that I appeal not to be misled by his seductive example. They have little chance of rivalling him as sociological essayists; but if they treat their art seriously, and as a pure art, they may easily surpass him as dramatists. By adopting his practice they will tend to produce, not fine works of art, but inferior sociological documents. They will impair their originality and spoil their plays in order to do comparatively badly what Mr. Shaw has done incomparably well.
The common-sense rule as to stage directions is absolutely plain; be they short, or be they long, they ought always to be impersonal. The playwright who cracks jokes in his stage-directions, or indulges in graces of style, is intruding himself between the spectator and the work of art, to the inevitable detriment of the illusion. In preparing a play for the press, the author should make his stage-directions as brief as is consistent with clearness. Few readers will burden their memory with long and detailed descriptions. When a new character of importance appears, a short description of his or her personal appearance and dress may be helpful to the reader; but even this should be kept impersonal. Moreover, as a play has always to be read before it can be rehearsed or acted, it is no bad plan to make the stage-directions, from the first, such as tend to bring the play home clearly to the reader's mental vision. And here I may mention a principle, based on more than mere convenience, which some playwrights observe with excellent results. Not merely in writing stage-directions, but in visualizing a scene, the idea of the stage should, as far as possible, be banished from the author's mind. He should see and describe the room, the garden, the sea-shore, or whatever the place of his action may be, not as a stage-scene, but as a room, garden, or sea-shore in the real world. The cultivation of this habit ought to be, and I believe is in some cases, a safeguard against theatricality.
CHAPTER V
DRAMATIS PERSONAE
The theme being chosen, the next step will probably be to determine what characters shall be employed in developing it. Most playwrights, I take it, draw up a provisional Dramatis Personae before beginning the serious work of construction. Ibsen seems always to have done so; but, in some of his plays, the list of persons was at first considerably larger than it ultimately became. The frugal poet sometimes saved up the characters rejected from one play, and used them in another. Thus Boletta and Hilda Wangel were originally intended to have been the daughters of Rosmer and Beata; and the delightful Foldal of John Gabriel Borkman was a character left over from The Lady from the Sea.
The playwright cannot proceed far in planning out his work without determining, roughly at any rate, what auxiliary characters he means to employ. There are in every play essential characters, without whom the theme is unthinkable, and auxiliary characters, not indispensable to the theme, but simply convenient for filling in the canvas and carrying on the action. It is not always possible to decide whether a character is essential or auxiliary--it depends upon how we define the theme. In Hamlet, for example, Hamlet, Claudius, and Gertrude are manifestly essential: for the theme is the hesitancy of a young man of a certain temperament in taking vengeance upon the seducer of his mother and murderer of his father. But is Ophelia essential, or merely auxiliary? Essential, if we consider Hamlet's pessimistic feeling as to woman and the "breeding of sinners" a necessary part of his character; auxiliary, if we take the view that without this feeling he would still have been Hamlet, and the action, to all intents and purposes, the same. The remaining characters, on the other hand, are clearly auxiliary. This is true even of the Ghost: for Hamlet might have learnt of his father's murder in fifty other ways.
Polonius, Laertes, Horatio, and the rest might all have been utterly different, or might never have existed at all, and yet the essence of the play might have remained intact.
It would be perfectly possible to write a Hamlet after the manner of Racine, in which there should be only six personages instead of Shakespeare's six-and-twenty: and in this estimate I assume Ophelia to be an essential character. The dramatis personae would be: Hamlet, his confidant; Ophelia, her confidant; and the King and Queen, who would serve as confidants to each other. Indeed, an economy of one person might be affected by making the Queen (as she naturally might) play the part of confidant to Ophelia.
Shakespeare, to be sure, did not deliberately choose between his own method and that of Racine. Classic concentration was wholly unsuited to the physical conditions of the Elizabethan stage, on which external movement and bustle were imperatively demanded. But the modern playwright has a wide latitude of choice in this purely technical matter. He may work out his plot with the smallest possible number of characters, or he may introduce a crowd of auxiliary personages. The good craftsman will be guided by the nature of his theme. In a broad social study or a picturesque romance, you may have as many auxiliary figures as you please. In a subtle comedy, or a psychological tragedy, the essential characters should have the stage as much as possible to themselves. In Becque's La Parisienne there are only four characters and a servant; in Rostand's Cyrano de Bergerac there are fifty-four personages named in the playbill, to say nothing of supernumeraries. In Peer Gynt, a satiric phantasmagory, Ibsen introduces some fifty individual characters, with numberless supernumeraries; in An Enemy of the People, a social comedy, he has eleven characters and a crowd; for Ghosts and Rosmersholm, psychological tragedies, six persons apiece are sufficient.
It can scarcely be necessary, at this time of day, to say much on the subject of nomenclature. One does occasionally, in manuscripts of a quite hopeless type, find the millionaire's daughter figuring as "Miss Aurea Golden," and her poor but sprightly cousin as "Miss Lalage Gay"; but the veriest tyro realizes, as a rule, that this sort of punning characterization went out with the eighteenth century, or survived into the nineteenth century only as a flagrant anachronism, like knee-breeches and hair-powder.
A curious essay might be written on the reasons why such names as Sir John Brute, Sir Tunbelly Clumsy, Sir Peter Teazle, Sir Anthony Absolute, Sir Lucius O'Trigger, Lord Foppington, Lord Rake, Colonel Bully, Lovewell, Heartfree, Gripe, Shark and the rest were regarded as a matter of course in "the comedy of manners," but have become offensive to-day, except in deliberate imitations of the eighteenth-century style. The explanation does not lie merely in the contrast between "conventional" comedy and "realistic" drama. Our forefathers (whatever Lamb may say) did not consciously place their comedy in a realm of convention, but generally considered themselves, and sometimes were, realists. The fashion of label-names, if we may call them so, came down from the Elizabethans, who, again, borrowed it from the Mediaeval Moralities.[37] Shakespeare himself gave us Master Slender and Justice Shallow; but it was in the Jonsonian comedy of types that the practice of advertising a "humour" or "passion" in a name (English or Italian) established itself most firmly. Hence such strange appellatives as Sir Epicure Mammon, Sir Amorous La Foole, Morose, Wellbred, Downright, Fastidius Brisk, Volpone, Corbaccio, Sordido, and Fallace. After the Restoration, Jonson, Beaumont and Fletcher, and Massinger were, for a time, more popular than Shakespeare; so that the label-names seemed to have the sanction of the giants that were before the Flood. Even when comedy began to deal with individuals rather than mere incarnations of a single "humour," the practice of giving them obvious pseudonyms held its ground. Probably it was reinforced by the analogous practice which obtained in journalism, in which real persons were constantly alluded to (and libelled) under fictitious designations, more or less transparent to the initiated. Thus a label-name did not carry with it a sense of unreality, but rather, perhaps, a vague suggestion of covert reference to a real person. I must not here attempt to trace the stages by which the fashion went out. It could doubtless be shown that the process of change ran parallel to the shrinkage of the "apron" and the transformation of the platform-stage into the picture-stage. That transformation was completed about the middle of the nineteenth century; and it was about that time that label-names made their latest appearances in works of any artistic pretension--witness the Lady Gay Spanker of London Assurance, and the Captain Dudley (or "Deadly") Smooth of Money. Faint traces of the practice survive in T.W. Robertson, as in his master, Thackeray. But it was in his earliest play of any note that he called a journalist Stylus. In his later comedies the names are admirably chosen: they are characteristic