Modal Indefinites: Cross-Linguistic Typology & Kratzer Anchoring #
@cite{alonso-ovalle-royer-2024} @cite{alonso-ovalle-menendez-benito-2010} @cite{alonso-ovalle-menendez-benito-2018} @cite{coon-2019} @cite{hacquard-2006} @cite{alonso-ovalle-royer-2021} @cite{chierchia-2013} @cite{jayez-tovena-2006} @cite{kratzer-shimoyama-2002}
Cross-linguistic typology of modal indefinites and bridge theorems connecting
the event-relative modality theory (@cite{hacquard-2006}, formalized in
Theories/Semantics/Modality/EventRelativity) to empirical observations.
Lexical entries are defined in Fragment files (single source of truth):
Fragments/Chuj/ModalIndefinites.lean: yalnhej, komonFragments/Spanish/ModalIndefinites.lean: algún, uno cualquieraFragments/German/ModalIndefinites.lean: irgendeinFragments/French/ModalIndefinites.lean: n'importe quelFragments/Italian/ModalIndefinites.lean: un qualsiasi
Three Dimensions of Variation (§6) #
- Status: Is the modal component at-issue or not-at-issue?
- Content: Which modal flavors does the component support?
- Upper-boundedness: Does the indefinite impose an anti-singleton inference (¬∀x[P(x) → Q(x)])?
Key Bridge Theorems #
- Position → Anchor: The syntactic position of the yalnhej DP determines which anchoring functions are semantically productive.
- Anchor → Flavor: Speech event → epistemic; described event → RC.
- Volitionality: RC requires a volitional verb (decision subevent).
- Voice → Position: Chuj voice system determines external/internal.
- Cross-linguistic predictions: Three-dimensional typology instantiated.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Chuj yalnhej and German irgendein share the same flavor inventory (epistemic + random choice) but differ in status.
The at-issue / not-at-issue split (§6.1).
Upper-bounded items are a proper subset: only algún and uno cualquiera impose anti-singleton inferences.
Yalnhej is the only item that is both at-issue AND has both epistemic and random choice flavors. This is the core empirical contribution of @cite{alonso-ovalle-royer-2024}.
The three dimensions are logically independent: we find items in multiple cells of the 2×2 (status × upper-bounded) matrix.
Syntactic positions for a DP in Chuj, cross-classified with verb volitionality (§3–4, Table 5).
The paper shows that RC availability depends on TWO factors: (1) structural position (external vs internal/adjunct), and (2) whether the verb describes a volitional event (one containing a decision subevent that can anchor RC modality).
- externalArg : ChujDPPosition
External argument (subject of transitive)
- internalArgVolitional : ChujDPPosition
Internal argument of a volitional verb (e.g., "buy")
- internalArgNonVolitional : ChujDPPosition
Internal argument of a non-volitional verb (e.g., "like")
- adjunctVolitional : ChujDPPosition
Adjunct of a volitional verb (e.g., "Malin ate where")
- adjunctNonVolitional : ChujDPPosition
Adjunct of a non-volitional verb (e.g., "it rained where")
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Which modal flavors are available to yalnhej in each position (Table 5, §3.2–4.2).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Position sensitivity: external ≠ volitional internal flavor sets.
Volitionality sensitivity: volitional ≠ non-volitional internal flavor sets (§4.1).
A Chuj yalnhej example sentence with empirical judgments.
- chuj : String
- gloss : String
- position : ChujDPPosition
- availableFlavors : List Core.Modality.ModalFlavor
- exampleNumber : String
Example number in @cite{alonso-ovalle-royer-2024}
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.
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.
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.
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Consistency: each example's available flavors match the position-based prediction from yalnhejFlavorsAt.
Yalnhej is compatible with partial-domain scenarios, unlike maximal free relatives (English whatever).
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.
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Some modal indefinites have "unremarkable" (plain existential) readings in addition to their modal readings. @cite{alonso-ovalle-royer-2024} (§5) correlate this with predicativity.
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.
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.
Instances For
Equations
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.
Instances For
Predicativity correlates with unremarkable readings across all data.
Cross-check: fragment entry fields agree with independent empirical data.
Predicativity correlates with unremarkable readings across all entries.
Under an external modal (imperative, deontic, attitude verb), the MI's anchor can be co-indexed with the modal's event, giving "any X is fine" readings.
- imperative : EmbeddingModal
- deontic : EmbeddingModal
- attitudeVerb : EmbeddingModal
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.
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.
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Same surface string, two readings: (82) non-harmonic and (85) harmonic share the same Chuj form but differ in anchor co-indexing.
Which anchor types are semantically productive given a syntactic position (cross-classified with verb volitionality).
This is A-O&R's core structural explanation for the position-sensitivity of yalnhej's modal flavor.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The predicted flavors for a position = the flavors of its available anchor types.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The anchoring theory predicts exactly the observed flavor distribution for ALL five position types in Table 5.
Active transitive voice (vØ): agent = external → epistemic only.
Passive voice (-ch): theme promoted, agent implicit → both flavors.
Upper-bounded modal indefinites impose an anti-singleton inference (§5).
Non-upper-bounded modal indefinites: no anti-singleton.
Yalnhej is not upper-bounded: compatible with partial-domain
scenarios. The EventRelativity worked example demonstrates this
concretely with yalnhej_nonmaximal_ab.