The Works of Seneca the Younger and their Dates

© 2014 Koninklijke Brill NV ISBN 978 90 04 15461 2

Brill’s Companionto Seneca

Philosopher and Dramatist

Edited by

Gregor DamschenAndreas Heil

With the assistance of

Mario Waida

LEIDEN • BOSTON2014


CONTENTS

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

PART ONE

LIFE AND LEGACY

Imago suae vitae: Seneca’s Life and Career . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Thomas Habinek

The Works of Seneca the Younger and Their Dates . . . . . . . . . . . . . . . . . . . . . 33C.W. Marshall

Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45Rolando Ferri

Seneca and Senecae: Images of Seneca from Antiquity to Present

Seneca the Philosopher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53Matthias Laarmann

Seneca the Dramatist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73Werner Schubert

PART TWO

PHILOSOPHY

Context: Seneca’s Philosophical Predecessors and Contemporaries . . . . 97John Sellars

Works

De providentia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115R. Scott Smith

De constantia sapientis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121R. Scott Smith


vi contents

De ira . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127Maria Monteleone

Consolatio ad Marciam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135Jochen Sauer

De vita beata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141Fritz-Heiner Mutschler

De otio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147R. Scott Smith

De tranquillitate animi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153Fritz-Heiner Mutschler

De brevitate vitae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161R. Scott Smith

Consolatio ad Polybium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167Jochen Sauer

Consolatio ad Helviam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171Jochen Sauer

De clementia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175Ermanno Malaspina

Naturales quaestiones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181Gareth D. Williams

Epistulae morales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191Aldo Setaioli

De beneficiis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201Mario Lentano

Lost and Fragmentary Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207Anna Maria Ferrero

Epistulae Senecae ad Paulum et Pauli ad Senecam . . . . . . . . . . . . . . . . . . . . . . . 213Alfons Fürst

Topics

Ontology and Epistemology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217Mireille Armisen-Marchetti


contents vii

Ethics I: Philosophy as Therapy, Self-Transformation, and“Lebensform” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239Aldo Setaioli

Ethics II: Action and Emotion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257Margaret R. Graver

Ethics III: Free Will and Autonomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277Aldo Setaioli

Ethics IV: Wisdom and Virtue. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301Jula Wildberger

Ethics V: Death and Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323Catharine Edwards

Physics I: Body and Soul . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343R. Scott Smith

Physics II: Cosmology and Natural Philosophy . . . . . . . . . . . . . . . . . . . . . . . . . 363Bardo Maria Gauly

Physics III: Theology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379Aldo Setaioli

PART THREE

TRAGEDY

Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405Wolf-Lüder Liebermann

Works

Hercules furens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425Margarethe Billerbeck

Troas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435Wilfried Stroh

Phoenissae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449Marica Frank

Medea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459Wolf-Lüder Liebermann


viii contents

Phaedra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475Roland Mayer

Oedipus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483Karlheinz Töchterle

Agamemnon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493Christoph Kugelmeier

Thyestes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501Chiara Torre

Dubious Works

Hercules Oetaeus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515C.A.J. Littlewood

Octavia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521Rolando Ferri

Topics

Space and Time in Senecan Drama . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531Ernst A. Schmidt

Vision, Sound, and Silence in the “Drama of the Word” . . . . . . . . . . . . . . . . . 547Andreas Heil

The Chorus: Seneca as Lyric Poet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561Giancarlo Mazzoli

The Rhetoric of Rationality and Irrationality . . . . . . . . . . . . . . . . . . . . . . . . . . . 575Gottfried Mader

Characters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593G.W.M. Harrison

Themes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615G.W.M. Harrison

Greek and Roman Elements in Senecan Tragedy . . . . . . . . . . . . . . . . . . . . . . . 639Sander M. Goldberg

Philosophical Tragedy? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653François-Régis Chaumartin


contents ix

PART FOUR

APOCOLOCYNTOSIS

Apocolocyntosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673Renata Roncali

PART FIVE

OTHER WORKS

Epigrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689Joachim Dingel

De vita patris . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695Michael Winterbottom

PART SIX

SYNTHESIS

Seneca’s Language and Style . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699Michael von Albrecht

Systematic Connections between Seneca’s Philosophical Works andTragedies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745Susanna E. Fischer

