It is hard to say whether Mr. Shaw is here writing seriously or in a mood of solemn facetiousness. Perhaps he himself is not quite clear on the point. There can be no harm, at any rate, in assuming that he genuinely believes the unity of Getting Married to be "a return to the unity observed in," say, the Oedipus Rex, and examining a little into so pleasant an illusion.
It is, if I may so phrase it, a double-barrelled illusion. Getting Married has not the unity of the Greek drama, and the Greek drama has not the unity of Getting Married. Whatever "unity" is predicable of either form of art is a wholly different thing from whatever "unity" is predicable of the other. Mr. Shaw, in fact, is, consciously or unconsciously, playing with words, very much as Lamb did when he said to the sportsman, "Is that your own hare or a wig?" There are, roughly speaking, three sorts of unity: the unity of a plum-pudding, the unity of a string or chain, and, the unity of the Parthenon. Let us call them, respectively, unity of concoction, unity of concatenation, and structural or organic unity. The second form of unity is that of most novels and some plays. They present a series of events, more or less closely intertwined or interlinked with one another, but not built up into any symmetrical interdependence. This unity of longitudinal extension does not here concern us, for it is not that of either Shaw or Sophocles. Plum-pudding unity, on the other hand--the unity of a number of ingredients stirred up together, put in a cloth, boiled to a certain consistency, and then served up in a blue flame of lambent humour--that is precisely the unity of Getting Married. A jumble of ideas, prejudices, points of view, and whimsicalities on the subject of marriage is tied up in a cloth and boiled into a sort of glutinous fusion or confusion, so that when the cloth is taken off they do not at once lose the coherent rotundity conferred upon them by pressure from without. In a quite real sense, the comparison does more than justice to the technical qualities of the play; for in a good plum-pudding the due proportions of the ingredients are carefully studied, whereas Mr. Shaw flings in recklessly whatever comes into his head. At the same time it is undeniably true that he shows us a number of people in one room, talking continuously and without a single pause, on different aspects of a given theme. If this be unity, then he has achieved it. In the theatre, as a matter of fact, the plum-pudding was served up in three chunks instead of one; but this was a mere concession to human weakness. The play had all the globular unity of a pill, though it happened to be too big a pill to be swallowed at one gulp.
Turning now to the Oedipus--I choose that play as a typical example of Greek tragedy--what sort of unity do we find? It is the unity, not of a continuous mass or mash, but of carefully calculated proportion, order, interrelation of parts--the unity of a fine piece of architecture, or even of a living organism. The inorganic continuity of Getting Married it does not possess. If that be what we understand by unity, then Shaw has it and Sophocles has not. The Oedipus is as clearly divided into acts as is Hamlet or Hedda Gabler. In modern parlance, we should probably call it a play in five acts and an epilogue. It so happened that the Greek theatre did not possess a curtain, and did possess a Chorus; consequently, the Greek dramatist employed the Chorus, as we employ the curtain, to emphasize the successive stages of his action, to mark the rhythm of its progress, and, incidentally, to provide resting-places for the mind of the audience--intervals during which the strain upon their attention was relaxed, or at any rate varied. It is not even true that the Greeks habitually aimed at such continuity of time as we find in Getting Married. They treated time ideally, the imaginary duration of the story being, as a rule, widely different from the actual time of representation. In this respect the Oedipus is something of an exception, since the events might, at a pinch, be conceived as passing within the "two hours' traffick of the stage"; but in many cases a whole day, or even more, must be understood to be compressed within these two hours. It is true that the continuous presence of the Chorus made it impossible for the Greeks to overleap months and years, as we do on the modern stage; but they did not aim at that strict coincidence of imaginary with actual time which Mr. Shaw believes himself to have achieved.[[56]] Even he, however, subjects the events which take place behind the scenes to a good deal of "ideal" compression.
