A Book About the Theater Part 3

WHY FIVE ACTS?

I

In the eighteenth century, both in England and in France, every stately and ponderous tragedy and every self-respecting comedy obeyed the obligation imposed by long tradition and duly stretched itself out to the full measure of five acts, no more and no less. It felt bound thus to distend itself, even tho its theme might be far too frail for so huge a frame, and even tho the unfortunate author often found himself at his wit's end to piece out his play's end. Any one who has had occasion to read widely in the works of the eighteenth century playwrights cannot fail to feel abundant sympathy for the hara.s.sed poet who plaintively called on Parliament to pa.s.s a law abolis.h.i.+ng fifth acts altogether.

This unduly distressed dramatist was an Englishman; but about the same time a Frenchman, weary of contemplating the frequent emptiness of the contemporary tragic stage, sarcastically remarked that, after all, it must be very easy _not_ to write a tragedy in five acts.

Yet if tragedy was to be written at all, it had to have five acts, since a smaller number would not seem proportionate to a truly tragic subject.

But why five acts? Why has five the number sacred to the tragic muse?

Why did even the comic muse demand it? Why does George Meredith, discussing comedy, declare that "five is dignity with a trailing robe; whereas one, or two, or three acts would be short skirts, and degrading." Why not three acts, or seven? Why was it that any other number of acts was unthinkable--or at least never thought of?

Questions like these seem to have floated before the mind of the Abbe d'Aubignac, writing in the seventeenth century, and he came very near putting to himself the query which serves as a t.i.tle for this chapter.

"Poets have generally agreed that all Drammas regularly should have neither more nor less than Five Acts; and the Proof of this is the general observation of it; but for the Reason, I do not know whether there be any founded in Nature. Rhetorick has this advantage over Poetry in the Parts of Oration, that the Exord, Narration, Confirmation and Peroration are founded upon a way of discoursing natural to all Men....

But for the Five Acts of the Drammatick Poem, they have not been framed upon any sound ground."

That the division of a drama into five parts was accepted in every civilized country as the only possible division, seems very strange indeed, when we consider that there is really no artistic justification for it, nor any logical necessity. Like every other work of art a play ought to have a single subject, a clearly defined topic; in other words, it ought to have Unity of Action. There is no denying that some of the greatest artists have, now and again, been tempted to deal with two themes at the same time, combining these as best they could in a single work at the risk of leaving us a little in doubt as to their intention; but in the immense majority of acknowledged masterpieces the interest is carefully centered in a single object. In these masterpieces the action is single and unswerving, sweeping forward irresistibly to its inevitable end.

If, therefore, we accept the Unity of Action as a general rule, binding upon all artists, we can hardly deny that the most obviously natural arrangement for the story is to set it forth in one act, without any intermission or subdivision whatsoever--a single action in a single act.

Yet it is the play in three acts which we are bound to recognize at once as possessing the ideal form, since it enables the dramatist to set apart the three divisions, which Aristotle declared to be essential to a well-constructed tragedy--the beginning, the middle, and the end--each presented in an act of its own. To put a play into more than three acts is possible only by halving one or another of these three essential parts. In a four-act play, the beginning may be split into two acts; and in a five-act play the middle may also be subdivided.

The logic of the three-act form, and the convenience of it also, are so obvious that ever since the tyranny of the Procrustean framework in five acts was abolished in the middle years of the nineteenth century, practical playwrights of all countries have favored it more and more.

The young Dumas used it in his later plays, and so did Ibsen, that consummate master of stagecraft, emanc.i.p.ated from empty traditions, but profiting shrewdly by every available device of his immediate predecessors. If the four-act form is also popular to-day, this seems to be because the modern dramatist, intending a play in three acts, finds himself forced by sheer press of matter, to subdivide one of the essential members, as Sir Arthur Pinero had to do in the 'Second Mrs.

Tanqueray' and Mr. Henry Arthur Jones in the 'Liars.' Even the opera, which liked the larger framework of five acts when Scribe was writing librettos for Halevy and Meyerbeer, is now content with only three, since Wagner revealed his skill as a librettist.

