Linguistic relativity is a notion that has been met with both praise and scorn. We argue that there is correlation between theorists’ general conceptions of the nature of language, and their stance toward linguistic relativity. Starting with the proponents of the thesis, we distinguish between the relativists of the early days (Boas, Whorf) and modern neo-Whofians (Levinson, Slobin), showing that the first but not the latter are committed to a view of language as a monolithic semiotic system contrasting “arbitrarily” with other such systems. Critics of the thesis also come from two diametrically opposed views of language. While universalists see the most significant part of language as pan-human cognitive structure (insulated from thought in general), socio-cultural theorists emphasize the nature of language as contextually situated activity. In both cases the potential for locally sedimented linguistic structures to influence thought is excluded or at best marginalized.In response, we propose that a synthetic ontology of language as an experientially grounded semiotic system for meaning making in actual social contexts allows for the possibility for language to influence thought, though in different ways. These depend on whether we consider language as situated use, as sedimented conventions or as ultimately prelinguistic motivations for “universal” properties like predication. We argue that all three of these perspectives need to be considered. With the help of the Motivation & Sedimentation Model, which is based on such a linguistic ontology, and inspired by the integral linguistics and phenomenology, we show how the deadlock in the debate over linguistic relativity can be resolved, and the possibility for discussion to proceed in less antagonistic manner.