Of course, when Mr. Shaw protests that, in Getting Married, he did not indulge in a "deliberate display of virtuosity of form," that is only his fun. You cannot well have virtuosity of form where there is no form. What he did was to rely upon his virtuosity of dialogue to enable him to dispense with form. Whether he succeeded or not is a matter of opinion which does not at present concern us. The point to be noted is the essential difference between the formless continuity of Getting Married, and the sedulous ordering and balancing of clearly differentiated parts, which went to the structure of a Greek tragedy. A dramatist who can so develop his story as to bring it within the quasi-Aristotelean "unities" performs a curious but not particularly difficult or valuable feat; but this does not, or ought not to, imply the abandonment of the act-division, which is no mere convention, but a valuable means of marking the rhythm of the story. When, on the other hand, you have no story to tell, the act-division is manifestly superfluous; but it needs no "virtuosity" to dispense with it.
It is a grave error, then, to suppose that the act is a mere division of convenience, imposed by the limited power of attention of the human mind, or by the need of the human body for occasional refreshment. A play with a well-marked, well-balanced act-structure is a higher artistic organism than a play with no act-structure, just as a vertebrate animal is higher than a mollusc. In every crisis of real life (unless it be so short as to be a mere incident) there is a rhythm of rise, progress, culmination and solution. We are not always, perhaps not often, conscious of these stages; but that is only because we do not reflect upon our experiences while they are passing, or map them out in memory when they are past. We do, however, constantly apply to real-life crises expressions borrowed more or less directly from the terminology of the drama. We say, somewhat incorrectly, "Things have come to a climax," meaning thereby a culmination; or we say, "The catastrophe is at hand," or, again, "What a fortunate dénouement!" Be this as it may, it is the business of the dramatist to analyse the crises with which he deals, and to present them to us in their rhythm of growth, culmination, solution. To this end the act-division is--not, perhaps, essential, since the rhythm may be marked even in a one-act play--but certainly of enormous and invaluable convenience. "Si l'acte n'existait pas, il faudrait l'inventer"; but as a matter of fact it has existed wherever, in the Western world, the drama has developed beyond its rudest beginnings.
It was doubtless the necessity for marking this rhythm that Aristotle had in mind when he said that a dramatic action must have a beginning, a middle and an end. Taken in its simplicity, this principle would indicate the three-act division as the ideal scheme for a play. As a matter of fact, many of the best modern plays in all languages fall into three acts; one has only to note Monsieur Alphonse, Francillon, La Parisienne, Amoureuse, A Doll's House, Ghosts, The Master Builder, Little Eyolf, Johannisfeuer, Caste, Candida, The Benefit of the Doubt, The Importance of Being Earnest, The Silver Box; and, furthermore, many old plays which are nominally in five acts really fall into a triple rhythm, and might better have been divided into three. Alexandrian precept, handed on by Horace, gave to the five act division a purely arbitrary sanction, which induced playwrights to mask the natural rhythm of their themes beneath this artificial one.[[57]] But in truth the three-act division ought no more to be elevated into an absolute rule than the five-act division. We have seen that a play consists, or ought to consist, of a great crisis, worked out through a series of minor crises. An act, then, ought to consist either of a minor crisis, carried to its temporary solution, or of a well-marked group of such crises; and there can be no rule as to the number of such crises which ought to present themselves in the development of a given theme. On the modern stage, five acts may be regarded as the maximum, simply by reason of the time-limit imposed by social custom on a performance. But one frequently sees a melodrama divided into "five acts and eight tableaux," or even more; which practically means that the play is in eight, or nine, or ten acts, but that there will be only the four conventional interacts in the course of the evening. The playwright should not let himself be constrained by custom to force his theme into the arbitrary mould of a stated number of acts. Three acts is a good number, four acts is a good number,[[58]] there is no positive objection to five acts. Should he find himself hankering after more acts, he will do well to consider whether he be not, at one point or another, failing in the art of condensation and trespassing on the domain of the novelist.
