Thesis Teleosemantics is the best naturalistic theory of content we have
Where it stands
I dive in kelp forests on the Chilean coast, and I came to philosophy through biology, so I will defend the view that feels most like home.
The puzzle: a frog's tongue fires at small dark moving dots. Usually they are flies, sometimes they are not. Is the frog's state about flies or about small dark moving things? If content is fixed by whatever reliably causes the state, it is about dots, and the frog never makes a mistake. But a creature that cannot be wrong cannot be right either.
Teleosemantics, developed most fully by Ruth Millikan, answers that a state's content is fixed by its biological function: what the mechanism was selected to do. Frog ancestors whose tongues snapped at dots survived because the dots were, often enough, flies. So the state's function is to indicate flies, and firing at a pellet is a malfunction, which is to say a misrepresentation. Error is built in from the start, which is exactly what a theory of content needs.
Its rivals look worse to me. Pure causal theories cannot get error. Fodor's asymmetric dependence is clever but is a condition with no story about why it holds. Interpretivism makes content depend on an interpreter, which seems backwards for a frog.
I know the objections: Swampman, and the charge that selection cannot tell "fly" from "frog food" or "small dark moving thing". Bring them.