Tacitus: Agricola 0521876877, 9780521876872, 9780521700290

The first work of any great historian has always commanded attention, and Tacitus was ancient Rome's very greatest

375 20 27MB

English Pages 374 [371] Year 2014

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Tacitus: Agricola
 0521876877, 9780521876872, 9780521700290

Citation preview

CAMBRIDGE

GREEK

AND

LATIN

CLASSICS

TACITUS AGRICOLA

EDITED

BY

WITH

A. ]. WOODMAN C. S. KRAUS

TACITUS

AGRICOLA EDITED

A.

J.

BY

WOODMAN

Basil L. Guldersleeve Professor of Classics, Untversity of Virginia

WITH

CONTRIBUTIONS

C.

S.

FROM

KRAUS

Thomas A. Thacher Professor of Latin, Yale Unersity

:

»

CAMBRIDGE UNIVERSITY

PRESS

CAMBRIDGE UNIVERSITY

PRESS

University Printing House, Cambridge cB2 8Bs, United Kingdom Cambridge University Press is part of the University of Cambridge. It furthers the University's mission by disseminating knowledge in the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/0780521700290 (O Cambridge University Press 2014 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2014 Printed in the United Kingdom by CPI Group Ltd, Croydon cRo 4vv A catalogue record for this publication is available from the British Library ISBN 978-0-521-87687-2 Hardback ISBN 978-0-521-70029-0 Paperback Cambridge University Press has no responsibility for the persistence or accuracy of URLS for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

CONTENTS List of maps and figures Preface

page vi vii

List of abbreviations

ΙΧ

Introduction

]

1 The author and hzs work

1

3 Imperialism, freedom and servitude

15

4 The Agricola as history

25

n Language and expression

30

6 The manuscripts CORNELII

TACITI

DE

VITA

35 IVLII

AGRICOLAE

Commentary

39

05

Appendixes 1 Roman legions in Britatn during Agricola’s career

331 331

Marks cited

333

Indexes General

332

EEE

2 Chronological tables for Agricola and Tacitus

MAPS

AND

FIGURES

MAPS Identifiable places and peoples mentioned in the Agricola Places and peoples mentioned in the Introduction and Commentary

page xii xiii

FIGURES The Rudge Cup (mid- or late 2nd century Ap ?). © The Trustees of the British Museum Lead pipe from Chester (/LS 8704a — IRB 25) showing Agricola's full name. Courtesy of The Grosvenor Museum, Chester (a) Bronze coin of Nero

14

(AD 50—54, RPC 2381), the reverse

showing a double axe (bzpennis). From coinarchives.com, Ex Lanz sale 117, 2003, 587 (b) Denarius of L. Procilius (80 Bc, RRC 379/2), the reverse

showing Juno Sospita with her scutulum. © The Trustees of the British Museum Claudius subduing Britannia (relief from the Sebasteion at Aphrodisias). Courtesv of the NYU Excavations at Aphrodisias

134 135 245

PREFACE For almost half a century the standard English commentary on Tacitus' Agricola has been that by R. M. Ogilvie and Sir Ian Richmond, which was published in 1967. It began life asa revision of the commentarv produced by J. G. C. Anderson in 1922, which itself was a revision of the commentary by H. Furneaux published in 1898, when the United States and Spain were at war with each other and the British fought at Omdurman. The present commentary differs from that of Ogilvie and Richmond in three principal ways. First, it is not a revision of any predecessor but is an entirely new and independent work. The text, for example, is different from, and considerably more open to conjecture than, others currently available. Second, the commentary lacks the heavy archaeological content which characterised their book and which was in many wavs intellectually misleading: Tacitus in his biography of Agricola provides very few specific details of events or localities which can be illustrated by reference to evidence on the ground; for the most part he talks in general terms, designed to portray his father-in-law as an ideal military commander and provincial governor. Readers should therefore not turn to the present book for the latest information on Roman Britain, which is in any case a scholarly field subject to rapid change and revision. Third, and most important, the main aim throughout has been to explain the nature and meaning of Tacitus' Latin. In keeping with the general principles of the series Cambridge Greek and Latin Classics, an effort has been made to provide ‘all the guidance with grammatical and syntactical matters' needed by today’s undergraduate and graduate students. At the same time, it is hoped that the work will not be deemed negligible by fellow scholars. If there has been no reluctance in quoting parallels, that is partly to illustrate the texture of Tacitus' language, partly to correct any misleading inferences that may be drawn from the commentary of H. Heubner (1984), who wrote without the benefit of modern computerised search programmes and the like. It was extremelv helpful to test-drive the commentary with an undergraduate Latin class in the spring of 2013 at the University of Virginia. Much gratitude is also owed to those scholarly friends from whom help and advice has been sought and received, notablyJ. N. Adams, S. Bartera, E. Courtney, C. H. George, P. R. Hardie, N. Holmes, T. A. Joseph, M. Lavan, J. E. Lendon, the late R. H. Martin, S. P. Oakley, C. B. R. Pelling, J. G. F.

Powell, B. D. Shaw,J. B. Solodow and gratitude is owed to A. R. Birley, who commentary and given the benefit of Britain; he is certainly not to be held conceptions which may remain.

vil

C. L. Whitton. An especial debt of has read through the whole of the his unrivalled knowledge of Roman responsible for any mistakes or mis-

viii

PREFACE

It was originally intended that this book be written jointly with C. S. Kraus. In the event, she was unable to write much

toward this collabora-

tion, which she greatly regrets. She is responsible only for the commen-

tary on 10—12, some occasional notes elsewhere, and some revisions. She

would like to thank the many friends who helped her talk through Agricolan matters, and the series editors for their forbearance.

A. J. W.

ABBREVIATIONS Tacitus 15 abbreviated as T., Agricola as A.; T.'s works are abbreviated as A. (Annals), Agr. (Agricola), D. (Dialogus), G. (Germania) and H. (Histo-

ries). References to Agr. normally omit the title; references within the same chapter normally omit the chapter-number.

A-G BMC BNP CCRH CCT CGRH CIL C-L CLE

Emp.

Allen and Greenough's New Latin Grammar (rev. A. Mahoney, Newburyport, MA 2001) H. Mattingly, Coins of the Roman Emprire in the British Museum, Vol. 3: Nerva to Hadrian (London 1930, corr. repr. 1966) Brill's New Pauly, Vols. 1—15

(Leiden 2002-10)

The Cambridge Companion to the Roman Historians (ed. A. Feldherr, Cambridge 2009) The Cambridge Companion to Tacitus (ed. A. J. Woodman, Cambridge 2009) A Companion to Greek and Roman Historiography (ed. . Marincola, Malden, MA/Oxford

2007)

Corpus Inscriptionum Latinarum (Berlin 1863-) J.-P. Chausserie-Laprée, L'expression narrative chez les historiens latins (Paris 1969) F. Buecheler and E. Lommatzsch,

Carmina Latina

Epigraphica, Vols. 1—g (Leipzig 1895—1926) A Companion to Tacitus (ed. V. E. Pagán, Malden 2012) H. Furneaux,

Corneliz Taciti Vita Agricolae (Oxford

1898)

H. Furneaux and J. G. C. Anderson, Corneliz Taciti De Vita Agricolae (Oxford 1922) S. Riccobono, Fontes zuris Romani anterustinian: (Florence

1941)

E. Courtney, The Fragmentary Latin Poets (2nd edn, Oxford 200%)

F. R. D. Goodyear, The Annals of Tacitus. Vol. 1 Annals I.1—54, Vol. 2 Annals 1.55—61 and Annals 2 (Cambridge 1972, 1981) A. Gerber and A. Greef, Lexicon Taciteum (Leipzig

1877-90)

B. L. Gildersleeve and G. Lodge, Latin Grammar (repr. 1992, Walton on Thames) H. Heubner: either P. Cornelius Tacitus: Die Historien. Kommentar. Vols. 1—5 (Heidelberg 1963-82) or Kommentar zum Agricola des Tacitus (Gottingen 1984) iX

LIST

OF

ABBREVIATIONS

Inscriptiones Latinae Selectae (Berlin 1892—1916) V. A. Maxfield and B. Dobson, /nscriptions of Roman Britain (LACTOR 4, 4th edn, 2006) C. S. Kraus, Lwy: Book VI (Cambridge 1994) H. G. Liddell and R. Scott, A Greek-English Lexicon, oth edn,

rev. H. S. Jones (Oxford 1940) R. H. Martin and A. J. Woodman, Tacitus: Annals Book IV (Cambridge 1989, rev. 1999) R. G. M. Nisbet and M. Hubbard, A Commentary on Horace. Vol. 1 OdesI, Vol. 2 Odes II (Oxford 1970, 1978) E. C. Woodcock, A New Latin Syntax (London 1959) R. G. M. Nisbet and N. Rudd, A Commentary on Horace. Odes Book III (Oxford 2004)

Oxford Classical Dictionary (ed. S. Hornblower, A. J. Spawforth and E. Eidinow, 4th edn, Oxford 2012) Oxford Latin Dictionary (Oxford

1968—82)

R. M. Ogilvie and I. Richmond, Tacitus: Agricola (Oxford

1967)

Ε. Malcovati, Oratorum Romanorum Fragmenta Liberae Rei Publicae, 4th edn (Turin/Milan/Padua 1976—79) Oxford Readings in Classical Studies: Tacitus (ed. R. Ash, Oxford 2012)

A. J. Woodman, From Poetry to History (Oxford 2011) A. R. Birley, The Roman Government of Britain (Oxford 2005) Roman Inscriptions of Britain, Vols. 1—3 (Oxford/ Gloucester/Stroud 1965-2009) A. J. Woodman, Rhetoric in Classical Historiography (London/Sydney/Portland

1988)

R. Syme, Roman Papers, Vols. 1—7 (Oxford 1979-91) A. M. Burnett, M. Amandry and P. P. Ripollés, Roman

Provincial Coinage. Vol. 1: From the death of Caesar to the death of Vitellius (44 BC-AD 69) (London/Paris 1992) M. Crawford, Roman Republican Coinage (Cambridge 1974) Senatus Consultum de Cn. Pisone Patre Supplementum Epigraphicum Graecum (Leiden 1923-) Thesaurus Linguae Latinae (Leipzig 1900-)

TR Tab.

W.

Vind.

A. . Woodman, Tacitus Reviewed (Oxford 1998) A. K. Bowman and J. D. Thomas, The Vindolanda Writing Tablets (Tabulae Vindolandenses), II (London 1994) A. J. Woodman, Velleius Paterculus: the Tiberian Narrative (2.94—131) (Cambridge 1977), Velletus Paterculus: the

LIST

OF

ABBREVIATIONS

xi

Caesarian and Augustan Narrative (2.41—93) (Cambridge

1983)

A. J. Woodman and R. H. Martin, The Annals of Tacitus Book 3 (Cambridge 1996)

KEY (alphabetical)

3

Bodotria / Firth of Forth Clota / River Clyde Colonia / Colchester

"

Mona / Isle of Anglesey

(j

. -äy'Thule?

Oceani fretum / English Channel

Orcades insulae / Orkney Islands Taus / Tay estuary

Thule? / Shetland

.

o

ο

« f E

ξ

^

M w

0575

BRITANNIA Coloni (Camtfio%?ll:um)o

SILURES

ζ» .

0

100

| A

S

L

0

200

i

50

OCEANI FRETUM

300

i

S

100

400 km

i

S

150

J

|

200 miles

a

ο

. o

TN

Map 1 Identifiable places and peoples mentioned in the Agricola

KEY (alphabetical) Aldborough/Isurium (13) Newstead/Trimontium (7) Caerleon/Isca (20) Richborough/Rutupiae (29) Camboglanna/Castlesteads (10) Silchester/Calleva Atrebatum (27) Camulodunum/Slack (15) Springhead (28) Carlisle/Luguvalium (9) St Albans/Verulamium (25) Chester/Deva (16) Strageath (3) Corbridge/Coria (12) Tassiesholm (8) Dalginross (2) Usk/Burrium (21) Elginhaugh (6) Vindolanda (11) e Exeter/Isca (30) Wroxeter/Viroconium (18) oo ; Frilford (23) York/Eboracum (14) — 2" Thule* o Gloucester/Glevum (22) Inchtuthil (1) Inveresk (5) Leicester/Ratae (19) Q.; o9 oO Lincoln/Lindum (17) Orcades insulae Lockleys (24) London/Londinium(2

Mollins (4)

0 [

100 I

0

50

200km nE

100 miles o

, o $ >

ooo ὡς

« .

9 \§ ζ

o3 qS 0

Qo°o°

conNovn\Ca'th"ess \)

QD

Y VENICONES

Taus

— 5 dotria

Ο

SILURES 210 209 30

22

23 @

279

25*

INOVAT 024

26 28°

Cantium

Col 3

> OCEANI

FRETUM L

Map 2 Places and peoples mentioned in the Introduction and Commentary

FRISII

INTRODUCTION 1 THE

AUTHOR

AND

HIS

WORK

The Agricola, as it is conventionally known,' is described by its author at the start as a biography (1.4 ‘narraturo mihi uitam defuncti hominis'),? a form of writing which even today is seen as problematic and hard to define. ‘Is biography a genre?’, begins a recent book on the subject, and, if it is a genre, what are its characteristics?? T. says that his purpose in writing is to honour his father-in-law (3.3 ‘liber honori Agricolae soceri mei destinatus’). Some readers have certainly seen a problem here, since in the modern world *we expect factual information' rather than encomium in a biography; but this is to misunderstand the ancient equivalent, since in the classical world ‘biographers were free to be encomiastic'.* Cicero makes this point in one of his letters to Atticus, when, discussing various autobiographical accounts of his consulship, he assures his friend in a playful paradox that they are ‘not encomiastic but historical' (A#. 1.19.10 ‘non ἐγκωμιαστικὰ

sunt haec sed ἱστορικά᾽)."

Indeed

T. himself alludes to the

connection between biography and encomium when he chooses 'criticising' as the term with which to describe the opposite of ‘biography’ (1.4 narraturo . .. uitam — incusaturus) and 'admiration and praise' asan implied description of the Agrzcola itself (46.2 'admiratione ...et laudibus’). A theoretical distinction between biography and history, such as that implied by Cicero, was famously mentioned in the biographical writings of Nepos in the first century Bc (Pelop. 1.1 ‘uereor...ne non uitam eius enarrare sed historiam uidear scribere’) and of Plutarch in the early second century AD (Alex. 1.2 *we are not writing histories, ἱστορίας, but lives,

Bious’).® Although T. defines the Agricola as biography, many readers have

' The manuscripts vary between de vita luli? Agricolae and de vita et moribus Iul Agricolae. ? Despite T.’s own description, ‘The precise nature of the literary genre into which the Agricola should be fitted has taxed scholars for generations' (Hanson

(1991) 1746).

? Hagg (2012) 2, with a helpful summary of the issues (2-8). * The quotations are from Momigliano (1993) 15. See also Hagg

(2012)

3: 'In

criticism of biography, modern and ancient, one often finds a naive demand that it should be "true", in the sense of verifiable and historically correct.' 5 Elsewhere (Brut.

112) Cicero uses the terms ‘Cyri uitam et disciplinam' to refer

to Xenophon's Cyropaed:ia, which Diog. Laer. describes as ἐγκώμιον (6.84). Cf. also

Plb. 8.8.6, 10.21.8.

? Duff (1999: 14-22) argues that the latter passage is applicable only to the Lives of Alexander and Caesar and should not be elevated into a general statement about generic differences, but he acknowledges both that the passage does distinguish between the two genres and that the distinction is also to be found at Plut. Fab. 1

2

INTRODUCTION

questioned this definition and have preferred to see the work as some kind of historical monograph not unlike, for example, Sallust's Bellum Tugurthinum. Reasons for this preference emerge from a consideration of the work's contents, which are conveniently illustrated by analvsis of the work's structure. The preface (1—3), which has the phrase posterss traderein its opening line, is balanced by a conclusion (44-6), which has the words posteritati . . . traditusin its last line;? and within this outer frame an account of A.'s earlier vears (4—0) is balanced by that of his final decade (39—43). But these seventeen framing chapters, amounting to 35 per cent of the whole work, are significantly outnumbered by the twenty-nine central chapters (10-38), which are devoted to Britain and especially to A.'s governorship of the province in the years AD 77—84. On the structure of these central chapters more than one view 15 possible. Chapters 10-- 7 comprise a two-part digression or excursus on Britain, dealing successively with its ethno-geography (10-13.1) and its earlier history (13.1—17.2), and it may be argued that this digression is an entirely self-contained section which is used to separate and set off the year-by-year narrative of A.’s governorship.? If that is the case, the narrative of A.'s seven-year governorship pivots around the central year (AD 80), which is the shortest of all (23), and cli-

maxes in the battle of Mons Graupius in AD 83 (29-38). Alternatively one may regard the digression on Britain as an integral part of the narrative, as Gudeman suggested (below, p. 4), and as balanced by the battle of Mons Graupius, with the vears AD 77—83 constituting a central panel: Preface

(1-3)

Farlier years (4—9) Britain: digression

(10-17)

Britain: the years 77—83 (18-28) Britain: Mons Graupius (29-38) Final decade (39-43) Conclusion (44-6)

..

On this view the work as a whole is an almost perfect example of circular or ‘ring’ composition,? the correspondence between the digression and Mons Graupius residing not only in length of treatment but also in theme. With its description of a foreign country, surrounded 16.6. It is conventional to refer also to Rhet. Herenn. 1.13, Cic. Inv. 1.27, Sext. Emp. Adv. Math. 1.253. 7 There are numerous other parallels between preface and conclusion (see nn.), of which the most obvious is 1.4 narrare — 46.4 narratus. $ So e.g. O-R 164. 9 See esp. Giancotti (1971) 253ff., though his scheme is slightly different from that adopted here. Scholarly views on the structure of the work are surveyed by Wille (1983) 5-8.

1

THE

AUTHOR

AND

HIS

WORK

3

by Ocean and characterised by recalcitrant barbarians, the digression symbolises the many-faceted task with which A. was confronted when he first landed;

by the

time

Mons

Graupius

is reached,

the Romans

have

con-

quered nature itself in their penetration of Caledonia (33.2, cf. 30.3) and have outdone the achievements of previous armies and commanders (33.3): with this final victory there is no one left to conquer, and they are free at last to embark on the circumnavigation which in the digression is invoked to confirm the insular status of the country (38.3—4 — 10.4). Whichever view of the structure one prefers, it is clear that 65 per cent of the work is less biographical than historical: the digression is of the kind which is inserted into major works of history by Sallust, Livy and T. himself elsewhere

(p. 12), while the remainder is a conventional historical narra-

tive of the years AD 77—84 and, arranged year by year as it is, it resembles the kind of narrative one meets routinely in the major Roman historians.'? Since so much of the Agricola therefore looks like ‘history’, some scholars have questioned whether the work as a whole is to be classed generically as biography. But this too is to misunderstand its nature. Ancient literary theorists laid down rules for various types of writing, including encomium. Thus Menander Rhetor, who wrote in the late third century AD, prescribed the following elements (368.8-377.2):'' preface (προοίμιον), native land (πατρίς), family (yévos), birth (yéveois), nature (φύσις), upbringing (ἀνατροφή), behaviour (ἐπιτηδεύματα), deeds (πράξεις) and finally comparison with others (σύγκρισις). Since the

genres of biography and encomium were so closely related, as we have seen, there is no reason why these prescriptions should not be applied to biography; and, as Gudeman was the first to point out, all of them are directly applicable to the Agricola.'* The correspondence between the first seven elements and chh. 1-9 of the biography (1-3 Preface, 4-ὁὁ Earlier years) is clear and does not require elaboration here; but the remaining two elements are less straightforward. Menander stipulated that ‘deeds’ should be divided into 'times of peace and times of war, and put war first, if the subject of vour praise has distinction in this' '? See e.g. C-L 5, where it is suggested that T. was providing his contemporaries with 'the measure of his genius' by displaying his intimate familiarity with earlier historians and with the fundamental techniques of historical narrative. ''

References and translations are taken from Russell and Wilson

(1981), who

use Spengel's numeration. Though Menander wrote much later than T., it is generally accepted that his recommendations are based on earlier examples and Cairns (1972) has argued that it is valid to use his work to help elucidate earlier texts: see Du Quesnay (1981) 53-62. References to Menander are given in the Commentary where they are most relevant. '* Gudeman 311-22. Menander's prescriptions are also applicable to Plutarch; for the war narrative embedded in Plutarch's Caesar, for example, see Pelling (20112) 219, 226—-7.

