Lest anyone suppose that I exaggerate the difficulty or attribute to Greek playwrights a perplexity which they did not experience, let me point out the confessed failure of a modern poet. Concerning the close of Act III in the second part of Faust, Goethe said: “You have observed the character of the chorus is quite destroyed by the mourning song: until this time it has remained thoroughly antique, or has never belied its girlish nature; but here of a sudden it becomes nobly reflecting, and says things such as it has never thought or could think.” And to this Eckermann, uncontradicted, replied: “These little inconsistencies are of no consequence, if by their means a higher degree of beauty is obtained. The song had to be sung, somehow or other; and as there was no other chorus present, the girls were forced to sing it.”[247] That Euripides was equally conscious of what he was doing is proven by the fact that in some cases he makes only too patent an attempt to gloss over the difficulty. Thus, he makes the chorus in the Electra explain that they had heard of Achilles’ shield in the nearby harbor of Nauplia “from one who had fared from Troy” (vss. 452-55); and the Tyrian maidens justify their knowledge of Theban history by saying that they “had received an account at home in an alien tongue” (Phoenician Maids, vs. 819). A curious self-consciousness seems to obsess dramatic poets and force them to call to the hearer’s attention the very difficulty that they are striving to avoid. Like some scientists who think they have explained a phenomenon if they have provided a name for it, playwrights sometimes act as if they had justified an incongruity if they mention it. An excellent modern illustration of this occurs in Twelfth Night, II, 5. In order to extract the full humor from the scene it is necessary that Malvolio read aloud the forged letter which he has just found. Therefore, Shakespeare makes Sir Toby say: “The spirit of humours intimate reading aloud to him!” Since these words are uttered in an aside, they can have no real effect. Nevertheless, the dramatist eased his conscience by inserting them.

Sometimes the difficulty of finding motifs suitable for the rôle of the chorus caused the playwrights to introduce a second chorus of a different type. Phrynichus seems to have done this in 476 B.C., bringing on a chorus of elders as well as one of Phoenician women.[248] Likewise, in Euripides’ Hippolytus that hero’s comrades in the chase appear and sing a short ode (vss. 61-72) before the arrival of the regular chorus. Several other instances are known of in Euripides’ lost plays. In Seneca’s Agamemnon there is a chorus of Mycenaean women and another of Trojan captives. In the same writer’s Hercules on Mt. Oeta, Dr. Fries (op. cit., p. 49) maintains that three choruses are introduced, one of Oechalian captives at vs. 104, another of Deianira’s companions at vs. 583, and a third of Hercules’ comrades at vs. 1031. The same sort of thing occurs also in comedy. Thus, from Terence’s Self-Tormentor, which is a Latin translation of Menander’s play of the same name, it would appear that in the Greek original a chorus of banqueting companions performed at vs. 171 and another chorus of maidservants at vss. 409 and 748.[249] Occasionally, before making its appearance, the chorus sings, from behind the scenes, in a different character from that which it later assumes. Aristophanes’ Frogs, for example, derives its name from a chorus which never is seen. At vs. 209 the chorus, from behind the scenes, delivers a batrachian strain as an accompaniment to Dionysus and Charon when they row across the subterranean lake (see [p. 90], above). It is not until after vs. 315 that this chorus actually appears and reveals its true character, that of men and women who had, when on earth, been initiated into the mysteries. This method of procedure gained one of two results—it obviated the necessity either of a lightning change of costume on the part of the chorus or that of hiring extra choreutae. As to the latter alternative, whatever may have been true of the tragic poets (see [p. 134], above), there is no reason to suppose that the comic poets always had spare choreutae at their disposal.

But not only should choral odes be appropriate to the dramatic character of the chorus; they ought also to be closely connected with the theme of the play. And this requirement is no less difficult than the other. The ode on the inventive spirit of man in Sophocles’ Antigone, vss. 334-75, is so vague that an audience might well be in doubt as to which one of the dramatic characters it was intended for. Verses 1115-52 in the same play, a hymn to Dionysus, is quite irrelevant, except in so far as that divinity was the patron of the dramatic festival. Other instances are found in Euripides. Verses 1301-68 of Helen deal with Demeter’s search for her lost daughter and are so alien to the subject of the tragedy that many have considered them an interpolation. An adventitious connection is sought, at the close, by the suggestion that Helen’s misfortunes are due to her neglect of Demeter’s worship (vss. 1355-57). Again, the chorus’ eulogy of Apollo in Iphigenia among the Taurians, vss. 1234-83, is so disconnected with the story that Professor Decharme (op. cit., pp. 312 f.) could defend it only by saying: “If, therefore, the chorus wishes not to rouse the suspicion of Thoas, it must speak of something else than that which really engrosses its attention. Hence the eulogy of Apollo that compromises nobody, whose purport Thoas would not understand were he to appear suddenly, but which the spectator comprehends, provided he reflects.” The description of Achilles’ armor in Euripides’ Electra, vss. 434-78, has already been mentioned ([pp. 139 f.], above). It is as little connected with the plot as it is appropriate to the chorus of that play. Nevertheless, Euripides brought the ode back to the theme with a jerk by saying: “The lord of such warriors didst thou slay, O Clytemnestra” (vss. 479 f.). There are but two things that can be said to palliate this offense. The first is to indicate the difficulty of the problem; the other, to point out that the ingenuity of the ancient playwrights fell short in only a few plays and seldom more than once in any one piece.

