Cassirer and Reichenbach on the conventionality of geometry

Much of the scholarship on the relativization of the a priori has taken its starting point from the way in which Reichenbach, in The Theory of Relativity and A Priori Knowledge (1920), accounts for the possibility of varying the coordinating principles linking abstract mathematical structures to empirical reality. Such are the principles that Reichenbach called a priori in the sense of constitutive of the objects of experience but no longer valid for all time as Kant thought. However, this poses the problem of distinguishing the relativization of the Kantian a priori from conventionalism. This paper aims to reconsider this issue in the example of Cassirer's and Reichenbach's discussions of the status of geometry in general relativity. It will be argued that both Cassirer and Reichenbach, when his work is read against the neo-Kantian background of the debate, offer a genuine option, according to which the relativization of the Kantian a priori depends on the interrelation of different levels of constitutivity.