List of Journal Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771

Andrea Balbo and Ermanno MalaspinaEditions of Seneca’s Works (Since Haase’s Opera Omnia) . . . . . . . . . . . . . . . 861Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865General Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873


THE WORKS OF SENECA THE YOUNGER AND THEIR DATES

C.W. Marshall

The works of L. Annaeus Seneca cannot be dated with any great preci-sion. This is frustrating, since the interpretation and understanding of hisimmense and wide-ranging output would benefit from a precise chrono-logical sequence. The works themselves resist any such systematization,however: Seneca makes very few references to his personal circumstances,which is appropriate considering his philosophical emphasis on the innerlife, and this reticence has led one scholar to ask ironically, “Est-il possible de‘dater’ un traité de Sénèque?”1 Nevertheless, some headway is possible, andGiancotti (1957) on the Dialogues, Abel (1967) and Griffin (1976) on all theprose works, and Fitch (1981) and Nisbet (1995: 293–311) on the tragedies havemade significant advances in understanding the dates of Seneca’s literarywritings. This chapter seeks to integrate the conclusions of these studies.When something can be said in relation to a landmark event in Seneca’slife, it is often limited to terminus ante or post quem: his exile to Corsicafollowing the accession of Claudius in ad41 (Dio 60.8.5); his recall to serveas personal tutor to Nero in ad49 (Tac. ann. 12.8.2); his rise to prominenceon Nero’s accession in ad54 (ann. 13.2.1); his diminished influence followingthe murder of Agrippina in ad59 (ann. 14.14.2); and his withdrawal from allinfluence with Nero in ad62 (ann. 14.52.1). Within the spans bounded bythese points, a generally coherent picture emerges.

There are of course many methodological issues associated with assigningdates (both relative and absolute) to literary works. Internal references,stylistic features, external testimonia, and other factors may be employedto argue for a date, and different types of argument will carry differentweights with different readers. Crucially, circularity must be avoided, andinterpretations of a work cannot presume a date for which evidence doesnot exist. It may be possible to perceive a development in thought from onework to the next, but that in itself cannot be used as an argument for therelative dates of the works in question. There is also a danger with this sort ofanalysis in assuming a tendency toward the limits: a given work that shows

1 Grimal 1949a, and see also Griffin 1976: 5 n. 2.


34 c.w. marshall

indications of being written before ad54, for example, does not need to havebeen written close to ad54; the limits identify boundaries, but in most casesdo not establish more or less likely dates within the possible range. Indeed,the opposite is true: as one approaches the limit, there is a greater need forindependent, unrelated points of reference. While any two arbitrary factstouching on an author’s life may be close in time to one another (e.g., whena given work was written and an event recorded by Tacitus), it is improbablethat such clusters will occur repeatedly, given how few data points survive.Agnosticism often remains the most prudent course. These issues are ofcourse further confused if works are re-worked or re-edited following theirinitial circulation.2 My hope here is not to overstate the case, but withineach section to describe works in what may reasonably be thought to bechronological order, given the appropriate cautions offered below.

Dialogues

Ten treatises in twelve books, as found in the eleventh-century Ambrosianmanuscript, are collectively known as Dialogues (Dialogi) and are num-bered 1–12. The earliest of these, Ad Marciam de Consolatione (= dial.6), probably dates to ad39 or 40, although a later date into the 40s is pos-sible.3 Seneca writes with authority to console Marcia, daughter of thehistorian A. Cremutius Cordus, on the death of her son Metilius threeyears earlier, and Seneca may have written works before this that areno longer extant. Reference to the republication of Cordus’s works (cons.Marc. 1.3), which occurred under Gaius (Suet. Cal. 16.1), establishes a ter-minus post quem.4 Further, praise of Tiberius (as is found in cons. Marc.3.2, 15.3) “would not have been prudent before 39” (Griffin 1976: 397, cit-ing Dio 59.16.4 and Suet. Cal. 30.2). For the upper limit of the range, Abelargued that it must be before Seneca’s exile, based on in qua istud urbe,di boni, loquimur? (cons. Marc. 16.2), which suggests both speaker andaddressee are in Rome.5 This is not convincing: loquimur could equally bean epistolary conceit, whereby the letter creates the air of intimate com-munication, regardless of where the sender is; indeed, this effect would

