The first printed editions of tragedy belong to the end of the fifteenth century. It has been well said that the age of the printed book dawned just in time to secure the survival of the genre of tragedy. From now on the inevitable process of corruption could be not only halted but also turned back. From the fifth century bce to the fifteenth century ce the text of the tragic poets underwent, as we have seen, a progressive deterioration. Some scholars, of course, such as Aristophanes of Byzantium, were aware of corruption and tried to correct it. But in the absence of printing their work could have little effect on the majority of copies. It was in any case, as far as we can tell, unsystematic and focused on particular cases, without any attempt to formulate a general plan of attack. With the Renaissance begins a new era in the history of the text, an era when improvements once made could be transmitted to future generations and piecemeal and scattershot methods were gradually replaced with a systematic and scientific attempt to view problems whole. Like the natural sciences, which made impressive progress in the same period, classical scholarship now tried increasingly to discover the regularities of its subject and use them to explain or correct the primary evidence presented to our senses, the text given in our manuscripts. It is an impressive achievement of the human spirit, and to it we owe the fact that our texts of the tragic poets, like those of all classical authors, are now intelligible in so many places where they were previously obscure. It is impossible in this brief treatment to touch on all the kinds of investigation that have contributed to this effort, and we will have to be content with illustrative synecdoche.
To collect all the variant readings for a given text and make them available for the reader was a big step, one already taken unsystematically in Alexandria. But it became gradually clear that textual witnesses, like witnesses in a court of law, had to be evaluated, not merely cited. By the nineteenth century scholars learned to eliminate manuscripts that were merely copies of other existing manuscripts (codices descripti):. such derivative witnesses will contain no new truth. Then they learned to sift through what was left, trying to arrive in each case at the earliest state of the text attested to by our manuscripts. This in turn they learned to subject to examination: is this what the author wrote, and if not can we make a guess, based on the regularities of the genre and the author’s style, as to what he did write? The principles on which this recensio or sifting of the evidence is based are commonsensical once stated, but it took a while for them to be formulated.
Our principal means of testing transmitted readings to see if they are genuine has to do with the observed regularities of Greek grammar, Greek meter, and tragic convention. The relationship between a ‘‘rule’’ of grammar, meter, or convention and the manuscripts is often paradoxical, for to almost every rule the manuscripts give exceptions, and often the critic finds himself deriving a rule from the witness of the manuscripts but then turning around and correcting the exceptions. This is paradoxical but is not as irrational as it sounds. Take, for example, the metrical example of Porson’s Law, which says that when in a tragic trimeter there is a word break before the third syllable from the end, the syllable before that word break, which could theoretically be long or short, must be short. How do we deal with the exceptions (for there are some)? Some of the exceptions, scholars have decided, require us to refine the rule: thus examination of a large category of exceptions leads to the conclusion that if the syllable before the word break is a monosyllabic word, then it need not be short. But in other cases we decide that a line that violates this rule cannot have been written as it stands by a tragic poet in the classical period: at Iphigenia at Aulis 1578-1614, where Porson’s Law, together with many other rules of tragic meter, is repeatedly broken, scholars diagnose later interpolation, whereas in Ion 1 they emend to produce regular meter. In the case of all such rules and apparent exceptions we must decide whether there is some good reason for the exception (in which case it may be allowed to stand) or whether, on the contrary, we can see other features that show the passage in question to have been interpolated or corrupted. By studying the regularities of tragic style and the typical ways the tragic poets dealt with recurring dramatic situations, or by paying attention to the habits of scribes or studying the habits of interpolators where interpolation has been convincingly diagnosed, other scholars have given their successors a body of evidence with which to judge the soundness of our texts.
The idea is occasionally expressed by scholars who should know better that after five centuries of assiduous effort by scholars in nearly all the countries of Europe and North America all the corruptions in our tragic texts are either healed or incurable. The truth is that students of tragedy who know Greek well, are willing to take the trouble to acquire the technical knowledge and the familiarity with the tragic texts possessed by their predecessors, and have some measure of their predecessors’ knack of seeing just where a text is ailing and what might set it right, need not despair of improving the text.
