Sentences like (1) present familiar puzzles for the familiar idea that declarative sentences of a natural language have truth conditions.

(1) The first numbered sentence in 'Framing Event Variables' is false.

Action reports like (2) and (3), which might be used to describe a scene in which two chipmunks chased each other, illustrate other (perhaps even harder) puzzles for this idea.

(2) Alvin chased Theodore gleefully and athletically but not skillfully.

(3) Theodore chased Alvin gleelessly and unathletically but skillfully.

I'll argue against various (broadly Davidsonian) attempts to reconcile intuitions regarding (2) and (3) with the claim that these sentences have truth conditions. In my view, the puzzles reflect deep "framing effects" – of the sort that Kahneman and Tversky made famous, though my central example is due to Thomas Schelling. If this is the right of diagnosis of the puzzles regarding sentences like (2) and (3), then I think we need a conception of linguistic meaning according to which sentence meanings do not determine truth conditions, not even relative to contexts. And as it happens, I've been peddling such a conception for a while now: it's better to think of meanings as instructions for how to build concepts, which might be used (when conditions allow) to form truth-evaluable judgments in contexts.