2 E.g., Schmidt 1961.3 For previous discussions, see Giancotti 1957: 45–73, Abel 1967: 159f., Griffin 1976: 397,

and references there.4 Bellemore’s argument (1992) for an earlier, Tiberian date for the work requires rejecting

Suetonius’s evidence.5 Abel 1967: 159 f., and references there.


the works of seneca the younger and their dates 35

be enhanced if Seneca were in exile. If this view is accepted, then Griffin’scalculation of Marcia’s age excludes only a date after ad49 (1976: 397). Whilea Gaian date is perhaps reinforced by the absence of any mention of Gaius inthe text, a date in the 40s remains possible, depending on the interpretationof cons. Marc.16.2.

Two other consolationes, Ad Polybium de Consolatione and Ad Hel-viamMatrem de Consolatione (= dial. 11 and 12), certainly belong to theperiod of Seneca’s exile.6 The first, to Polybius, a freedman and secretaryto Claudius a studiis (Suet. Claud. 28), offers consolation on the death ofhis brother. It was evidently written before Claudius’s conquest of Britainin ad43, which remains an anticipated event in cons. Pol. 13.2 Britanniamaperiat (“may [Claudius] open up Britain,” though the verb could conceivablybe used for a short while thereafter). The letter is an unsuccessful effort towin Seneca’s recall. The date of his letter to his mother, offering her comforton his own exile, cannot be circumscribed so precisely: Abel (1967: 163) tookthe ten months mentioned in cons. Helv. 16.1 literally; Griffin cautions againstthis, emphasizing instead the length of his absence, stressed in cons. Helv. 1.2and 2.5. Neither of these need be determinative, however, and it is safer toaccept a larger range, between ad41 and 49.

The three books of the treatise De Ira (= dial. 3–5) likely also belong toSeneca’s exile.7 They are addressed to Seneca’s older brother L. AnnaeusNovatus, whose name was changed to L. Junius Gallio Annaeus (PIR2 I 757),by ad52 at the latest, perhaps due to a testamentary adoption.8 Referencesto Gaius make it clear that he is dead (de ira 1.20.8, 2.33.3–4, 3.18.3–4), butthe force of modo (“recently”: 3.18.3) to describe an action of Gaius cannot bepressed unduly, as Seneca himself noted in epist. 49.4. It is not necessary tofollow Coccia and Abel in asserting that the work must be confined to theperiod following Gaius’s death and before Seneca’s exile; any time between41 and, at the latest, early ad52 is possible.

Griffin argues forcefully for a date of ad55 for De Brevitate Vitae (= dial.10), although she insists on nothing more specific than a date between ad

6 For previous discussions, see Giancotti 1957: 74–92, Abel 1967: 163f., Griffin 1976: 397f.,and references there.

7 For previous discussions, see Giancotti 1957: 93–150, Abel 1967: 159, Griffin 1976: 398, andreferences there.

8 Griffin 1976: 48, n. 2. The so-called “Gallio inscription” (see Plassart 1967, Oliver 1971,and Hemer 1980), found at Delphi, provides the lynchpin for dating the New Testament (cf.Acts 18.12–17), and was written in the first half of ad52, and it shows that he was at that timeproconsul of Achaia and was using the name Gallio. Since the proconsulship would havebegun in ad51, it is likely that the adoption occurred before this.


36 c.w. marshall

mid-48 and mid-55.9 This was the period during which the work’s addressee,Pompeius Paulinus, Seneca’s father-in-law (see ann. 15.60.4), was praefectusannonae. Brev. 18 f. urge Paulinus to retire from his administrative dutiesmanaging Rome’s grain supply to more important things (maiora: brev. 18.2).In this period at least some of Seneca’s tragedies are likely to have beencomposed (see below), a fact that bears on the interpretation of the lifeSeneca recommends.

De Vita Beata (= dial. 7) is dedicated to Seneca’s brother, to whom De Irahad also been dedicated, after he had been adopted by Gallio and had takenhis name.10 The Gallio inscription (see n. 8) attests his name in ad52, so theadoption must have taken place before this. Vit. beat. was written after theadoption, but conceivably still written before ad52. While Gallio did not dieuntil ad66, the subject matter, in which Seneca justifies his great wealth andprosperity, strongly suggests a Neronian date before ad62. It does not follow,however, that the work was composed as a direct response to the attacks inad58 by P. Suillius Rufus (ann. 13.42.1–43.5), as Griffin argues (1976: 19 f. and306–309).

