NOTE ON THE TEXT

We have no manuscripts in Euripides’ hand, or going back anywhere near his own time. If we had, they would be difficult to decipher, and would lack many aids which the modern reader takes for granted: stage directions, punctuation, clear indications of change of speaker, regular divisions between lines and even between words. In fact, although some parts of his plays, mostly short extracts, survive in papyri from the earliest centuries AD, our complete manuscripts of the plays translated in this volume go back no further than the tenth century. Moreover, the textual evidence for the various plays differs greatly in quantity. Three plays were especially popular in later antiquity, namely Hecabe, the Phoenician Women and Orestes (the so-called ‘Byzantine triad’). These survive in more than 200 manuscripts, and modern editions of the last two rest on a selection of a significant number of these. The opposite situation prevails with Bacchae and Iphigenia at Aulis. Here we are essentially dependent on a single primary manuscript of the early fourteenth century, known as L (short for Laurentianus 32.2); in the case of the Bacchae L fails us after line 755, and recourse must be had to another manuscript of the same period known as P (Palatinus Gr. 287), whose relation to L is debated. A slightly larger range of manuscripts offers independent evidence for Rhesus. Papyrus fragments survive for all five plays in this volume: again most numerous are those for the Phoenician Women and the Orestes, reflecting the popularity of these two.

This situation is not unusual in the history of classical authors. No ancient dramatist’s work survives in his own hand: in all cases we are dealing with a text transmitted by one route or several, and copied many times over. In an age which knew nothing of the printing-press, far less the Xerox machine, all copying had to be done by hand, every copy in a sense a new version. The opportunities for corruption of the text – that is, the introduction of error – were numerous. The reasons for such corruption include simple miscopying or misunderstanding by the scribe, omission or addition of passages by actors in later productions, efforts to improve the text by readers who felt, rightly or wrongly, that it must be corrupt, accidental inclusion of marginal notes or quotations from other plays, and very occasionally bowdlerization of ‘unsuitable’ passages. Problems of this kind were already recognized in antiquity: efforts were made to stabilize the texts of the tragedians in fourth-century BC Athens, and the ancient commentaries or ‘scholia’ to some of Euripides’ plays make frequent comments on textual matters, for instance remarking that a line is ‘not to be found’ in some of their early manuscripts, now lost to us. In the same way, when a modern scholar produces an edition of a Euripidean play, there are many places where he or she must decide between different versions given in different manuscripts. Sometimes the choice will be easy: one version may be unmetrical, ungrammatical or meaningless. But often the decision may be more difficult, and in many cases it is clear that no manuscript preserves the lines in question in the correct form. Hence the editor must either reconstruct Euripides’ authentic text by ‘conjecture’, or indicate that the passage is insolubly corrupt, a conclusion normally signalled by printing daggers (‘obeli’) on either side of the perplexing passage.

A translator is in a slightly more fortunate position than an editor. The editor must make a decision what to print at every point, and uncertainty may prevail as to the exact wording even when the overall sense is fairly clear. In this translation James Diggle’s excellent Oxford Classical Text has normally been followed. When he has marked a word or phrase as probably or certainly corrupt, we have usually adopted a conjectural reading, whether made by him or by a previous editor, even though we often agree that there can be no certainty that this is what Euripides actually wrote. In cases where the corruption is more extensive, we have tried to give a probable idea of the train of thought. These problems arise particularly in choral and other lyric passages, where the language is less close to everyday speech, and where unusual metre and dialect often misled copyists.

Many of the smaller problems involving variations of words or uncertainty over phrasing will be unlikely to cause difficulties to users of this translation. More noticeable are the occasional places where it seems that something has dropped out of the text; usually this can be explained by the accidents of miscopying or by damage to some of the manuscripts from which our texts descend. Scholars refer to a gap of this kind as a lacuna. The most important example of this in the plays translated here occurs in the closing scenes of the Bacchae. Here we are dependent on manuscript P, but there are two points at which the text of P’s source was evidently defective: we have lost a passage of uncertain length after line 1300 and another, probably longer section after 1329 (see the Notes on these lines). Besides this, there are other passages where a lacuna has been detected or suspected: Bacchae 652 and 1036 are certain examples, 200 and 843 highly probable. It is also possible, though by no means certain, that an initial prologue has been lost at the start of the Rhesus.

A much more serious problem which affects criticism of Euripides is that of interpolation. This is the term used to describe the inclusion of alien material in the original text, expanding and elaborating on the author’s words. Sometimes the new material betrays itself by its very unsuitability to the context, and we may suppose that it has been included by accident (for instance, parallels from other plays were sometimes copied out in the margin, then found their way into the text in subsequent copies). Sometimes lines may be present in one manuscript but omitted in others: if they seem superflous in themselves, they may well be a later addition. Sometimes a speech may seem unnecessarily wordy, and we may suspect without feeling certain that it has been expanded; here textual criticism merges with literary judgement. It has often been suggested that some passages in the plays have been ‘padded out’ by actors seeking to improve their parts: although this phenomenon has probably been exaggerated, it would be a mistake to rule it out altogether. One speech which has fallen under suspicion on these grounds is Medea’s famous soliloquy as she wavers over the killing of her children (Medea 1019–80: the boldest critics would excise all of 1056–80). In the translation our normal policy is to follow Diggle’s text, and therefore we normally omit passages which he brands as interpolated. Already in volume 1 of this translation, we made an exception for this speech of Medea, because the speech is so distinguished and the case remains controversial.

In the last plays of Euripides, however, there are special problems, which vary from play to play. In the Orestes, Bacchae and Rhesus (in the last of these the issue seldom arises) we have continued our former practice, but upon occasion have been more inclusive than Diggle. In particular we retain, though with cautionary comments in the Notes, a few passages in Orestes and Bacchae where his determination to excise the irrelevant may be thought to have gone too far (especially Orestes 895–7, 904–13, Bacchae 199–203). In the Phoenician Women it is widely held that extensive passages have been added or altered by later hands: scholars are in general agreement that the last 150 lines or so of the play are spurious, but there is little consensus on earlier parts of the drama. Under these circumstances we have presented the reader with a translation of most of the text, including the conclusion, while continuing to exclude shorter passages of one or two lines which seem to mar the sense or to be clearly intrusive. This may seem inconsistent, but it at least offers those using this volume a version which could be performed as a whole. Many specific difficulties are discussed further in the Notes.

The most recalcitrant case is the Iphigenia at Aulis. It is clear from ancient testimony that this play was posthumously produced, and it seems certain that Euripides did not finish it. Probably the greater part was complete, but there has been some additional material included at the beginning and elsewhere, either for the first performance or in the next century. At a later stage the ending has been tampered with: one hypothesis is that the original ending was lost and replaced, but by a greatly inferior composer, whose intervention is betrayed by metrical and linguistic blunders. In these circumstances to provide a version consisting only of certainly Euripidean material is hardly possible, and to do so would mean truncating the play. We therefore present a translation of the entire text, despite the manifest inconsistencies of plot, style and dramatic technique. (See further the Preface and Notes to this play.)

These complications should not deter the student or general reader. The text of Euripides is much better preserved than that of Aeschylus, and the magisterial Oxford edition of James Diggle has placed it on a much firmer foundation. Knowledge of the nature of the textual tradition helps the modern reader to understand why certain difficulties occur, but despite the depredations of the centuries, Euripides’ authentic text survives in great part, and whether in the original or in translation will long continue to exercise a fascination on reader and audience.