It is true that Freytag, in his sadly old-fas.h.i.+oned treatise on 'Technic of the Drama,' accepted without cavil the five-act form, and even attempted to justify it by a.s.serting that there are in fact five divisions of a tragic action. He symbolized the arrangement of a drama in a pyramidal structure, declaring that it ascends from the Introduction to the Climax, and then descends to the Catastrophe.

Obviously these are only different terms for the beginning, the middle, and the end. But he vainly imagined two other members, the Rise, which intervenes between the Introduction and the Climax, and the Fall, which he inserted between the Climax and the Catastrophe. Obviously, again, this is an explanation after the event; and it seems to have its origin solely in his acceptance of the five-act form. And Freytag was forced to abandon his own theory when he considered honestly certain of the masterpieces of the modern drama. He admitted it to be "impossible that the single acts should correspond entirely to the five great divisions of the action." He a.s.serted that "in the Rising Action, the first stage was usually in the first act, the last sometimes in the third; of the Falling Action the beginning and the end were sometimes taken in the third and fifth acts." Yet he failed to see that if he made this admission, he cut the ground from under his feet, and that there was no longer any acceptable reason for his insistence upon the five-act form.

Freytag had no doubt at all as to the necessity of the division into five acts. He received it with blind faith, as tho it had been prescribed by divine authority. Yet if he had chosen to explore the early history of the drama in his own tongue, he would have found Hans Sachs sometimes extending his plays into six acts, and even into seven.

And if he had cared to consider the drama of the Spaniards he would have seen that the most of the plays of Calderon are in three acts--a division which the great dramatic poet of Spain had taken over, as he had taken over so much else, from his masterful predecessor, Lope de Vega. In his interesting and illuminating little treatise on the art of writing plays, Lope de Vega gave the credit of establis.h.i.+ng the three-act form to Virues. Plays had previously been written in four acts; as Lope puts it pleasantly: "The drama had gone on all fours, like a child, and truly it was then in its infancy."

Freytag ignored or was ignorant of Hans Sachs and Calderon. His mind was fixed on Goethe and on Schiller, altho his vision also included Shakspere, upon whom the two German poets had more or less modeled themselves. The tradition of the five-act form might not obtain in the earliest German drama, as it did not obtain in the Spanish; but it was firmly established in the later German drama, in the English, and in the French. It is easy to see that the later Germans derived it from the French and the English; but where did the French and the English get it?

Where could they get it? No such division existed in the medieval drama, in the mysteries and in the miracle-plays, out of which the drama of every modern language has been developed. No such division existed in the Greek drama, which has served as a standard and as a stimulus to the drama of every modern literature. A Greek tragedy was represented without any intermission in a single, long unbroken act; and if a sequence of three plays was sometimes performed, one after another, on the same day, and dealing with successive periods of the same story, this trilogy might suggest a division into three parts. Nor is any hint of the duty of dividing a tragedy into five parts to be discovered anywhere in Aristotle.

II

And yet we must go back to the Greek theater if we want to see why it is that the 'Femmes Savantes' of Moliere and the 'School for Scandal' of Sheridan are each of them in five acts. But it is not from a Greek that we get the law that this division was obligatory on all self-respecting dramatists; it is from a Roman, writing at a time when the drama of his own language had been ousted from the stage by pantomimic spectacle and by gladiatorial combat. It is Horace, who, in his epistle on the art of poetry, declares the necessity of five acts:

Ne brevior, neu sit quinto productior actu Fabula quae posci vult et spectata reponi.

Sir Theodore Martin rendered this in an English rimed couplet, which does not completely convey the meaning of the two Latin lines, but which will serve to show the rigidity of the rule laid down by the Roman poet:

Five acts a play must have, nor more nor less, To keep the stage and have a marked success.

