Aeschylus

Three Athenian tragedians have plays that survive: Aeschylus, Sophocles and Euripides. When their metrical practice is compared, Aeschylus is again and again the strictest or the earliest of the three.^[1]

His life is known only roughly. The ancient Life of Aeschylus was built on anecdotes collected by Ion of Chios and on the caricatures of the comic poets, and it “contains few reliable facts”.^[2]

Dates. He was born in 525/4 BC, though the date is uncertain, and he died in 456.^[3] He produced his first play in 498 and won his first victory in 484. His rivals were Choerilus, Pratinas, Phrynichus and, at the end of his life, Sophocles.^[4] The surviving plays that can be dated run from the Persians of 472 to the Oresteia of 458.^[5]

The fixed points come from official records. The Parian Marble is an inscription from Paros, carved in the middle of the third century BC, that lists dated events from 1581 to 264. It says that Aeschylus fought at Marathon in 490 at the age of thirty-five, and that he died at Gela at sixty-nine. The birth in 525/4 and the death in 456/5 are worked out from those ages. The Suda says he first competed at twenty-five, and its entry on Pratinas puts that contest in the years 500–496.^[6]

Output and success. The ancient figures for the number of his plays disagree. They range from seventy-three to ninety.^[7] One source gives seventy tragedies and “around five” satyr-plays. Another gives ninety plays, and ninety is “the more plausible” figure.^[8] About sixty titles of his tragedies are known.^[9]

He won thirteen victories in his lifetime and more after his death. This may explain the Suda’s figure of twenty-eight.^[7] After his death the Athenians voted that anyone who wished to put on one of his plays should be granted a chorus.^[10] Centuries later, Quintilian says that “revised” versions of his plays won prizes.^[11]

The Life and the Suda agree on thirteen first prizes at the Great Dionysia. Between seventy and ninety plays is as close as the evidence allows. If he competed about every other year, his thirteen wins came from some twenty attempts.^[12]

Soldier and guest of Hiero. The epitaph preserved in the Life remembers him as a soldier at Marathon, not as a poet. He is also said to have fought at Salamis and Plataea.^[13] Hiero of Syracuse invited him to compose plays for festivals under his patronage.^[14] A tradition says that he, like Pindar, “ate tyrannical bread” in Sicily, but the biographical tradition may not deserve our trust.^[15]

He went to Sicily at least twice, and he died there. For Hiero’s founding of the city of Aetna in 476/5 he wrote the Women of Aetna.^[16] Later sources give hostile reasons for his leaving Athens, such as defeat by the young Sophocles. But Simonides and Pindar were also honoured by Hiero, so the visits are better read as a sign of his fame abroad. The people of Gela gave him a rich burial and honoured him as a hero.^[17]

Like Pindar, he seems to have been given the status of a classic early, perhaps in his lifetime.^[18] One source says that Anacreon lived long enough to know his early tragedies.^[19]

Among the lyric poets. Aeschylus is said to have called his plays “slices” from the great banquets of Homer (Athenaeus 347e).^[20] Yet much of his material and manner came from lyric. His Oresteia owes a great deal to Stesichorus. An ancient commentator noted this debt in the recognition by a lock of hair.^[21] He also imitated some characteristic rhythms of Anacreon, such as the mixture of glyconics and pherecrateans.^[22]

In the Hellenistic period, collections of epigrams went under his name, as they did under Anacreon’s and Simonides’. Their attributions are hard to trust.^[23] Later still, Timotheus’ Persians quotes him in its monodies.^[24]

Where he stands. Stesichorus, Simonides and Pindar began to mix different metres in one song. In the fifth century the choral songs of Attic tragedy carried this mixing a stage further.^[25] Four measures of style show where Aeschylus stands:

The plays, their dates and the dispute over Prometheus Bound are on the Plays tab; how his texts reached us is on the Transmission and Manuscripts tabs.

Seven tragedies survive under Aeschylus’ name, out of some seventy to ninety plays.^[1] The text quoted on this site is the Perseus Digital Library’s digitisation of Herbert Weir Smyth’s Loeb edition (1922–1926). Five of the seven are dated by the production records: the Persians, the Seven and the three plays of the Oresteia. A papyrus fixes the decade of a sixth.

