Kant’s Critique of Pure Reason contains an original and powerful semantics of singular, specifically cognitive reference which has important implications for epistemology and for philosophy of science (and which is entirely independent of Transcendental Idealism). Here I argue that Kant’s cognitive semantics directly and strongly supports Newton’s Rule 4 of Experimental Philosophy in ways which support Newton’s realism about gravitational force. I begin with Newton’s Rule 4 and its role in Newton’s justification of realism about gravitational force (§ 1). I then briefly summarize Kant’s semantics of singular cognitive reference (§ 2), and show that it is embedded in, and strongly supports, Newton’s Rule 4, and that it rules out not only Cartesian physics (per Harper) but also Cartesian, infallibilist presumptions about empirical justification generally (§ 3). This result exposes a key fallacy in Bas van Fraassen’s main argument for his anti-realist Constructive Empiricism, and in many common objections to realism (§ 4). These problems reveal yet a further important regard in which Constructive Empiricism is not (so to speak) ‘empirically’ adequate, not even to Classical Newtonian Mechanics (§ 5). This inadequacy of Constructive Empiricism highlights a chronic empiricist misunderstanding of Newton’s mechanics (§ 6). Finally, Kant’s cognitive semantics improves upon the semantic interpretation of scientific theories, and rectifies the presumption that laws of physics literally ‘lie’ (§ 7). Thus Kant and Newton still have invaluable lessons for contemporary philosophy and history of science (§ 8).