Using empirical research on pain, sound and taste, I argue against the combination of intentionalism about consciousness and a broadly ‘tracking’ psychosemantics of the kind defended by Fodor, Dretske, Hill, Neander, Stalnaker, Tye and others. Then I develop problems with Kriegel and Prinz's attempt to combine a Dretskean psychosemantics with the view that sensible properties are Shoemakerian response-dependent properties. Finally, I develop in detail my own 'primitivist' view of sensory intentionality.