The earliest surviving plays open the second third of the fifth century.^[3] The Eumenides belongs to 458.^[4]

A papyrus redates a play. The Suppliant Maidens has a chorus of Danaids. It used to be read as the earliest surviving tragedy, a “primitive” piece of the 490s. A hypothesis is an ancient summary of a play, often with a note on its first production. In 1952 a scrap of the hypothesis to this play was published (P.Oxy. 2256 fr. 3). It showed that the play was produced in competition with Sophocles. So it belongs in the 460s, late in Aeschylus’ career.^[5] Hall puts it in “the later 460s”.^[6]

Sophocles first competed in 468, so the play cannot be earlier than that. The papyrus also has the letters ἐπὶ Ἀρ[. If they begin the name of the archon Archedemides, the year is 463.^[7]

The lost Egyptians belonged to the same trilogy. Whether it came first or second is disputed. Del Grande and Sommerstein put it first. Winnington-Ingram, Garvie, Herington and, Saïd says, most scholars put it second.^[8]

Trilogies. The Oresteia is the only surviving trilogy. The trilogy itself was most likely Aeschylus’ own invention.^[9] A tetralogy is a set of three tragedies and a satyr-play, performed together. Aeschylus’ connected tetralogies certainly include the Theban one of 467, the Danaid one, the Oresteia and an undated Lycurgus group. The plays of 472 seem to have been unconnected.^[10]

The Persians is the only surviving tragedy on recent history. Its subject is the Persian defeat at Salamis eight years before.^[11] Phrynichus’ Phoenician Women was apparently its model.^[12] Its ionic lyrics are commonly thought to be designed to sound exotic. Its trochaic tetrameters (226–229) are a typical specimen of that metre in tragedy.^[13] The ionics of 65–114 are among the orientalizing uses of the metre.^[14]

Actors and chorus. Aristotle credits Aeschylus with the second actor (Poetics 1449a15–16). The third actor was Sophocles’ innovation, but Aeschylus used it in the Oresteia.^[15] The ancient Life (5) gives him the third actor as well.^[16]

The tragic chorus had twelve members at first, and fifteen from the mid-fifth century.^[17] Hammond and Taplin argue for fifteen in the Agamemnon. Their evidence is the three single lines and twelve couplets of 1344–1371.^[18] The Persians and Suppliant Maidens are the only extant fifth-century tragedies that open with the entry of the chorus.^[10]

Prometheus Bound. The ancient sources unanimously attribute the play to Aeschylus, but its authenticity is disputed. Schmid (1929), Griffith (1977) and West (1979, 1990) deny it, chiefly on grounds of style and metre. Herington (1970, 1986), Saïd (1985) and Pattoni (1987) defend it.^[19] Battezzato dates it between the Oresteia and Euripides’ Medea.^[20]

Two metrical facts set the play apart. Dactylo-epitrite never appears in the undisputed plays, but it appears several times in Prometheus.^[21] Choral lyric fills only 13% of its lines, against 35–55% in the other plays.^[22] It is among the plays “now widely believed” to be by someone other than their nominal author.^[23] This site records the dispute and does not decide it.

The play has no recorded date. Most scholars today do not think it is Aeschylus’ work, and a growing number would give it to his son Euphorion. The strongest case for authenticity is Herington’s, who places it at the very end of Aeschylus’ career, in Sicily.^[24]

The end of the Seven. Lines 1005–1078 bring in the ban on burying Polynices. Many scholars think them spurious, among them Winnington-Ingram, Herington and Sommerstein.^[25] Kovacs explains them as an addition made to fit the plot of Sophocles’ Antigone.^[26]

Aeschylus’ text is followed here from the poet’s script to the Alexandrian edition. The late-antique selection, the Byzantine manuscripts and the modern editions are on the Manuscripts tab.