I end with two examples of textual argument chosen because they can be made intelligible to a Greekless reader. There are some technical points at issue in each of them, but they also show that it is at times useful to be conscious of one’s own literary presuppositions and those of one’s predecessors. For literary presuppositions play an important role in the decisions editors make, and it is much better if editors are conscious of these presuppositions and aware that there may be evidence against them (an argument made at greater length in Kovacs 1987). To put this another way, when examining any particular passage in a play it is a good idea to have the entire play in view.
My first example demonstrates that not even well-established conjectures should be taken on authority: all must be scrutinized again from time to time. At the end of Euripides’ Electra, when Electra and Orestes, having killed Clytemnestra, have fallen into despair, Castor and Polydeuces arrive by the stage crane ( mechane) and announce the future for the two siblings: both must leave Argos, Electra to marry Pylades and Orestes to be pursued by the Erinyes until, having finally been acquitted in Athens, his troubles are ended and he lives happily ever after. Immediately after Castor’s speech, there is the following interchange:
Orestes: Sons of Zeus, may we have speech with you?
Castor: Yes, you may, for you are not polluted [masculine adj.] by this bloodshed.
Electra: May I too speak to you?
Castor: You too may do so: for it is Apollo I blame for this murderous affair.
The first printed edition of the play, edited by Petrus Victorius (Pietro Vettori), appeared in 1545. (Electra was somehow overlooked in the first printed edition of the extant plays of Euripides, the 1503 Aldine.) Feeling that Orestes was indeed ‘‘polluted by this bloodshed’’ Victorius gave Orestes’ line to the leader of the chorus of Argive women, which means Castor used that phrase of her. As previously noted, speaker indications do not go back to the author’s autograph, so changing them with good cause is a sensible thing to do. In this case, however, Victorius also had to change the masculine adjective for ‘‘polluted’’ to feminine. Two slight changes, therefore, that seemed to give better sense.
In fact, however, the sense was decidedly worse. (1) When a deus ex machina appears at the end of a play, he comes to address the principals of the drama, and it is without parallel in tragedy for a chorus to push themselves forward to meddle in a conversation that does not directly concern them. (2) Castor, as his preceding speech has made clear, has come to console his kinsman, not to chastise him, but with Victorius’ emendation Castor’s speech makes a gratuitous jab at Orestes: ‘‘Yes, you may, for you (unlike Orestes) are not polluted by this bloodshed.’’ (3) After such a comment to the chorus-leader there is no reason for Electra to ask, ‘‘May I too speak?’’: Castor has allowed the chorus-leader to speak only because she is innocent of Clytemnestra’s murder, and Electra is not. (4) Castor’s reply to Electra makes no sense either since his leniency toward her because of the command of Apollo applies no less to Orestes than to her, and if she is free to address the gods, so should he be. By contrast, if we take the parallelism of the two answers at face value and assign the two questions to the two siblings, this suggests that both address Castor and both are excused by him. In effect he says to both, ‘‘I do not regard you as polluted by this bloodshed since it is Apollo I blame for this murderous affair,’’ this single statement being parceled out between his two interlocutors. There is a parallel for Castor’s view of pollution (if your actions were not morally blameworthy you are not polluted) at Orestes 75-76. In my view, the transmitted assignments are correct, and every editor from Victorius in 1545 until Basta Donzelli’s Teubner fascicle of 1995 has got the thing wrong, exactly 450 years of error. (It must be noted, however, that some had argued against Victorius: see Stoessl 1956, 82-85; Kovacs 1985, 310-14.) The persistence of such an error owes much, in my judgment, to the belief that Euripides’ deus ex machina scenes are not really intended to resolve anything and are not to be taken seriously. By contrast, an editor who is willing to look independently at the tone of the final scene and is not ready to believe on very little evidence that Euripides’ dialogue will be incoherent can restore Euripides’ meaning.