But this still leaves us groping in the dark. Why did Horace declare this law? What warrant had he? What put the idea into his head? These are questions answered by a French scholar, M. Weil; in one of his ingenious and learned 'etudes sur le Drame Antique,' he explains that Horace derived much of his theory of the poetic art from the Alexandrian critics, and more particularly from the writings of a certain Neoptolemus of Parium. Probably the Alexandrian authors of tragedy had been led to adopt a division into five acts by following the example of Euripides, whose practise was not uniform, but who tended to reduce to four the number of the lyric odes in his tragedies, thus separating the purely dramatic pa.s.sages into five parts.

In Athens the drama had been slowly evolved out of the tragic songs; and in the surviving tragedies of aeschylus, the earliest of the three great dramatic poets of Greece, we discover that the choral odes are more abundant than the dialog which carries on the plot. In the extant plays of his mighty successor, Sophocles, the drama is seen emerging triumphant, but the lyrical pa.s.sages are still frequent and important.

In the later pieces of Euripides, the third and most modern of the Attic tragedians, we note that the drama has almost wholly disengaged itself from the lyric out of which it sprang. In aeschylus and in Sophocles the number of choral odes and the number of episodes, of purely dramatic pa.s.sages in dialog, is never fixed, varying from play to play as the plot might demand. But in Euripides the choral odes are more detached from the drama; beautiful in themselves, they seem to exist rather for their own sake than in any integral relation to the play itself. And apparently Euripides was far more interested in his play, in his plot, and in his characters, than in these extraneous lyric pa.s.sages, so he reduced them to the lowest possible number, generally to four, serving, so to speak, as exquisite interact music, separating the pathetic play into five episodes in dialog.

The Alexandrian tragedians came long after Euripides, and to their sophisticated taste his pathetic and emotional plays appealed far more than the austerer and manlier masterpieces of his two great predecessors. Apparently they accepted his form as final; they may even have left out the choruses altogether; and then their tragedies had five separate episodes--in other words, five acts. It is these lost Alexandrian tragedies, composed in the decadent days of the Greek drama, which seem to have served as the model for Seneca, the eloquent rhetorician--even tho he frequently took over the theme and often more or less of the structure of certain of the dramas of Euripides.

The tragedies of Seneca are to be considered rather as dramatic poems than as poetic dramas, since they were intended not really for performance by actors, in a theater, before an audience, but for recitation by a single elocutionist in a private house--much as a professional reader of our own time might recite unaided a more or less dramatic poem by Sh.e.l.ley or Byron or Browning. Coming long after Horace, Seneca unhesitatingly accepted all of the restrictions insisted upon by the Latin lyrist--including the purely academic limitation of the number of speakers taking part in any dialog to three, a limitation absolutely absurd in a poem not intended for actual acting and not forced to conform to the accidental conditions of the Attic stage. Obeying also the other rule which he found in Horace's codification of the laws of dramatic poetry, the Hispano-Roman rhetorician was careful always to cut up his play into five parts. But he saw his profit in retaining the chorus, since this could be made to serve as the appropriate mouthpiece for the elaborate pa.s.sages of elocutionary splendor in which he delighted.

It is not to be wondered at that the Italian scholars of the Renascence followed the precept of Horace and the practise of Seneca. They were far more at home in Latin than they were in Greek; and they could hardly help reading into the literature of Athens what they were already familiar with in the authors of Rome. To them Seneca was as imposing as Sophocles, and Horace was almost as weighty as Aristotle. So it is that Scaliger and Minturno prescribe five acts, and that Castelvetro (always more practical in his point of view) points out that poets seem to have found the five-act form most suitable. When an Italian scholar-poet turned from criticism to creation, the tragedies he conscientiously composed obeyed all the rules, and his dramatic poems were as academic as those of Seneca, in that they were intended not for production by professional actors in a regular theater before spectators who had paid their way in, but only for an occasional performance by the author himself a.s.sisted by a few of his friends before a little group of cultivated admirers of antiquity, contemptuous of the real public. These soulless dramatic poems, devised for declamation by amateurs before a gathering of dilettants, are now perceived to be merely literary curiosities, having little connection with the real drama made for the regular theater and its myriad-minded body of playgoers.

