‘To speak, to reason, to form general or abstract ideas, are then in fact the same thing: and this truth, simple as it is, might pass for a discovery. Certainly, men in general have not had any notion of it; this is evident from the manner in which they speak and reason; it is evident from the abuse which they make of abstract ideas; finally, it is evident from the difficulties which those persons confessedly find in conceiving of abstract ideas who have so little in speaking of them. ‘The art of reasoning resolves into the construction of languages, only because the order of our ideas itself depends entirely on the subordination that subsists between the names given to _genera_ and _species_; and as we arrive at new ideas only by forming new classes, it follows that we can only determine or define our ideas by determining their classes. In this case we should reason well, because we should be guided by analogy in our conclusions as well as in the acceptation of words. ‘Convinced, therefore, that classes or sorts of things are pure denominations, we shall never think of supposing that there exist in nature _genera_ or _species_; and we shall understand by these words nothing but a certain mode of classing things according to the relations which they have to ourselves and to one another. We shall be sensible that we can only discover those relations, and not what the things truly are.’ Berkeley handled his subjects with little tenderness, and he has perfectly anatomised this subject of abstract ideas. In choosing to answer the objections to this doctrine as stated by him, I shall not be accused of wishing to encounter a mean adversary. I can only trust to the goodness of my cause. I hope I shall be excused for going at some length into the argument, because it is one of the most difficult and complicated in itself, and is of the most extensive application to other questions relating to the human understanding. If we can come to any satisfactory issue to it, it will be worth the pains of enquiry.