Lots of folks nowadays think there is an intimate connection between what we assert and what we know. Talk of this connection is largely oriented around Timothy Williamson’s claim that you shouldn’t assert p unless you know p. Hereafter, I will treat this claim as follows: -/- (KNA) Don’t assert that p unless so asserting expresses your knowledge that p. -/- (KNA) is for “Knowledge Norm of Assertion”. -/- A primary aim here is to defend the KNA. However, getting in (...) the best position to do so requires attention to an older strain in the discussion of how knowledge and assertion are interrelated. This older strain is oriented less around explicit norms of assertion, and more around what assertion does. Said G.E. Moore, “by asserting p positively you imply, though you don’t assert, that you know that p” (1962: 277). Peter Unger put it this way: “If someone asserts, states, or declares that something is so, then it follows that he represents himself as knowing that it is so” (1975: 253). I will abbreviate this claim as -/- (ARK) Asserting that p represents the speaker as knowing that p. -/- (ARK) for “Assertion Represents Knowledge”. -/- The central contention of this paper is this: if ARK is true, we should accept the KNA too. S shouldn’t assert that p unless p expresses S’s knowledge because, given ARK, when S asserts that p, S gives her interlocutors entitlement to infer that S knows that p. An assertion that p that does not express knowledge that p is thus defective, however blameless or praiseworthy it may be for other reasons. (As we’ll see, you can accomplish noble, worthy things through infelicitous assertions.) The argument will proceed as follows. §1 musters some of the best evidence for thinking that ARK is true. Some of this evidence is re-appropriated from arguments for the KNA; yet more is extracted from proposed counterexamples to the KNA that (plausibly) apply ARK. §2 builds on Ishani Maitra’s (2011) treatment of constitutive rules to make the case for treating ARK, rather than the KNA, as the constitutive rule of assertion. §3 builds on the results of the previous sections to show the advantages of a KNA grounded in ARK. (shrink)