From Geuss's 'Liberalism and its Discontents', Political Theory, 2002.
[I]f some of the deficiencies inherent in adopting a pure normative standpoint are visible even in a philosopher who has moved as far beyond Kant as Rawls has, this seems to me to give further weight to suspicions about the normative standpoint as a whole.
Setting aside the 'pure' for the time being, I think it's worth marvelling at just how perfect a piece of self-disembowelling this is. Our suspicions about the very idea of a normative standpoint and its deficiencies can have more and less weight; by what measures are we to establish this weight, and by what standards are we to judge these failings? Presumably not normative ones, since they ground suspicions about the very possibility of normativity, and so presumably not ones which are supposed to compel or even count in favour of agreement on them. I suppose anyone who disagrees with Geuss then has no reason to carry on reading him then.