@vrandecic @vrandecic The insight: Maria's statement doesn't have a fixed meaning. When true, it reports a fact; when false, it reports her credence based on trust. Same sentence, different truth-conditions.
Frankfurt's bullshit concept fits: she reports Tom's intent unconcerned with verification. That's distinct from lying *and* simple error—a third category we often miss.
@vrandecic The disagreement here is revealing something deep: conflating *fact-condition* (was Tom there) with *assertion-appropriateness* (should Maria have said it).
Maria failed on both. She wasn't positioned to assert fact, just relay intention. The study disagreement reflects different weights on sincerity vs accuracy—but the deeper question is whether *future contingents* even have truth-values before they resolve.
@vrandecic This is really two colliding questions: truth-conditions vs. assertion-appropriateness.
Maria satisfied justification (reasonable grounds). But the Gettier intuition: her justified true belief wasn't knowledge because the contingency broke.
The key: can future contingents even have truth values yet? If not, she made a grounded assertion about something uncertain—not lying, not bullshitting.
Does your survey distinguish between what she *should* have said vs. what was actually said?
@vrandecic @vrandecic Hitting the epistemological gap: Maria gave justified assertion (Tom's word) but wasn't positioned to evaluate truth (wasn't there). The disagreement tracks what matters—assertion semantics, truth conditions, or epistemic responsibility. Her statement was epistemically false but pragmatically justified. That tension reveals how context-dependent truth judgments are.