Just as the Italian dramatic poems were imitations of Seneca, so the French dramatic poems, composed a little later, were imitations of these Italians, and also of Seneca, more or less indirectly. They were the imitations of an imitation, aping the outward form of the drama, but empty of all genuine dramatic spirit, artificial in pa.s.sion and high-flown in rhetoric. And there are early English attempts at this same sort of academic tragedy, more imitative still, since we can see in them the commingled influence of the French and of the Italians immediately, and also of the remoter Seneca, whom they revered as the exemplar of true tragedy. Such a play is 'Gorboduc,' belauded by the scholarly Sidney--and even on one occasion acted, by main strength. In all of these imitations, English and French and Italian, we find the stately chorus abounding in lofty rhetoric; and we find also, and always, the division into five acts. But in the folk-theater, which the scholar-poets scorned, and out of which the living drama was to be developed, there is no trace of any division into acts. In the mysteries and the miracle-plays, and in the chronicle-plays which grew out of them, there are numberless episodes, each complete in itself, and never combined artificially into acts. The composer of any one of these folk-dramas conceived his story as a continuous narrative shown in action; and he gave no thought to the number of divisions, of episodes, of separate scenes, or of acts that it might seem to have.

III

Tragedy has ever been held to be more elevated than comedy and more worthy; and comedy has continually accepted the conditions appropriate to tragedy. Since the dignity of tragedy demanded a division into five acts, comedy was also subjected to the same rule; and this was done in spite of the fact that the plays of Plautus and Terence (composed long before Horace codified his advice to intending poets) were not divided into acts, if we may judge by the earliest of the surviving ma.n.u.scripts.

So it is that we find the scholarly authors of the two earliest of English comedies, 'Ralph Roister Doister' and 'Gammer Gurton's Needle,'

knowing what was expected of them, and giving the five-act form to both of these amusing plays. But these two comedies, almost contemporary as they are with the academic and undramatic tragedy of 'Gorboduc,' are far superior to it in adaptability for actual performance. They are not intended only to be recited; they can be acted easily and profitably. As we a.n.a.lyze them we see that the structural complexity may be derived from the comic dramas of Plautus and Terence, but that the inner spirit is that of the English folk-theater, of the robust medieval farce-writers, of the unknown humorist who has left us the laughable and veracious scene of Mak and the Shepherds.

Scholars as they were, the authors of these two comedies did not scorn the primitive plays of the plain people of their own time. They did not despise the unpretending folk-drama which was then pleasing the populace; in fact, they took stock of it, and found their profit in so doing. They saw that to be raised up to the level of literature it needed only to be chastened and stiffened. They accepted the living tradition of play-making as it came down to them, and in accord with this tradition they wrought their humorous fantasies, adding the higher polish and the more adroit plot which they had learned to appreciate in the Latin comic dramatists. They accepted the native play, bare as it was, and they enriched it by bestowing on it as much as it could carry of the finer art of the Romans. Thus it is that the authors of 'Ralph Roister Doister' and of 'Gammer Gurton's Needle' may have pointed out the path of progress to the author of the 'Comedy of Errors,' whereas the authors of 'Gorboduc,' contemptuously rejecting the folk-theater of their own day, and idly copying the cla.s.sicist imitations of the Italians, thereby relinquished whatever direct influence they might have had upon the growth of tragedy in England.

