RB 65

This does not imply that empirical knowledge based upon and synthesized through inductions is worthless, on the contrary, for if perceptions are assumed not to provide a progressively growing body of knowledge of the external world, then knowledge in all of its forms becomes impossible, including knowledge of the subject solus ipse.304 In fact, Hägerström argues that it, contrary to the epistemological absolutism of metaphysics, is the relativity of the inductive inference itself that makes external knowledge possible. One can thus argue that this is the best method at hand when attempting to determine external knowledge, because the alternative method, deduction, because of its analytic nature, will only estrange us from the aforementioned form of knowledge.While induction provides us with knowledge of empirical affairs, induction simultaneously invalidates the assumed necessary certainty of absolute or objective concepts.The reason is that empiricism itself only provides concepts that strictly speaking are accidentally valid, but never apodictically valid, and these empirical concepts, once established, will thus only have conditional objectivity.305 Hence, from a transcendental vantage point what is problematic is that the universality of synthetic concepts, strictly speaking, is accidental. Hägerström’s view on induction can be summarized as follows. It is through observations and inductive inferences that we determine the physical world to which we belong, and which we necessarily must presuppose in the quest for object-oriented knowledge. In essence, our notion of the physical world has been formed through observations and inferences drawn from the former, whereby any attempt to disregard empirical facts and realities, as well as direct sensory observations of the physical world is contradictory and nonsensical, in fact the usefulness of induction is repeatedly demonstrated by the fact that it allows us to a ca l l f o r s c i e n t i f i c p u r i t y 137 304 Ibid., p. 206. 305 Hägerström, Stat och rätt, p. 82.