Source of information as a category
Evidentiality concerns the source or kind of evidence associated with a statement. A system may distinguish direct experience from indirect evidence, or make more specific distinctions such as visual, inferred and reported information. These category names are comparative approximations. Their use in a particular language can involve perspective, conventional expectations and interactions with other grammatical features.
Evidence is not certainty
A report can be trusted, and an eyewitness can be mistaken. Evidential marking therefore should not be collapsed into a probability scale from doubtful to certain. Some forms do have conventional epistemic effects, but those effects require their own analysis. Languages without grammatical evidential categories can still state how information was obtained through independent expressions, such as saying that someone reported an event.
A source switch has a limited job
The experiment changes the displayed information path while keeping the event proposition fixed. It illustrates why “I saw it” and “I heard about it” add different commitments without deciding whether the event is true. It does not judge which evidential a real speaker must choose. Published examples and their discourse context are needed for that question, including whether evidential marking is obligatory in the construction under discussion.
Evidentiality categorizes information sources; it is not a universal truthfulness or confidence meter.
Try it for yourself
Keep the event fixed and switch between direct observation, inference and report.
- Semantic Distinctions of EvidentialityFerdinand de Haan · 2013; WALS Online v2020.6Rights & edition ↗
- Coding of EvidentialityFerdinand de Haan · 2013; WALS Online v2020.6Rights & edition ↗