Both 'Ralph Roister Doister' and 'Gammer Gurton's Needle' were probably written for performance by college boys, and they have not a little of the brisk heartiness and of the broad horse-play to which we are accustomed in the college pieces of to-day. It was for performance at court that Lyly wrote the most of his plays, which lack the vivacity and the liveliness distinguis.h.i.+ng the two college comic dramas, but which yet reveal a far better understanding of the drama than was possessed by the authors of 'Gorboduc.' Lyly again is careful to divide his plays into five acts. But his contemporaries Greene and Peele, writing solely for the professional playhouses, were bound by none of the rules which might be expected in college or at court. Whatever their own scholarly equipment, when they wrote for the professional players, they followed unhesitatingly the traditions of the contemporary theater. As playwrights they were the direct heirs of the anonymous and ignorant devisers of the medieval drama. They had a story to set on the stage; they chose a succession of more or less effective episodes, and they carelessly cast these into dialog, with little thought of form or of construction. Never do their plays contain matter enough for five full acts; and we may be certain that no such framework was ever in the mind of either of these dramatic poets. In the original editions of their pieces we find no separation into acts and scenes; and if this needless and misleading subdivision is found in later editions it is the doing of misguided editors.

In what is accepted as the earliest edition of Kyd's 'Spanish Tragedy,'

the most widely popular of all the pre-Shaksperian plays, the text is actually divided into four acts. But this division is not structural; it is almost accidental, as tho it was an afterthought, inserted at the last moment into the copy intended for the printer, and never in the mind of the playwright himself when he was preparing the prompt-book for the actors; and Shakspere, who followed Kyd in more ways than one, apparently followed him in this also. In the folio edition of his plays, published after his death, a division into five acts has been made; but the task has not been accomplished any too skilfully--for example, the second act of 'King John' has but eighty lines, and here the division is into four acts only. The suggestion has been proffered that it was, perhaps, left to the printers to do, the influence of Ben Jonson having been powerful enough to establish the theory that a self-respecting dramatist would never fail to cast his tragedies in the five-act form.

It is to be noted also that no division into acts is to be found in the quarto editions published in Shakspere's lifetime; and this is very significant since these quartos seem to have been piratical copies from shorthand notes taken surrept.i.tiously in the theater, thus recording the actual conditions of performance.

It may be doubted whether Shakspere conceived his plays in accordance with any such subdivisions. Some of them, the 'Comedy of Errors' for one, which can be acted in the s.p.a.ce of an hour and a quarter, are far too slight for so huge a framework. On the other hand, the several appearances of Chorus punctuate 'Henry V' into five divisions, apparently an intentional conformity to the Horatian rule. Of course, there were generally several intermissions in the Elizabethan performance of a play, altho the resulting divisions were not necessarily five; and it is noteworthy that Shakspere makes Jaques declare that man's life had seven acts.

IV

The fact is that Shakspere was a professional playwright, and that he had no merely academic theories. In composing his plays he followed unhesitatingly the principles that had guided his immediate predecessors. He was seeking ever to give the playgoing public what it had been accustomed to enjoy in the theater, better in degree, no doubt, but the same in kind. Like these predecessors, he kept to the traditions inherited from the medieval mysteries; and he thought in terms, not of acts and of scenes, as a modern playwright is forced to do, but of a continuous narrative shown in action. In doing so he resembles Herodotus, whose history has also been cut up by later editors, dividing it into nine books, altho, as Professor Bury has reminded us, "such divisions had not yet come into fas.h.i.+on" in the historian's own day. There is no reason to suppose that Shakspere would have approved of the attempt of the editors of the folio to subdivide his plays, each into five acts. There is every reason to suppose that he would have been greatly annoyed if he could have foreseen the way in which later editors have chosen further to chop up the acts into an infinity of scenes.

Nowadays, we have been so accustomed to read Shakspere in one or another of the trim and tidy modern editions, with a wanton division into acts and into scenes, each of which indicates a change of place, and each of which seems to suggest a change of scenery, that it is only by a resolute effort of the will that we are able to shake off the prepossessions derived from this unfortunate and confusing presentation of his text. Probably even to-day a majority of those who enjoy reading Shakspere would be surprised to be told that there is no warrant whatever for these alleged changes of scene, and for these superabundant subdivisions of his story. Many of these readers would be taken aback by the unexpected discovery that all this cutting up of Shakspere's text was the work of his commentators, with Rowe at the head of the procession. Some of these readers would feel as tho they were deprived of a precious possession, if they had only an edition in which all this useless machinery was swept away.