There are certain ways, however, in which the lack of an organic relationship between chorus and actors or the failure of the odes to spring naturally from the dramatic situation may be compensated for or glossed over. One is by giving the choreutae an active participation in the action. The scene of Euripides’ Helen is laid in Egypt and the chorus consists of Greek slaves, who assist the heroine in her deception mainly because she is a fellow-Greek and her victim a barbarian. Their connection, therefore, is only moderately close and, as we have seen ([p. 142], above), one of their odes is by some considered an interpolation. Yet, apart from their choral songs, they take an active and important part in the play. It is they who persuade Helen not to believe Teucer’s announcement of her husband’s death but to consult the seeress Theonoe concerning the matter (vss. 306 and 317). Again, it is they who, when the Egyptian king avows his intention of murdering Theonoe for abetting his deceivers, grasp his garments and declare: “We are your slaves and you can slay us, but slay us you must ere you can kill Theonoe” (vss. 1629 ff.). Similarly, in Euripides’ Orestes the chorus of Argive women is friendly toward Electra and her brother but does not share the danger which threatens them. Yet when Helen is being murdered behind the scenes, at Electra’s request, in order to guard against surprise, it divides into semi-choruses, which picket the two roads leading before the palace (vss. 1251 ff.). A little later they attempt to make noise enough to prevent the tumult from within the palace attracting the notice of the Argive citizens (vss. 1353 ff.). Thus, a chorus may actively participate in a plot to which it is but loosely joined. In fact, Professor Capps has boldly declared: “In every play whose chorus has been criticized for the irrelevancy of its songs, whether the criticisms have been just or not, are found indications of direct participation in the action” (op. cit., p. 295).

In this connection certain words of Aristotle[250] are usually cited: “The chorus ought to be regarded as one of the actors; it ought to be an integral part of the whole and take a share in the action, in the manner, not of Euripides but of Sophocles. The choral songs of the successors of Euripides and Sophocles have no more to do with the subject of the piece than with that of some other tragedy. They are therefore sung as mere intercalary numbers (ἐμβόλιμα), a practice first begun by Agathon. Yet this is no more justifiable than to transfer a speech or a whole act from one tragedy to another.” Aristotle’s praise of Sophocles at the expense of Euripides probably refers to the choice and setting of Sophoclean choruses and to the relevancy of their songs—points in which Sophocles usually surpassed his rival. Aristotle failed to notice or did not value the other characteristic of Euripidean choruses, viz., that they have more effect upon the plot and come into more direct contact with the actors, that is to say, that they really “act” more, than is the case in Sophocles. In fact, it is Sophocles’ use of the chorus which is mainly responsible for the modern notion that the Greek chorus was merely the “ideal spectator.”

The precise meaning of the latter part of this passage from the Poetics has not until recently become clear. It is evident that Aristotle brings the same charge, that of irrelevancy, against the choruses of both Euripides and Agathon. But if the difference between them were merely one of degree, he would hardly have said that Agathon “began” a practice which he had really borrowed from Euripides and only “developed” or “extended.” Therefore, Aristotle must mean that Agathon was guilty of a different kind of irrelevancy than Euripides, and we are now in a position to see whereof this consisted. Recently discovered fragments of Menander show that often in New Comedy the chorus did not appear in the course of the action at all, but only between acts, and that the poets did not write down the words of these entr’actes but simply indicated where they should come by writing the word ΧΟΡΟΥ (“of the chorus”) at the places required. To the stage manager ΧΟΡΟΥ in the manuscript would be simply a hint to use anything he chose or to refer to the poet or that he could rely upon the latter to provide the choreutae with a libretto, according to whatever arrangement they had between them on the subject. To the reader it was convenient, as marking off the divisions of the play. A parallel to this custom is found in Greene’s James the Fourth, where at the beginning of Act IV the stage directions read “Enter certain huntsmen (if you please, singing),” and again at the close of the same act, “Enter a round, or some dance at pleasure.” A passage in the ancient Life of Aristophanes had already mentioned this practice of the writers of New Comedy but had received scant consideration until substantiated by the Menander fragments.

