Scientific Representation — Stanford Encyclopedia of Philosophy

Supports treating representations as tools about targets, not as identical with their targets; also establishes that representation can remain epistemically useful despite incompleteness.