On the Rational Reconstruction of our Theoretical Knowledge