4

INTRODUCTION

(372.25-8). The prominence which T. gives to A.'s military exploits in Britain (18-38) exactly fulfils this prescription. But Menander also said that, when treating deeds of war, 'you should describe the natures and situations of the places where the wars took place, rivers, harbours, mountains, plains, and whether the country was bare or wooded,

rocky' (373.17—20). This description is precisely what T. has provided in the ethno-geographical part of his digression (10-13.1); in the remaining part (13.1—17.2) his survey of the earlier history of Britain encompasses A.'s predecessors from Julius Caesar to Julius Frontinus and 'dwells as far as possible on the unsuccessful features of the Roman administration. It is difficult to avoid the conclusion that the survey of previous governors is included for purposes of synkruszs.”'3 Thus T. elevates by contrast the achievements which are about to be attributed to his father-in-law. In other words, the biography as a whole conforms remarkably to the rules for encomium as laid down by Menander Rhetor, and its ‘historical’ features are to be regarded in this light. Some might still object that the presence of formal speech, both indirect (15.1—5) and direct (30—4), crosses the boundary into historiography proper.'* Yet the boundary was never absolute in practice: Nepos, who in his Life of Pelopidas expressed the distinction between biography and historiography (above, p. 1), proceeded to imply that he was going to combine both genres (cf. Pelop. 1.1 ‘utrique rei occurram’). Such generic combinations are, after all, a regular feature of verse literature, whether we are talking about genres in the conventional sense (such as epic or elegy) or in the 'Cairnsian' sense (such as ‘propemptikon' or ‘kletikon’); and the regularity of their appearance is explained precisely because poets liked to surprise or frustrate the literary expectations of their readers.'> There 15 certainly plenty of evidence that biography and historiography were closely related.'? In his prescriptions for historiography, for example, Cicero said that an author should deal with the life and nature of any famous protagonist in the narrative (De or. 2.63 ‘de cuiusque uita atque natura’). The factis that in the ancient world '3 McGing (1982) 16-17. It should not be thought that comparison is restricted to this section of the work; as McGing comprehensively shows, it is a technique which T. uses throughout. '+ Speeches are found in Plutarch’s biographies but they are not common and tend to be short. 'Extended speeches always illustrate important themes' (Pelling on Plut. Ant. 84.4—7, q.v.). '5 For a recent study of the former type, associated above all with W. Kroll (‘Kreuzung der Gattungen’), see Harrison (2007); for 'sending-off poems' and 'summoning hymns' and the like see Cairns (1972). !9 See e.g. Stadter (2007), Kraus (2004) and (2010); further bibliography in Duff (1999) 17 n. 13. For historiography note Marincola (1999), esp. 318-20 for the Agricola.

THE

AUTHOR

AND

HIS

WORK

Q

1

biography and historiography were both rhetorical genres and authors would resort to any appropriate technique to move and impress their readers. A good example of this is to be found in the conclusion to the work (44—6). Here T. deploys many of the commonplaces which are to be found ΪΠ various types of ‘death literature' such as the conquestio or consolatio and which are vet again to be found in theoretical form in Menander.'7 Yet T.'s deployment is hardly different from that of the Tiberian historian Velleius Paterculus, who resorted to exactly similar motifs in order to make his account of Cicero's murder as moving and effective as possible (66.9—5).'? Velleius' history affords a further comparison as well. His concentration on the imperial figures of Caesar, Augustus and Tiberius means that he is an excellent representative of 'biographical history', a trend which inevitably became common under the empire;!'? the Agricola, beinga ‘historical biography’, is the converse of this and can even be seen as a textual as well as a moral challenge to the princeps under whom A. achieved his highest distinction: in the ‘historical’ section of the work it is A. whose role is central,

while Domitian is not mentioned. By the time T. came to write the Agrzcola, he was about forty vears old and an experienced orator. He had been born early in the principate of Nero at some time between the years 56 and 58. It is usually thought that his birth-place was southern Gaul, but Trier in Gallia Belgica has been suggested as an alternative:?? it is generally accepted that the Cornelius Tacitus who is mentioned by the elder Pliny as procurator in Gallia Belgica (NH 7.76) was the historian's father. In 76 T. was engaged and subsequently married to A.’s daughter, who was then about twelve or thirteen years old and a girl ‘of exceptional promise' (Agr. 9.6 n. egregiae). If a fragmentary inscription (CIL 6.41106) has been correctly identified as T.'s epitaph, and if its last two lines have been correctly restored, he will have served as military tribune in a province. 'Why not suppose’, suggests A. R. Birley, 'that Tacitus served in Britain, in one of the four legions in the army of his father-in-law? He could well have staved there for two to three years, from 77—79.?' However that may be, the inscription next tells us that T. was quaestor August?, probably in 81: the quaestorship brought with it entry to the senate, and it was a mark of special favour to be one of the emperor's two personal quaestors. It is noteworthy that for the last third of the year 81 the emperor was Domitian, who draws such fierce criticism from T. in the Agricola. According to T.'s own later account (A. 11.11.1),

'7 See Gudeman 321-2, and nn. in the Commentary. '9 See Woodman

(1977)

51-6, Pelling (1997),

(2011a)

!35 See W. ad loc.

32-3 and (2011b).

^" For a full study of T.’s life and career see Birley (20002). ?' Birley (20002) 237-8, cf. RGB 281.

6

INTRODUCTION

he was praetor in 88 and, having already had the distinction of appointment to one of the priestly colleges (the quindecimuir: sacris factundis) , he helped supervise the Secular Games which Domitian put on in that year. We know that T. and his wife then soon left Rome,

since he tells us that,

when A. died on 23 August 93, they had already been awav four years and did not have the comfort of paying their last respects to him before his death (Agr. 45.5). T.'s absence from Rome meant that he also missed not only the trial of Baebius Massa (Agr. 45.1n. Massa), of which an account was later sent to him by Pliny (Ep. 7.33), but — almost certainly — the condemnation and execution of Arulenus Rusticus and Herennius Senecio as well (2.1n. legimus, 45.1n. mox), events which marked a dramatic development in Domitian's reign of terror. Four vears later, in 97, T. was suffect consul, possiblv holding office in the autumn of that year.?? Although Domitian had been assassinated in September of 96, T.'s appointment had almost certainly been approved by Domitian: T. thus continued to the last the favour he had enjoyed from the princeps throughout his reign. We know from Pliny that during his consulship T. delivered the funeral oration for the aged Verginius Rufus. Pliny describes T. on this occasion as 'laudator eloquentissimus' (Ep. 2.1.6), just as elsewhere he describes him in a court case as having spoken 'eloquentissime et — quod eximium orationi eius inest — σεμνῶς᾽

(Ep. 2.11.17): the Greek adverb is difficult to translate but

suggests solemnity and majesty. It was during or immediately after his consulship that T. embarked on writing the Agricola.?? (If the argument of his preface has been correctly interpreted,?* he had planned to write the biography as soon as A. died in the summer of 93 but was obliged to defer his project when news reached him at his posting abroad that Arulenus Rusticus and Herennius Senecio had verv recently been executed for having written the biographies of Thrasea Paetus and Helvidius Priscus respectively and had had their books burned.) Biographical writing, like almost all other Greek and Latin genres, had a long tradition behind it and would-be authors were extremely conscious of the tradition in which their work would take its place.*5 ?? See Birley (2000a) 238. Under the empire consules suffecti or ‘replacement consuls' regularly took over from the consules ordinarii who had begun the year. The ‘ordinary consuls' gave their name to the year (cf. 44.1), enjoyed more prestige than the suffects, and remained in office for a month or two before being replaced;

there could be several pairs of suffect consuls in any one year, as was the case in g7. *3 See the Commentary on 3.1. ** See the Commentary on 1.4-2.1. *5 Forstudies of classical biography see Leo (1901), Stuart (1928), Dihle (1970)

and (1987), Geiger (1985), Moles Burridge

(2004),

Kraus

(2010),

(1989), Momigliano

Pelling

(2011a)

13-25,

(1993), Ehlers (1998), Hagg

(2012);

scholars

tend to distinguish between biographies of historical or political figures and those

1

THE

AUTHOR

AND

HIS

WORK

7

Isocrates and Xenophon, both born in the 430s BC, wrote encomiastic accounts of Evagoras and Agesilaus which are normally invoked as models for the Agricola: indeed it has been remarked that ‘one should compare the epilogue chapters of the Agrzcola (44—40) sentence by sentence with the concluding chapters of the Ageszlaos (10—11) in order to see clearly the close dependence of Tacitus on Xenophon in detail'.?? Tacitus thus exemplifies the influence which Cicero attributed to Xenophon's work over a century and a half earlier (Fam. 5.12.7 'unus enim Xenophontis libellus in eo rege laudando facile...superauit').?^7 The works of only two Latin biographers have survived, and that in part. Cornelius Nepos, the friend and dedicatee of Catullus, wrote the lives of several hundred indi-

viduals according to category: of this vast output we possess in its entirety the De excellentibus ducibus exterarum gentium and two lives (the attenuated

Cato and the more substantial Atticus) from the De historicis Latinis.?? Sue-

tonius, a younger contemporary of T., wrote lives of from Julius Caesar to Domitian;?? he also wrote a various literary men: we possess the De grammaticis vidual lives of a very few poets including Horace.?? works is at all comparable with the Agrzcola. Almost

the Roman emperors De urris illustribus on et rhetoribus and indiNone of these Latin all of them are brief

and sketchy; the exception is Suetonius' Caesares, but Suetonius' narrative mode, unlike that of T., is less chronological than thematic, as he

explains in his Life of Augustus (g.1): ‘proposita uitae eius uelut summa partes singillatim neque per tempora sed per species exequar, quo distinctius demonstrari cognoscique possint’ (‘having displayed his life's (so to speak) totality, I shall go through its elements individually, not chronologically but thematically, so that they can be demonstrated and understood more clearly’). Another contemporary was Plutarch, whose biographies are different again: he writes paired lives, in which a Roman is compared with a Greek counterpart.?! of philosophers or writers. For a brief introduction to the modern genre see Lee

(2009). *^ Münscher

(1920) 92-3. ?7 Cicero is here talking of ‘clari uiri’, the same two words with which T. opens the Agricola. *5 For the latter see Horsfall (1989). ?9 The standard study of Suetonius is Wallace-Hadrill (1983). Suetonius is invoked as a model by the so-called Historia Augusta or Scriptores Historiae Augustae, a series of imperial biographies from AD 117 to 284. Almost everything about this weird collection is controversial, though most scholars now agree that it was written by a single author in the late fourth century. 3? See Kaster (1995) for the former. 3 For Plutarch see Duff (1999) and Pelling (2002) and (2011a). Burridge (2004: 124-84) compares various Greek and Latin biographical writings, including the Agricola, with one another on the basis of a series of generic features.

8

INTRODUCTION

There is no reason to doubt T.'s own statement that he was inspired by pietas to write the biography of his father-in-law (3.3). Though it was natural that a devoted son-in-law with literary aspirations would want to commemorate his wife's beloved father after his death, especially when that father had performed deeds worthy of commemoration, it is very likely that T. was encouraged in this direction by a sub-genre known as the exitus llustrium wirorum, which flourished in the first century Ap.?* Whether or not that was the case, the biography was written against a background of rapid change and considerable political uncertainty. Nerva had been przncepsfor little more than a year when he was compelled by events - including ΔῊ attempt at rehabilitating Domitian - to adopt Trajan as his successor in the autumn of 97; by late January 98 Nerva himself had died and been succeeded by Trajan (3.1nn.), who was away on the Rhine at the time of his succession.?? None of these vicissitudes is mentioned in the text of the biography. On the contrary, the preface depicts Nerva and Trajan as presiding over a new era of hope, promise, confidence and security (3.1). This sunlit scene naturally emphasises bv contrast the dark account of Domitian's reign by which the preface is framed (1.4-2.3 — 3.2) and which occupies the concluding chapters of the work (39—45); but it also serves another purpose. If Domitian's reign had been as detestable as depicted by T. here and by Pliny in his letters and Panegyricus, which adopts exactly the same contrast, there arose in retrospect the question of guilt by association: what of those numerous men whose careers had not only prospered during Domitian's reign but had been actively promoted by the princeps himself? Not least among these men was the new biographer, whose every senior magistracy had been held under Domitian or achieved with his support. T.'s strategy in confronting this problem is twofold. On the one hand he uses all his rhetorical skill to underline the tyranny and terror of Domitian's fifteen years, impressing upon his readers that contemporaries of the emperor had little choice but to accede to his wishes. On the other hand, the support expressed for Nerva and Trajan served as a reminder that T. was not alone in having received honours from Domitian. They too had held consulships during Domitian's reign, Nerva in 9o and Trajan in g1, and the first-person plurals by which their naming in the preface is surrounded (2.3 dedimus,

nos, perdidissemus,

nostra, 3.2

nostri, sumus,

uenimus)

implies

that T.'s numerous fellow collaborators included the two przncipes who now ruled the Roman world. The assassination of Domitian marked a break with the past and a new start; but the new start was in the hands of those

3*

See Plin. Ep. 3.10.1, 5.5.3, 8.12.4 and Sherwin-White's nn.

33 See Eck (2002) for these events.

1

THE

AUTHOR

AND

HIS

WORK

9

who had enjoyed power and success in the past. Nerva and Trajan represented continuity, as if Domitian himself had been an awful aberration. Itis one of the tenets of Agricolan scholarship that, in writing the biography of his father-in-law, T. was at the same time presenting a defence of his own career under Domitian. This feature illustrates the phenomenon called 'transference' or ‘automimesis’, which is very well recognised in literature and art and 'isin factso common that one might argue that biographies differ only in the degree to which this happens, or 15 observable’:34 One of the reasons is evidently that biographers from the start tend to choose figures that resemble themselves (in occupation, temperament, situation in life). Novelists and poets depict great literary figures, politicians trace political careers, women prefer women. A related factor is that the biographer has to rely on introspection to reconstruct the inner life of the subject. Primarily what he or she recognises from self-experience is likely to be included in the characterisation. Sympathy and empathy are key concepts. Although this analysis does not fit all cases by any means, its relevance to the Agricola is clear. Born in AD 40, A. had the misfortune to spend his young adulthood and much of his maturity under the emperors Nero (54—68) and Domitian (81—96).?5 His progress through the cursus honorumto a suffect consulship in 76 and the governorship of Britain thereafter was uninterrupted and, if T. is to be believed, almost pre-ordained; more

significant, as T. presents it, was his style of life and in particular the way in which the gloria and fama of his military achievements were counterbalanced by compliance, self-effacement and quietism.9? The problem is raised in A.'s very first posting as military tribune in Britain, where 'there entered his soul a desire for military glory which was unwelcome in times when a sinister construction was put on eminent men and there was no less danger from a big reputation than a bad one' (5.3). He spent 65, the vear after his quaestorship, and 66, the year of his tribunate of the plebs, ‘quiete et otio’, being well aware of the times under Nero when 'inertia pro sapientia fuit’ (6.3). There was the same 'silent tenor' to his praetorship in (probably) 68 (6.4 ‘idem...tenor et silentium’). Sent to Britain again during the civil wars of 69—70, he displayed obsequium and uerecundia (8.1, 3) and, having learned by the example of 34 Hagg (2012) 5-6. Remarkably he does not comment on this aspect of Agr. 35 Almost all of the known details of his life are owed to T.’s narrative. For modern accounts see Raepsaet-Charlier (1991), with a summary at 1856—7, and RGB 71—95. À. naturally features prominently in the many books on Roman Britain (below, n. 98); there are also more specialised monographs such as Hanson (1987). 39 For A. and fama see Hardie (2012) 273-84.

10

INTRODUCTION

his earlier posting (5.3 'gloria in ducem cessit’), he allowed his commanding officer to take all the credit (8.3 'ad ...ducem...fortunam referebat’); nor did he seek fama later when governor of Aquitania (9.4). He continued the same policy in 77, his first campaigning season as governor of Britain

(18.6 ‘dissimulatione famae’), when

the reticence of his

official reports (‘ne laureatis quidem gesta prosecutus est’) was repeated at the end of his tour of duty after the victory of Mons Graupius (~ 39.1 'nulla uerborum iactantia epistulis Agricolae auctum’). His return to Rome in 84 was the very opposite of that expected of a great man (40.3—4): ‘he drew deep breaths of calm inactivity' (40.4 'tranquillitatem atque otium penitus duxit’). When the time came for him to be considered for the proconsulship of Africa or Asia, men arrived from Domitian to praise quzesand otium (42.1): as a result, Domitian was assuaged 'because neither by truculence nor by the empty flaunting of freedom did he invite his fame and his fate' (42.3). There follows a famous verdict (42.4): sciant, quibus moris est inlicita mirari, posse etiam sub malis principibus magnos uiros esse, obsequiumque ac modestiam, si industria ac uigor adsint, eo laudis excedere quo plerique per abrupta, sed in nullum rei publicae usum ambitiosa morte inclaruerunt. Those who are accustomed to admire illegality should know that even under evil principes there can be great men, and that compliance and modesty, if accompanied by industry and energy, can reach the same level of praise as many persons do by precipitous routes; but the latter, with no benefit to the commonwealth, have achieved their

distinction by an ostentatious death.

obsequium and modestia aptly summarise the guiding principles of A.'s life, while their counterpoise bv zndustria and wigor epitomises the balance which he retained from his very earliest years (4.3 ‘retinuitque...ex sapientia modum’). The challenging and defensive tone 15 unmistakable but invites speculation about why it is there, since men of this balanced type were no new phenomenon.?7 Perhaps it 15 the suggestion that such men qualify for greatness which is the explanation: none of the previous exponents of this lifestyle could boast of a similarly impressive military record. But it is also the case that, by the very attributing of greatness to A., his biographer was defending a fortior: the similar behaviour of lesser achievers such as himself.?? 37 See W. on Vell. 88.2. 35 See also Sailor (2008)

critical reference loc.).

112-13. It is usually thought that plerique at 42.4 is a

to the Stoics, but whether

that is the case is uncertain

(see ad

2

TACITUS'

BRITAIN

11

It is an interesting feature of T.'s authorial career that his reticence about his own role in Domitian's reign is graduallv replaced by acknowledgement and then by forthrightness. Though there is no call for him to make any statement of his personal politics in the ethno-geographical Germania, which was published the year after the Agricola, or in the oratorical treatise Dialogus, which may have been published in 102,39 in the preface to the Histories he famously addresses the issue with these words (H. 1.1.3):

"That my political standing [ dignitatem] was begun by Vespasian, enhanced by Titus and advanced still further by Domitian I would not deny.' We infer from Plinv's letters that T. was engaged with the Histories around the year 107;*? mid-way through the next decade, about 114-116, he was writing the Annals, his most famous work, and there he seems to go out of his way to draw his readers’ attention to the privileged role he played as quzndecimutr and praetor when Domitian presented the Secular Games in 88 (A. 11.11.1).*! Itisa far cry from the silence of his first work almost twenty years before. T. evidently embarked on the Annals after returning to Rome from the governorship of the province of Asia, a position of the utmost distinction which he held in (almost certainly)

when the Annals further about T. ble for erecting questions which answer.'44

112/13.**

We do not know

was completed, if it ever was,*? nor do we know anything himself. 'Exactly when he died and who was responsihis funerary monument with its massive inscription, are another fortunate epigraphic discovery might one day

2 TACITUS'

BRITAIN

The lure of foreign countries and peoples is in evidence as early as the Odyssey and may be traced through various genres of Greek and Latin literature down to the novel. Historical writing was no exception: since Rome was an imperialist power, the operations narrated bv its historians often took place in distant localities whose landscape and inhabitants 39 For discussion of the date see Mayer (2001) 22-7. 1^ The date is inferred from Ep. 6.16 and 6.20, the two 'Vesuvius' letters where Pliny responds to T.’s request for information about the elder Pliny's death, and 7.33, where Pliny volunteers some information which he hopes will find its way into the historiae which T. is then writing. For the dating of the letters see SherwinWhite (1966) 36-8. ' See Woodman (2000b) 37-9. For the dating of the Annals see M-W on A. 4.5.2 (pp. 102-3 and n. 1). 4* Birley (2000a) 235-6. 15 The text breaks off half-way through Book 16. Whether that was the final book, or whether T. intended a work of eighteen books in total, is unknown.

