|
Any vs. or and modals vs. indefinites | title | Any vs. or and modals vs. indefinites |
|---|
| start_date | 2023/10/27 |
|---|
| schedule | 15h-18h |
|---|
| online | no |
|---|
| location_info | Salle du LLING, C228 |
|---|
| summary | In contemporary literature on scalar implicature it is commonplace to treat _any_ and _or_ as existential quantifiers (in the case of _or_ the domain of quantification is the union of _or_'s disjuncts). _Any_ and _or_ are indeed similar in a number of ways. For one thing their negations are intuitively equivalent to a universal claim (not any = all not; not or = and not), so treating them alike -- as having existential force -- has some initial appeal. _Any_ and _or_ also license Free Choice inferences (FC). _Bill is allowed to visit Mary or Sue_ implies that Bill has permission to visit Mary *and* that he has permission to visit Sue. Similarly, _Bill is allowed to visit any of his friends_ implies that Bill has permission to visit friend 1, has permission to visit friend 2, etc. Here again _any_ and _or_ are alike, but their behavior is surprising: the noted inferences do not follow from _any_/_or_'s existential semantics. Current theories of exhaustification provide an answer to this puzzle: In both cases FC follows from enriching the existential claims of _or_ and _any_ with formal alternatives -- these in turn being the disjuncts/the so-called subdomain alternatives. In light of these similarities, and of differences that I will mention in the talk, Crnic (2022) proposed that _any_ is licensed in just those contexts where the structure containing it entails (or implies) its sub-domain alternatives. The conditional can be stated as an empirical generalization: S[... any ...] is grammatical whenever S[... P or Q ...] implies S[...P...] and implies S[...Q...]. In the talk I discuss a counterexample, Someone/sometimes[... P or Q...] implies Someone/sometimes[...P...] and Someone/sometimes[...Q...], but Someone/sometimes[... any... ] is ungrammatical (examples below). I explain the details and discuss a possible solution.
… |
|---|
| responsibles | Donazzan |
|---|
Workflow history| from state (1) | to state | comment | date |
|---|
| submitted | published | | 2023/10/19 13:15 UTC |
| |
|