There is a fundamental disagreement about which norm regulates assertion. Proponents of factive accounts argue that only true propositions are assertable, whereas proponents of nonfactive accounts insist that at least some false propositions are. Puzzlingly, both views are supported by equally plausible (but apparently incompatible) linguistic data. This paper delineates an alternative solution: to understand truth as the aim of assertion, and pair this view with a non-factive rule. The resulting account is able to explain all the relevant linguistic data, and finds independent support from general considerations about the differences between rules and aims