What distinguishes full‐on stating a proposition from merely communicating it? For instance, what distinguishes claiming/asserting/saying that one has never smoked crack cocaine from merely implying/conveying/hinting this? The enormous literature on ‘assertion’ provides many approaches to distinguishing stating from, say, asking and commanding: only the former aims at truth; only the former expresses one's belief; etc. But this leaves my question unanswered, since in merely communicating a proposition one also aims at truth, expresses a belief, etc. My aim is not to criticize extant accounts of the state‐versus‐merely‐convey contrast, but rather to draw on clues from Dummett, functional linguistics and moral theory, to offer a novel one. The main idea is that full‐on stating is distinctively conventionalized in a way that conversationally implicating, hinting, giving to understand, etc., are not. Specifically, full‐on stating is constitutively tied to a particular conventional, linguistic, function‐bearing device, the declarative sentence. To full‐on state that p is to hit that ‘target speech act’ which owes its existence to that special‐purpose device. It is therefore also to make one's action lie‐prone. Nonetheless, once that sui generis target is there to be aimed for, a person may reach it without using the special‐purpose tool—e.g. one may full‐on state using a mere word or phrase, or coded hand signals, or semaphore. I end by considering several philosophical implications of this means of capturing the contrast.