Scripts for performance. Aeschylus wrote for actors and chorus. Late in the century Sophocles and Euripides may have had readers in mind as well, but Aeschylus did not.^[1] The earliest dramatic papyri suggest what such scripts looked like. They gave the words only, without stage directions and probably without the speakers’ names. A change of speaker was marked by a stroke in the margin.^[2] Lyric was written out as if it were prose, without metrical division.^[3]

Revivals and revisions. A law passed shortly after his death allowed anyone who wished to produce his plays to be granted a chorus. Quintilian says that “revised plays” of his won prizes. At some point the Seven was given a new ending, which alludes to the burial edict of Sophocles’ Antigone. This may have happened at a fifth-century revival.^[4]

From 386 one “old” tragedy was revived each year at the City Dionysia. The revivals were mostly of Euripides and sometimes of Sophocles. On the surviving records, a play of Aeschylus was never chosen.^[5]

An official copy. Pseudo-Plutarch credits the orator Lycurgus (c. 390–324 BC) with an official copy of the three tragedians. Actors were not to depart from it.^[6] The measure is dated to 330.^[7]

Whether official copies existed before Lycurgus is disputed. Reynolds and Wilson think it likely that copies were kept at the theatre or the record office. Griffith and Mastronarde doubt it. Kovacs argues that Pseudo-Plutarch’s phrase means a single combined copy, not the first copy ever deposited.^[8]

Galen tells a story about the Athenian copy. A Ptolemy (Ptolemy III, in Kovacs’s account) borrowed it against a deposit of fifteen talents. He sent back a fine copy instead of the original and forfeited the deposit.^[9]

Alexandria. Under Ptolemy Philadelphus, Alexander of Aetolia is said to have edited the tragedies, but not one of his readings survives.^[10] The text of tragedy was probably established by reference to the Athenian copy.^[11] The Alexandrians had two plays called Women of Aetna under Aeschylus’ name, and they judged one of them spurious.^[12]

Colometry. Lyric verse is made of short metrical units called cola. A manuscript sets out each colon on its own line, and this division of the text is called its colometry. The colometry of the lyric texts is usually credited to Aristophanes of Byzantium (c. 257–180 BC).^[13]

Whether his was the only ancient colometry is disputed. A papyrus of Stesichorus (P.Lille 76a + 73) is older than his career, and it is already divided into cola. The medieval manuscripts agree in easy metres such as anapaests and dochmiacs, but they differ widely in harder ones. So Kovacs does not rule out several different ancient colometries.^[14]

The music. After Lycurgus’ decree, copies were stored in Athens. They may have kept the musical notation and reached Alexandria with the books. Whether the Alexandrian editors still had the music is disputed. Fleming and Kopff argue that musical and metrical knowledge was handed down to them. Parker and Prauscello argue that it was lost.^[15] Wilson thinks it “may well be” that knowledge of the music went to Alexandria with the texts.^[16]

The metrician’s view. The text is often corrupt. A strophe and its antistrophe should match in metre, line for line; this match is called responsion. Where corruption breaks it, we may have no means of telling which of the two keeps the true scheme, or whether either does. No strophe in tragedy recurs more than twice, so the signs that mark the end of a period appear only now and then.^[17]

There are a handful of seeming exceptions to Porson’s Law in tragedy, and most of them are suspect on textual grounds.^[18] The language can show how a passage was delivered. At Agamemnon 1136–1145 Cassandra speaks trimeters among the dochmiacs, and these trimeters keep the Attic vocalisation. This may be a sign that for a moment she falls into speech.^[19]

The Transmission tab ends with the Alexandrian edition. Here the seven plays are followed through late antiquity and Byzantium to the printed editions. The general history is on How the texts reached us.

A narrowed corpus. Until about AD 250, writers quote from many of the roughly three hundred plays of the three tragedians. After that, they quote only seven plays each of Aeschylus and Sophocles and ten of Euripides.^[1]

How this came about is disputed. Wilamowitz supposed that an influential schoolmaster fixed a syllabus in the second or third century. Reynolds and Wilson find the theory too neat. No such schoolmaster can be named, though Eugenius, who wrote on the colometry of fifteen plays in the fifth century, is one possible candidate. Texts outside the syllabus were still being read in late antiquity. So they would abandon the idea that one person’s conscious choice decided what survived.^[2]