** Birley (2000a) 247.

12

INTRODUCTION

required description for their readers. Cicero defined historical writing as the genre ‘in which a region is often described' (Orat. 66 ‘in qua...regio saepe ...describitur’), and, when he outlined his famous instructions for how history should be written, he said that historical works require 'the description of regions' (De or. 2.63 'rerum ratio ...desiderat regionum descriptionem’). T. in his later Annals acknowledged that foreign lands and peoples ‘retain and reinvigorate the attention of readers' (4.33.3 ‘retinent ac redintegrant legentium animos’). Thus both Africa and the Black Sea were described by Sallust ([ 17-19, H. 3.61-80), Gaul and Germany bv Livy (5.33.4—35.3, per. 103, 104), and Judaea by T. himself (H. 5.2-10). Awhole sub-genre of the ethno-geographical monograph was developed, of which T.'s own German?a, published shortly after the Agr:cola, is a classic illustration. Yet there is no evidence that Livy, for example,

attempted to visit the many distant points of the compass, from Africa via Spain, the Alps and Macedonia to Asia, where his story is set. As a general rule this was not how

the ancient world worked,

and

the universal

assumption of modern scholars is that Livy never left Italy: he relied on the transferable motifs or commonplaces (called Wandermotive by German scholars) which, in various combinations, were used to describe most for-

eign peoples and countries.*5 ‘Details and traits originally inherent in one native people were conveyed to another, from Thracians and Scythians to Gauls, and, with enhancement,

from

Gauls

to Germans

- the Germans

were larger and fiercer, with redder hair.'4? As Syme's words imply, the method is true even of an apparently specialist work like the Germanza. A recent scholar, after performing ‘a close reading of the text against the background of the ethnographical tradition', concluded that, 'although the work does contain a few verifiable observations, it is so shaped by ethnographical preconceptions as to be virtually unusable as a historical source'.*? The geography of foreign countries was conceptual rather than actua]l.4? This way of writing and reading is so utterly different from that of today that we may find it difficult to credit, but it nevertheless lasted through the centuries.The Liber Chronicarum ox 'Nuremberg Chronicle' isan illustrated history of the world which was published in 1493:*?

15 For discussion of these commonplaces see Rives (1999) 56-66. 39 Syme (1958) 126. The standard work on this subject is Dauge (1981); see also Balsdon (1979). +7 Rives (1999) 57, reporting on the work of A. A. Lund, ‘Zur Gesamtinterpretation der “Germania” des Tacitus’, ANRW 2.33.2 (1991) 1857-1988; Rives himself,

however,

is less sceptical

59-72. u See Horsfall (1985).

(above,

n. 45).

See also R. F. Thomas

19 Cramsie (2010) 79.

(2009)

2

TACITUS'

BRITAIN

13

if we come to the book expecting a faithful representation of how the world looked at the end of the fifteenth century, we shall be somewhat disappointed. The total number of illustrations is over 1800, almost three times the number of woodcuts. This is due to the large number of repeated images. The view of Mainz, for example, with its solid walls and flowing river, looks exactly like the views of Bologna and Lyon. These last two cities were too far from Nuremberg for people to have much information about their actual appearance. Thus, rather than being topographically correct depictions, they served as symbolic shorthand, indicating to the reader that each place was a city. Few people had travelled or would travel to such far flung places, so it was enough to show that they were ofa certain size and import. Exactly the same could be said, mutatis mutandis, about ethno-geographical writing in the classical world. It may be objected that such considerations do not apply to T.'s Agricola: even if he did not spend the years of his own military tribunate in Britain (above, p. 5), his marriage to A.'s daughter presumably meant that, in theorv at least, he was able to benefit from A.'s own personal experience of the country (below, p. 25). But here the case of Sallust serves as a warning. Sallust not only saw military service in Africa but was appointed as governor of the province of Africa Nova for 46/45 BC. Some years later he came to write his monograph on the Jugurthine War, and, though the modern reader might expect his digression on Africa to have been informed by his personal experience of the country, nowhere does it seem 'to benefit from that fact’, as Syme remarked. ‘A man’s own experience might seem less attractive and convincing than what stood in literary tradition, guaranteed by time and famous names.'?? Like Sallust, T. goes out of his way to give his readersan ethno-geographical account of the place where his narrative 15 set (10—13.1). A.'s penetration to the north of Scotland enables his son-in-law to pronounce more authoritatively on the shape of Britain than earlier writers (10.3), while the circumnavigation of the country confirms the notion (in fact long since established) that Britain is an island (10.4); but these are meagre advances in knowledge, and they are in no way offset by information provided elsewhere. In the main narrative T. mentions for certain only three tribes (Brigantes 17.1, cf. 31.4; Ordovices 18.1—2; Silures 11.2, 17.2);?' three rivers (Bodotria

9" Syme

(1958)

126. 566 also Thomas

(1982). The case of Caesar remains puz-

zling: though Posidonius and other writers may lie behind much of his ethnograph-

ical material, he did himself travel to Gaul, Germany, and Britain. See Krebs (2006) and Aili (2009).

5! The paradosis at 38.2 also mentions the Boresti but the text is likely to be

corrupt (see n.).

14

INTRODUCTION

Figure 1 The Rudge Cup (mid- or late 2nd century AD ?) 23, 25.1, 25.3; Clota 23; Taus 22.1);?* three island names (Mona 18.3; Orcades 10.4; Thule 10.4); one mountain (Graupius 29.2); and one harbour (Trucculensis 38.4).5? Although we know that Roman agrimensores were capable of the kind of field-survey which their name implies, and although we are amazed at the skilful plotting of Hadrian's Wall as it sweeps along the crest of the Whin Sill above Crag Lough on its way from Wallsend to Carlisle, maps were almost certainly unknown:** in practice it was therefore impossible for authors to provide the kind of geographical detail which a modern historian takes for granted, nor, if thev had provided it, would their readers have been any the wiser.> Since such ?? There is some textual uncertainty over the last of these (see n.). 55 The harboury, like the mountain, remains unidentified and there are again doubts about the text (see n.).

?** For surveying 566 Lewis (2001), although he does not mention Hadrian's Wall; for maps see Bertrand (1997) 108-11, Mattern (1999) 41-3 and Brodersen

(2004) and (2012), though the matter is controversial: see e.g. Talbert (2008). On the other hand, the Romans did use itineraries, in which places and the distances between them were represented in a linear fashion (see Dueck (2012) 60-1): a

good example is the inscribed itinerarium from Patara which marked the establishment of the province of Lycia in Asia Minor in AD 43 (Brodersen (2004) 185-6; Dueck

(2012)

116). Also interesting is a series of vessels, including the so-called

'Rudge Cup', which are inscribed with the names of some of the forts on Hadrian's Wall (see e.g. Tomlin (2004) 344-—5 and n. 47), as 'Camboglan (na)' in Figure 1. 55 Editors (e.g. O-R g6ff.) sometimes adduce 'Ptolemy's map' as indicating the kind of cartographical information to which T. would have had access: Claudius

3

IMPERIALISM,

FREEDOM

AND

SERVITUDE

15

information was impossible to acquire and pointless to convey, it is hardly surprising that, whether dealing with the geography of Germany or Britain, ‘Tacitus shows little interest in it.’5°

The past mav be a foreign country but Roman Britain is remarkably familiar today, and not only to those fortunate enough to have been born and brought up within reach of a Roman site.57 Ordnance Survey, the mapping agency for Great Britain, produces a superb historical map and guide entitled ‘Roman Britain'. The map is crowded with Roman settlements of various sizes and kinds, many of them linked by an impressive system of Roman roads. In addition there are temples and shrines, fortresses and camps, signal stations and watch towers, and 'other substantial buildings'. Most of the places are named and identified and, since the Roman is superimposed upon the modern, one can find one's way to almost any of the sites with the greatest of ease. None of this extensive detail 15 owed to T.’s Agricola, whose lack of information is only emphasised thereby. The limitations discussed above, combined with the deplovment of transferable motifs, mean that the work offers little help to those modern readers hoping to glean from it some useful knowledge. ‘For the topography of Roman Britain the Agricola is of little use.'5? 3 IMPERIALISM,

FREEDOM

AND

SERVITUDE

During his first tour of duty in Britain as a young man in the late 50s, A. was seized with the desire for military glory (5. ‘militaris gloriae cupido’). When he returned to Britain twenty vears later as governor in 77, his first act was to massacre almost the entire tribe of the Ordovices in North Wales (18.3 ‘caesa...prope uniuersa gente’). The next vear he followed this up with a policy of general harassment and calculated levels of terror (20.2 ‘nihil interim apud hostes quietum pati...ubi satis terruerat’), and in 79 he crushed the nations as far north as the Tay estuary (22.1 ‘uastatis usque ad Taum ...nationibus'). After a period of consolidation in 8o (23), he meted out the same treatment to the peoples living on the opposite coast of Scotland (24.1 'gentes crebris simul ac prosperis proeliis Ptolemaeus was writing in the second half of the second century AD but 'the earliest preserved maps based on his data ...date from a millennium later' (Brodersen (2012) 107); and there is no indication earlier version, had one existed.

that T. would

have cared to consult an

50 Syme (1958) 126. For discussion of the ancient attitude to Britain and its inhabitants, often influenced by T., see e.g. Walser (1951) 28—-42, Stewart (1995), Braund (1996), Günnewig (1998) 255-307, Clarke (2001) = ORT g7-71. 57 For more than twenty years (1929—50) Eric Birley, the distinguished historian of Rome, its army, and Roman Britain, actually owned the land which incorporates

the site of Vindolanda (see e.g. Birley (2002) 18-20).

55 F-A xxxix. See Map 1.

16

INTRODUCTION

domuit') and had dreams of conquering Ireland (24.3). His exploits in 82 encouraged in his troops the determination of reaching the northernmost point of Britain (27.1), and the dénouement came in 83 at Mons Graupius, where roughly ten thousand of the enemy were massacred, with the loss of only 360 on the Roman side (37.6). Well might Domitian reflect that, after the slaughter of so many thousands, this was truly a great victory which was being celebrated (39.1 'ueram magnamque uictoriam tot milibus hostium caesis ingenti fama celebrari'): A. had achieved the military glory which had been his youthful ambition (39.2 ‘militarem gloriam’), and the prznceps duly rewarded the returning conqueror with the trzumphalia ornamenta and an honorific statue (40.1).

In the twenty-first century the Western world tends to take a dim view of mass slaughter, but there 15 not the slightest indication in any of the above passages that T. was at all troubled bv his father-in-law's genocidal advance across Britain. On the contrary, the positioning of the narrative immediately after the digression on the island's previous history (13.1—17.2) suggests that A. 15 no different from, but merely more successful than, a whole series of previous commanders in Britain as far back as Julius Caesar (above, p. 4). He qualified to be counted amongst Rome's greatest generals, on each of whose monuments, according to Cicero, were

inscribed

the words

‘He

extended

the boundaries

of the

empire' (Rep. 3.24 ‘fines imperii propagauit’). It was no doubt precisely such inscriptions which Augustus restored when in his forum he set up statues of those ‘who had transformed the empire of the Roman people from its smallest to its greatest size' (Suet. Aug 31.5 'qui imperium populi Romani ex minimo maximum reddidissent’). Augustus' own Res Gestae is a hymn to imperialism, paying equal attention to enemy slaughter (e.g. 26.5 *magnae . hostium ...copiae caesae sunt in acie’) and territorial expansion (e.g. 26.1 'omnium prouinciarum populi Romani quibus finitimae fuerunt gentes quae non parerent imperio nostro fines auxi’), while the victory monument which the first princeps erected at La Turbie in France listed by name the more than forty Alpine tribes he had conquered ('gentes Alpinae deuictae’: see Plin. NH 3.136—7, CIL 5.7817). Those who extended the empire had the privilege of enlarging the pomerium of Rome to symbolise their achievement, as T. himself explains when recording that the emperor Claudius performed the ancient ritual after his invasion of Britain (A. 12.23.2 ‘more prisco quo iis qui protulere imperium etiam terminos Vrbis propagare datur’). So successful was Rome in extending the boundaries of her empire that by the second century Bc it was realistic for Romans to be portrayed, or to portray themselves, as ‘masters of the world' — rerum dom?nos in Virgil's famous phrase - and this role was emphasised by the symbolism of the submissive globe which appeared on coins from the late first century Bc

3

IMPERIALISM,

FREEDOM

AND

SERVITUDE

17

onwards.5? Since Romans were the masters, it followed that the conquered peoples of the world were their slaves, and it is well known that Rome's rise to world dominion was accompanied by justifications of imperialism which relied on ancient theories of the relationship between superior masters and inferior slaves. Aristotle in the first book of his Politics had argued at length that some people are natural slaves, that barbarians and slaves are the same by nature, and thatit was in the slaves' own interests to be ruled by a master; and similar arguments were taken up with enthusiasm by Roman writers and thinkers from Cicero down to Augustine.?? At the same time it was recognised that imperialism was not without 115 responsibilities: ‘just as masters were bound to give slaves just treatment, so an imperial power had a duty to care for the ruled'.?' This too finds its most memorable expression in Virgil, in the famous words with which Anchises addresses Aeneas (Aen. 6.851—3):9? 'tu regere imperio populos, Romane, memento (hae tibi erunt artes) pacique imponere morem, parcere subiectis et debellare superbos.' ‘Remember, Roman, to rule nations by means of vour power (these will be vour skills) and to crown peace with civilisation, to spare the humbled and to war down the haughty.’ It 15 striking that the policy attributed to A. in the Agrzcola seems almost to be a paraphrase of these lines. When in AD 78 he deemed that he had caused sufficient terror amongst the inhabitants of Britain, he turned to forgiveness and enticements to peace (20.2): ‘ubi satis terruerat, parcendo rursus inuitamenta pacis ostentare'. Explaining that the winter of the same year was taken up with 'the most salutary policies’ (21.1 'saluberrimis consiliis’), T. proceeds to list the construction of temples, market-places and housing, the introduction of education, and the learning of Latin (21.1-2). A.'sintention, we are told, was 'that men who were scattered and

crude and thus prone to warfare should become pleasurably accustomed to quiet and peace' (21.1 ‘ut homines dispersi ac rudes eoque in bella 59 Virg. Aen. 1.282, cf. Liv. 37.45.8 'dominos orbis terrarum', 38.51.4 'ciuitatem

dominam

orbis terrarum', Ti. Gracch.

ORF 34.13 κύριοι τῆς oikoupévns. Early in the

Annals T. reports that Tiberius began his reign with a recitation of the imperial possessions (1.11.4), and the historian himself provides a comparable inventory under the year 23 (A. 4.4.3—5.4). For the importance of the geographical inventorying of Rome's power 566 in general Nicolet (1991); for the globe see Weinstock (1971) 42ff. 9" See e.g. Cic. Rep. 3.37, Off. 2.26, Aug. Civ. Dei 4.15; Strasburger (1965) and in general Lavan (2013).

?' Brunt (1990) 317.

— "? See also the parallels in Austin ad loc.

18

INTRODUCTION

faciles quieti et otio per uoluptates adsuescerent'). T.'s choice of terminology captures exactly the perceived natural inferiority of the inhabitants of Britain and the superiority of their Roman masters. His father-in-law was both civiliser and conqueror, fulfilling to the letter Anchises' instructions to his son. *

In the biography, A.'s appointment as governor of Britain (9.6) is separated from his arrival in the province (18.1) by the two-fold digression on the country (10-17): we might therefore have expected the digression to describe conditions as they were in AD 77 on the eve of his governorship. This indeed is the perspective of the second, historical, section of the digression (13.1—-17.2), as we have just noted; but the ethnogeographical section, as we have also noted (above, p. 3), is written from the superior perspective of A.'s conquest of the entire island seven years later in AD 84 (10.1 ‘Britanniae situm populosque ...referam ...quia tum primum perdomita est', cf. 10.3—4). It therefore comes as something of a surprise to learn, in the course of this section of the digression, that the Britons are showing significant defiance (11.4 'plus tamen ferociae Britanni praeferunt’) and that, although some Britons have been conquered, others remain unconquered (11.4 'Britannorum olim uictis...; ceteri manent...’). There seems to be a fundamental contradiction between the

perspective promised at 10.1 and that provided at 11.4: in the former passage Britain is a conquered nation, in the latter it is not.®3 How are these two perspectives to be reconciled? The statement at 10.1 seems able to be reconciled with the tenses of 11.4 only ifwe assume that the digression asa whole 15 written from the perspective of T.'s own day, namely, AD 97/98.94 T.'s assertion that by the end of A.'s governorship Britain was completely conquered (10.1 ‘perdomita est’) is echoed and qualified in the preface to his later Hustorzes, where he says ‘perdomita Britannia et statim missa'

(H.

1.2.1). This famous

state-

ment sees Britain metaphorically as an animal which, having been completely tamed, was then released back into the wild.95 Scholars assume that T. is referring to the period when, five vears after A.'s departure from the province, Domitian transferred troops from Britain for operations on the Rhine and Danube. It was ‘a massive withdrawal'.9? In retrospect the years of A.’s governorship represented the high point of the Roman occupation of Britain; Domitian's actions meant that the Romans' territorial gains 93 See also Sailor (2008) 88-9. ?* The possibility that the verbs in 11.4 are historic present seems highly unlikely;

see also below, n. 68.

95 See OLD mitto 2a, cf. 3a.

99 Jones (1992) 132.

3

IMPERIALISM,

FREEDOM

AND

SERVITUDE

19

reverted to the smaller scale which had been imposed on Britain in the years before A. became governor.?7 Some areas of the country (for example, the south east) continued to experience the progress of humanitas which A. had started (21.2: p. 23), whereas the northern territories which A. had overrun were reclaimed by the native inhabitants. Butin the Agricola itself no mention is made of this reversion, presumably because it would have detracted from A.’s achievement; on the contrary, the very insertion of the ethno-geographical digression into the narrative of AD 77 seems designed to distract from the realitv that, more than twenty years later, conditions in Britain were not very different from when A. first arrived on the island as governor.?? *

T. ends the ethno-geographical section of his digression with the epigrammatic assertion that the British provincials are ‘already tamed to obey, but not yet to be slaves' (13.1 ‘iam domiti ut pareant, nondum ut seruiant’). An intermediate stage between freedom and servitude, such as appears to be indicated here, seems borne out by T.’s earlier comparison between Britons and Gauls. Having been pacified since the mid-first century Bc, the Gauls ‘have lost their uzrtus along with their freedom' (11.4 'amissa uirtute pariter ac libertate’), and the same thing 'is happening to those of the Britons who were conquered some time ago' (‘quod Britannorum olim uictis euenit'). On the assumption that euenztis present tense and not past,?? this passage confirms that the Britons who were conquered in the Claudian invasion of AD 43 (‘olim’) are currently engaged in making the transition from freedom to servitude. Two further passages are not incompatible with the notion of an ongoing process of enslavement. When the client king Togidumnus was given charge of certain communities in the late 40s AD, T. comments that it was a long-standing Roman practice to use kings as ‘the instruments of servitude' (14.1 'instrumenta seruitutis’). Writing about the introduction into Britain of porticoes and baths and 9?7 This pattern of occupation is seen clearly in the series of maps 3-5 in Salway (1993) 100-1. 98 There is perhaps a parallel in the ‘blurred temporalities’ of the Dialogus, ‘straddling its dramatic date in the mid-70s and the date of Tacitus' staged report of the conversation to Fabius Justus twenty-five years later' (Whitton (2012) 358). It should be added that the verbs at Agr. 12.1—2, like those at 11.4, are also present tense: one of them

(12.1

'nunc... trahuntur’)

seems clearly to refer to 'the

time of Tacitus' (F.); the others, while compatible with the hypothesis outlined above, could perhaps be 'timeless' or generalised presents (like those used of the island itself), providing characteristics of the inhabitants (e.g. 12.2 'ita singuli pugnant, uniuersi uincuntur') rather than describing them as they actually were at the time. For this kind of present tense see Pinkster (1999) 714-16.

99 See n. ad loc.

20

INTRODUCTION

elegant dining about thirty years later, T. producesa famous epigram about servitude (21.2 'idque apud imperitos humanitas uocabatur, cum pars seruitutis esset').7? If it is right that the ethno-geographical digression is written from the perspective of T.'s own day in AD 97/98, as has just been argued, that process has still not been completed fifty-five years after the Claudian invasion — a relatively short period compared with Gaul. Naturally the Britons themselves are depicted as unaware of such niceties: in the governorship of Suetonius Paulinus thev discuss amongst themselves 'the evils of servitude'

(15.1

‘mala seruitutis’) and in AD 60, under the lead-

ership of Boudica, they attack Camulodunum servitude'

(16.1

‘as being the seat of their

‘ut sedem seruitutis’).

Freedom or servitude is mentioned, on average, once on every page of the Agricola;7' yet this statistic, though telling, is misleading. Nearly 75 per cent of these references are applied to non-Romans, and the majority of them are put into the mouth of the Caledonian chief, Calgacus, in his speech before the battle of Mons Graupius (30-2). His message is simple: freedom is represented by his fellow Caledonians, servitude by the Romans (30.1 libertatis, seruitutis, 30.2 seruienttum, 30.3 libertatis, 31.1 serutturi,

31.2 seruituli, seruitutem, seruorum,

conseruis, famulatu, 31.4

libertatem,

32.1 seruos, 32.3 libertatem, 32.4 seruienttum, cf. 20.3 seruitium). He assures his warrior audience that compliance and self-effacement would not afford them an escape from savage treatment if they were to give in to the Romans (30.3 ‘quorum superbiam frustra per obsequium ac modestiam effugias’), whose references to 'empire' and 'peace' are merely false labels for the atrocities which thev inflict on those over whom they lord it (30.5 'falsis nominibus’).

