II
From this question of Stevenson’s conviction, however (the question of the inevitable as opposed to the practicable), arises a further question. I have said earlier that in the case of a work of art there is left with the reader some abiding emotion, an evocation, as it were, of emotion distinct from all incidental emotions, excitements, dreads, or anxieties aroused in the course of the book. In that pervading and prevailing emotion, it seems to me, lies the particular quality which distinguishes a work of art from a work of merely consummate craft. If I question whether such abiding emotion is evoked by the longer stories of Stevenson, I am bound to answer that these do not arouse in me any emotion greater than that of interest, the consequence of a succession of pleasant excitements. The romances as a whole have great ingenuity, many scenes to which all readers must look back with recollected enjoyment. In no case does the book reappear as a whole. The recollection is a recollection of “plums.” That they are good plums does not affect the validity of the argument if once the specific test suggested above is accepted. In the case of Weir of Hermiston the recollection is obviously difficult, because the book is a fragment: it is, however, perfectly clear and level in performance, which leads to the supposition that Weir, as it stands, will actually bear whatever test is applied to it. For that reason Weir is truly regarded as Stevenson’s masterpiece among the longer stories.
With the short stories I have already dealt in considerable detail; to the remaining creative works there is no need to refer on these grounds, for the plays are admittedly poor. And indeed, I should not have raised the question about the romances if it had not been the case that very considerable claims have been made on behalf of the permanent value of Stevenson’s work by many writers whose opinions ordinarily command respect. The truth is probably that all good novels, of whatever kind, whether modern or historical, must be based upon idea and upon character. To Stevenson, character was incidental. To Stevenson incident, picturesque or exciting, and the employment of an atmosphere, or appropriate “style,” were the most important things in romance. That was perhaps the grave mistake which made his romances what they are, and which has very considerably affected the romantic novels published since Stevenson’s time and written in accordance with his conventions. The use of conventional characters, easily-recognisable romantic types, has for twenty years and more been accepted by English romantic novelists as a legitimate evasion of the need for creating character. Thus it happens that so few modern romantic novels have at this time any standing. Their names are forgotten (except, possibly, by their authors, and by some sections of the public only if the novels have been made into stage plays). If Stevenson’s romances had enjoyed the strength of definite themes, and if they had been based upon character, the whole position of the romantic novel in England at the present day might have been different. As it is, the romantic novel is a survival. The freshness of Stevenson’s manipulated convention is stale, and the imitators of Stevenson have forsaken romance for the writing of detective mystery stories. They still have popularity; but they have no status.