@cite{izvorski-1997}: The Present Perfect as an Epistemic Modal — Data @cite{izvorski-1997} #
Empirical data from @cite{izvorski-1997}. In Bulgarian, Turkish, Norwegian, and other languages, present perfect morphology doubles as an indirect evidential (the "Perfect of Evidentiality" = PE). The paper's central proposal (8):
The indirect evidential Ev is an epistemic modal which: (i) has universal quantificational force, (ii) has a presupposition that the evidence for the core proposition is indirect.
The key empirical contrasts establishing (8):
- Ev vs. must ((10)–(13)): Both are epistemic necessity modals (same □ force), but Ev restricts the modal base to indirect evidence only. Must allows any epistemic base. The difference is in the base, not the force.
- Presupposition diagnostics ((14)–(16)): The indirect-evidence requirement is a presupposition (not an implicature) — it resists cancellation (14), projects past negation (15), and denial targets the assertion (16).
Languages exhibiting the Perfect of Evidentiality (@cite{izvorski-1997}, fn. 1). The paper's body text discusses Bulgarian, Turkish, and Norwegian; footnote 1 lists ~25 languages across 6 families.
- bulgarian : PELanguage
- turkish : PELanguage
- norwegian : PELanguage
- macedonian : PELanguage
- albanian : PELanguage
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
Instances For
Binary evidence basis: Izvorski's central contrast variable. The paper argues that Ev and must have the same quantificational force (□) but differ in whether the modal base is restricted to indirect evidence only.
- direct : EvidenceBasis
- indirect : EvidenceBasis
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
Instances For
A data point from the Ev/must paradigm. The paper's argument (§3, pp. 227–229):
- (10)–(11): With indirect evidence, both Ev and must are felicitous
- (12)–(13): Ev + "I have no evidence" → contradictory; must + "I have no evidence" → acceptable (must doesn't presuppose indirect evidence)
- Prose (p. 228): With direct evidence (speaker witnessed the event), Ev is infelicitous; must is fine
- evidenceBasis : EvidenceBasis
- evFelicitous : Bool
- mustFelicitous : Bool
- label : String
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
- One or more equations did not get rendered due to their size.
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.instBEqEvMustDatum.beq x✝¹ x✝ = false
Instances For
Indirect evidence context: both Ev and must felicitous. Paper (10)–(11): "Knowing how much John likes wine..."
Equations
- One or more equations did not get rendered due to their size.
Instances For
Direct evidence context: Ev infelicitous, must fine. Paper prose (p. 228): when speaker has direct evidence (witnessed the event), PE is infelicitous but must is acceptable.
Equations
- One or more equations did not get rendered due to their size.
Instances For
All Ev/must data points.
Equations
Instances For
Standard presupposition diagnostics applied to the evidential.
- cancellation : PresupDiagnostic
- projection : PresupDiagnostic
- denial : PresupDiagnostic
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
A presupposition diagnostic datum.
- diagnostic : PresupDiagnostic
- evidentialSurvives : Bool
- label : String
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
- One or more equations did not get rendered due to their size.
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.instBEqPresupDiagnosticDatum.beq x✝¹ x✝ = false
Instances For
(14): Cancellation fails — "Maria apparently kissed Ivan. # I witnessed it." The indirect-evidence requirement cannot be cancelled, so it is a presupposition, not an implicature.
Equations
- One or more equations did not get rendered due to their size.
Instances For
(15): Projection under negation — "Apparently, Ivan didn't pass the exam." The indirect-evidence requirement projects past negation: the speaker still has indirect evidence; what's negated is that Ivan passed.
Equations
- One or more equations did not get rendered due to their size.
Instances For
(16): Denial targets assertion — "Ivan passed-PE the exam. That's not true." The denial targets p (Ivan passed), not the evidential content (that the speaker has indirect evidence).
Equations
- One or more equations did not get rendered due to their size.
Instances For
All presupposition diagnostic data.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Ev requires indirect evidence: felicitous with indirect, infelicitous with direct. This captures (8ii).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Must allows both evidence bases — no presupposition on evidence type.
Instances For
All data points satisfy the indirect-evidence generalization.
All data points satisfy the must-allows-both generalization.
All diagnostics confirm presupposition status (not implicature).
Izvorski's EV operator (formalization of (17)–(19) + (8ii)).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.johnDrank Core.Proposition.World4.w0 = true
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.johnDrank Core.Proposition.World4.w1 = true
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.johnDrank Core.Proposition.World4.w2 = false
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.johnDrank Core.Proposition.World4.w3 = false
Instances For
Equations
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.bottlesEmpty Core.Proposition.World4.w0 = true
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.bottlesEmpty Core.Proposition.World4.w1 = false
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.bottlesEmpty Core.Proposition.World4.w2 = true
- Phenomena.TenseAspect.Studies.Izvorski1997.Data.bottlesEmpty Core.Proposition.World4.w3 = false