In 1958 Syme treated Calgacus' speech in a chapter entitled ‘Tacitean opinions', and, although 'criticism has rightly adopted a more cautious approach in recent years',? the chief's eloquence 15 so powerful that it tempts some readers into the belief that T. must sympathise with the man into whose mouth he has put such moving sentiments (see further below, pp. 22—5). Yet this belief is hard to reconcile both with ancient rhetorical practices and with T.'s description of another foreign chief, the Gallic leader Julius Valentinus, in the Histories (4.68.5): ‘meditata oratione cuncta magnis imperiis obiectari solita contumeliasque et inuidiam in populum Romanum effudit’, ‘in a practised speech he poured forth all the customary objections to great empires, as well as insults and resentment 79 For the meaning of pars here see n. ad loc. ' There are roughly ten references to freedom and almost twice as many to servitude; the classic discussion 15 Liebeschuetz (1966), updated in ORT 73-94. In general see Wirszubski (1950) 124-71, Vielberg (1987) 113-34,150-68, Rutledge (2000), Oakley

(200ga)

184-94.

7^* Rutherford (2010) 314, cf. Syme (1958) 528-9.

3

IMPERIALISM,

FREEDOM

AND

SERVITUDE

2]

against the Roman people'. And, just as Valentinus' speech is explained bv his role as 'the fiercest agitator for war'

('acerrimo instinctore belli’),

so the same is true of Calgacus. It is important to bear in mind that the battle of Mons Graupius is presented as the climax of A.'s governorship of Britain. It is easily the longest narrative episode (35.2—38.2), it forms the triumphant conclusion to the British section (above, pp. 2—-3), and it features a comprehensive array of dramatic and literary devices (see ad loc.). Ancient historians regularly enhanced military encounters of such importance by prefacing them with speeches of exhortation from the opposing protagonists, spelling out exactly what was at stake in the forthcoming battle. Calgacus would necessarily expatiate on the notions (however conventional) of freedom and servitude because no subject had

a greater guarantee of motivating the listening Caledonians.7? As Caesar said with reference to the Gauls, 'all men by nature strive for freedom and hate the condition of servitude' (G. 3.10.3). The more eloquent Calgacus' speech, the more critical the consequent battle; and, the more critical the battle, the greater A.'s victory. Calgacus's eloquent but ultimately futile views on Roman imperialism maximise the conqueror's glory. But the Agricola is neither simplistic nor (in our terms) jingoistic. Throughout the biography key words and phrases recur, sometimes strengthening its architecture (above, p. 2), sometimes setting up resonances between characters and episodes that invite the reader to consider carefully the relationships among them. Sometimes the repeating language crosses boundaries. The terminology deployed by Calgacus is similar, and in some cases identical, to that with which T. portrays political life at Rome. When in AD 93 Domitian and his associates burned the biographies of Thrasea Paetus and Helvidius Priscus, the intention was to destroy the freedom of the senate (2.2 ‘libertatem senatus’); it was a time when T.

and his contemporaries saw the ultimate in servitude, just as their ancestors had

seen

the ultimate

in freedom

(2.3).7* After Domitian's death,

Nerva blended what had once been irreconcilable, the principate and liberty (3.1 ‘principatum ac libertatem"), with the result that T. can now look forward to writing a history which will include a record of their former servitude under Domitian (3.3 ‘memoriam prioris seruitutis'). That servitude was characterised by savagery on the part of the princeps (3.2 ‘saeuitia principis’) and by silence from his subjects (3.2 ‘per silentium’), a silence 73 See further the Commentary ad loc. 74 ultimum in libertate is evidently a reference to the civil war which had accompanied the last years of the republic and serves as a useful reminder of the point often made - that T. saw no simple dichotomy between a ‘good and free' republican past and a 'bad and servile' imperial present.

22

INTRODUCTION

which testified eloquently to their submissiveness (2.3 ‘grande patientiae documentum...tacere’).

This similarity of terminology has suggested to some that T.’s domestic and foreign narratives should be seen in terms of each other: "Tacitus' presentation of senatorial seruitus cannot be understood in isolation from the British narrative.'75 The obsequium ac modestia which Calgacus scorns at 30.3 are 'the very values for which Agricola is praised' at 42.4 ('obsequiumque ac modestiam');?? and the quies et otium to which A. accustoms the Britons at 21.1, as we have just seen, constitute the same conditions as those undertaken by A. himself under Nero, the proto-Domitian (6.3 ‘quiete et otio'), and as will be pressed upon him by Domitian's henchmen (42.1 ‘quietem et otium’).77 Finally, since T. is himself alert to distinguish the reality of enslavement from

the labels used to describe it (21.3), ‘it

is...possible to recoup Calgacus even as a surrogate for the authorial persona' 79 Just as biographers may be drawn to particular subjects in an 'automimetic' way (above, p. 9), so readers will tend to read texts in wavs congenial to themselves. T.'s own critical stance has always made him seem more analytical and authoritative than less critical historians.’® Since the 1960s there has been in the West a general antagonism towards political and military authority which, especially amongst intellectuals, has made dissent a strongly preferred mode of discourse. Since few today would regard imperialism or slavery as acceptable, T.'s acknowledged status as a brilliant and authoritative writer is thus regarded as in jeopardy if he is not thought to be likewise opposed to what we see as deplorable about the Roman empire. Putting it more crudely, we want T. to be ‘like us'.?? Yet such a reading goes too far afield from T.’s own explicit statement about the purpose of the biography, and fails to take account of ancient rhetorical decorum. If we are to imagine Calgacus as somehow representing the author himself, we would contradict the explicit statement ofT. in the preface that the biography was a work of pietas towards his father-in-law (3.3). Calgacus' speech is a famous representative of the convention whereby Roman writers put anti-Roman sentiments into the mouths of their barbarian characters (see 30—33.1, intro. n.). On the one hand, rhetorical training enabled these authors to write convincing and emotionally true speeches for all

75 Lavan (2011) 304; his discussion has now been reworked as Lavan (2013) 127—42. 7° Lavan (2011) 304. 77 Whitmarsh (2006) 319-20, Lavan (2011) 305. 7* Whitmarsh (2006) 318. 79 See RICH 202—0. 8¢ The debate over the interpretation of the Aeneid well exemplifies the problem: see now the excellent discussion in Tarrant (2012) 33-7.

3

IMPERIALISM,

FREEDOM

AND

SERVITUDE

23

their characters: Calgacus sounds no less real, or persuasive (to his audience), than does A. And there is a degree - which will be felt differently by different readers — to which the anti-Roman sentiments in these speeches resonate with criticisms, especially of Roman social and imperialist policies, that seem less dependent on rhetoric: in the Agrzcola one might single out the famous remark about luxury (21.2), where the practices that the Britons misunderstand are exactly the same as those that T. and other authors elsewhere identify as problematic in Roman imperial society. The Britons, then, subscribe to practices that T. says are part of their slavery; but so do decadent or thoughtless Romans.?' Yet it seems very unlikely that T. intends his readers to condemn A.'s project of conquest; indeed, he elsewhere appears to regret the absence of foreign conquest (e.g. À. 4.4.3, 4.32.1, cf. 2.61.2). A closer look at the repetition of obsequium ac modestia, on which much weight 15 placed by scholars who see Calgacus as a mouthpiece for T.,?? shows that this echo is more contrastive than confirmatory. Calgacus' point is that, since these qualities will be of no use against the Romans' saeuitza, the alternative 15 to avoid saeu:tza by fighting (30.3); but, when the Britons are persuaded by his arguments and enter battle, thev are annihilated. In other words, Calgacus has offered the Britons only two alternatives: one is saeuitza, the other results in their death (the third possibility, that obsequzum ac modestia might be to their advantage, is not mentioned). In the case of A., on the

other hand, obsequium ac modestia mean that he avoids Domitian's saeuztia and contribute to his designation as a great man (42.4). The repetition of the same phrase underscores Calgacus' misreading of his own situation, and, for the reader, suggests that Calgacus might make more common cause with those preferring a death useless to themselves and to the state (42.4), rather than with A. That Calgacus' views on obsequium ac modestia differ from those of A. and his biographer is a difference one might expect between Roman exconsuls and a Highland chief. Within the biography two value-systems are operating in parallel, and the power which A. exercises in Britain is regarded in a different light from that which Domitian exercises at Rome. Calgacus sees the Britons' enslavement in terms of hard labour (g1.1-2), whereas the Roman view is of temples, market-places and learning Latin (21.1—2). We

today see here the operation

of double

standards, but the

Roman view would have occasioned little surprise in the British at the height of their Empire: one and a half centuries ago few saw any inconsistency between freedom for themselves and virtual enslavement for the *! See for example O'Gorman (1993), now reprinted in ORT 95-118. 82 See Lavan (2011) 304 and n. 31 ('the striking echo has not always received

the attention it deserves’).

24

INTRODUCTION

many peoples of the globe who were subject to them.53 With the words of Rule, Britannia ringing in their ears ('Britons never, never, never shall be slaves’), imperial administrators travelled out to India, where, in much

of the sub-continent, 'British authority resided solely in a single district officer, who spent each day in the saddle, moving from village to village, adjudicating disputes, levving taxes, enforcing the rule of law, and maintaining order across thousands of square miles.' The power of the ruling nation 'was reinforced on a daily basis by ten thousand acts of domination and will intended to inculcate in the Indian people a sense of their own inherent inferiority’.®4 Such inferiority meant that individuals were treated like slaves. When one particular imperial representative went on his travels, *his personal kit alone filled some twenty-nine cases, including two large steel trunks and a container exclusively for hats .. . With such baggage to carry, it was no wonder that eighty-eight porters died of exhaustion on the march.'55 An episode like this strikes us as appalling today and may seem almost incredible, but it more nearly approximates to the world of ancient Rome than do our contemporary sensibilities. No one disputes that T. is a subtle and complex writer or that he is capable of seeing and presenting two sides of a question; and Roman literature often creates pictures of admirable - though fatally flawed — enemy leaders.? But the biography of his father-in-law presents a figure in the best traditions of the Roman state, a consular and general who had exhibited not only the severity required for expanding the empire and taming barbarians but 4150 the humanity required for their cultural development and continued pacification. T. begins his biography by saying that it is a work of pietas (3.3); he ends it with these words (46.2): admiratione te...et laudibus et, si natura suppeditet, similitudine colamus: is uerus honos, ea coniunctissimi cuiusque pietas. Letus...revere you with admiration, praise and, if nature should permit, imitation. This is true honour, this the devotion from everyone

closest to you.

53 There are, of course, exceptions, most famously Kipling, who could to some

extent put himself into the shoes of the colonised, as in 'Gunga Din' or The man who would be king; cf. also Pratt (2008) 77—9 on Mungo Park and ‘the mystique of reciprocity', and Bebbington (2008) 95-6 on Gladstone's humanitarianism. Earlier in the century, the British abolitionist movement was extremely active, passing (znter alia) the Slavery Abolition Act of 1833; but slavery in the territories controlled by the East India Company was not made illegal until 1870. 5* Davis (2011) 41-2.

55 Davis (2011)

57.

*9 Sallust and Livy, T.'s great models, were experts at this, often seeming to prefer the complexity of Rome's enemies (so e.g. Catiline and Jugurtha in Sallust, Hannibal in Livy).

4

THE

AGRICOLA

AS

HISTORY

25

This exhortation loses all point if T.'s admiration and devotion have been directed primarily not at Agricola but at his enemy Calgacus.57 4 THE

AGRICOLA

AS

HISTORY

How does 'history', in the sense of events which took place in the past, become ‘history’, in the sense of an account of events which is readable at a later time? Obviously there needs to be some link between the two ‘histories': perhaps the author himself took part in the events about which he writes (like Thucydides and Xenophon, or Caesar and Velleius); but more usually the author relies on one or more intervening accounts. In the latter case the link may be fairly direct: thus many believe that in the Annals T. relied on the acta senatus, which in turn are believed to be an accurate

record of events that took place in the senate roughly a century before he wrote. Or the link may be more complicated: when writing about the Hannibalic War, Livy relied not onlv on Polybius (who himself relied on Silenus, a member of Hannibal's entourage) but also on other predecessors and was therefore faced with choosing between his sources when they differed from one another. In writing the biography of his father-in-law T. was in a privileged position since his link to most events, including the crucial events in Britain,

was the very individual about whom he was writing. From our modern point of view it is therefore somewhat disappointing that all three occasions on which T. refers to his subject's testimony are relatively trivial. The first relates to a frequent reminiscence about a youthful enthusiasm for philosophy (4.3 ‘memoria teneo solitum ipsum narrare...’); the second relates to an often expressed belief that it would have been easy to conquer Ireland (24.3 'saepe ex eo audiui’); the last relates to a prediction that Trajan would become princeps (44.5 ‘quod ...apud nostras aures ominabatur'). We are free to assume that T. relied on his father-in-law also

for his narrative of the years 77—84; butitis an assumption and there is no actual evidence for it.9? If T. relied on A. for information on the years of the latter's governorship, the resulting account in chapters 18-38 is liable to the various 57 See also Rutherford (2010) 316, 318 ‘The length and eloquence of the concluding encomium resists any kind of ironic reading.' 88 ‘It is worth noting, however, that the work was written several years after the death of its subject, whom Tacitus had not seen for four years before that. Thus, unless the work had been conceived at least ten years before its publication, there would have been no opportunity for Tacitus to interrogate his father-in-law about specific events. It would be a mistake to assume, therefore, that the Agricola is necessarily an account based solely upon first-hand information which can be relied upon unquestioningly and implicitly’ (Hanson (1991) 1742).

26

INTRODUCTION

problems to which oral history, eye-witness accounts and battlefield reports are notoriously prone.® For example, there is a recognised tendency for participants to exaggerate the importance both of the events in which they participated and of their own role in those events. Whether or not A. himself was susceptible to this tendency, the encomiastic nature of ancient biography almost guaranteed that he would be portraved in heroic fashion. One manifestation of this is the way in which T. repeatedly describes his fatherin-law in terms of 'the ideal general', exhibiting the various virtues and qualities which were outlined in military handbooks and elsewhere.9? Another example is the attribution of pioneering exploits to A.;?' these have caused difficulty in recent years, since archaeological investigations have shown that A. seems not necessarily to deserve the priority which on the basis of the Agricola modern scholars had previously taken for granted.?? If T. relied on some

other source or sources, we do not know who

or

what they were; like other ancient historians, he hardly ever mentions the earlier texts on which his works of history depend: the single reference in the Agricola (22.4 ‘apud quosdam ...narrabatur') is not only unhelpful but clearly designed for the purposes of apologia rather than informativeness. Where, then, could T. have found out about this exciting engagement in AD 82, sixteen years before he set pen to paper (25.3—20.2)?

253

Ad manus et arma conuersi Caledoniam incolentes populi, magno paratu, maiore fama (uti mos est de ignotis) oppugnare ultro castellum adorti, metum ut prouocantes addiderant. regrediendumque citra Bodotriam et cedendum potius quam pellerentur ignaui specie prudentium admonebant, cum interim cognoscit hostis pluribus agminibus inrupturos. ac, ne superante numero et peritia locorum circumiretur, diuiso et ipse in tres partes exercitu incessit. quod ubi cognitum hosti, mutato repente consilio uniuersi nonam legionem ut maxime inualidam nocte adgressi, inter somnum ac trepidationem caesis uigilibus inrupere. iamque in ipsis castris pugnabatur, cum Agricola, iter hostium ab exploratoribus edoctus et uestigiis insecutus, uelocissimos equitum peditumque adsultare tergis pugnantium iubet, mox ab uniuersis adici clamorem; et propinqua luce fulsere signa. ita ancipiti malo territi Britanni; et nonanis rediit animus, ac securi pro salute de gloria certabant. ultro quin etiam erupere, et fuit atrox in ipsis portarum angustiis proelium donec pulsi %9 See e.g. RICH 15-22. 9? See General Index s.v. ‘ideal general’. 9! 566 e.g. the sequence at 20.3-22.2. ?* See e.g. Hanson (1991) 1757-77; also McGing (1982) 19-20o, Hanson (1987), Woolliscroft and Hoffmann (2006) 185-93. At the head of this trend was Birley (1946).

4

THE

AGRICOLA

AS

HISTORY

27

hostes, utroque exercitu certante, his ut tulisse opem, illis ne eguisse auxilio uiderentur. quod nisi paludes et siluae fugientes texissent, debellatum illa uictoria foret. Turning to armed combat, the peoples inhabiting Caledonia embarked - with great preparation, and attended (as is the custom with the unknown) by still greater rumour - on an unprovoked assault on a fort, and the fact that they were the challengers had caused additional dread. Cowards in the guise of the sensible were advising retreat to this side of the Bodotria and withdrawal rather than be beaten, when he learned in the meanwhile that the enemy would

be bursting upon

them

in several columns. And,

to prevent

encirclement by superior numbers with experience of the locality, he too divided his army into three sections and advanced. But, when the enemy learned of this, with a sudden change of plan their entire forces mounted a nightattack on the Ninth Legion on the grounds of its being the most ineffective, slaughtering the watches and bursting upon the men as they hovered between sleep and fright. And already there was fighting within the camp itself when Agricola, informed by scouts about the enemy's route and following in its tracks, ordered the swiftest of the cavalry and infantry to assail the rear of the fighters and then a shout should be raised by the entire forces; and in the approaching daylight their standards gleamed. Thus the Britons were terrified by trouble on two fronts; and spirit returned to the Ninths

and,

unconcerned

for their salvation,

they strove

only for

glory. In fact, now taking the offensive, they actually burst out and there was a fierce battle in the very narrows of the gates until the

enemy was beaten, each army striving to be seen, on the one side, to

have brought aid and, on the other, not to have needed help. And, had not marshes and woods protected the fugitives, the conquest would have been completed by that victory.

The details of this account seem derived from earlier accounts of other operations in earlier authors,?? and this is true both of the preliminary manoeuvres: 25.9 oppugnare...castellum adorti Liv. 35.51.8 ‘castellum...oppugnare est adortus', 43.21.4 regrediendumque...et cedendum Sall. H. fr. incert. 4 ‘regressi... cedebant' 93 The following is derived from the Commentary ad loc., where more details and information may be found.

28

INTRODUCTION

25.4 pluribus agminibus inrupturos Liv. 1.14.11 'agmine uno inrumpit', 2.90.14 peritia locorum Sall. | 46.8 ‘peritia locorum’; and

of the

dénouement:

26.1 adici clamorem Liv. 37.41.11 'clamore...adiecto' 26.2 ancipiti malo Sall. C. 29.1 ‘ancipiti malo’, J. 67.2 ultro ... erupere Liv. 23.18.6 ‘ultro erumpentis’ atrox in ipsis portarum angustiis proelium Liv. 10.5.10 portarum', 30.5.10; atrox proeltum is frequently (19x) exclusivelv Livian before T. his ut tulisse opem, illis ne eguisse auxilio uiderentur ‘digni estis qui pauci pluribus opem tuleritis, ipsi ilio egueritis'. quod nisi paludes et siluae fugientes texissent Liv. 3.22.9 exercitus

foret,

ni

fugientes

siluae

texissent'.

paludes

'in angustiis and almost Liv. 7.35.4 nullius aux'deletusque

and

siluae are

esp. combined in Caesar (9x) to designate the landscape of northern Europe. In addition, the warning figures of 25.3 (‘ignaui...admonebant’) are a topos of ancient historiographical writing; the gleaming standards (‘fulsere signa’) are 'a commonplace in the description of legions';?* and the contrast between salus and gloria ('securi pro salute de gloria certabant’) 15 extremely widespread (e.g. Sall. [ 94.5 'pro gloria ...his, illis pro salute certantibus', 114.2 ‘pro salute, non pro gloria certare', Liv. 7.30.7, 22.60.13). ‘Strictly speaking', wrote Paul Fussell in his exceptional book on the Great War, ‘it would seem impossible to write an account of anything without some "literature" leaking in. Probably only a complete illiterate who verv seldom heard narrative of any kind could give an "accurate" account.’9 It is important to recognise, however, that the above similarities with Sallust and Livy cannot be dismissed as 'stylistic touches’: it is not the case that T. (as it were) had in his mind some pre-existing battle scenario (‘whatreally happened’) which he then 'clothed' in the phraseology of his great predecessors. Rather, like other Greek and Latin authors, he had read extensively and had retained a great deal of what he had read; from his reading he knew what battle descriptions looked like and he would

have made

inferences about the nature

of actual battles; and,

when it came to writing his own description, elements from those descriptions were instantly available to him as ready-made 'props', somewhat like ethno-geographical Wandermotive (above, p. 12), on which to base his own 9+ Wellesley on H. 3.18.1.