Two, perhaps three works were dedicated to Annaeus Serenus, who diedas a prefect of the nightwatch (Plin. nat. 22.96) sometime before ad64, whenit is probable Seneca wrote epist. 63.14 (and perhaps before the beginningof ad62, when Tigellinus became praefectus vigilum).11 The first of these, DeConstantia Sapientis (= dial. 2), seems to have been written after the deathof Valerius Asiaticus in ad47 (ann. 11.3.2), given the reference to him in const.18.2. De Tranquillitate Animi (= dial. 9), in which Serenus is a Stoic (tranq.1.10), is later than const., in which he is still an Epicurean (15.4; the sequence ofSerenus’s philosophical development is secured by const. 3.2, and see Griffin1976: 316). The fragmentary De Otio (= dial. 8) is also addressed to a Stoic,and there is reason to believe this too is Serenus, although this is conjectural.If so, it also postdates const., but the relative position between it and tranq.cannot be determined.12 Many have believed De Otio to be the last of thethree, but Seneca’s “shifting positions […] may well be more experimental

9 For previous discussions, see Giancotti 1957: 363–445, Abel 1967: 162 f., Griffin 1962 and1976: 398 and 401–407, and references there.

10 For previous discussions, see Giancotti 1957: 310–362, Abel 1967: 160–162, Griffin 1976:399, and references there.

11 Griffin 1976: 447 f. and Giancotti 1957: 153–157. See also Giancotti 1957: 151–177, Abel 1967:159 and 162, Griffin 1976: 316 f. and 399, and references there for the relationship between thethree works. For const., see also Giancotti 1957: 178–192; for tranq., see also Giancotti 1957:193–224; for de otio, see also Giancotti 1957: 225–243 and Williams 2003: 12–16.

12 Griffin 1976: 316 f.



than a ‘sincere’ reflection of his own beliefs on either or both occasions”(Williams 2003: 16). All three works were then composed between ad47 and(probably) ad62, with const. being composed first.

There is no solid basis for establishing the date of De Providentia (= dial.1), a work dedicated to Lucilius, who is also the dedicatee of works composedin the 60s, except to say that it was written after the death of Gaius (see prov.4.4).13 Abel placed it securely in ad64 because the work presents Luciliusas an avowed Stoic, but the literary nature of this (or any) philosophicalwork makes such an inference untenable. Tying the work to the period of theLetters (Epistulae) however, is not unreasonable, even if it cannot be proved.

Tragedy, Verse, and Satire

Of the more than seventy epigrams that have been ascribed to Seneca, nonecan be dated securely. The Anthologia Latina ascribes only three to himexplicitly (Anth. Lat. 232, 236, 237), the last two of which concern exile onCorsica. This, I suggest, undermines the attribution rather than reinforcesit: Seneca was Corsica’s most famous refugee, and the subject is a likelyone for someone appropriating Seneca’s voice. The reference to Seneca’spoetry as a model for the versiculi of Pliny (epist. 5.3.2–5) is not a statementabout the genre Seneca employed, as other names in the list demonstrate.14

Nevertheless, if any epigrams are authentic, some may date to the period ofhis exile.

It is to this period that one may also place the first group of Seneca’stragedies, Agamemnon, Phaedra, and Oedipus. Eight authentic tragediesexist,15 and the echoes of Hercules Furens in Apocolocyntosis (see n. 20), ifvalid, set a terminus ante quem for that play of ad54. The earliest certainreference to Seneca’s tragedies is Quintilian, inst. 8.3.31, in which Quintilianrecalls an exchange between Pomponius Secundus and Seneca about atragedy in his youth (iuvenis admodum) that must date soon after the return

13 For previous discussions, see Giancotti 1957: 244–309, Abel 1967: 158, Griffin 1976: 400 f.,and references there.

14 Note also that Pliny is not certain that Seneca’s poetry was recited, at least by the author:recito tamen, quod illi an fecerint nescio (epist. 5.3.7, “But I recite [my verses], though I do notknow whether they [my predecessors] did”).