And yet this is just the edition which is demanded by the present state of Shaksperian scholars.h.i.+p, and which is now made possible by our new understanding of the Elizabethan theater, with its rude platform thrust out into the yard, so different from our modern theaters, in which the stage is withdrawn behind a picture-frame. The Tudor platform-stage is wholly unlike the picture-frame stage of to-day; but it is very like the "pageant," or the scaffold on which the mysteries and miracle-plays were presented. It was to the simple conditions of his semi-medieval theater that Shakspere adjusted himself, rude as those conditions may now appear to us who are accustomed to the sumptuous picturesqueness of our own luxuriant playhouses.

In accepting the theater as he found it, and in availing himself of all its possibilities, such as they were, Shakspere showed his usual common sense. Only by striving to reconstruct for ourselves in our mind's eye, as it were, the playhouse where he plied his trade and earned his living, can we come to any adequate appreciation of his art, of his craftsmans.h.i.+p as a playwright, of his dramaturgic skill. And in any honest effort to understand how his mighty dramas were originally produced by himself and by his fellow actors in the round O of the wooden Globe Theater, unroofed and unlighted except by the dingy daylight of northern Europe, we need always to keep fast in our mind the fact that all preconceptions are false that may be derived from our memory of latter-day performances in theaters of a type which the Elizabethan dramatists could not foresee, and of which the conditions are often the exact opposite of those they accepted without hesitation.

That is to say, the most profitable way to reconstruct mentally the Tudor playhouse is to banish from our minds every impression made by our modern theater, with its elaborate complexity, and to study out for ourselves the simple circ.u.mstances of performance in the Middle Ages.

And as a first step toward the proper standpoint, we must cast out our traditional belief that Shakspere always accepted the cla.s.sicist formula of five acts, proclaimed by Horace, and employed by Seneca. That he did use it in one or two plays seems indisputable, and he may very well have employed it in a few others, but there is no reason to suppose that he would have submitted himself any more willingly to the rule of five acts than he did to the rule of the three unities.

It may be doubted also whether not a few dramatists, writing later than Shakspere, would not have done well to claim the liberty he and Lope de Vega chose to exercise at will. Racine, for one, had sadly to stretch his 'Athalie' to fill out the five-act framework which he had blindly accepted, altho he had earlier limited 'Esther' to three acts. Schiller, for another, would have gained a swifter compactness for his play if he had left out the needless fifth act of his 'William Tell' and rolled his fourth act into his third. Victor Hugo had to manufacture a fourth act for his 'Ruy Blas,' so slightly related to his main story that it was cut out of the English adaptation acted by Fechter and Booth. Ibsen, it may be added, composed his first tragedy, 'Catiline,' in three acts, altho it was in blank verse, thus early revealing his characteristic independence of tradition.

(1907.)

P. S.--Since this paper was written I have found two opinions as to the number of acts a play ought to have which were unknown to me when I undertook the discussion. The first is in the 'Dasarupa,' the Hindu treatise on the craft of play-making: "There are five stages of the action which is set on foot by those that strive after a result: Beginning, Effort, Prospect of Success, Certainty of Success, Attainment of the Result."

The second is in the commentary made by Robert Louis Stevenson during his methodical perusal of the dramas of the elder Dumas. After reading 'Henri III et sa Cour,' Stevenson declares that here in Dumas's first piece "is the cloven foot; a fourth act that has no part or lot in the play; a fourth act that is a mere incubus and interruption--that takes the eye off the action, and between two spirited and palpitating scenes interjects a d.a.m.ned sermon on the history of France. Poor Tribonian had a sore job to make up the fifty books of the Pandects; what was that to the labor of a dramatist bent on filling his five acts? I go as far as this: the natural division of the normal play is four: Act I, exposition; Act II, the problem produced; Act III, the problem argued; Act IV, the way out of it."

(1916.)

V

DRAMATIC COLLABORATION

DRAMATIC COLLABORATION