Kovacs leaves the question open. The selection may have been deliberate, perhaps for teaching. Or it may have been the chance result of popularity and of the commentaries available. A book could once be a roll, a long strip wound on itself, or a codex, a book of bound leaves like ours. The change from roll to codex may have helped, since one codex could hold seven or ten plays.^[3] The move to the codex was “the first major bottle-neck” through which classical literature passed.^[4]

Byzantium. Learning revived in the ninth century, and the plays were then recopied in the new minuscule script. Nearly all Greek texts depend on such ninth-century transliterations.^[5] The ancient commentaries were cut down to fit the margins. Notes written in the margins of a manuscript are called scholia. The scholia vetera (“old scholia”) are remnants of ancient scholarship; the scholia recentiora (“more recent scholia”) were written for medieval readers.^[3] Copious marginal scholia are not common before the ninth century.^[6]

In the ninth century Photius had read the Prometheus as a school text.^[7] Only one medieval manuscript contains all seven plays. It is Laur. 32.9 in Florence, written in the tenth century, and it is fundamental for Sophocles too.^[8]

The triad. By the late thirteenth century, and perhaps earlier, schools read three plays of each tragedian. These three are called the “triad”, and most later manuscripts contain only them.^[9] For Aeschylus the three were the Persians, Seven against Thebes and Prometheus Bound.^[10] So there are many more copies of these three than of the other four. The other four survived just long enough to be taken to Italy by Renaissance collectors.^[11]

Triclinius and metre. Demetrius Triclinius was a schoolmaster at Thessalonica c. 1305–1320. He worked on the poets of the school curriculum, which included the tragic triads. He was the first Byzantine to grasp classical metre and to use it. He learnt it from Hephaestion’s handbook. He knew that the lyrics of tragedy and comedy should respond exactly, and he emended the text to restore responsion. Some of his emendations have had lasting success; many were only stopgap words. He also wrote separate metrical commentaries on many plays.^[12]

Printed editions. The first printed editions of tragedy appeared at the end of the fifteenth century.^[13] The early editions of Aeschylus left out lines 323–1050 of the Agamemnon. The first to print the whole play was the edition of Pier Vettori and Henri Estienne.^[14] The sixteenth-century editions are described on Printing.

The modern texts are D. Page’s Oxford text (1972) and M. L. West’s Teubner (1990). West’s Studies in Aeschylus (1990) discusses the text, and S. Radt’s Tragicorum Graecorum Fragmenta vol. 3 (1985) collects the fragments.^[15] West’s edition has a metrical appendix.^[16] Dawe (1964) studies the manuscript tradition.^[17] A. M. Dale analyses the lyrics in Metrical Analyses of Tragic Choruses (1971–1983) and Lyric Metres of Greek Drama (2nd ed., 1968).^[18]

This site quotes Smyth’s Loeb text (1922–1926), because the Perseus Digital Library makes it available under an open licence. Where West prints a different reading, the Metres tabs say so.

Drama used three modes of delivery: speech without music, recitative (παρακαταλογή, parakatalogē), and song. A stichic metre repeats the same line again and again. The recited parts are always in stichic metres, or in long runs of one uniform rhythm.^[1]

Landels divides the delivery of tragedy in the same way. He argues that its marching anapaests must have been chanted, since twelve or fifteen men speaking together would have been drowned by the pipe.^[2] The spoken trimeter and the recited metres are described on this page; song is on the next tab.

The spoken trimeter

Dialogue is in the iambic trimeter, × — ⏑ — × — ⏑ — × — ⏑ —. A caesura is a regular word-end inside the line. The trimeter usually has its caesura after the first position of the third foot; this is the penthemimeral caesura. Failing that, the caesura comes after the first position of the fourth foot; this is the hephthemimeral caesura.^[3] Here is the opening of the Agamemnon:

All four lines have the penthemimeral caesura and no resolution. In line 3 the first syllable of Ἀτρειδῶν is short before τρ. This is the usual Attic treatment of a stop followed by a liquid.^[4]

Resolution is the replacement of a long position by two short syllables. Aeschylus uses it sparingly. In early iambus and in Aeschylus and Sophocles there is on average no more than one resolved line in every ten.^[5]