?5 Fussell (1975)

173.

4

THE

AGRICOLA

AS

HISTORY

29

account. One finds Tacitus’ details summarised in standard handbooks on Roman Britain,?? but the fact is that we cannot tell what is ‘literary’ from what is 'actual'. Itis striking that Caesar features hardly at all amongst the authors from whom T. has borrowed phraseologv in this passage, but in fact the night battle itself is merely part of a larger Caesarian patterning. Just as A.'s seven-year campaign climaxed in the decisive battle of Mons Graupius (29.2—38.2), at which a barbarian chief delivered a speech about freedom and slavery (30-2), so the genuinely Caesarian books of the De bello Gallico describe a seven-year campaign climaxing in the decisive battle of Alesia (7.75—89), at which the barbarian Critognatus delivers a speech about freedom and slavery (7.77.3-16). Just as A.'s first campaign in Britain began late in the season (18.1), so Caesar's first expedition to Britain began late in the season (4.20.1); and, just as Britain was an island too far for Caesar, so Hibernia was an island too far for A. (24.1—3). And, although A.'s night-time arrival at the camp of the depleted and terrified Legio IX exhibits almost no Caesarian phraseology, as we have just seen, the episode took place in the sixth year of campaigning, as did its counterpart in the De bello Gallico when Caesar arrives by night at his depleted and terrified camp (6.41.2—4, cf. 6.39.4). It seems difficult not to conclude that T. has modelled his whole account of A. with Caesar's narrative in mind.?7 We have already observed that the Agricola, though described by its author as a biography of his father-in-law, exhibits elements which are more normally associated with mainstream historiography (p. 3); and it is certainly the case that modern scholars have tended to rely on the Agrz cola for the history of Roman Britain during the years of its subject's governorship. In 1967 Ogilvie and Richmond devoted thirty pages of their commentary to A.’s ‘conquest of Britain'.9? But in the last three decades or so scholars have directed our attention to the literary nature of Roman historical writing, to its very great differences from its modern namesake, and to the problems of using historical texts as 'sources'.?? Indeed so successful has this re-direction been that in some scholarly quarters it is now 9% Frere (1999) 97-8. 97 See Lausberg (1980) 411-16 and Kraus (forthcoming), with references to earlier scholarship; Nutting (1929) suggested that T.’s entire battle narrative of Mons Graupius is modelled on Caesar's account of the battle of Pharsalus (C. 3.85-99). For echoes of Caesar in the text see e.g. 10-13.1n., 19.3n., 26.2n., 35.2—38.2n., 44.3n., and General Index. 95 O-R 46—76 (exactly double the length of T.’s Latin text on the same period). Standard histories of Roman Britain are Salway (1993), Frere (1999), Mattingly (2006).

99 See e.g. Wiseman

(1979), RICH 70-116, Kraus and Woodman

(1997)

1-9.

30

INTRODUCTION

regarded as the current orthodoxy, against which it has become convenient to rail;'?^? but the present analysis of Agr. 25.3—20.2 suggests that it is no less difficult to use T.'s text for the history than for the geography of Roman Britain. His is the history which, as he predicted (46.4), survives; but 115 relationship to ‘what really happened' is by no means straightforward.'?! 5 LANGUAGE

AND

EXPRESSION

Ancient literary theory was greatly concerned with the suitability (aptum, decorum, 16 πρέπον) of an author's expression to his subject matter (e.g. Cic. Orat. 69—71). In the preliminary moves of the battle of Mons Graupius, A.'s actions are described in a perfect periodic sentence (35.4):'?* tum

Agricola

frontem,

simul

superante hostium in

latera

suorum

multitudine

wueritus ne

pugnaretur,

diductis

simul

in

ordinibus,

quamquam porrectior acies futura erat et arcessendas plerique legiones admonebant, promptior in spem et firmus aduersis, dimisso equo pedes ante uexilla constitit. " Then Agricola, the enemies' superior numbers making him afraid that his men might be drawn to fight in the front and on the flanks simultaneously, spaced out the ranks and, although the line would be rather extended and there were numerous warnings that the legions should be summoned, he was more enthusiastic in his expectations and strong in adversity: dismissing his horse, he took up position on foot in front of the banners. The sentence is framed by the subject's name at the start and the main verb denoting his final action at the end. Within this frame a series of subordinate actions is described in a succession of participial or adjectival phrases (alternately ablative absolutes), varied by a subordinate clause, and each element represents a response to its predecessor. The overwhelming enemy numbers (abl. abs.) lead A. to fear a three-pronged attack (nom. part. 4- neclause), and so he spaces out the ranks (abl. abs.);

although this produces a more extended line and some unwelcome advice

(clause), his own confidence (parallel nom. adjs.) leads to his dismissing his horse (abl. abs.) and finally he takes up his stance on foot (main verb).

Co-ordination is achieved within the neclause by anaphora of szmul and '9" For orthodoxy see e.g. Grethlein

Lendon

(2009).

and Krebs

"1 So too the conclusion by Hanson (1991) '?? Noted by e.g. C-L 661.

1778.

(2012)

8; for a protest see

5

LANGUAGE

AND

EXPRESSION

31

within the quamquam-clause by et (with varied word order on either side); and, though etis repeated in the next colon, variation is maintained by the matching of in spem by aduersis. It would be difficult to find a more harmonious adaptation of expression to subject matter; the systematic actions of the ideal general are perfectly mirrored in the controlled development of the Tacitean sentence.'?3 Rather different is the description of A. at the start of AD 78 (20.2): Sed, ubi aestas aduenit, contracto exercitu multus in agmine, laudare modestiam, disiectos coercere; loca castris ipse capere, aestuaria ac

siluas ipse praetemptare; et nihil interim apud hostis quietum pati quominus subitis excursibus popularetur. atque, ubi satis terruerat, ærcmdo

rursus inuitamentá pacis ostentare.

But, when the summer season arrived, after assembling his army he was much in the column, praising discipline and rounding up stragglers. He personally chose the positions for camp, personally reconnoitred estuaries and woods, and in the meanwhile allowed no rest

on the enemy side but pillaged them in sudden raids; and conversely, when he had terrorised them enough, he would be sparing, demonstrating thereby the inducements of peace.

The initial ubzclause, followed by an ablative absolute, perhaps leads one to expect a periodic sentence, but instead we are given three pairs of historic infinitives. The first pair is arranged chiastically to highlight the contrasting actions of laudareand coercere, whereas the second pair is arranged in parallel to underline A.'s parallel initiatives (acc. — 2256 + inf.); the third pair continues the same word order in the main clauses (acc. —^ inf.) but combines it with a chiastic arrangement of subordinate clauses (dependent quominus — introductory ubz). And the very succession of historic infinitives is

a mannerism

imitated

from

Sallust,'?*

to whom

T.

indeed alludes directly throughout (cf.J. 96.3 ‘in agmine. . . multus adesse', ]. 66.1 * nihil... quietum patz , J. 88.2 ‘nzhil... apudillos tutum pat?). It will be noticed too that T. has sharpened the conclusion of his sentence by alliteration (pointing the etymological connection between parcereand pax) and by paronomasia or word play, a device of which he is exceptionally fond (e.g. 13.1 'potest uideri ostendisse posteris, non tradidisse', 29.3 'omnium ciuiatium ues exciuerant’).'® This fondness helps to explain the '^3

For

discussion

'^4

See 5.1n. (noscere), 38.1n.

esp. 172—-88.

of

periodic

sentences

(Britann?).

see

Wilkinson

(1963)

167-88,

!*5 "Iypographical devices (underlining, italics) will be found in the lemmata in the Commentary to indicate various kinds of paronomasia; for more on Tacitean word play see TR 222-6.

32

INTRODUCTION

profusion of aphoristic expressions or sententiae, many of which deploy the same techniques (e.g. 42.4 ‘posse etiam sub malis principibus magnos uiros esse’).'°®

The ideal writer was supposed to adapt his expression to the context and naturally T. has a great variety of sentence structures. Consider 12.1: In pedite robur. quaedam nationes et curru proeliantur. honestior auriga, clientes propugnant. Their hard-core lies in their infantrv; some nations do battle also from a chariot. The driver is the more high-ranking and is defended by his clients. Ogilvie and Richmond maintain that here 'the structure of the sentences conforms to a standard pattern', implying that T. is writing in a peculiarly 'ethnographical' manner appropriate to the ethno-geographical digression; yet their only evidence comprises similar short sentences in T.'s own Germania and Histories. The fact 15 that T. uses short sentences in a variety of different contexts (e.g. 4.2, 5.1—2, 6.2, 8.2, 44.2), while elsewhere he writes with considerable complexity (see esp. 3.1—2, 18.2, §7.4, 40.2).'°7 Sometimes the sentence structure is inextricably associated with the text to which T. is alluding: thus the brevity of 37.2 ("Tum uero patentibus locis grande et atrox spectaculum: sequi, uulnerare, capere’) derives from the allusion to Sallust's description of the battle of Cirta (J. 101.11 'tum spectaculum horribile in campis patentibus: sequi, fugere, occidi, cap?’). In speeches we expect such devices as anaphora (e.g. 33.2 'tot...tot', 33.4 ‘quando...quando’), but T. uses it in narrative too (e.g. 9.4 ‘procul...procul’, 18.4 'hostes qui classem, qui naues, qui mare expectabant', 35.4 above, 40.3 'noctu...noctu', 41.1 ‘absens...absens’). T. is almost as likely to place the main verb at the start of a sentence as at the end (e.g. 3.1 subit, 4.2 arcebat, 4.3 mitigauit, 5.3 intrauitque, 8.1 praeerat, 8.2 habuerunt); sometimes this ‘fronting’ (as it is called) is used to generate a chiasmus (e.g. 8.1 'temperauit Agricola uim suam ardoremque compescuit’), and it should be noted that chiastic arrangements of various types and complication are omnipresent throughout the Agricola.'® One of T.'s most characteristic structures is to continue an otherwise completed sentence by means of an 'appendix ; very often the appendix is expressed or introduced by an ablative absolute, which either adds a further

fact or offers an

explanation

of, or comment

on,

the action

196 For a study of epigrams in Agr. see Sa&el Kos (1 990); note also Damon (2003) 15—16, 302-4, Ash (2007b) 21—4 and (2012) 14, Oakley (2009b) 202-3. '7 In general on T.’s different types of sentence see Oakley (2009b) 203-6. !95 Exs. include 1.2, 2.3, 3.1 (multiple), 3.2, 5.3, 8.3, 13.2, 20.2, 22.4, 29.1, 29.3, 30.4. See General Index s.v. 'chiasmus, chiastic arrangement'.

5

LANGUAGE

AND

EXPRESSION

33

of the main clause. The appended element may be very short (e.g. 9.6 'adiecto pontificatus sacerdotio') or of considerable length (e.g. 16.2 ‘tenentibus...consuleret’) and is a very useful device for giving either men's responses to events or their motives for their actions; it clearly appealed to T. as the ideal vehicle for the cynical psychology which he so often attributes to his characters, and, although its potential is exploited nowhere more fully than in the Annals, itis 4150 extremely common in the Agricola.'*9 'The device 15 used twice in successive sentences, to great effect, when T. is discussing Domitian's reaction to the dying hero (43.3): supremo quidem die momenta ipsa deficientis per dispositos cursores nuntiata constabat, nullo credente sic adcelerari quae tristis audiret. speciem tamen doloris animi uultu prae se tulit, securus iam odii et qui facilius dissimularet gaudium quam metum. It was agreed that at least on the final day the actual crises in his deterioration were announced by means of a relay of couriers, no one believing that he was sad to hear news which was speeded up in this way. Nevertheless he carried before him on his countenance a look of heart-felt pain, being now unconcerned in his hatred and one who dissembled joy more easily than dread. In the second of the above appendixes the co-ordination of securus not with a second adjective but with a relative clause illustrates the stylistic phenomenon known as uariat?o.''? This feature, which is associated particularly with Sallust and Thucydides, becomes perhaps the most distinctive element of T.'s writing and, although there is no doubt that it is exploited most frequently and idiosyncratically in the Annals, it is mistaken to think that it is absent from the Agricola. There are many different forms of Tacitean uariatio but most often it consists either in a rejection of parallelism, as here, or in a deviation from some linguistic or syntactical 'norm'. At its simplest a sequence of place names is varied by the personal (e.g. 10.1 'Germaniae . Hispaniae ... Gallis') or abstract by concrete (e.g. 24.2 ‘per commercia et negotiatores’); more complex is the variation of prepositional phrase by a clause (e.g. 10.1 ‘non in comparationem ...sed quia...’) or of an ablative absolute by a finite verb (e.g. 11.2 'seu durante originis ui seu...positio...dedit’). Correspondingly T. prefers the rare and poetic fidus to the more normal fidelis (14.1n.), exterreo !'^* Exs. include 2.1, 3.2, 7.2, 9.5 (double), 11.4, 13.3, 14.1—-3 (four), 18.6, 19.4, 22.1, 22.3, 23, 20.2, 30.1, 39.1, 41.3. For discussion of the phenomenon in T.’s work see e.g. M-W 23-4, Martin (1981) 221-3, Damon (2003) 16-19, Oakley (2009b)

205-6.

"1 The standard discussion

(2003)

19-20.

15 Sórbom

(1935); see also M-W

24-5, Damon

34

INTRODUCTION

to the more normal simple form (36.3n.); he likes intrinsically rare words, such as dissociabilis (3.1n.), and unprecedented constructions (e.g. uelox-gen. at 19.2) or usages (e.g. erga at 5.3). With his inclination towards inconcinnity and his delight in difference, T. is always a challenge for his readers.''! Several of the examples quoted in this section have involved allusions to Sallust, and itis probably true that Sallustis T.'s principal model even in the early Agrzcola: indeed it has been argued that A. combines the virtues which are attributed to Caesar and Cato by Sallust in his Bellum Catilinae (54).''* Yet we have also seen evidence of T.'s thorough familiarity with the work of Livy (above, pp. 27-8); the influence of Cicero is all-pervasive;''? and it was indeed inevitable that T.’s writing should attest to the vast amount of earlier and contemporary literature which he, like other ancient writers of verse and prose, had read and retained. Sometimes the evidence is so striking and exact that the passages in question seem to deserve being called quotations: thus cruda ac uiridis senectus strikinglv recalls the strange cruda . . . utridisque senectus at Aen. 6.304, while ?ncerta fugae uestigia at 38.2 is an exact rendering of Luc. 8.4. On the other hand there seems nothing exceptional about the Britons' flight from the massacre of Mons Graupius (37.5 ‘longinqua atque auia petiere’), yet it is perhaps more than coincidental that Silius not only combines the same two adjectives (13.92 'auia tunc longinqua placent') but also makes one of them the object of the same verb (4.177-8 ‘auia...| dum petit’): this looks like an allusion to the Punica. If it seems improbable that T. should be echoing a contemporary epicist, his exhortation at 46.2 ('is uerus honos, ea...pietas’) seems taken straight from the Thebaid of Statius (10.711 'haec pietas, hic uerus honos’). It is of course true that the loss of so much Latin literature means that we can rarely be sure just how exclusive these verbal similarities are;''4 and it is also true that a similarity, if accepted as genuine, will raise questions of significance and intention. We must just consider each example on its merits.' ? Allusions to Cato and Xenophon at the start of the biography may involve play with A.'s name

(see 1.1n. Clarorum), while at the conclusion

there is a possible allusion to the viticultural writings of his father which explained the origin of his son's name (see 46.4n. Agricola). T. is alwavs

"1

!3

For many exs. of uariatio see General Index s.v.

?

Lausberg (1980).

See General Index s.v. ‘Cicero, Ciceronian language and expressions'.

''* One must always bear in mind the possibility that a lost or unknown predecessor was the common source of verbal similarities in later texts. ''5 Forsome recent discussion of intertextuality in T. see e.g. Woodman (20092) 1—7, 14 and (2009b) 35-7, Ash (2012) 12.

6

THE

MANUSCRIPTS

35

alert to the significance of names,''® and at 19.1—4 a series of agricultural metaphors has Agricola 'rooting out’, ‘pruning’ and ‘lopping off' as if he were acting out his name (excidere, coercuit and . circumcisis). Metaphors are in fact a major source of T.'s narrative power and are drawn from a wide spectrum, for example, travel (1.1n. magna), disease and medicine (3.1n. tardiora, 29.1n. in luctu), and fire (4.3n. studzum, 8.1n. temperauit). Particularly noteworthy are the concept of ‘living death' at 3.2 (n. pauci) and the succession of images drawn from wild animals and the arena during the battle of Mons Graupius (32.2n. clausos, 33.4n. e latebris, 34.2n. quomodo, 34.3n. ?n quibus, 37.2n. tum uero). T.’s Agricola is a brilliant text, which ranges from the dazzling dialectic of the preface (1—3) to the emotional eulogy of the conclusion (44-6); it incorporates speech both indirect (15) and direct (30—4) and it balances the elements of biography (4—9, 39—43) with those of historiography proper (10-38), pages of ‘an astonishing virtuosity' climaxing in the superlative battle of Mons Graupius.''? Few historians can have made such a spectacular debut.''?

6 THE

MANUSCRIPTS

In his book Landscape and Memory, Simon Schama tells the dramatic story of how in the autumn of 1943 a detachment of the SS arrived at villa Fontedamo, just west of Ancona

on the northern Adriatic coast of Italy,

and conducted a violent search of the villa and then of a palazzo of the Balleani family.'!'? The men were searching for a manuscript known as the Codex Aesinas, Aesis being the ancient Roman name of the neighbouring town of Jesi. Their pursuit of the manuscript, which had been discovered in the Balleani library in 1902, was explained by the fact that it comprises three Latin works (the Bellum Troianum by the so-called 'Dictys of Crete','?? and the Agricola and Germania of Tacitus), the last of which was

regarded by leading Nazis as the foundational text of the German race.'?!

!!9 See e.g. W-M on A. 3.75.1; TR 219-22. "7 C-L 5 'une étonnante virtuosité'. ε18 For a useful survey of T.’s language and expression in the Agricola see Bews

(1987).

''9 Schama (1995) 75-81. :7? "The Bellum Troianum, which occupies fos. 1—51 of the MS, is probably a thirdor fourth-century AD translation (by one L. Septimius) of an earlier work by the pseudonymous 'Dictys'. '?' 566 Krebs (2009) 295-9 and (2011a), esp. 214—44.

36

INTRODUCTION

Nazi interest in the manuscript was well represented by a recent monograph, published in 1943, in which R. Till discussed its readings and characteristics and provided a photographed facsimile of the text.'?? However, the SS failed to find the manuscript, which, after further vicissitudes, was

acquired fifty vears later by the Biblioteca Nazionale in Rome, where it is now Cod. Vitt. Em. 1631. The Agricola occupies folios 52-65 of the manuscript,'*? but the copying is of two radically different dates. The central portion of the text, from 13.1 munia to 40.2 missum (= fos. 56—63), was written in the ninth century in the so-called Carolingian minuscule hand; the two outer portions of text (fos. 52—55 and 64-65) were written in the fifteenth century by the Renaissance scholar Stefano Guarnieri, who also wrote all of the Ger-

mania.'** The different styles of handwriting, six centuries apart, can be seen in the photographs of the manuscript which are appended to Till’s book.'*5 Itis now generally agreed that the four ninth-century pages (technically called a ‘quaternion’) were part of a manuscript which was brought back in 1455 by the Renaissance scholar Enoch of Ascoli from the abbey of Hersfeld in Germany and which, in addition to the three works mentioned, contained Tacitus' Dzalogus and Suetonius' De grammaticis. In order to distinguish the ninth-century portion of the Codex Aesinas from the rest and to acknowledge its probable origin, it has become conventional to use the letter H to designate fos. 56—63, while E 15 used for the remaining folios.'?*? H and E alike exhibit corrections both within the text and in the margins, which it is conventional to designate by superscript c and m respectively (thus e.g. H" or E^). There are two late-fifteenth-century manuscripts in the Vatican which are also cited by editors. A comprises only the Agrzcola; B additionally con-

tains various other works, including the Dzalogus, Germania, Frontinus' De

aquaeductu urbis Romae, Suetonius' De grammaticis and Seneca's Apocolocyntosis. The relationship between A and B and between both of them and the Codex Aesinas is disputed, but it seems likely that they each derive

'?? Till (1943), complete with epigraph from 'Heinrich Himmler, ReichsführerSS'. :?3 The Germania occupies fos. 66—75 of the MS. '*3 The Nazis were therefore wrong to believe that the text of the Germania as it survives today in this manuscript was the source of all other copies of this work. '?5 It should be noted that fo. 69 is a palimpsest which originally continued the text of the Agricola from 40.2 ad Agricolam to 43.4 testamento and that fo. 76 15 now blank but originally contained the remainder of the text of the Agricola. The few traces which can still be read on these folios (see Till (1943) 21—5), although not on their photographic reproductions, are cited in the apparatus to Delz' edition. !?9 "Thisisthe practice adopted by Delz (1983) and Soverini (2004); O-R (1967) and Ogilvie in the OCT (1975) used E and eto designate the ninth- and fifteenthcentury folios respectively.

