But having written so much for imitation of what is excellent, in that part of the preface which related only to myself, methinks it would neither be unprofitable nor unpleasant to inquire how far we ought to imitate our own poets, Shakespeare and Fletcher, in their tragedies; and this will occasion another inquiry, how those two writers differ between themselves:  but since neither of these questions can be solved, unless some measures be first taken, by which we may be enabled to judge truly of their writings, I shall endeavour, as briefly as I can, to discover the grounds and reason of all criticism, applying them in this place only to Tragedy.  Aristotle with his interpreters, and Horace, and Longinus, are the authors to whom I owe my lights; and what part soever of my own plays, or of this, which no mending could make regular, shall fall under the condemnation of such judges, it would be impudence in me to defend.  I think it no shame to retract my errors, and am well pleased to suffer in the cause, if the art may be improved at my expence:  I therefore proceed to

THE GROUNDS OF CRITICISM IN TRAGEDY.

Tragedy is thus defined by Aristotle (omitting what I thought unnecessary in his definition).  It is an imitation of one entire, great, and probable action; not told, but represented; which, by moving in us fear and pity, is conducive to the purging of those two passions in our minds.  More largely thus:  Tragedy describes or paints an action, which action must have all the properties above named.  First, it must be one or single; that is, it must not be a history of one man’s life, suppose of Alexander the Great, or Julius Caesar, but one single action of theirs.  This condemns all Shakespeare’s historical plays, which are rather chronicles represented, than tragedies; and all double action of plays.  As, to avoid a satire upon others, I will make bold with my own “Marriage A-la-mode,” where there are manifestly two actions, not depending on one another; but in “OEdipus” there cannot properly be said to be two actions, because the love of Adrastus and Eurydice has a necessary dependence on the principal design into which it is woven.  The natural reason of this rule is plain; for two different independent actions distract the attention and concernment of the audience, and consequently destroy the intention of the poet; if his business be to move terror and pity, and one of his actions he comical, the other tragical, the former will divert the people, and utterly make void his greater purpose.  Therefore, as in perspective, so in tragedy, there must be a point of sight in which all the lines terminate; otherwise the eye wanders, and the work is false.  This was the practice of the Grecian stage.  But Terence made an innovation in the Roman:  all his plays have double actions; for it was his custom to translate two Greek comedies, and to weave them into one of his, yet so, that both their actions were comical, and one was principal, the other but secondary or subservient.  And this has obtained on the English stage, to give us the pleasure of variety.