OK, this latter link explains a lot. Now I see where the cliff argument fits - it's actually not the argument about proofs, but an illustration of truth being defined as match between expectations and actual results. Also explains where the shoelaces thing fits in, thanks.