15 Despite attempts of its revival (Kohn 2003), the idea that the plays are not by L. AnnaeusSeneca has not won general approval. Quintilian, inst. 9.2.8, quotes Medea 453 and attributesit to Seneca, but that single reference is sufficient for the attribution. Two other plays in themanuscripts are spurious and post-Neronian. They are discussed at the end of this section.


38 c.w. marshall

of Pomponius to Rome at the end of ad51: this suggests a public presentationof a play in some form in the early 50s, but says nothing about this being thetime of composition.16 Stylistic features discussed by Fitch (1981), particularlythe increased incidence of sense-pause mid-line in the iambic verses (whichhe demonstrates is also a useful diagnostic for Sophocles and Shakespeare,but not for Euripides) and the increased incidence of shortened final -o,identify three clear chronological groupings for the plays. Fitch reckons thepercentage of sense-pauses that occur within the line as measured againstthe total number of sense-pauses within an iambic section: Agamemnon(32.4 %), Phaedra (34.4 %), and Oedipus (36.8 %) form a coherent group, butcontextual variation should create a margin of error that does not place theworks within this group in a certain order. The play with the next smallestratio, Medea (47.2 %), clearly belongs to a different cluster. The plays in thisgroup predate Hercules Furens (which belongs to the middle group) andtherefore all were also composed before ad54.17 The nature of the evidencedoes not allow any conclusion more precise than this: the three plays werecomposed (in whatever order) in a cluster, but no certain sequence can bedetermined. A Claudian date seems reasonably certain for these first threeplays, but they could conceivably date to the exile, or before (into the reignsof Gaius or even Tiberius), or after.18

This metrical approach is to be preferred to those who seek covert anti-Neronian messages in the tragedies, placing some or all in the 60s (e.g., Bishop1985). Töchterle (1994: 44–48 and infra, p. 483) believes verbal parallels,particularly with nat., point to a composition date for Oedipus betweenad62 and 65. The implications are significant: in addition to vitiating Fitch’sconclusions, the date would point to a particular political purpose forthe tragedies, whereby, in this case, Nero, Claudius, and Agrippina are tobe mapped onto Oedipus, Laius, and Jocasta. Accounts of Jocasta’s death(hunc pete | uterum capacem: Oed. 1038f., hunc petite uentrem: Pho. 447)and Agrippina’s death as presented in post-Neronian sources (Oct. 368–372;uentrem feri: Tac. ann. 14.8.4) would therefore allude to Agrippina’s death inad59. It is easier, if less scandalous, to believe that Tacitus and the author ofOct. “saw in Seneca’s presentation of Jocasta [in Oed.] a suitable model fortheir account of Nero’s mother, one which carried implications of incest and

16 See Tarrant 1985: 12 and Fitch 1987a: 50 f., and references there.17 This date suggests that the apparent allusions to Ag. 330–341 in Einsiedeln Eclogue 1.22–33

are real and not incidental; see Tarrant 1985: 11.18 Coffey 1957: 150 adopts this agnostic position, followed by Tarrant 1976: 6 f.



moral perversion at the heart of the late Julio-Claudian court” (Boyle 1997:102). This process may have begun with Seneca himself, when writing Phoen.(see infra, p. 40).

The second group of tragedies, Troades, Medea, and Hercules Furens,possibly centers around ad54, but again it need not, and all three couldwell predate this. As with the first group of plays, incidences of sense-pausewithin a line strongly suggests they were composed within a short time ofone another: Medea (47.2 %), Troades (47.6 %), and Hercules Furens (49.0 %).As with the first group, a reasonable margin of error does not allow for theorder of works within this cluster to be established. Nisbet argues that Med.,which likely alludes to Claudius’s invasion of Britain in ad43, is unlikely tohave been composed after Claudius’s death, and suggests that ad51 to 52,when Gallio was proconsul in Corinth, “would be quite a good moment forSeneca’s Corinthian play” (1995: 295). This is appealing, but hardly certain.Verbal echoes have been detected between Herc. f. and apocol., and thegratuitous nature of the tragic Hercules in apocol. argues that the tragedyis the earlier work.19 Fitch believes “the tragedy was fresh in his mind andhad either been written, or at any rate presented in recitatio, within a year ortwo of 54” (1987: 53), and this certainly could be the case. That would alsomean that De Ira had probably been composed by the time Seneca beganworking on these plays, and this has some bearing on the interpretationof these works. Strictly speaking, however, though the relative position ofHerc. f. and apocol. seems probable, there is nothing to require compositionnear this date, and it is conceivable that all the plays in these first two groupscome from a period much earlier in Seneca’s career. The Apocolocyntosis, aprosimetric Menippean satire on the death of Claudius, almost certainly datesto November or December of ad54: Eden accepts Furneaux’s hypothesis thatit was produced for the Saturnalia, which began on December 17, at whichNero was rex (ann. 13.15.2).20 This would also establish some time between themore traditional encomium for Claudius’s funeral that Seneca wrote (ann.13.3.1), to which it would have been compared in any case at the time. At leastone play in this middle cluster (Herc. f.) predates apocol., and possibly allthree do.

