salt to remove salt humours," adding "the homoopathic comparison shows how near he was to the correct notion." Bernays concludes that by Katharsis is denoted the "alleviating discharge" of the emotions themselves. In other words, pity and fear are bad, and it is a good thing to get rid of them in a harmless way, as it is better to be vaccinated than to have small pox.

Now this alleviating discharge is pleasurable (μeľ ¿dovŷs), and the pleasure seems, from allied passages, to arise not in the accomplished relief from oppression, but in the process itself. This becomes intelligible from the point of view of Aristotle's definition of pleasure as an ecstatic condition of the soul. For every emotion contains, according to Aristotle, be it ever so painful, an ecstatic, and hence a pleasurable element; so that the excitement of pity and fear in the ecstatic degree would effect, at the same time with an alleviating discharge, a pleasure also. Pity and fear are aroused to be allayed, and to give pleasure in the arousing and the relief.

Such, approximately, is Aristotle's view of the Tragic Emotion, or Katharsis. Is it also our own? To clear the field for this inquiry, it will be well first of all to insist on a distinction which is mostly discounted in significance because taken for granted. We speak of Aristotle's Katharsis as the Tragic Emotion, forgetting that to-day Tragedy and the Tragic are no longer identical. Aristotle

conceives himself to be dealing with the peculiar emotion aroused by a certain dramatic form, the name of which has nothing to do with its content. For Tragedy is literally goat-song, perhaps from the goat-skins worn by the first performers of tragedy disguised as satyrs. Since then we have borrowed the name of that dramatic form to apply to events which have the same type or issue as in that form. In popular speech to-day the word tragic attaches itself rather to the catastrophe than to the struggle, and therefore, I cannot but think, modern discussion of "the tragic" is wrong in attempting to combine the Aristotelian and the modern shades of meaning, and to embody them both in a single definition. Aristotle is dealing with the whole effect of the dramatic representation of what we should call a tragic occurrence. It is really the theory of the dramatic experience and not of the tragic, in our sense, which occupies him. Therefore, as I say, we must not assume, with many modern critics, that an analysis of the tragic in experience will solve the problem of the Katharsis. Our 'tragic event," it is true, is of the kind which dramatically treated helped to bring about this peculiar effect. But the question of Aristotle and our problem of Katharsis is the problem of the emotion aroused by the Tragic Drama. What, then, is the nature of dramatic emotion?

66

II

The analogy of Aristotle's conception of the emotion of tragedy with certain modern views is evident. To feel pain is to live intensely, it is said; to be absorbed in great, even though overwhelming, events is to make us realize our own pulsing life. The criticism to be made on this theory is, however, no less simple: it consists merely in denying the fact. It does not give us pleasure to have painful emotions or to see other people's sorrows, in spite of the remains of the "gorille féroce" in us, to which Taine and M. Faguet attribute this imputed pleasure. And if we feel pleasure, excitement, elevation in the representation of the tragic, it must be due to some other element in the experience than the mere self-realization involved in suffering. It is indeed our first impulse to say that the painful quality vanishes when the exciting events are known to be unreal; pity and fear are painful because too intense, and in the drama are just sufficiently moderated. The rejoinder is easy, that pity and fear are never anything but painful down to the vanishing point. The slight pity for a child's bruised finger is not more pleasurable because less keen; while our feeling, whatever it is, for Ophelia or Gretchen, becomes more pleasurable in proportion to its intensity.

It is clear that the matter is not so simple as

Aristotle's psychology would make it. Pity and fear do not in themselves produce pleasure, relief, and repose. These emotions as aroused by tragedy are either not what we know as pity and fear in real life, or the manner of their undergoing brings in an entirely new element, on which Aristotle has not touched. In some way or other the pity and fear of tragedy are not like the pity and fear of real life, and in this distinction lies the whole mystery of the dramatic Katharsis.

But there is an extension of Aristotle's theory, lineally descended from that of Lessing, which professes to elucidate this difference and must be taken account of, inasmuch as it represents the modern popular view. Professor Butcher, in his edition of the "Poetics," concludes, on the basis of a reference in the "Politics" implying that the Katharsis of enthusiasm is not identical with the Katharsis of pity and fear, that the word is to be taken less literally, as an expulsion of the morbid elements in the emotions, and these he takes to be the selfish elements which cling to them in real life. Thus "the spectator, who is brought face to face with grander sufferings than his own, experiences a sympathetic ecstasy, a lifting out of himself. It is precisely in this transport of feeling, which carries a man outside his individual self, that the distinctive tragic pleasure resides. Pity and fear are purged of the impure element which clings to them in life. In the glow of tragic excitement these feel

ings are so transformed that the net result is a noble emotional satisfaction."

In spite of our feeling that the literal and naïve reading of the analogy was probably after all nearer Aristotle's meaning, we may accept the words of Professor Butcher as its modern formulation. They sound, indeed, all but a truism: yet they are seen on examination to glide lightly over some psychological difficulties. Firstly, the step is a long one from the pity and fear felt by the Greek toward or about the actors, to a sharing of their emotion. The one is a definite external relation, limited to two emotions; the other, the "sympathetic ecstasy," opens the door to all conceivable emotions, and needs at least to be justified. But, secondly, even suppose the step taken; suppose the "sympathetic imitation" conceded as a fact: the objections to Aristotle's interpretation are equally applicable to this. Why should this "transport of sympathetic feeling" not take the form of a transport of pain? Why should the net result be "a noble emotional satisfaction?" If pity and fear remain pity and fear, whether selfish or unselfish, it doth not yet appear why they are emotionally satisfactory. The "so transformed" of the passage quoted assumes the point at issue and begs the question. That is, if this transformation of feeling does indeed take place, there is at least nothing in the nature of the situation, as yet explained, to account for it. But explanation there must be. To this, the lost