Both the theories coming out of the linguistic turn and those running away from it have placed special emphasis on human language (or human symbolic thinking) as a matter of convention and shared meanings. Yet there are other histories that link language and humanness through invention, deceit, and secrecy rather than through convention and publicness. These alternate models have been used as diagnostic of humanness in a range of contexts, from the colonial past into the technologized present. I examine here the ways in which the unshared, non-public symbol has stood at the center of two disparate contexts in which the humanness of speakers of novel languages are put in question. The first case examines the ways in which Christian missionaries started to see Tok Pisin, a novel pidginized language spoken by indentured laborers in colonial Papua New Guinea, as a possible language of evangelism when it became associated with deceit and moral dissolution. The second case examines a 2017 moral panic in the United States about two chatbots that were reported to have invented their own language and then used it to lie to one another. In contrast to the first case, one of the ways that bots get figured as beyond-human is in the fear that there is no way to impose a moral order, no colonial evangelism that could be used to encompass them. By taking on the symbolic while withholding public meanings, the speakers of these unshared symbols sit at the boundaries of humanness.