I agree that in far mode we are more likely to accept propositions without sufficient evidence, and that much of our effective learning of procedures happens in near mode. Nevertheless, we can and do create valid propositions in far mode, and we do sometimes learn far mode procedures.