TRANSCENDENTAL AND EMPIRICAL
'All this clearly points to the conclusion that transcendental questions allow only of transcendental answers . . .' (B 665).
IT has been already suggested (Ch. 1, p. 16) that the distinction between transcendental and empirical bears directly on Kant's aim in the Deduction; and many commentators have recognised the importance of this distinction in the Critique. One of the earliest, Mellin, understood Kant to mean that the objects said to affect our senses do so in a dual way. On his view Kant thought that although these objects were themselves empirical, yet they had an effect both upon our empirical and our transcendental sensibility. Much more recently a similar position was reached by Adickes, who suggested a complex 'double affection' between appearances and things in themselves on one side and empirical and noumenal selves on the other. Such views are more simply illustrated by diagrams such as those which appear in Kemp Smith Commentary (p. 281) or Weldon Kant's Critique of Pure Reason ( 2nd ed., p. 253). Yet such diagrams presuppose that the objects drawn in them can be easily distinguished, when in reality nobody has been able to give a clear and unobjectionable sense to the notions of a transcendental object or a transcendental self.
There is at least one passage in Kant's writings where he notices and firmly rejects duplications of entities in this way. At the time when he was writing the Prolegomena he criticised Berkeley for needlessly duplicating mental phenomena, by bringing them into relation both with appearances and things in themselves. He said