstance is salient in the discourse; it is infelicitous as a response to a neutral wh-question where no specific stance is at issue.

theorem Semantics.Lexical.Expressives.OutlookMarker.outlook_allows_perspective_shift : outlookMarkerProfile.allowsPerspectiveShift = true ∧ expressiveProfile.allowsPerspectiveShift = false

Outlook markers ALLOW perspective shift (unlike default expressives).

): "My advisor seems to have thought I wouldn't possibly get accepted at SALT" — the negative evaluation is the advisor's, not the speaker's.

theorem Semantics.Lexical.Expressives.OutlookMarker.outlook_vs_hardPresup_ineffability : outlookMarkerProfile.descriptivelyIneffable = true ∧ hardPresupProfile.descriptivelyIneffable = false

Hard presuppositions differ from outlook markers in descriptive ineffability.

Mata 'again' presupposes "happened before" — this IS paraphrasable. Nanka conveys negative stance — this is NOT paraphrasable.

Modal Selectional Restrictions

–(46)) shows that outlook markers interact differently with different modal flavors. This connects to Kratzer's conversational backgrounds.

structure Semantics.Lexical.Expressives.OutlookMarker.ModalCompatibility : Type

Modal flavors that an outlook marker is compatible with.

semete 'at least' is compatible with deontic -beki and desiderative -tai but NOT with epistemic hazu or ability -eru (: (46)).

This reflects the fact that minimum-standard outlook markers require that the ordering source involve subjective preferences (deontic/bouletic), not objective evidence (epistemic) or factual circumstances (circumstantial).

• epistemic : Bool

  Compatible with epistemic modals (hazu 'supposed', -eru 'can/ability')

• deontic : Bool

  Compatible with deontic modals (-beki 'should')

• circumstantial : Bool

  Compatible with circumstantial/ability modals

instance Semantics.Lexical.Expressives.OutlookMarker.instReprModalCompatibility : Repr ModalCompatibility

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instReprModalCompatibility = { reprPrec := Semantics.Lexical.Expressives.OutlookMarker.instReprModalCompatibility.repr }

def Semantics.Lexical.Expressives.OutlookMarker.instReprModalCompatibility.repr : ModalCompatibility → ℕ → Std.Format

Equations

• One or more equations did not get rendered due to their size.

instance Semantics.Lexical.Expressives.OutlookMarker.instBEqModalCompatibility : BEq ModalCompatibility

Equations

• Semantics.Lexical.Expressives.OutlookMarker.instBEqModalCompatibility = { beq := Semantics.Lexical.Expressives.OutlookMarker.instBEqModalCompatibility.beq }

def Semantics.Lexical.Expressives.OutlookMarker.instBEqModalCompatibility.beq : ModalCompatibility → ModalCompatibility → Bool

Equations

• One or more equations did not get rendered due to their size.
• Semantics.Lexical.Expressives.OutlookMarker.instBEqModalCompatibility.beq x✝¹ x✝ = false

def Semantics.Lexical.Expressives.OutlookMarker.ModalCompatibility.allows (mc : ModalCompatibility) : Core.Modality.ModalFlavor → Bool

Check if a modal flavor is compatible with an outlook marker's selectional restrictions.

Equations

• mc.allows Core.Modality.ModalFlavor.epistemic = mc.epistemic
• mc.allows Core.Modality.ModalFlavor.deontic = mc.deontic
• mc.allows Core.Modality.ModalFlavor.circumstantial = mc.circumstantial

def Semantics.Lexical.Expressives.OutlookMarker.nanka_modalCompat : ModalCompatibility

nanka is compatible with all modal flavors, but the evaluative force varies: deontic/bouletic → pejorative, epistemic → more neutral (45).

Equations

• Semantics.Lexical.Expressives.OutlookMarker.nanka_modalCompat = { epistemic := true, deontic := true, circumstantial := true }

def Semantics.Lexical.Expressives.OutlookMarker.semete_modalCompat : ModalCompatibility

semete 'at least' selects for deontic/desiderative ordering sources. Incompatible with epistemic hazu and ability -eru (46a,b).

Equations

• Semantics.Lexical.Expressives.OutlookMarker.semete_modalCompat = { epistemic := false, deontic := true, circumstantial := false }

theorem Semantics.Lexical.Expressives.OutlookMarker.semete_rejects_epistemic : semete_modalCompat.allows Core.Modality.ModalFlavor.epistemic = false

semete rejects epistemic modals.

theorem Semantics.Lexical.Expressives.OutlookMarker.semete_accepts_deontic : semete_modalCompat.allows Core.Modality.ModalFlavor.deontic = true

semete accepts deontic modals.

theorem Semantics.Lexical.Expressives.OutlookMarker.nanka_accepts_all_modals : nanka_modalCompat.allows Core.Modality.ModalFlavor.epistemic = true ∧ nanka_modalCompat.allows Core.Modality.ModalFlavor.deontic = true ∧ nanka_modalCompat.allows Core.Modality.ModalFlavor.circumstantial = true

nanka accepts all modal flavors (though evaluation force differs).

Projection Behavior

Outlook markers exhibit both presuppositional and CI-like projection. The counterstance requirement projects (like a presupposition), while the evaluative stance projects (like a CI). This dual behavior is what makes them interesting.

theorem Semantics.Lexical.Expressives.OutlookMarker.counterstance_projects_through_neg {W : Type u_1} (om : OutlookMeaning W) : om.toPrProp.neg.presup = om.counterstance

The counterstance requirement projects through negation.

If an outlook marker appears under negation, the counterstance is still presupposed. This follows from the presuppositional layer (PrProp.neg preserves presupposition).

theorem Semantics.Lexical.Expressives.OutlookMarker.stance_projects_through_neg {W : Type u_1} (om : OutlookMeaning W) : om.toTwoDimProp.neg.ci = om.counterstance

The evaluative stance projects through negation (CI-like behavior).

This follows from the CI layer (TwoDimProp.neg preserves CI content).