An anceps is a position that may be filled by a long or a short syllable. Porson’s Law holds in Aeschylus’ trimeters. Under this law a long anceps in the third metron is not followed by word-end, unless one of the two syllables is a monosyllable.^[6]

The law holds for the trimeter and the tetrameter of tragedy alike, with a single exception: a Persian name at Persians 321. Comedy breaks it about once in five lines. Allen explains the law by stress. A heavy syllable at the end of a word would carry stress, and here that stress would fall in the weak half of the foot. The monosyllables that are allowed are proclitics and postpositives, which lean on their neighbours.^[7]

Aeschylus’ trimeters show some rarer features too. Sometimes the caesura falls at the end of the third foot, nearly always with elision (Ag. 20). At Sept. 488 the first metron has the form of a choriamb, to fit the name Ἱππομέδοντος (Hippomedontos), whose first four syllables, Ἱπ-πο-μέ-δον-, make — ⏑ ⏑ —; this licence occurs only three or four times in tragedy. At Eum. 238 f. a prepositive is left at line-end. At Cho. 181 a postpositive stands immediately after the caesura.^[8]

The trochaic tetrameter

The trochaic tetrameter catalectic runs — ⏑ — × — ⏑ — × | — ⏑ — × — ⏑ — ‖. In tragedy it belongs chiefly to Aeschylus and to Euripides’ late plays.^[9] Persians 226–229 is a typical example. In tragedy Porson’s Law holds for the end of the line, and Havet’s Law (also discovered by Porson) for the beginning. Persians 165 is one of only two tragic lines without the normal caesura.^[10] The chorus greets the queen in this metre ten lines earlier.

Marching anapaests

An anapaestic metron runs ⏔ — ⏔ —. In marching anapaests the metra run on in stretches of varying length. Each stretch is a period, and it closes with catalexis: its last metron is cut short. Editors print the metra in pairs, as dimeters. Yet the whole period is metrically continuous until the catalexis; this continuity is called synapheia.

Word-end normally follows every metron except the one before the catalexis. Where it does not, it comes after the first short of the next biceps.^[1] Here is Ag. 60–67:^[11]

In lines 62 and 63 both the first and the second princeps are resolved (— ⏖ — ⏖). In 64 the caesura comes one short later than usual. At the end of 66 the period ends, although the sense runs on.^[12]

Line 66 is the catalectic dimeter that closes the period. This line is called the paroemiac. In these schemes ⏖ is a resolved princeps, and ⏑ ⏑ is an uncontracted biceps. A metron with resolution normally has the form — ⏖ — —.^[13] West prints Δαναοῖσιν in 66 where Smyth has Δαναοῖσι; the scansion is the same.

The lyric cola of Aeschylus are set out below, with analyses of two passages. How much of each play is sung, and by whom, is on the Chorus and song tab.

Iambics. Lyric iambics play an especially large part in Aeschylus. They are made of syncopated and unsyncopated metra, mostly grouped in dimeters and trimeters. Ag. 184–191 is an example. Aeschylus prefers to fill every anceps but the last with a short syllable. Where the strophe has a long anceps, the antistrophe usually has one too. The opening of the same song shows the lekythion, the syncopated iambic colon that runs through it.

Resolution is rare in his lyric iambics. The exceptions are some unsyncopated dimeters and trimeters (Eum. 385 ~ 393, Sept. 850). In these, resolution goes much further than he permits in his spoken trimeters.

In the last verse but one of a strophe he likes to change the rhythm. He does so with a dochmiac or dactylic colon, or with a choriamb or two. The final colon is often an aristophanean, — ⏑ ⏑ — ⏑ — —, or sometimes a pherecratean. Sophocles does not close his strophes in these ways.^[1]

Dochmiacs. The basic rhythm of the dochmiac is ⏑ — — ⏑ —. Any of its three longs may be resolved. Either of its shorts may be replaced by a long, and this is called “drag”. The dochmiac is the characteristically tragic metre. Every surviving tragedy has some, and they always carry urgency or strong feeling.^[2]