6

THE

MANUSCRIPTS

37

ultimately from H via a shared lost ancestor.'?7 A fourth manuscript (T) 15 a direct copy of both the ninth- and fifteenth-century portions of the Codex Aesinas and thus has no independent authority.!?? '?7 In the apparatus criticus A and B are cited only very occasionally. The ed;tio princeps of Puteolanus (c. 1475—80) derives independently from the lost ancestor of A and B and therefore in theory has authority equivalent to A and B; in practice, its evidence adds virtually nothing of value and it needs to be cited only for corrections made by Puteolanus himself. 128 While the circulation of the Agricola was restricted to the witnesses mentioned here, many MSS of the Germania, Dialogus, and De grammaticis survive. Quite how the survival of ninth-century leaves only for the Agricola (and ‘Dictys’) is connected with this fact is uncertain; perhaps the Hersfeldensis was split up shortly after its arrival in Rome, and the rest of the MS was lost. For further details of the Agricola’s transmission see O-R 84—qo, Martin (2009) 245-8, Murgia (2012) 16—-17;for early commentaries and editions of Agricola see Ulery (1986) 99-109, 159-04.

DE

CORNELII TACITI VITA IVLII AGRICOLAE

3

L

-

SIGLA codex Hersfeldensis (= folia 56-6g codicis Aesinatis), saec. IX codicis Aesinatis folia 52—55 et 64—65, a Stephano Guarnieri scripta ante a. 1474 codex Vaticanus latinus 3429, saec. XV codex Vaticanus latinus 4498, saec. XV codex Toletanus 49.2, a. 1474 corrector cuiusque codicis lectiones in margine cuiusque codicis

CORNELII TACITI VITA IVLII AGRICOLAE

DE

Clarorum uirorum facta moresque posteris tradere, antiquitus usitatum, ne nostris quidem

temporibus

(quamquam

incuriosa suo-

rum) aetas omisit quotiens magna aliqua ac nobilis uirtus uicit ac supergressa est uitium paruis magnisque ciuitatibus commune: ignorantiam recti et inuidiam. sed apud priores ut agere digna memoratu

pronum

quisque

ingenio

plerique

suam

aut ambitione

magisque

ad

bonae

in

prodendam

tantum

ipsi uitam

aperto

erat,

uirtutis

conscientiae

narrare

ita

celeberrimus

memoriam

sine

gratia

potius morum

quam

pretio ducebatur.

fiduciam

ac

adrogantiam arbitrati sunt, nec id Rutilio et Scauro citra fidem aut obtrectationi

fuit: adeo

uirtutes iisdem

temporibus

optime

aesti-

mantur quibus facillime gignuntur. at nunc narraturo mihi uitam

defuncti hominis uenia opus fuit quam non petissem incusaturus:

tam saeua et infesta uirtutibus tempora. legimus, cum Aruleno Rustico

Paetus

Thrasea,

dati

essent,

capitale

Herennio fuisse,

Senecioni

neque

in ipsos

Priscus modo

Heluidius auctores

lau-

sed

in

libros quoque eorum saeuitum, delegato triumuiris ministerio ut monumenta clarissimorum ingeniorum in comitio ac foro urerentur. scilicet illo igne uocem populi Romani et libertatem senatus et conscientiam

generis humani

insuper

professoribus

sapientiae

ium acta, ne quid usquam grande

patientiae

aboleri arbitrabantur, expulsis

atque

honestum

documentum;

omni

bona

arte

in

exil-

occurreret. dedimus profecto

et, sicut

uetus

aetas

uidit

quid

ultimum in libertate esset, ita nos quid in seruitute, adempto per inquisitiones etiam loquendi audiendique commercio. memoriam

quoque ipsam cum uoce perdidissemus, si tam in nostra potestate esset obliuisci quam tacere. Nunc

demum

redit animus;

et, quamquam

primo

statim beatis-

simi saeculi ortu Nerua Caesar res olim dissociabiles miscuerit principatum ac libertatem - augeatque cotidie felicitatem temporum Nerua Traianus, nec spem modo ac uotum securitas

1.4 incusaturus tam saeua E, interpunxit Wex 2.1 legimus E: lugemus Courtney: legimus Leeman: uidimus Lipsius: meminimus Peerlkamp: fleuimus Baehrens: scimus Watt 3.1 rediit Spengel et E: set ed. Bipont.: sed Croll 41

42

CORNELII

publica,

sed

tamen

ipsius

infirmitatis

uoti

TACITI

fiduciam

humanae

ac

robur

tardiora

aetas

sunt

suaserit,

remedia

natura

quam

mala;

et, ut corpora nostra lente augescunt, cito extinguuntur, sic inge-

nia studiaque oppresseris facilius quam etiam

amatur.

ipsius

quid

spatium,

inertiae

dulcedo,

si per

multi

quindecim

fortuitis

reuocaueris: subit quippe

et inuisa

casibus,

annos,

primo

desidia

grande

promptissimus

postremo

mortalis

quisque

principis interciderunt, pauci et (ut ita dixerim)

non

rum

€ media

sed etiam

annis,

quibus

nostri superstites sumus, iuuenes

ad senectutem,

exemptis senes

prope

aeui

saeuitia

modo

alio-

uita tot

ad ipsos

exac-

tae aetatis terminos per silentium uenimus? non tamen pigebit uel incondita ac rudi uoce memoriam prioris seruitutis ac testimonium praesentium bonorum

composuisse. hic interim liber, honori Agri-

colae soceri mei destinatus, professione pietatis aut laudatus erit

aut excusatus.

Gnaeus

Iulius Agricola,

nia ortus,

utrumque

uetere

auum

et inlustri Foroiuliensium

procuratorem

Caesarum

colo-

habuit,

quae

equestris nobilitas est. pater illi Iulius Graecinus senatorii ordinis, studio eloquentiae sapientiaeque notus, iisque ipsis uirtutibus iram Gai Caesaris meritus: namque Marcum Silanum accusare ius-

sus

et,

quia

abnuerat,

interfectus

est.

mater

Iulia

Procilla

fuit,

ab illecebris peccantium, praeter ipsius bonam

inte-

rarae castitatis. in huius sinu indulgentiaque educatus per omnem honestarum artium cultum pueritiam adulescentiamque transegit. arcebat eum

gramque naturam, quod statim paruulus sedem ac magistram studiorum Massiliam habuit, locum Graeca comitate et prouinciali parsimonia mixtum ac bene compositum. memoria teneo solitum ipsum

narrare

ultra quam

se prima

concessum

tia matris incensum

sublime

et erectum

in iuuenta Romano

ac flagrantem

ingenium

studium

ac senatori,

animum

philosophiae hausisse,

pulchritudinem

acrius,

ni pruden-

coercuisset. scilicet ac speciem

mag-

nae excelsaeque gloriae uehementius quam caute appetebat. mox

mitigauit ratio et aetas, retinuitque (quod est difficillimum) ex sapientia modum. Prima

castrorum

diligenti ac moderato

rudimenta

aetas suaserit Woodman: assumpserit £

ut sic Wolfflin

in

Britannia

Suetonio

duci, adprobauit, electus quem 3.2

Paulino,

contubernio

ut ita Urlichs (post Rhenanum):

4.3 ultra«que» quam Lipsius

uti E:

DE

VITA

IVLII

AGRICOLAE

43

aestimaret. nec Agricola licenter, more iuuenum qui militiam in lasciuiaam uertunt, neque segniter ad uoluptates et commeatus titulum nosci in

tribunatus

exercitui,

lactationem,

ius et intentus

et inscitiam

discere nihil

agere.

ob

rettulit;

a peritis,

formidinem

non

sane

sed

sequi

noscere

optimos,

recusare,

prouinciam,

nihil

adpetere

simulque

alias exercitatior

et anx-

magisque

in

ambiguo Britannia fuit: trucidati ueterani, incensae coloniae, intercepti

exercitus;

tum

de

salute,

mox

de

uictoria

certauere.

cuncta, etsi consiliis ductuque alterius agebantur ac summa et recuperatae

prouinciae

et stimulos addidere cupido,

ingrata

gloria in ducem

iuueni,

temporibus

intrauitque quibus

cessit, artem

animum

sinistra erga

quae rerum

et usum

militaris gloriae eminentes

inter-

pretatio nec minus periculum ex magna fama quam ex mala.

Hinc ad capessendos magistratus in urbem degressus Domitiam

Decidianam,

splendidis

natalibus

ortam,

sibi

iunxit;

idque

mat

rimonium ad maiora nitenti decus ac robur fuit. uixeruntque mira concordia, per mutuam caritatem et inuicem se antepo-

nendo, nisi quod in bona uxore tanto maior laus quanto in mala plus culpae est. sors quaesturae prouinciam Asiam, proconsulem Saluium Titianum dedit, quorum neutro corruptus est, quamquam et prouincia diues ac parata peccantibus, et proconsul in omnem auiditatem pronus quantalibet facilitate redempturus esset mutuam dissimulationem mali. auctus est ibi filia, in subsidium simul ac solacium: nam filium ante sublatum breui amisit. mox inter quaesturam ac tribunatum plebis atque ipsum etiam tribunatus annum

quiete

et otio

transiit, gnarus

sub

Nerone

temporum,

quibus inertia pro sapientia fuit. idem praeturae tenor et silentium: nec enim iurisdictio obuenerat. ludos et inania honoris medio atque

abundantiae

duxit,

uti

longe

a luxuria,

ita famae

propior. tum electus a Galba ad dona templorum recognoscenda diligentissima conquisitione fecit ne cuius alterius sacrilegium res publica quam Neronis sensisset.

Sequens annus graui uulnere animum domumque eius adflixit.

nam classis Othoniana licenter uaga pars est) hostiliter populatur, matrem 5.2 excitatior Buchner filia E"AB: nactus...filiam B: transit EA 6.4 tenor 7.1 Intimilios Gudeman: in

dum Intimilios (Liguriae Agricolae in praediis suis

intercepti Puteolanus: intersepti E 6.2 auctus est ibi E 6.3 inter«uallum inter> Koestermann transiit Rhenanus: certior Ε: al alia 6.5 fecit Heinsius templo E

τ

rationis

44

CORNELII

TACITI

interfecit, praediaque ipsa et magnam patrimonii partem diripuit, quae causa caedis fuerat. igitur ad sollemnia pietatis profectus Agricola,

nuntio

adfectati

a Vespasiano

imperii

deprehensus

ac

sta-

tim in partes transgressus est. initia principatus ac statum urbis Mucianus

regebat, iuuene

admodum

Domitiano

et ex paterna for-

tuna tantum licentiam usurpante. is missum ad dilectus agendos

Agricolam integreque ac strenue uersatum uicesimae legioni tarde ad

sacramentum

agere narrabatur: formidolosa

erat,

transgressae

quippe

nec

praeposuit,

ubi

legatis quoque

legatus

praetorius

decessor

seditiose

consularibus nimia ac ad

cohibendum

potens,

incertum suo an militum ingenio. ita successor simul et ultor electus rarissima moderatione maluit uideri inuenisse bonos quam fecisse. Praeerat tunc Britanniae Vettius Bolanus, placidius quam

feroci

prouincia dignum est. temperauit Agricola uim suam ardoremque compescuit ne incresceret, peritus obsequii eruditusque utilia hon-

estis miscere.

breui

deinde

alem accepit: habuerunt

Cerialis labores modo bat: saepe

Britannia

consularem

uirtutes spatium

Petilium

exemplorum,

Ceri-

sed primo

et discrimina, mox et gloriam communica-

parti exercitus

in experimentum,

aliquando

maioribus

copiis ex euentu praefecit. nec Agricola umquam in suam famam gestis exultauit;

ad auctorem

ac ducem

ut minister fortunam

ref-

erebat. ita uirtute in obsequendo, uerecundia in praedicando extra

inuidiam nec extra gloriam erat. Reuertentem

patricios

ab

adsciuit;

splendidae

ac

inprimis

legatione

legionis

deinde

prouinciae

dignitatis

diuus

Vespasianus

Aquitaniae

administratione

ac spe

inter

praeposuit, consulatus,

cui destinarat. credunt plerique militaribus ingeniis subtilitatem deesse quia castrensis iurisdictio secura et obtusior ac plura manu

agens

calliditatem

fori non

exerceat; Agricola naturali

prudentia,

quamuis inter togatos, facile iusteque agebat. iam uero tempora

curarum remissionumque diuisa: ubi conuentus ac iudicia poscerent, grauis, intentus, seuerus, et saepius misericors; ubi officio satis factum,

nulla

ultra

potestatis

persona;

[tristitiam

et adrogantiam

et auaritiam exuerat] nec illi (quod est rarissimum) aut facilitas auctoritatem aut seueritas amorem deminuit. integritatem atque 7.3 ubi...narrabatur secl. Wex

8.1

quii £ (corr. in obsequi): obsequi AB tiam ... exuerat secl. Peerlkamp

Bolanus B: Vol- EA

(item 16.5 bis)

9.1 legionis secl. Koestermann

obse-

Q.3 tristi-

abstinentiam

famam

in

DE

VITA

IVLII

tanto

uiro

AGRICOLAE

referre

iniuria

45

uirtutum

fuerit.

ne

quidem, cui saepe etiam boni indulgent, ostentanda uir-

tute aut per artem quaesiuit. procul ab aemulatione aduersus collearbitrabatur. minus triennium in ea lega-

tione detentus ac statim ad spem consulatus reuocatus est, comi-

tante opinione sermonibus,

aliquando

Britanniam

sed

quia

ei prouinciam

par

uidebatur.

et eligit: consul

egregiae

despondit ac post consulatum

dari, nullis in hoc ipsius

haud

tum

Britanniae

situm

populosque

in comparationem

errat

fama;

iuueni

mihi

collocauit, et statim Britanniae prae-

positus est, adiecto pontificatus sacerdotio. non

semper

spei filiam

Q

rium et atteri sordidum

τ

gas, procul a contentione aduersus procuratores, et uincere inglo-

multis

scriptoribus

memoratos

10

curae ingeniiue referam, sed quia tum pri-

mum perdomita est. ita quae priores nondum comperta eloquentia percoluere rerum fide tradentur. Britannia,

insularum

quas

Romana

notitia

complectitur

maxima, spatio ac caelo in orientem Germaniae, in occidentem Hispaniae obtenditur. Gallis in meridiem etiam inspicitur; septentrionalia eius, nullis contra terris, uasto atque aperto mari pulsantur.

formam

totius

Britanniae

Liuius

ueterum,

Fabius

citra

Caledoniam,

unde

Rusticus recentium eloquentissimi auctores oblongo scutulo uel adsimulauere.

spatium cuneum

procurrentium extremo iam litore terrarum tenuatur. hanc oram nouissimi maris tunc

Romana

classis

et in uniuersum

fama;

et

est

sed

circumuecta

ea

transgressis

insulam

ac simul incognitas ad id tempus inuenit

et hiems

domuitque.

adpetebat.

dispecta

sed

facies

esse

insulas

est et Thule,

mare

pigrum

immensum

et enorme

Britanniam

adfirmauit,

uelut in primum

(quas Orcadas quia

uocant)

hactenus

et graue

iussum

remigantibus

perhibent ne uentis quidem proinde attolli, credo quod rariores terrae montesque, causa ac materia tempestatum, et profunda moles continui maris tardius impellitur. naturam Oceani atque aestus neque addiderim,

quaerere nusquam

huius latius

operis

est, ac multi

dominari

mare,

rettulere.

multum

unum

fluminum

huc atque illuc ferre, nec litore tenus adcrescere aut resorberi, sed 10.3 oblongo scutulo Lacey: -gae -lae EA: -gae scupulae B fama; sed transgressis Peerlkamp: fama est transgressis unde et uniuersis fama sed (uerba unde ...sed perperam scripta indicauit librarius) Ε: fama est transgressis sed AB 10.5 proinde £: perinde Grotzus

Q

bipenni

46

CORNELII

TACITI

influere penitus atque ambire et iugis etiam ac montibus inseri uelut in suo. 11

Ceterum Britanniam qui mortales initio coluerint, indigenae an

aduecti,

ut inter barbaros

uarii atque

parum

ex eo argumenta.

compertum.

namque

habitus

corporum

rutilae Caledoniam

habi-

tantium comae, magni artus Germanicam

originem

adseuerant;

Silurum

crines

posita

colorati

tra Hispania fidem

faciunt.

uultus,

Hiberos

torti

plerumque

ueteres

traiecisse

easque

Gallis

et similes

sunt,

proximi

et

sedes seu

con-

occupasse

durante

orig-

inis ui seu procurrentibus in diuersa terris positio caeli corporibus habitum dedit. in uniuersum tamen aestimanti Gallos uicinam insulam occupasse credibile est. eorum sacra deprehendas

superstitionum

persuasionem;

sermo

haud

multum

diuer-

sus; in deposcendis periculis eadem audacia et, ubi aduenere, in detrectandis eadem formido. plus tamen ferociae Britanni praeferunt, ut quos nondum longa pax emollierit. nam Gallos quoque in

bellis floruisse accepimus; mox segnitia cum otio intrauit, amissa uirtute

pariter ac libertate.

quod

Britannorum

olim

uictis euenit;

ceteri manent quales Galli fuerunt. 12

In pedite robur. quaedam nationes et curru proeliantur. honestior auriga, clientes propugnant. olim regibus parebant,

nunc

per

principes

factionibus

et studiis

trahuntur.

aliud aduersus ualidissimas gentes pro nobis utilius quam

(nec

quod

in commune non consulunt.) rarus duabus tribusue ciuitatibus ad propulsandum commune periculum conuentus: ita singuli pugnant, uniuersi uincuntur. Caelum

crebris

imbribus

ac

nebulis

foedum;

asperitas

frigo-

rum abest. dierum spatia ultra nostri orbis mensuram; nox clara et extrema Britanniae

parte breuis, ut finem

atque initium

lucis

exiguo discrimine internoscas. quod si nubes non officiant, aspici per noctem solis fulgorem, nec occidere et exsurgere sed transire

adfirmant. scilicet extrema et plana terrarum humili umbra non erigunt tenebras, infraque caelum et sidera nox cadit. Solum praeter oleam uitemque et cetera calidioribus terris oriri sueta patiens frugum

fecundum

«que»

; tarde mitescunt,

cito

11.2 11 Rhenanus: usu E 11.3 «ac» s- persuasionem Glick: s- persuasione £E 12.1 trahuntur Zeinsius 12.3—13.1 Caelum...seruiant wel 12.3-6 Caelum...auaritiam fost caput 10 uel 11 uarie transpos. Reifferscheid, Wex, Schiitz, Baehrens 12.3 dierum Peerlkamp 12.5 fecundum Hedicke: fecundum E"AB: pecudumque E

DE

VITA

IVLII

AGRICOLAE

47

proueniunt. eademque utriusque rei causa, multus umor terrarum caelique. fert Britannia aurum et argentum et alia metalla, pretium uictoriae. gignit et Oceanus margarita, sed subfusca ac liuentia. quidam artem abesse legentibus arbitrantur: nam in rubro mari

uiua ac spirantia saxis auelli, in Britannia, prout expulsa sint col-

ligi. ego facilius crediderim naturam margaritis deesse quam nobis

auaritiam. munia

ipsi Britanni

impigre

dilectum

obeunt,

ac

si iniuriae

domiti ut pareant, nondum

Igitur primus omnium

tributa

absint:

et iniuncta

has aegre

imperii

tolerant,

iam

13

ut seruiant.

Romanorum

Britanniam ingressus, quamquam

diuus Iulius cum

exercitu

prospera pugna terruerit incolas

ac litore potitus sit, potest uideri ostendisse posteris, non tradidisse.

mox

bella ciuilia et in rem

publicam

uersa principum

arma,

ac

longa obliuio Britanniae etiam in pace: consilium id diuus Augustus

uocabat,

intranda

Tiberius

Britannia

praeceptum.

satis

constat,

agitasse

ni

uelox

Gaium ingenio

Caesarem mobili

de

paeni-

tentiae, et ingentes aduersus Germaniam conatus frustra fuissent. diuus Claudius auctor tanti operis, transuectis legionibus auxili-

isque et adsumpto

in partem rerum Vespasiano, quod initium uen-

turae mox fortunae fuit: domitae gentes, capti reges et monstratus fatis Vespasianus.