A number of factors suggest instead that Seneca returned to the tragicform late in his career, and that the last two plays (the third group) are to be

19 See Mesk 1912, Weinreich 1923, and Fitch 1987a: 51–53.20 Eden 1984: 4 f., esp. 5 n. 11. See also Griffin 1976: 129 n. 3. Tacitus surprisingly accepts that

the position of rex fell to Nero by lot.


40 c.w. marshall

dated to the 60s, even though, strictly speaking, only their position relativeto the second group is established. Thyestes exhibits another jump in itsuse of mid-line sense-pause (54.5 %) and a significant increase in the use ofa shortened final -o (Fitch 1981: 303–305). Tacitus, ann. 14.52.3, describes how,before ad62, Seneca’s poetic output (carmina) had increased once Nero hadtaken a liking to it. While carmina may refer to epigrams and other poeticforms, it could equally refer to a return to tragedy (see Tarrant 1985: 12 f.), andTarrant uses this comment to give an approximate range for the play of ad60–62. Nisbet denies this—it is “a false clue” (1995: 296)—but nevertheless sees inThyestes a series of historical allusions that put its composition at ad62 (1995:300–309). Given that Seneca’s retirement may have been a gradual process(cf. ann. 14.53.1–57.2), distinguishing between these positions is difficult, andany date close to ad62 remains possible.

Perhaps the most distinctive result of Fitch’s metrical analyses is theconfluence of two measures in determining a late date for Phoenissae.The play has the highest incidence of sense-pause within a line (57.2%),and, by some margin, the highest incidence of shortened final -o. Both ofthese strongly indicate Phoenissae was the last tragedy Seneca composed, aconclusion corroborated by its apparent incompleteness. If the argument forascribing Thyestes to ca. ad62 is accepted, then Phoenissae would date to thefinal years of Seneca’s life, as Seneca chose a mythical subject that attractedboth Euripides and Sophocles in the final years of their lives (Nisbet 1995:309). This has bearing on the date of Hercules Oetaeus, a play includedamong the Senecan tragedies alongside the certainly spurious Octavia. Theplay has been defended as authentic by Rozelaar (1985) and Nisbet (1995:209–212), who treat it as a late play, composed “shortly before Seneca’s deathin 65; that would explain the anomalies, the verbosity, the other signs ofhaste” (Nisbet 1995: 210). Several indications tell against this. The incidence ofshortened final -o is very small, as Nisbet admits (1995: 310), and the verbosenature of Oetaeus, and its sheer length, suggest a fundamentally differentapproach to playwriting than that suggested by Phoenissae: both are unlikelyto have developed from the same author composing at the same time in hislife. Hercules Oetaeus is therefore not by Seneca,21 and may date as late as theearly second century, as suggested by Zwierlein (1986b: 313–343).