Now, since embolimon means “something thrown in,” it seems clear that the songs of the chorus in the intermissions marked by ΧΟΡΟΥ (if songs not recorded in the text were sung) would be embolima in Aristotle’s use of the term. I believe that this was the innovation which Agathon introduced. This conclusion will be strengthened if we ask ourselves what sort of evidence enabled Aristotle to attribute the invention of embolima to Agathon. It is fairly certain that he never saw one of Agathon’s tragedies actually performed in the theater. Then his knowledge of Agathon’s dramatic art must have depended upon the latter’s published works. Therefore, if Agathon’s choral numbers were notable rather for the music than for the libretto, or consisted of music and dancing without words, or were borrowed from other poets, or if for any reason whatsoever Agathon preferred not to copy them down with the rest of the text, but merely to mark their location by ΧΟΡΟΥ or some other symbol, then we can understand how Aristotle could know that Agathon had inaugurated something new in dramatic technique. Whatever their defects of irrelevancy, Euripides’ odes were not “thrown in” in this sense; they were right there in the text. But in Agathon’s manuscripts, on the other hand, there were gaps indicated between acts. In actual performance suitable odes were “thrown in.” A “thrown-in” ode then would be one not appearing in the text. It is self-evident that this interpretation throws a flood of light upon Aristotle’s statements.

That ΧΟΡΟΥ was so used in tragedy prior to the time of New Comedy is attested by its occurrence in a recent fragment of a fourth-century Medea.[251] Moreover, by inference its use can be safely traced still further back, even close to the period of Agathon. We have seen that tragedy exercised a profound influence upon Old Comedy (see [pp. 49 f.], above); and Professor Navarre[252] has correctly pointed out that the influence of tragedy was more quickly and strongly felt in the second half of a comedy (that after the parabasis or, when that is lacking, after the agon; see [p. 41], above). Accordingly a strong reason for believing that this use of ΧΟΡΟΥ originated in tragedy is found in the fact that ΧΟΡΟΥ occurs in this part of Aristophanes’ last two (extant) comedies; cf. Women in Council, vss. 729 and 876 (393-392 B.C.), and Plutus, vs. 770 (388 B.C.). It is significant that Aristophanes’ use of embolima is still embryonic, has not yet been carried to the logical issue found in New Comedy. That is to say, the chorus of these two plays still figures in the action and converses with the actors. In the Women in Council it even has, in addition to embolima, several choral songs, the words of which are preserved. The fragments of the fourth-century Medea, scanty as they are, nevertheless suffice to indicate that its author employed embolima and the chorus in the same fashion as Aristophanes.

But by the time of New Comedy a great change had taken place. In comedies of this period, or at least in many of them, the chorus appeared only to furnish entertainment between acts, withdrawing again at the end of its performance. It bore no speaking part and from the nature of the case could exercise no influence upon the plot. Occasionally it was brought into formal relationship with one of the actors. For example, in Menander’s Girl with the Shorn Locks the chorus seems to consist of Polemon’s boon companions, who took breakfast with him in the country and have now come to his house in the city to be on hand for the dinner in the evening. This is the most frequent type of chorus in New Comedy. The approach of these intermezzic choruses is often mentioned by the actors who thus motivate their own withdrawal from the scene during the choral entr’acte. For instance, in one case[253] ΧΟΡΟΥ is prefaced by one character remarking to another: “Let us withdraw into Charisius’ home, for a throng of tipsy youths is approaching whom it is inadvisable to provoke.” Such an introduction occurs also in a fragment of Alexis, a poet of Middle Comedy,[254] but the quotation is not long enough to determine whether Alexis resembled Aristophanes or the New Comedy in his use of embolima and of the chorus. Racine’s Athalie, which has been pronounced[255] the “one thoroughly satisfactory choric drama” that modern art has produced, presents several points of likeness to the later Greek chorus. The Levite maidens do not appear until just before the close of the first act and are withdrawn several times subsequently, being thus absent from the scene during long stretches of the dialogue. Their entrances, also, are sometimes alluded to by the actors. Their songs, however, are not embolima, but constituent parts of the text.