Consularium primus Aulus Plautius praepositus ac subinde Osto-

rius Scapula, uterque bello egregius; redactaque paulatim in formam prouinciae proxima pars Britanniae, addita insuper ueteranorum colonia. quaedam ciuitates Togidumno regi donatae (is ad

14

nostram usque memoriam fidissimus mansit), uetere ac iam pridem recepta populi Romani consuetudine ut haberet instrumenta

seruitutis et reges. mox Didius Gallus parta a prioribus continuit,

paucis admodum castellis in ulteriora promotis per quae fama aucti officii quaereretur.

Didium

Veranius

excepit,

isque

intra annum

extinctus est. Suetonius

hinc

Paulinus

biennio

prosperas

nationibus firmatisque praesidiis; quorum

ut uires rebellibus ministrantem

res habuit,

subactis

fiducia Monam

insulam

adgressus terga occasioni pate-

fecit. namque absentia legati remoto metu Britanni agitare inter se mala seruitutis, conferre iniurias et interpretando accendere:

13.1

munia

77

(qui

igitur secl. Baehrens

ab

hoc

uerbo

incipit)

et (ut. uid.:

expunctum)

Ε:

nihil

munera

AB

13.3 auctor tanti Bezzenberger. auctoritate H: auctor tan-

dem Frohlich: alii alia 14.1 togiduno Rhenanus: ut uetere ... haberet H

A™: cogidumnoH

uetere ...ut haberet

15

48

CORNELII

TACITI

profici patientia nisi ut grauiora tamquam imperentur;

singulos sibi olim

ex facili tolerantibus

reges fuisse, nunc

binos

imponi,

e

quibus legatus in sanguinem, procurator in bona saeuiret; aeque discordiam praepositorum, aeque concordiam subiectis exitiosam:

alterius manus centuriones, alterius seruos uim et contumelias mis-

cere; nihil iam cupiditati, nihil libidini exceptum: in proelio fortiorem esse qui spoliet; nunc ab ignauis plerumque et imbellibus

eripi domos, tum

pro

abstrahi liberos, iniungi dilectus, tamquam

patria

nescientibus;

51 et se Britanni numerent!

quantulum

enim

sic Germanias

mori

transisse

excussisse

tan-

militum,

iugum

- et

flumine, non Oceano defendi; sibi patriam, coniuges, parentes, illis auaritiam

et luxuriam

causas belli esse; recessuros,

ut diuus Iulius

recessisset, modo uirtutem maiorum suorum aemularentur; neue proelii unius aut alterius euentu pauescerent: plus impetus felicibus, maiorem constantiam penes miseros esse. iam Britannorum etiam

tum

deos

misereri,

exercitum

qui Romanum

ducem

absentem,

in alia insula detinerent; iam ipsos

qui relega-

(quod difficil-

limum fuerit) deliberare; porro in eiusmodi consiliis periculosius

esse deprehendi quam audere. 16

His atque talibus inuicem instincti, Boudicca generis regii fem-

ina duce (neque enim sexum in imperiis discernunt) sumpsere uni-

uersi bellum; ac sparsos per castella milites consectati, expugnatis

praesidiis ipsam coloniam inuasere ut sedem seruitutis, nec ullum

in barbaris ingeniis saeuitiae genus omisit ira et uictoria. quod nisi Paulinus cognito prouinciae motu propere subuenisset, amissa Britannia foret; quam

unius proelii fortuna ueteri patientiae restituit,

tenentibus arma plerisque quos conscientia defectionis et proprius

ex legato timor agitabat ne (quamquam egregius cetera) adroganter in deditos et ut suae cuiusque iniuriae ultor durius consuleret.

Missus igitur Petronius Turpilianus tamquam exorabilior et delictis hostium nouus eoque paenitentiae mitior, compositis pri-

oribus nihil ultra ausus Trebellio Maximo prouinciam tradidit. Tre-

bellius segnior et nullis castrorum experimentis, comitate quadam

curandi prouinciam tenuit. didicere iam barbari quoque ignoscere uitiis iustam

blandientibus, segnitiae

et

interuentus

excusationem;

sed

ciuilium

armorum

discordia

praebuit

laboratum,

cum

15.2 manus Α΄ : manum HAB: in manu Watt: alii alia 15.3 εἴ se H: sese H'AB 15.5 in alia insula exercitum H, transpos. Woodman 16.1 Boudicca Haase: uoadicca H: bouid|icta H™ ingeniis om. AB 16.2 cuiusque Wex: eiusque H

DE

VITA

IVLII

AGRICOLAE

49

adsuetus expeditionibus miles otio lasciuiret. Trebellius, fuga ac latebris uitata exercitus ira, indecorus atque humilis precario mox ac

uelut

pacta

exercitus

licentia

et

ducis

salute

sedi-

tio sine sanguine stetit. nec Vettius Bolanus, manentibus adhuc ciuilibus bellis, agitauit Britanniam

disciplina:

eadem

τ

praefuit,

inertia erga

hostes, similis petulantia castrorum, nisi quod innocens Bolanus et nullis delictis inuisus caritatem ubi cum duces,

parauerat loco auctoritatis. sed,

cetero orbe Vespasianus et Britanniam recuperauit, magni egregii

exercitus,

minuta

hostium

spes.

et

terrorem

17

sta-

üm intulit Petilius Cerialis, Brigantum ciuitatem (quae numerosissima prouinciae totius perhibetur) adgressus: multa proelia, et aliquando

non

incruenta; magnamque

Brigantum

partem

aut uic-

toria amplexus est aut bello. et Cerialis quidem alterius successoris

curam famamque obruisset; subiit sustinuitque molem Iulius Fron-

tinus, uir magnus (quantum licebat), ualidamque et pugnacem Silurum gentem armis subegit, super uirtutem hostium locorum

quoque difficultates eluctatus. Hunc tate

Britanniae

transgressus

statum,

has

Agricola

bellorum

inuenit,

cum

uices et

media

milites

iam

uelut

aes-

omissa

expeditione ad securitatem et hostes ad occasionem uerterentur. Ordouicum ciuitas haud multo ante aduentum eius alam in finibus suis

agentem

prouincia.

prope

et quibus

uniuersam bellum

ac recentis legati animum

obtriuerat,

uolentibus

illius anni

quies

initio

probare

erecta

exemplum

opperiri, cum Agricola - quamquam

transuecta aestas, sparsi per prouinciam militem

erat,

eoque

numeri, praesumpta apud

tarda et contraria bellum

incohaturo,

plerisque custodiri suspecta potius uidebatur - ire obuiam

et

discri-

mini statuit. contractisque legionum uexillis et modica auxiliorum manu, quia in aequum agmen,

caesaque

quo

degredi Ordouices non audebant, ipse ante

ceteris par animus

prope

ac, prout prima

simili periculo

uniuersa gente, non

cessissent,

terrorem

esset, erexit aciem.

ignarus instandum

ceteris fore, Monam

famae

insulam

(cuius possessione reuocatum Paulinum rebellione totius Britanniae supra memoraui) redigere in potestatem animo intendit. sed, ut in subitis consiliis, naues deerant; ratio et constantia ducis transuexit.

depositis

omnibus

sarcinis

lectissimos

auxiliarium,

16.4 εἴ ducissalute seditio Woodman: ducis salute et seditio 77 (et secl. John) subiit sustinuitque Π΄: sustinuitque HAB 18.3 «a» cuius ed. Bipont.

quibus

17.2

18

50

CORNELII

TACITI

nota uada et patrius nandi usus quo simul seque et arma et equos regunt,

ita repente

inmisit

Q

qui nauis, qui mare

»

hostes

— qui

cola, quippe

cui ingredienti prouinciam, et officiorum

ambitum

classem,

aut inuictum

diderint sic ad bellum uenientibus. Ita petita pace ac dedita insula clarus ac magnus ostentationem

quod

haberi Agri-

tempus

transigunt,

cre-

alii per

labor et pericu-

lum placuisset. nec Agricola prosperitate rerum in uanitatem usus expeditionem aut uictoriam uocabat uictos continuisse. ne laureatis

quidem

famam

19

ut obstupefacti

expectabant - nihil arduum

gesta

prosecutus

est, sed

ipsa

dissimulatione

auxit, aestimantibus quanta futuri spe tam magna

Ceterum

aliena

animorum

experimenta

prouinciae

parum

prudens,

simulque

armis

51 iniuriae

profici

famae

tacuisset.

doctus

per

sequeren-

tur, causas bellorum statuit excidere. a se suisque orsus primum

domum

quam

suam

coercuit,

prouinciam

rei, non

studiis

quod

regere.

priuatis

plerisque

haud

minus

arduum

nihil per libertos seruosque

nec

ex

commendatione

est

publicae

aut precibus

centurionem militesue adscire, sed optimum quemque fidissimum putare. omnia scire, non omnia exsequi. paruis peccatis ueniam,

magnis seueritatem commodare. nec poena semper sed saepius paenitentia contentus esse: officiis et administrationibus potius non

peccaturos

praeponere,

frumenti et tributorum

quam

exactionem

damnare

cum

peccassent.

aequalitate munerum

mollire,

circumcisis quae in quaestum reperta ipso tributo grauius tolerabantur. namque per ludibrium adsidere clausis horreis et emere

ultro frumenta ac luere pretio cogebantur. diuortia itinerum et longinquitas 20

regionum

indicebatur,

ut ciuitates proximis

hibernis

in remota et auia deferrent, donec quod omnibus in promptu erat paucis lucrosum fieret. haec primo statim anno comprimendo egregiam famam paci circumdedit, quae uel incuria uel intolerantia priorum haud minus quam bellum timebatur. Sed,

ubi

aestas

aduenit,

contracto

exercitu

multus

in agmine,

laudare modestiam, disiectos coercere; loca castris ipse capere, aestuaria

ac

quietum

siluas

ipse

praetemptare;

pati quominus

et nihil

subitis excursibus

18.4 inuium Weidner 19.2 primum £: -am H iam Puteolanus): milites scire H (ne suprascr. H*)

Ritter: proximis Bezzenberger entur Dahl

interim

apud

popularetur.

hostes

atque, ubi

militesue ascire Wex (ascire 19.4 proximis H: proximae

20.2 popularetur Π: popularet Β: popular-

DE

VITA

IVLII

AGRICOLAE

51

satis terruerat, parcendo rursus inuitamenta pacis ostentare. quibus

rebus multae ciuitates quae in illum diem ex aequo egerant, datis obsidibus iram posuere et praesidiis castellisque circumdatae, tanta

ratione curaque transierit.

Sequens

homines

ut nulla fante Britanniae

hiems

dispersi

saluberrimis ac rudes

consiliis

eoque

noua

parsj inlacessita

absumpta.

in bella faciles

namque quieti

ut

21

et otio

per uoluptates adsuescerent, hortari priuatim, adiuuare publice ut templa,

fora,

domos

extruerent,

laudando

promptos,

castigando

segnes: ita honor et aemulatio pro necessitate erat. iam uero principum filios liberalibus artibus erudire et ingenia Britannorum studiis Gallorum anteferre, ut qui modo

linguam Romanam

abnue-

bant eloquentiam concupiscerent. inde etiam habitus nostri honor et frequens toga; paulatimque discessum ad delenimenta uitiorum, porticus et balinea et conuiuiorum elegantiam. idque apud imper-

itos humanitas uocabatur, cum pars seruitutis esset. Tertius expeditionum annus nouas gentes aperuit, uastatis usque

22

ad Taum (aestuario nomen est) nationibus. qua formidine terῸ hostes quamquam conflictatum saeuis tempestatibus exerci-

tum lacessere non ausi; ponendisque insuper castellis spatium fuit. adnotabant periti non alium ducem opportunitates locorum sapientius legisse. nullum ab Agricola positum castellum aut ui hostium expugnatum aut pactione ac fuga desertum. crebrae eruptiones: nam aduersus moras obsidionis annuis copiis firmabantur. ita intrepida ibi hiems et sibi quisque praesidio, inritis hostibus eoque desperantibus quia soliti plerumque damna aestatis hibernis euentibus pensare tum aestate atque hieme iuxta pellebantur. nec

Agricola umquam

per alios gesta auidus intercepit: seu centurio

seu praefectus incorruptum

facti testem habebat. apud quosdam

acerbior in conuiciis narrabatur; malos iniucundus.

ceterum

ut erat comis

bonis, ita aduersus

ex iracundia nihil supererat secretum,

ut silentium eius non timeres: honestius putabat offendere quam odisse. Quarta uirtus

aestas

exercitus

obtinendis et

Romani

inuitamenta Acidalius: inrit- H

quae

percucurrerat

nominis

gloria

insumpta;

pateretur,

ac,

si

inuentus

20.3 ante...pars obelis notauit Woodman: post pars

add. pariter Frohlich, perinde Ritter, sic Ernesti 21.1 absumpta H honoris aemulatio 7* 22.1 Taum H™A": Tanaum HAB eruptiones /ranspos. post ita Perret, bost hiems Halm

Puteolanus: ads22.2 crebrae

23

52

in

CORNELII

ipsa

Britanniae

terminus.

maris aestibus per immensum dirimuntur;

quod

tum

namque

Clota

et Bodotria,

diuersi

reuectae, angusto terrarum spatio

praesidiis firmabatur

sinus tenebatur, summotis

24

TACITI

atque

uelut in aliam insulam

omnis

propior

hostibus.

Quinto expeditionum anno naue primum transgressus ignotas ad id tempus gentes crebris simul ac prosperis proeliis domuit. eamque

partem Britanniae quae Hiberniam aspicit copiis instruxit,

in spem

magis quam

ob formidinem,

si quidem

Hibernia - medio

inter Britanniam atque Hispaniam sita et Gallico quoque mari opportuna - ualentissimam imperii partem magnis inuicem usibus miscuerit. (spatium eius — 51 Britanniae comparetur, angustius - nostrl maris insulas superat. solum caelumque et ingenia cultusque hominum haud multum a Britannia differunt; 71 meliusT

aditus portusque per commercia et negotiatores cogniti. Agricola

expulsum seditione domestica unum ex regulis gentis exceperat ac specie amicitiae in occasionem retinebat. saepe ex eo audiui legione una et modicis auxiliis debellari obtinerique Hiberniam posse;

idque

etiam

aduersus

Britanniam

profuturum

si Romana

ubique arma, et uelut e conspectu libertas tolleretur.)

25

Ceterum aestate qua sextum officii annum incohabat, amplexus ciuitates trans Bodotriam sitas, quia motus uniuersarum ultra gentium

et infesta

hostili

exercitu

itinera

timebantur,

portus

terra,

mari

classe

explorauit. quae ab Agricola primum adsumpta in partem uirium

sequebatur

egregia

specie,

cum

simul

simul

impellitur, ut saepe isdem castris pedes equesque mixti copiis et laetitia sua quisque

bellum

et nauticus miles

facta, suos casus adtollerent, ac

modo siluarum ac montium profunda, modo tempestatum ac fluctuum

aduersa,

hinc

terra

et hostis,

hinc

lactantia compararentur. Britannosque uisa classis obstupefaciebat, tamquam

mum uictis perfugium clauderetur. uli,

Ad

manus

magno

et

arma

paratu,

conuersi

maiore

fama

uictus

Oceanus

aperto maris sui secreto ulti-

Caledoniam

(uti

mos

incolentes

est

de

23 Britanniae Shaw-Smith: Britannia H 24.1 primum Boot: prima prima obelis not. Delz: alii alia) 24.2 differunt Rhenanus: differt H H, obelis notauit Woodman,

militari

(ut ex captiuis audiebatur)

secl. Wex: melius uel eius Rhenanus: alii alia

pop-

ignotis)

77 (naue in melius

25.1

hos-

tili exercitu H™: hostilis exercitus H: hostibus exercitus Becker impellitur H: impelleretur Rhenanus ut Woodman: ac H 25.2 Britannosque Gudeman: Britannos quoque H

DE

VITA

IVLII

AGRICOLAE

53

oppugnare ultro castellum adorti, metum ut prouocantes addiderant. regrediendumque citra Bodotriam et cedendum potius quam pellerentur ignaui specie prudentium admonebant, cum interim cognoscit hostis pluribus agminibus inrupturos. ac, ne superante numero

et

peritia

consilio

uniuersi

locorum

circumiretur,

diuiso

et

ipse

in

tres

partes exercitu incessit. quod ubi cognitum hosti, mutato repente nonam

adgressi, inter somnum

legionem

ut maxime

ac trepidationem

inualidam

26

nocte

caesis uigilibus inrupere.

iamque in ipsis castris pugnabatur cum Agricola, iter hostium ab exploratoribus edoctus et uestigiis insecutus, uelocissimos equitum

peditumque

adsultare

adici clamorem;

tergis pugnantium

et propinqua

iubet, mox

ab uniuersis

luce fulsere signa. ita ancipiti malo

territi Britanni; et nonanis rediit animus, ac securi pro salute de gloria certabant. ultro quin etiam erupere, et fuit atrox in ipsis portarum angustiis proelium donec pulsi hostes, utroque exercitu cer-

tante, his ut tulisse opem, illis ne eguisse auxilio uiderentur. quod nisi paludes et siluae fugientes texissent, debellatum illa uictoria

foret. Cuius conscientia ac fama ferox exercitus nihil uirtuti suae inuium et penetrandam Caledoniam inueniendumque tandem

27

Britanniae terminum continuo proeliorum cursu fremebant. atque illi modo cauti ac sapientes prompti post euentum ac magniloqui erant. (iniquissima haec bellorum condicio est: prospera omnes

sibi uindicant, aduersa uni imputantur.) at Britanni, non uirtute se uictos sed occasione et arte ducis rati, nihil ex adrogantia remit-

tere quominus iuuentutem armarent, coniuges ac liberos in loca tuta transferrent,

coetibus

et sacrificiis

conspirationem

ciuitatum

sancirent. atque ita inritatis utrimque animis discessum. Eadem aestate cohors Vsiporum per Germanias conscripta et in Britanniam transmissaa magnum ac memorabile facinus ausa est. occiso centurione ac militibus qui ad tradendam disciplinam

tres

uno tis),

inmixti

Liburnicas remig«i nondum

manipulis

adactis

per

exemplum

uim

et

rectores

gubernatoribus

imper-ante (suspectis duobus uulgato rumore ut miraculum

habebantur,

ascendere.

et

eoque interfecpraeuehebantur.

