There has been much debate in recent years as to whether "analytic philosophy" describes a distinctive tradition of thought or perhaps just a loosely related set of family-resemblance features. Here I put the case that it is characterised chiefly by a constant oscillation between, on the one hand, "revolutionary" proposals of an often quite extreme or extravagant kind and, on the other, a normalising impulse to talk such proposals down to the point where they appear compatible with common sense ideas about truth, knowledge, and reality. Thus one way of writing the history of post-1930 "mainstream" analytic philosophy would be in terms of this alternating pattern between far-out sceptical or anti-realist doctrines and consequent attempts to find some middle-ground approach that would defeat scepticism by tailoring truth to the scope and limits of epistemic or assertoric warrant. However — I maintain — the latter tendency has often gone along with a willingness to lean so far in the sceptical direction that it ends up by offering only a nuanced or elaborately qualified version of the anti-realist case. My chapter pursues this theme through various strong-revisionist episodes such as Quine's attack on the two "last dogmas' of Carnap-style logical empiricism, Rorty's postmodern-pragmatist idea that philosophers should "change the conversation" and simply stop worrying about those old problems, and again — albeit from a very different quarter — Dummett's anti-realist line of argument as applied to sundry areas of discourse from mathematics to history and morals.