He demonstrated that the general type of reasoning is neither from generals to particulars, nor from particulars to generals, but from particulars to particulars.