On Induction as Inversion of Deduction