A primary intension is the component of a term’s meaning that determines its referent based on how things turn out to be — the epistemic dimension of intension. David Chalmers introduced the concept as part of his two-dimensional semantic framework, which separates two ways a term can determine reference.

The primary intension of a term is a function from scenarios (considered as possible ways the actual world might be) to extensions. It captures what a competent speaker would recognize as the referent given full information about the world. For “water,” the primary intension maps each scenario to whatever substance fills the water-role in that scenario — the clear, drinkable liquid that falls from the sky, fills the lakes, and so on.

This means the primary intension is sensitive to empirical discovery. Before anyone knew the chemical composition of water, a speaker’s primary intension for “water” would pick out whatever turned out to play the water-role. In a scenario where the lakes and rain consist of XYZ rather than H2O, the primary intension of “water” would pick out XYZ.

Primary intension contrasts with secondary intension, which fixes reference based on how the world actually is. Once we discover that water is H2O in the actual world, the secondary intension of “water” picks out H2O in every counterfactual world, regardless of what plays the water-role there.

The distinction helps explain why “water is H2O” is epistemically contingent (we had to discover it) but metaphysically necessary (it holds in every possible world). The primary intension leaves open what water turns out to be; the secondary intension locks it down.