Three dates have been argued for the inauthentic Octavia: severalscholars have argued for a date in ad68 and the reign of Galba;22 Ferri cites

21 See Leo 1878: vol. 1, 48–74, Friedrich 1954, and Axelson 1967.22 Kragelund 1982: 38–52 and 1988, Barnes 1982, and Wiseman 2001: 10 and 14.



parallels between the play and certain poems of Statius, which, if thedirectionality of the reference is correct, would argue for a date in the 90s(2003: 5–30). The early years of Vespasian’s reign (ad69 to the mid-70s) is alsopossible (Junge 1999: 199 f., Smith 2003: 426–430, and Boyle 2008: xiii–xvi).23

Longer ProseWorks

There are four longer prose works that survive, all of which are Neronianand date to the last decade of Seneca’s life. The first, De Clementia, isdated to Nero’s nineteenth year (clem. 1.9.1–2), i.e., sometime betweenDecember 15, ad55 and December 14, ad56, and is dedicated to the newprinceps.24 Originally in three books, only the first and part of the secondsurvive. Seneca had composed Nero’s speech at Claudius’s funeral (ann.13.3.1–2), as well as other speeches critical to securing his authority (Dio61.3.1), and early in ad55 he wrote several speeches for Nero that hadthe princeps clementiam suam obstrigens (ann. 13.11.2, “pledging himself tocompassion”)—the less respectful apocol. having been composed at exactlythis time. It was early in ad55, of course, that Britannicus was murdered(ann. 13.15.1–17.3), which makes the historical situation of clem. roughly oneyear later all the more interesting for an understanding of Seneca’s purpose.25

Clem. is an overtly political work, and probably represents a consolidation ofthe ideas adumbrated in these speeches, a coherent policy statement for thenew regime.

The seven books of DeBeneficiis are dedicated to Aebutius Liberalis, whois the subject of epist. 91 (from book 14, about which see below). The workwas written between ad56 and ad64 (Griffin 1976: 399). Seneca returnedto the subject of favors and ingratitude in epist. 81 (from book 10): epist. 81.3makes clear that benef. predates the letter. As with any multivolume work,composition over time must be considered. In this case, books 1–4 appear toform a cohesive unit, and benef. 5.1.1 begins with an acknowledgment thatthe remaining books are of a different character. While they have the same

23 Tanner 1985 offers a radically different division of the plays based primarily on hisperceptions of the performance demands. He suggests that Thyestes, Medea, Agamemnon,Oedipus, and Phaedra (and Phoenissae if produced as it survives) were composed by Senecaduring his exile from ad41–49, and that the remaining plays (Octavia, Troades, Hercules Furens,and Hercules Oetaeus) are not by Seneca and were composed between ad70 and 80. Dingel2009 appeared as this volume was due to go to press, and could not be taken into account.

24 Griffin 1976: 407–411.25 See Griffin 1976: 134–139 for a discussion of this tension.


42 c.w. marshall

addressee, there is no reason they could not have been composed after thespace of several years, even following the composition of epist. 81.

Indeed, several works may be ascribed to the period following Seneca’sretirement in ad62, which was a particularly prolific period: we have seenthat De Otio and De Providentia might belong to this period, as might Thyestesand some or all of De Beneficiis. Almost certainly Phoenissae comes fromthese final years, as do Naturales Quaestiones and the Letters to Lucilius(Epistulae). Hine (2006: 71) provides a conservative back-of-the-envelopecalculation, suggesting a rate of composition during this period of (at least)one book every forty days, even suggesting that this was “a very modest rateof composition compared to what Cicero achieved at the end of his life” (71,note 124), who achieved an average of about one book every twenty-fourdays. There are reasons to believe that Seneca was in fact composing fasterthan this, which demonstrates that much of this final period was dedicatedto literary endeavors.

The text of Naturales Quaestiones is corrupt, but the work may orig-inally have had eight books addressing various natural phenomena: sevenbooks survive, with a clear break evident in book 4.26 The date of the work isestablished by reference to specific natural phenomena within it. Mention ismade of an earthquake in Greece that preceded an earthquake in Pompeiiby one year (nat. 6.1.13). Tacitus (ann. 15.22.2) dates the Italian earthquake toad62, which is close to the date of ad63 suggested by Seneca (nat. 6.1.2).27

This could provide a context for the mention of earthquakes in Greece fol-lowing the appearance of a comet (nat. 7.28.3), which would therefore referto a comet that was visible from August to December of ad60 (Ramsey 2006:140–146). Similarly, when Seneca says “two such [comets] have been seen inour lifetime” (quales duo aetate nostra visi sunt: nat. 7.6.1), it seems certain heis referring to this comet and Claudius’s comet, visible for a month in ad54(Ramsey 2006: 136–140). This passage, therefore, was not revised after theappearance of another comet, in May of ad64.28 These details combine tosuggest that nat. was completed sometime betweenad61 and early 64. Seneca

26 Hine 1996: xxiv argues that the original order was 3, 4a, 4b, 5, 6, 7, 1, 2.27 Hine 1996 assumes nat. 6.1.2 Regulo et Virginio consulibus to be a gloss from Tacitus. See

also Wallace-Hadrill 2003: 182, who argues for a date of late ad61 for the Greek earthquake,and Hine 2006: 68–72, who summarizes the issue and emphasizes that there is no reason todoubt the Tacitean text.

