This question of naturalness, which is germane to the whole subject of dialogue and not merely to stage dialogue, is one of the most difficult things to lay down any rule about. It is not easy even to get any working formula which is helpful. On the one side there seems to lie the obvious rule――that all dialogue ought to be natural, appropriate to the person in whose mouth it is put――not merely what in substance he would say, but also said in the way he would say it. On the other side is the obvious fact that no two writers of any considerable merit do, as a fact, write dialogue in the same way, even when they are presenting the same sort of characters. Comparatively impersonal as the dialogue form is, when set beside the narrative, yet the writer’s idiosyncrasy will have its way, and in greater or less degree the author’s accent is heard from the lips of his imaginary interlocutors――and of each and all of them, however widely different they may be supposed to be, and really are, from one another. This appears to land us in an _impasse_; the obvious fact seems to conflict with the obvious rule. If it be so, I suppose the rule must go to the wall, for all its obviousness. But I fancy that some approach to a solution may be found in the suggestion that no two authors of creative power do, in fact, ever create characters of quite the same sort, and that we got into a seeming _impasse_ by being guilty of a fallacy. When an author sits down at his desk to contemplate, criticize, and reproduce the world about him, it is natural at the first thought to regard the author as subject contemplating and reproducing the world as object――pure subject as against pure object. Here is the fallacy as I conceive. The author as subject does not and cannot contemplate the world as pure object. What he sees is object-subject――that is to say, he consciously sets himself to contemplate and describe a world which is already modified for him by the unconscious projection of his own personality into it――or, in more homely language, he always looks through his own spectacles. It follows that when two creative minds――say Dickens and Thackeray――both set out to describe a duke or a costermonger, it is never the same duke or costermonger――it is not the abstract idea of duke or costermonger, laid up in heaven――but it is a duke-Dickens or a duke-Thackeray――a costermonger-Dickens or a costermonger-Thackeray. Consequently again it is not in the end natural――and, therefore, as the Admirable Crichton would remind us, it is not in the end right――that these two dukes or these two costermongers should speak in exactly the same way――though no doubt both of the pairs ought still to speak as dukes and costermongers of some sort――be it Dickensian or Thackerayan as the case may be. Of course, if an author’s idiosyncrasy is so peculiar that the subjective infusion of himself which he pours into the objective costermonger is so powerful as to cause the human race at large to object that no