The dochmiac is the quintessential metre of dramatic lyric. Its rapid, stuttering movement has been heard as a rhythm made for scenes of high excitement.^[3] Aeschylus seems to have invented the dochmiac, or at least to have been the first to use it in the theatre. It is brief in the Persians (472) and frequent from the Seven (467) on.^[4] Here is an example, Sept. 79–82:

The first metron of 80 has two resolutions, — ⏖ ⏖ ⏑ —. The forms ⏑ ⏖ — ⏑ — and — ⏖ — ⏑ — are among the commonest. Metron-end coincides with word-end throughout.^[5] In 79 -δον is long because it is closed by the λ of λιπών. In 80 the first syllable of πρόδρομος is short before δρ.^[6]

These lines come from the entrance-song of the Seven (78–107). Its dochmiacs, its many resolutions and its lack, at first, of any strophic structure all convey the women’s panic as the Argive army approaches. Eteocles’ spoken reply, urging calm, stands in sharp contrast. In the Agamemnon the roles are reversed. Cassandra sings her prophecies in a mostly dochmiac rhythm (1072–1177). The elders first answer in spoken iambic trimeters, then are drawn into dochmiacs of their own. At the end Cassandra drops song and speaks in trimeters.^[7]

Ionics are commonest in Aeschylus. The ionic metron is ⏑ ⏑ — —. Straight ionics are more frequent than anaclastic ones. Quite often a short metron, ⏑ ⏑ — (io∧), stands in place of ⏑ ⏑ — —. But io∧ cannot respond with io.

Many have heard an exotic, Asiatic colouring in the ionic songs of the Persians.^[8] The passage (65–114) is one of the orientalizing uses of the metre.^[9] Here is Pers. 68–72, set out in West’s three cola (Smyth divides the same words into five lines, 68–72):

The ᾱ is long in σχεδίᾳ, ἀμείψας and Ἕλλας.^[10]

Other rhythms. The only wholly cretic strophe in tragedy is Supp. 418–422 ~ 423–427.^[11] Pure aeolic strophes are uncommon in Aeschylus.^[12]

In the Furies’ Binding Song, Eum. 372–375 run in a “tripping” paeonic rhythm: each line is ⏑ ⏑ ⏑ — ⏑ ⏑ ⏑ —, three shorts and a long, twice over. The chorus sing of leaping high and bringing the foot down. The three quick shorts and the heavy long seem to act out the leap and the stamp.^[13] In the Libation Bearers the chorus describe beating themselves in lament (423–428). Their resolved iambics, with a marked increase in short syllables, match the blows.^[7] A fragment of the lost Bassarids seems to preserve part of a song by Orpheus, in bacchiacs, ⏑ — — (fr. 23 Radt).^[14]

In Aeschylus’ dactylic strophes the dactyls mostly run from four to seven metra. Unlike Homer, he stresses contractions by placing word-end after them. Ag. 104–121 ~ 122–139 is an example.^[15] In this song the spondees change the rhythm so much that dactyl and spondee rarely fail to respond. The refrain αἴλινον αἴλινον εἰπέ (ailinon ailinon eipe), “cry sorrow, sorrow”, ends in a double spondee. At 109 an iambic metron is welded to the dactyls as a “base”: ὅπως Ἀχαι|ῶν δίθρονον κράτος (hopōs Achai|ōn dithronon kratos).^[16]

Long dactylic periods like these, above all in the parodos of the Agamemnon, seem to call up the solemnity of prophecy and of heroic deeds. They show the influence of the epic hexameter and of the dactylic cola of Stesichorus. In the Frogs, “Euripides” mocks Aeschylus’ lyrics as overwhelmingly dactylic.^[17]

Dactylo-epitrite is absent from the genuine plays.^[18] In lyric anapaests, a pair of tribrach words, each proparoxytone, can take the place of an anapaestic metron. This happens at Pers. 985 ἔλιπες ἔλιπες (elipes elipes), “you left, you left”, ~ 1000 ἔταφον ἔταφον (etaphon etaphon): ⏑ ⏑ ⏑ ⏑ ⏑ ⏑ in place of ⏔ — ⏔ —.^[19]