28 The language of the sources (sidus cometes: Tac. ann. 15.47.1; stella crinita: Suet. Nero36.1) shows that this object was thought to be a comet, even though it may have been a nova.See Ramsey 2006: 146–148.



may have begun it earlier, and some relationship doubtless exists betweenpassages here and several lost works, including De Situ et Sacris Aegyptiorum(T 19 Vottero), De Situ Indiae (T 20–21 Vottero), De Motu Terrarum (T 55Vottero), and De Forma Mundi (T 56 Vottero). Nat. is dedicated to LuciliusJunior (PIR2 L 388), as was prov. Lucilius was a slightly younger contemporaryof Seneca (he is in senectute in epist. 19.1, and see 96.3), who in the early 60sattained a procuratorship in Sicily.

Lucilius is also the recipient of the collection of Stoic letters, Epistulae,which survive as 124 letters divided into twenty books, although AulusGellius 12.2.3 refers to a non-extant letter in book 22, which shows that thesurviving collection is incomplete and that its precise original size cannotbe determined. The book format is important for interpreting the letters,although modern editions typically obscure this aspect (Wilson 2001). Severalletters mention real events, and while some historical details may be includedto create a sense of dramatic moment, isolating the dramatic date for theletter from the actual time it was composed in the tradition of Athenianphilosophical texts is problematic. Many details refer to comparativelypersonal issues in the lives of Seneca and Lucilius that cannot serve thisfunction for a broader readership: a lawsuit, a new book by Lucilius, personalillness, retirement, and so on. These are of a different order than the referenceto the fire at Lugdunum (Lyons) in epist. 91 (dated to July 64 by Tacitus, ann.16.13.3), for example, which could more easily be used to establish a dramaticdate, if that were Seneca’s intention. Nevertheless, the rate of compositionsuggested by the letters clearly points to an inherent artificiality in the natureof the correspondence between the two, indicating “not only that he failedto wait for a reply before writing (as he does in 118.1 as a concession), but thathe sometimes sent letters, not individually, but in packets” (Griffin 1976: 418).Seneca knew he was writing for publication (epist. 21.5), and it is likely thatthere was some editorial work introduced either by him or by someone elsesoon after his death, which means that any apparent allusion to real eventsmay serve multiple unrecoverable purposes. References to the passing ofseasons do coincide with a relatively tight sequence for the letters betweenthe autumn of 63 and Seneca’s death in April 65 (Griffin 1976: 347–353 and400), with some books appearing for the public “perhaps in the latter part of64” (Griffin 1976: 349). From this Griffin concludes that Lucilius’s spiritualdevelopment, described in the letters, must be fictional, although this cannotbe taken as certain.

There are, of course, other works of Seneca that no longer survive. Martial7.45 implies that there was a collection of letters sent to Caesonius (orCaesennius) Maximus, a friend who had accompanied Seneca during his exile


44 c.w. marshall

(see also 7.44). There are no means by which to date the lost De Matrimonio(T 22–F 54 Vottero), De Superstitione (T 64–F 75 Vottero), or the life of hisfather (De Vita Patris, F 97 Voterro). In his account of Seneca’s suicide, Dio(62.25.2) has him revising a book and leaving it, and others, with his friends.It is not known which works these are. Tacitus, too, emphasizes that Senecawas composing to the very end, even though he had cut himself repeatedlyto increase the flow of blood: et novissimo quoque momento suppeditanteeloquentia advocatis scriptoribus pleraque tradidit—“and, even at the verylast moment, his eloquence in full supply, he called his scribes and dictateda great many things” (ann. 15.63.3).

The Works of Seneca the Younger and their Dates

Documents

Transcript of The Works of Seneca the Younger and their Dates