Recent years have seen a growing interest in language among political philosophers. While being, on the whole, a positive development, the considerable focus of this literature on the issue of linguistic justice paradoxically limits the breadth and depth of that emerging inquiry, by reducing the conceptual vocabulary intrinsic to the much broader notion of "language ethics". The paper examines the general theoretical question of the primacy of justice, and questions the adequacy of its prioritisation of the right (i.e. justice) over the good (i.e. benevolence) with regard to language. Such prioritising, it is suggested, does not only diminish the importance of the good in a moral philosophy of language, but also risks, in certain cases, its challenging and even rejection. The paper then moves on to explore the notion of the "dialogical self" as a preferable alternative for theorising language ethics (including linguistic justice) in a manner that is more commensurable with the simultaneous function of language as both a medium and a message of moral relationality. The intrinsic plurality of a dialogical self-based notion of language ethics, it is proposed, calls for a wider moral vocabulary that, alongside justice, comprises other key elements, such as friendship, empathy and hope.