ccess. Why, then, should we assume, in the ideal spectator to whom we
address ourselves, a state of mind which, we hope and trust, will not be
the state of mind of the majority of actual spectators?

To this question there are several answers. The first and most obvious
is that to one audience, at any rate, every play must be absolutely new,
and that it is this first-night audience which in great measure
determines its success or failure. Many plays have survived a
first-night failure, and still more have gone off in a rapid decline
after a first-night success. But these caprices of fortune are not to be
counted on. The only prudent course is for the dramatist to direct all
his thought and care towards conciliating or dominating an audience to
which his theme is entirely unknown,[1] and so coming triumphant through
his first-night ordeal. This principle is subject to a certain
qualification in the case of historic and legendary themes. In treating
such subjects, the dramatist is not relieved of the necessity of
developing his story clearly and interestingly, but has, on the
contrary, an additional charge imposed upon him--that of not flagrantly
defying or disappointing popular knowledge or prejudice. Charles I must
not die in a green old age, Oliver Cromwell must not display the manners
and graces of Sir Charles Grandison, Charles II must not be represented
as a model of domestic virtue. Historians may indict a hero or whitewash
a villain at their leisure; but to the dramatist a hero must be (more or
less) a hero, a villain (more or less) a villain, if accepted tradition
so decrees it.[2] Thus popular knowledge can scarcely be said to lighten
a dramatist's task, but rather to impose a new limitation upon him. In
some cases, however, he can rely on a general knowledge of the historic
background of a given period, which may save him some exposition. An
English audience, for instance, does not require to be told what was the
difference between Cavaliers and Roundheads; nor does any audience, I
imagine, look for a historical disquisition on the Reign of Terror. The
dramatist has only to bring on some ruffianly characters in Phrygian
caps, who address each other as "Citizen" and "Citizeness," and at once
the imagination of the audience will supply the roll of the tumbrels and
the silhouette of the guillotine in the background.

To return to the general question: not only must the dramatist reckon
with one all-important audie