The paper argues that an internal debate within Wittgensteinian philosophy leads to issues associated rather with the later philosophy of Martin Heidegger. Rush Rhees's identification of the limitations of the notion of a “language game” to illuminate the relation between language and reality leads to his discussion of what is involved in the “reality” of language: “anything that is said has sense-if living has sense, not otherwise.” But what is it for living to have sense? Peter Winch provides an interpretation and application of Rhees's argument in his discussion of the “reality” of Zande witchcraft and magic in “Understanding a Primitive Society”. There he argues that such sense is provided by a language game concerned with the ineradicable contingency of human life, such as (he claims) Zande witchcraft to be. I argue, however, that Winch's account fails to answer the question why Zande witchcraft can find no application within our lives. I suggest that answering this requires us to raise the question of why Zande witchcraft “fits” with their other practices but cannot with ours, a question of “sense” which cannot be answered by reference to another language game. I use Joseph Epes Brown's account of Native American cultures (in Epes Brown 2001) as an exemplification of a form of coherence that constitutes what we may call a “world”. I then discuss what is involved in this, relating this coherence to a relation to the temporal, which provides an internal connection between the senses of the “real” embodied in the different linguistic practices of these cultures. I relate this to the later Heidegger's account of the “History of Being”, of the historical worlds of Western culture and increasingly of the planet. I conclude with an indication of concerns and issues this approach raises, ones characteristic of “Continental” rather than Wittgensteinian philosophy.