There is undoubted convenience in the rule of the modern stage: "One act, one scene." A change of scene in the middle of an act is not only materially difficult, but tends to impair the particular order of illusion at which the modern drama aims.[[59]] Roughly, indeed, an act may be defined as any part of a given crisis which works itself out at one time and in one place; but more fundamentally it is a segment of the action during which the author desires to hold the attention of his audience unbroken and unrelaxed. It is no mere convention, however, which decrees that the flight of time is best indicated by an interact. When the curtain is down, the action on the stage remains, as it were, in suspense. The audience lets its attention revert to the affairs of real life; and it is quite willing, when the mimic world is once more revealed, to suppose that any reasonable space of time has elapsed while its thoughts were occupied with other matters. It is much more difficult for it to accept a wholly imaginary lapse of time while its attention is centred on the mimic world. Some playwrights have of late years adopted the device of dropping their curtain once, or even twice, in the middle of an act, to indicate an interval of a few minutes, or even of an hour--for instance, of the time between "going in to dinner" and the return of the ladies to the drawing-room. Sir Arthur Pinero employs this device with good effect in Iris; so does Mr. Granville Barker in Waste, and Mr. Galsworthy in The Silver Box. It is certainly far preferable to that "ideal" treatment of time which was common in the French drama of the nineteenth century, and survives to this day in plays adapted or imitated from the French.
I remember seeing in London, not very long ago, a one-act play on the subject of Rouget de l'Isle. In the space of about half-an-hour, he handed the manuscript of the "Marseillaise" to an opera-singer whom he adored, she took it away and sang it at the Opera, it caught the popular ear from that one performance, and the dying Rouget heard it sung by the passing multitude in the streets within about fifteen minutes of the moment when it first left his hands. (The whole piece, I repeat, occupied about half-an-hour; but as a good deal of that time was devoted to preliminaries, not more than fifteen minutes can have elapsed between the time when the cantatrice left Rouget's garret and the time when all Paris was singing the "Marseillaise.") This is perhaps an extreme instance of the ideal treatment of time; but one could find numberless cases in the works of Scribe, Labiche, and others, in which the transactions of many hours are represented as occurring within the limits of a single act. Our modern practice eschews such licenses. It will often compress into an act of half-an-hour more events than would probably happen in real life in a similar space of time, but not such a train of occurrences as to transcend the limits of possibility. It must be remembered, however, that the standard of verisimilitude naturally and properly varies with the seriousness of the theme under treatment. Improbabilities are admissible in light comedy, and still more in farce, which would wreck the fortunes of a drama purporting to present a sober and faithful picture of real life.
Acts, then, mark the time-stages in the development of a given crisis; and each act ought to embody a minor crisis of its own, with a culmination and a temporary solution. It would be no gain, but a loss, if a whole two hours' or three hours' action could be carried through in one continuous movement, with no relaxation of the strain upon the attention of the audience, and without a single point at which the spectator might review what was past and anticipate what was to come. The act-division positively enhances the amount of pleasurable emotion through which the audience passes. Each act ought to stimulate and temporarily satisfy an interest of its own, while definitely advancing the main action. The psychological principle is evident enough; namely, that there is more sensation to be got out of three or four comparatively brief experiences, suited to our powers of perception, than out of one protracted experience, forced on us without relief, without contrast, in such a way as to fatigue and deaden our faculties. Who would not rather drink three, four, or five glasses of wine than put the bottle to his lips and let its contents pour down his throat in one long draught? Who would not rather see a stained-glass window broken into three, four, or five cunningly-proportioned "lights," than a great flat sheet of coloured glass, be its design never so effective?
It used to be the fashion in mid Victorian melodramas to give each act a more or less alluring title of its own. I am far from recommending the revival of this practice; but it might be no bad plan for a beginner, in sketching out a play, to have in his mind, or in his private notes, a descriptive head-line for each act, thereby assuring himself that each had a character of its own, and at the same time contributed its due share to the advancement of the whole design. Let us apply this principle to a Shakespearean play--for example, to Macbeth. The act headings might run somewhat as follows--