My second example shows the converse duty of skepticism toward what is transmitted. Immediately after the parodos of Euripides’ Bacchae, the prophet Tiresias comes to Cadmus’ house to take him to the mountains to worship Dionysus. Their dialogue at 195-203 runs as follows:
Cadmus: Shall we alone dance in the god’s honor?
Tiresias: Yes, we alone have sense, the others none.
Cadmus: The wait is long. But take hold of my hand.
Tiresias: There, clasp my hand and pair it with your own.
Cadmus: I do not despise the gods, mortal that I am.
Tiresias: In no way do we play the sophist to the powers above.
The traditions of our fathers, which we possess coeval with time,
No argument shall overthrow them,
No matter what subtleties have been invented by deep thinkers.
There are problems here that many editors have chosen to ignore. (1) The dative toisi daimosi, ''to the powers above,’’ has no obvious construction in Greek. It could be accommodated by redividing ouden sophizomestha as oud' ensophizomestha, “nor do I play the sophist among,’’ as suggested by Musgrave. (2) In a continuous argument it is customary to connect each sentence with the previous one by means of a connecting particle, just as it is in English to start it with an uppercase letter. Only when the second sentence is explicative of the first (and we would punctuate at the end of the first with a colon) is it usual to leave out the connective. Between 200 and 201 there is no connective, an effect I have approximated by printing 201 without its initial capital. (3) In 202 the superfluous “them” is as clumsy in Greek as in my English translation, indeed more so since it does not agree in gender with “traditions.” (4) The appeal to traditions coeval with time seems an odd way to promote the new god Dionysus. The previous Oxford editor, Gilbert Murray, was untroubled by these problems and printed the text as it is transmitted. The latest Oxford editor, James Diggle, rightly thinks they cannot be ignored. His way of fixing them is to bracket 199-203 (an earlier scholar had already proposed bracketing 201-2).
Now the absence of a connective and the inept pronoun might be thought to point to a clumsy writer of a later age, but no diagnosis of interpolation is fully plausible unless we can see some motive for it. What could have motivated an interpolator here? In particular what person of any age could have thought of the idea of describing the worship of the new god Dionysus as “coeval with time’’?
Once we ask this question, it answers itself: this is precisely what Euripides does elsewhere in the play, and the paradoxical idea of connecting the worship of Dionysus with age-old sanctities could only be his. In 96-97 he has his chorus say, “I shall hymn the god with songs that have ever been in use,’’ and at 890-96 the chorus, urging piety toward Dionysus, praise as sovereign “what through long ages has ever been lawful and upheld by nature.’’ It is in this same vein that Euripides gives Dionysus’ cult many of the trappings of the ancient religion of Cybele, the Great Mother Goddess. Why he does this can be briefly put: Pentheus in this play is to be a fighter not merely against novel divinities but against divinity in general. (On this point see Kovacs 1994-2004, 6: 6-8.) Once we realize that this idea is central to the whole play, our duty is plain: we must emend, not delete. The absence of a connective could be a sign that a line has dropped out (as suggested by Kirchhoff: there are other such lacunae [gaps] at 652, 843, and 1036). The lost line could have had a verb to govern “traditions,” and then we could repunctuate before “no argument’’ with a perfectly idiomatic colon (in Greek a raised period) and put the offending pronoun in a different sentence. Here is a restoration (with the second speaker change moved down a line and Musgrave’s redivision of ouden sophizomestha as oud' ensophizomestha adopted):
Cadmus: I do not despise the gods, mortal that I am,
Nor do I play the sophist among the powers above.
Tiresias: <No, it is not the part of a wise man to look down on>
The traditions of our fathers, which we possess coeval with time: no argument shall overthrow them,
No matter what subtleties have been invented by deep thinkers.
Progress is indeed possible. Neither the scribes who copied our manuscripts nor their many predecessors in the chain of transmission deserve our uncritical faith; likewise even the most well-established conjectures of our scholarly predecessors can sometimes be reexamined with profit. We can judge for ourselves the regularities of the Greek language and tragic convention and the practice of the individual playwrights. What is needed is knowledge of the requisite technicalities joined with a sense for the work as a whole.
World History