25.3 castellum H:-a H™ 26.2 nonanis H (rom suprascr. H^) rediit Wex: redit H 27.1 proeliorum Rhenanus: proelium Π 28.1 uno remig«i imper- ante Woodman: uno remigante H: alii alia praeuehebantur H: praebebantur H"

28

54

CORNELII

mox

ad aquam

plerisque saepe

atque

Britannorum

uictores,

ut infirmissimos ita circumuecti

utensilia

sua

aliquando

suorum,

raptissent,

proelio

eo ad extremum

mox

sorte

ductos

congressi

inopiae

regendi

ac

uenere

uescerentur.

amissis per inscitiam

cum

atque

nauibus

pro praedonibus habiti, primum a Suebis, mox a Frisiis intercepti

29

sunt. ac fuere quos per commercia uenumdatos et in nostram usque ripam mutatione ementium adductos indicium tanti casus inlustrauit. Initio aestatis Agricola domestico uulnere ictus anno ante natum filium

amisit.

ambitiose

quem

neque

casum

per

neque,

lamenta

tulit, et in luctu bellum

ut plerique

rursus

ac

inter remedia

fortium

maerorem

uirorum,

muliebriter

erat. igitur praemissa

classe

quae pluribus locis praedata magnum et incertum terrorem faceret, expedito exercitu, cui ex Britannis fortissimos et longa pace exploratos addiderat, ad montem Graupium peruenit, quem iam hostis insederat. nam Britanni nihil fracti pugnae prioris euentu et ultionem

aut seruitium

expectantes,

tandemque

docti

commune

periculum concordia propulsandum, legationibus et foederibus omnium ciuitatium uires exciuerant. iamque super triginta milia

armatorum quibus

aspiciebantur, et adhuc adfluebat omnis iuuentus et

cruda

ac

uiridis

senectus,

clari

bello

et sua

quisque

dec-

ora gestantes, cum inter plures duces uirtute et genere praestans nomine Calgacus apud contractam multitudinem proelium poscentem in hunc modum

locutus fertur:

‘Quotiens causas belli et necessitatem nostram intueor, magnus

30

mihi animus est hodiernum diem consensumque uestrum initium libertatis toti Britanniae fore: nam et uniuersi coistis et seruitutis expertes,

et

imminente

nullae nobis

ultra

classe

terrae

Romana.

ac

ne

mare

quidem

ita proelium

atque

securum,

arma,

quae,

fortibus honesta, eadem etiam ignauis tutissima sunt. priores pugnae,

quibus

aduersus

ac subsidium

Britanniae

Romanos

in nostris manibus

eoque

uaria fortuna habebant,

in ipsis penetralibus

certatum

est, spem

quia nobilissimi

siti nec

totius

ulla seruientium

28.2 adquam (a suprascr. H') H: ad aquandum Halm utensilia Selling: utilla H rapt«um ubi deuert>issent, cum Woodman: raptis secum H: raptissent,

cum

Anderson:

(mox

ubi...)

rapt«um

ex>issent,

cum

7Tl

20.1 aestatis Ernesti: alii alia 29.2 Grampium Puteolanus 29.4 adfluebat /IAB: -bant H uiridis /7T AB: uiris H: uirens H"A™ 30.1 coistis et H (sed coistis incertum), om. AB

VITA

IVLII

AGRICOLAE

ι Q

DE

litora aspicientes, oculos quoque a contactu dominationis inuiolatos habebamus. nos terrarum ac libertatis extremos recessus ipse ac

sinus

a fama

in

hunc

diem

defendit;

nunc

terminus

Britan-

niae patet, atque omne ignotum pro magnifico est; sed nulla iam ultra gens, nihil nisi fluctus ac saxa, et infestiores Romani,

quorum

superbiam frustra per obsequium ac modestiam effugias. raptores Orbis, postquam

cuncta

tur:

hostis

51 locuples

Oriens, non

Occidens

uastantibus est,

auari,

defuere

terrae, mare

51 pauper,

ambitiosi,

satiauerit: soli omnium

scrutanquos

non

opes atque inopiam

Auferre,

trucidare,

rapere

falsis

nominibus

imperium,

atque,

Q

pari adfectu concupiscunt. ubi solitudinem faciunt, pacem appellant. liberos cuique ac propinquos suos natura carissimos turi auferuntur;

coniuges

esse uoluit; hi per dilectus alibi serui-

sororesque

etiam

si hostilem

libidinem

effugerunt, nomine amicorum atque hospitum polluuntur. bona fortunaeque in tributum, ager atque annus in frumentum, corpora ipsa ac manus siluis ac paludibus emuniendis inter uerbera et contumelias conteruntur. nata seruituti mancipia semel ueneunt atque ultro

a dominis

aluntur;

Britannia

seruitutem

suam

cotidie

emit,

cotidie pascit. ac, sicut in familia recentissimus quisque seruorum

etiam conseruis ludibrio est, sic in hoc orbis terrarum uetere famulatu noui nos et uiles in excidium petimur: neque enim arua nobis aut metalla aut portus sunt quibus exercendis reseruemur. uirtus

porro ac ferocia subiectorum ingrata imperantibus; et longinquitas ac secretum

ipsum

quo tutius, eo suspectius. ita sublata spe ueniae

tandem sumite animum, tam quibus salus quam quibus gloria caris-

sima est. Brigantes femina duce exurere coloniam, expugnare castra ac, nisi felicitas in socordiam

uertisset, exuere

iugum

potuere;

nos integri et indomiti et in libertatem non in paenitentiam «arma in>laturi primo statim congressu ostendamus quos sibi Caledonia uiros seposuerit.

‘An eandem Romanis in bello uirtutem quam in pace lasciuiam adesse creditis? nostris illi dissensionibus ac discordiis clari uitia hostium

in

diuersissimis

gloriam

exercitus

gentibus,

sui

ut secundae

uertunt; res

quem

tenent,

contractum ita aduersae

ex dis-

soluent - nisi si Gallos et Germanos et (pudet dictu) Britannorum 30.3 sinus a fama Constans: sinus famae /T: sinus fama Boxhorn

entiam

Wolfflin

«arma

in>laturi

Woodman

Meister. «bel-laturi Weinkauff. alii alia

(

iam

Mohr):

31.4 in pati-

laturi

32

56

CORNELII

TACITI

plerosque, licet dominationi alienae sanguinem commodent, diutius

tamen

metus

hostes

ac terror

quam

seruos,

est, infirma

fide

uincla

et

adfectu

caritatis;

teneri

quae

putatis.

ubi remoueris,

qui timere desierint odisse incipient. omnia uictoriae incitamenta

pro

nobis

sunt:

nullae

Romanos

coniuges

accendunt;

nulli

par-

entes fugam exprobraturi sunt; aut nulla plerisque patria aut alia est. paucos et siluas,

numero,

ignota

trepidos

omnia

ignorantia,

caelum

ipsum

circumspectantes,

clausos

quodam

ac

mare modo

ac uinctos dii uobis tradiderunt. ne terreat uanus aspectus et auri fulgor

atque

argenti,

quod

neque

tegit

neque

uulnerat.

in

ipsa

hostium acie inueniemus nostras manus: adgnoscent Britanni suam causam, recordabuntur Galli priorem libertatem, tam deserent illos

ceteri Germani quam nuper Usipi reliquerunt. nec quicquam ultra formidinis: uacua castella, senum coloniae, inter male parentes et iniuste

imperantes

aegra

municipia

et discordantia.

hic

dux,

hic

exercitus; ibi tributa et metalla et ceterae seruientium poenae, quas

in aeternum

perferre aut statim ulcisci in hoc campo est. proinde

ituri in aciem et maiores uestros et posteros cogitate!' 33

Excepere

orationem

alacres,

ut

barbaris

moris,

fremitu

can-

tuque et clamoribus dissonis. iamque agmina et armorum fulgores audentissimi cuiusque procursu, simul instruebatur acies, cum

Agricola quamquam accendendum

adhuc

laetum et uix munimentis coercitum militem ratus, ita disseruit:

‘Septimus annus est, commilitones, ex quo uirtute et auspiciis

populi

Romani,

fide

atque

opera

nostra

Britanniam

uicistis.

tot

expeditionibus, tot proeliis, seu fortitudine aduersus hostes seu patientia ac labore paene aduersus ipsam rerum naturam opus

fuit, neque me militum neque uos ducis paenituit. ergo egressi, ego ueterum legatorum, uos priorum exercituum terminos, finem Britanniae

non

fama

nec

rumore

sed

castris

et armis

tenemus:

inuenta Britannia et subacta. equidem saepe in agmine, cum uos paludes montesue

et flumina fatigarent, fortissimi cuiusque

uoces

audiebam: "Quando dabitur hostis? Quando animus ?" Q

ueniunt,

e

omniaque

latebris prona

32.2 'est H: sunt Beroaldus lius: octauus H

imperii Η

suis

extrusi,

uictoribus

atque

et

uota eadem

uirtusque uictis

uobis H: nobis Puteolanus

ante uel post uirtute edd. nonnulli

nostra /T: uestra Puteolanus

33.4

in

aduersa.

aperto, nam,

33.2 septimus Acidapopuli Woodman:

Woodman: alii alia

DE

ut

superasse

VITA

tantum

IVLII

itineris,

AGRICOLAE

euasisse

57

siluas,

transisse

aestuaria

pulchrum ac decorum in frontem, ita fugientibus periculosissima quae

hodie

eadem

sunt:

neque

enim

notitia aut commeatuum

prosperrima

eadem

abundantia,

arma et in his omnia.

quod ad me

nobis

aut

locorum

sed manus

attinet, iam pridem

et

mihi decre-

tum est neque exercitus neque ducis terga tuta esse. proinde ut

honesta

mors

turpi

uita

potior,

ita incolumitas

ac

decus

eodem

loco sita sunt; nec inglorium fuerit in ipso terrarum ac naturae fine

cecidisse. 'Si nouae

gentes atque

ignota acies constitisset, aliorum

exerci-

tuum exemplis uos hortarer; nunc uestra decora recensete, uestros oculos interrogate. hi sunt quos proximo anno unam legionem furto

noctis

tannorum

adgressos

fugacissimi

clamore

debellastis;

ideoque

tam

diu

hi

ceterorum

superstites.

34

Bri-

quomodo

siluas saltusque penetrantibus fortissimum quodque animal contra ruere,

sic

pauida

acerrimi

numerus

et

inertia

ipso

Britannorum

ignauorum

agminis

iam

sono

pridem

et timentium.

pelli

bantur,

ceciderunt,

quos

quod

reliquus

tandem

est

inuenistis,

non restiterunt sed deprehensi sunt: nouissimae res et extremo metu torpor defixere aciem in his uestigiis in quibus pulchram et spectabilem uictoriam ederetis. transigite cum expeditionibus, imponite quadraginta annis magnum diem, adprobate rei publicae numquam exercitui imputari potuisse aut moras belli aut causas rebellandi!' etadloquente adhuc Agricola militum ardor eminebat, etfinem orationis ingens alacritas consecuta est, statimque ad arma discursum.

Instinctos ruentesque

octo

milium

cornibus toriae

erant,

adfunderentur.

decus

citra

si pellerentur. editioribus

ita disposuit ut peditum

mediam

firmarent,

legiones

Romanum

Britannorum

locis

aciem

constiterat

pro

uallo

sanguinem ut

stetere,

bellandi,

acies in speciem ita

auxilia, quae

equitum

primum

simul agmen

tria milia ingens

et

uic-

auxilium,

ac terrorem in

aequo,

ceteri per adcliue iugum conexi uelut insurgerent; media campi 33.5 in frontem Shaw-Smith ita Rhenanus: item H 33.6 ut...ita Woodman: et...et H 34.1 unam H: nonam Hiibner (cf. 26.1) 34.2 pelli bantur Woodman: pellebantur H: pelluntur Puteolanus: pelli solent Wex et timentium 7 : dementium Π 34.3 torpor Ritter: corpora H quadraginta W. Heraeus: quinquaginta H 35.3 agminae quoceteri H, corr. Bekker conexi Puteolanus: conuexi H

35

58

CORNELII

couinnarius

eques

cola superante

strepitu

hostium

simul in latera suorum

porrectior

acies

admonebant,

ac

discursu

multitudine

primo

complebat.

promptior

erat

et arcessendas

in spem

congressu uitare

Agri-

in frontem,

eminus

plerique

uel

excutere,

legiones

et firmus aduersis, dimisso certabatur;

simulque

simul arte Britanni ingentibus gladiis et breuibus nostrorum

tum

ueritus ne simul

pugnaretur, diductis ordinibus, quamquam

futura

pedes ante uexilla constitit. Ac

TACITI

atque

ipsi magnam

equo

constantia,

caetris missilia uim

telorum

superfundere, donec Agricola quattuor Batauorum cohortes ac Tungrorum duas cohortatus est ut rem ad mucrones ac manus adducerent — quod et ipsis uetustate militiae exercitatum et hostibus inhabile parua scuta et enormes gladios gerentibus (nam Britannorum gladii sine mucrone complexum armorum et in arto pugnam non tolerabant). igitur ut Bataui miscere ictus, ferire umbonibus,

ora fodere

in colles aciem

et, stratis qui in aequo

adstiterant, erigere

coepere, ceterae cohortes aemulatione

et impetu

conisae proximos quosque caedere; ac plerique semineces aut inte-

gri festinatione uictoriae relinquebantur. interim equitum turmae

[fugere couinnariif peditum se proelio miscuere; et quamquam recentem terrorem intulerant, densis tamen hostium agminibus et inaequalibus

locis haerebant:

minime

enim

equestris

ea pug-

nae facies erat, cum aegre cliuo stantes simul equorum corporibus

impellerentur; ac saepe uagi currus, exterriti sine rectoribus equi, ut quemque

37

formido

tulerat, transuersos aut obuios incursabant.

Et Britanni qui adhuc pugnae expertes summa collium inseder-

ant et paucitatem nostrorum uacui spernebant degredi paulatim et

circumire terga uincentium coeperant, ni id ipsum ueritus Agricola quattuor

equitum

suisset, quantoque

fugam

disiecisset.

alas, ad subita belli retentas, uenientibus

ferocius adcucurrerant,

ita consilium

oppo-

tanto acrius pulsos in

Britannorum

in ipsos uersum,

35.3 eques Rhenanus 35.4 in latera Cartault: et latera H (ne [simul] in frontem simul et latera Frohlich) 36.1 parua...gerentibus secl. Wex, parua...tolerabant sec. Nipperdey nam...tolerabant secl. Haase arto De Medicis: aperto H 36.2 fodere Gesner: foedare H 306.3 fugere couinnarii obelis notauit Woodman: fugere couinnarii Wex: fugere couinnarii Doederlein minime 7 minimeque Π equestris ea enim H, transpos. Mackay aegre cliuo stantes Triller: aegradiu aut stante /T: aegre in gradu stantes OgilvieRichmond: alii alia

DE

VITA

IVLII

AGRICOLAE

59

transuectaeque praecepto ducis a fronte pugnantium alae auersam

hostium aciem inuasere.

Tum

uero

sequi,

iam

patentibus

uulnerare,

hostium,

paucioribus morti

prout

terga praestare,

locorum

arma

et aliquando

adpropinquauerunt. gnari

etiam

nam

grande

atque

cuique

offerre. passim

humus;

locis

capere,

et

eosdem

ingenium

quidam

et corpora uictis

atrox

oblats

erat, cateruae

inermes

ira

trucidare.

armatorum

ultro ruere ac se

et laceri artus et cruenta

uirtusque

primos sequentium

circumueniebant;

spectaculum: aliis

quod

postquam

siluis

incautos collecti et

ni frequens

ubique

Agri-

cola ualidas et expeditas cohortes indaginis modo et sicubi artiora equitum

dimissis equis, simul rariores siluas partem fiduciam

foret. ceterum, ubi compositos firmis ordinibus sequi rursus uidere, in fugam

uersi: non agminibus

τ

erant partem

persultare iussisset, acceptum aliquod uulnus per nimiam

(ut prius) nec alius alium respec-

tantes, rari et uitabundi inuicem longinqua atque auia petiere. finis sequendi nox et satietas fuit. caesa hostium ad decem milia; nostrorum trecenti sexaginta cecidere, in quis Aulus Atticus praefectus

cohortis, iuuenili ardore et ferocia equi hostibus inlatus. Et nox quidem gaudio praedaque laeta uictoribus; Britanni palantes mixto uirorum mulierumque ploratu trahere uulneratos, uocare integros, deserere domos ac per iram ultro incendere,

eligere

rum,

saepius

latebras

et statim

relinquere;

miscere

inuicem

con-

silia aliqua, dein separare; aliquando frangi aspectu pignorum suoconcitari,

satisque

constabat

saeuisse

quosdam

in

coniuges ac liberos, tamquam misererentur. proximus dies faciem

uictoriae

latius aperuit:

uastum

ubique

silentium,

secreti

colles,

fumantia procul tecta, nemo exploratoribus obuius. quibus in omnem partem dimissis, ubi incerta fugae uestigia neque usquam conglobari hostes compertum, et exacta iam aestate spargi bellum nequibat, in fines boreos totum exercitum deducit. Ibi acceptis praecipit.

obsidibus,

datae

ad

praefecto

id uires,

classis circumuehi

et praecesserat

atque equites lento itinere, quo nouarum

terror.

Britanniam ipse

peditem

gentium animi ipsa tran-

situs mora terrerentur, in hibernis locauit. et simul classis secunda

37.4 gnari Dronke: ignari H "partem Woodman: equite HA: equites B persultare Rhenanus: -ari HA: perlustrari B 38.1 aliqua secl. Classen 38.2 secreti H: deserti Ernesti boreos totum Wolfson: borestorum H

38

60

CORNELII

TACITI

tempestate ac fama Trucculensem portum tenuit, unde proximo

Britanniae latere praelecto omnis redierat.

39

Hunc

rerum

cursum, quamquam

nulla uerborum

iactantia epis-

tulis Agricolae auctum, ut erat Domitiano moris, fronte laetus, pec-

tore anxius excepit. inerat conscientia derisui fuisse nuper falsum

e Germania

triumphum,

emptis

tus et crinis in captiuorum

magnamque celebrari!

uictoriam

per commercia

speciem

tot milibus

id sibi maxime

quorum

formarentur;

hostium

formidolosum,

caesis

priuati

habi-

at nunc

ueram

ingenti

hominis

fama

nomen

supra principem adtolli: frustra studia fori et ciuilium artium decus in

silentium

utcumque

acta,

si

militarem

facilius dissimulari,

gloriam ducis

boni

alius

occuparet;

imperatoriam

cetera

uirtutem

esse. talibus curis exercitus, quodque saeuae cogitationis indicium

erat, secreto suo satiatus, optimum in praesentia statuit reponere

40

odium donec impetus famae et fauor exercitus languesceret: nam etiam tum Agricola Britanniam obtinebat. Igitur triumphalia ornamenta et inlustris statuae honorem et quidquid pro triumpho datur, multo uerborum honore cumu-

lata, decerni in senatu iubet additque insuper opinionem, Syriam prouinciam Agricolae destinari, uacuam tum morte Atli Rufi con-

sularis et maioribus reseruatam. (credidere plerique libertum ex secretioribus ministeriis missum ad Agricolam codicillos quibus ei Syria dabatur tulisse cum eo praecepto ut, si in Britannia foret, traderentur; colae,

ne

eumque

appellato

libertum quidem

in ipso freto Oceani eo

ad

Domitianum

obuium

Agri-

remeasse,

siue

uerum istud siue ex ingenio principis fictum ac compositum est.) tradiderat interim Agricola successori suo prouinciam quietam tutamque.

ac, ne notabilis

celebritate

introitus esset, uitato amicorum Palatium, et nullo

ita ut praeceptum sermone

militare nomen tranquillitatem mone

turbae

(graue atque

et frequentia

erat, uenit; exceptusque

seruientium

inter otiosos) otium

occurrentium

officio noctu in urbem, noctu in

penitus

facilis, uno aut altero amicorum

inmixtus

est.

breui osculo ceterum,

uti

aliis uirtutibus temperaret, duxit,

cultu

modicus,

ser-

comitatus, adeo ut plerique,

38.4 trucculensem /f: trutulense : Rutupensem Rhenanus, -iensem Lipsius: trux Tulensem Wolfson proximo Madvig prelecto A™T: prelecta H omnis /T: omni H* 39.1 auctum Lipsius: actum H domitiano moris H": domitianus crinis HAB: crines Puteolanus 39.3 praesentia H: prae-

sens H™

40.1

additque

(sed cf. p. 36 n. 125)

Muretus: addique

H

40.2

missum

40.4 duxit Woodman: auxit E: hausit Wex

hic finis codicis H

DE

VITA

IVLII

AGRICOLAE

61

quibus magnos uiros per ambitionem aestimare mos est, uiso aspectoque Agricola quaererent famam, pauci interpretarentur. Crebro

per eos dies apud Domitianum

absolutus est. causa periculi non cuiusquam,

sed

pessimum

infensus

inimicorum

absens accusatus, absens

crimen

uirtutibus

ullum

princeps

genus, laudantes.

41

aut querela laesi et

gloria

uin

ac

et ea insecuta sunt rei

publicae tempora quae sileri Agricolam non sinerent: tot exercitus in Moesia Daciaque et Germania et Pannonia temeritate aut

per ignauiam ducum amissi, tot militares uiri cum tot cohortibus expugnati

et capti; nec iam

nis legionum

de limite imperii et ripa sed de hiber-

et possessione dubitatum. ita, cum

damna damnis

continuarentur atque omnis annus funeribus et cladibus insignire-

tur, poscebatur ore uulgi dux Agricola, comparantibus

cunctis uig-

orem, constantiam et expertum bellis animum cum inertia et formidine ceterorum. quibus sermonibus satis constat Domitiani quoque aures uerberatas,

dum

optimus

quisque

pessimi malignitate et liuore pronum

libertorum

amore

et fide,

deterioribus principem exs-

timulabant. sic Agricola simul suis uirtutibus, simul uitiis aliorum

in ipsa gloria praeceps agebatur.

Aderat iam annus quo proconsulatum Africae et Asiae sortiretur, et occiso Ciuica nuper nec Agricolae consilium deerat nec Domitiano exemplum. accessere quidam cogitationum principis periti, qui iturusne esset in prouinciam ultro Agricolam interrogarent. ac primo occultius quietem et otium laudare, mox operam

iam

suam

obscuri

tianum.

in adprobanda

suadentes

simul

excusatione

terrentesque

offerre,

postremo

pertraxere

qui paratus simulationi, in adrogantiam

ad

non

Domi-

compositus,

et

audiit preces excusantis et, cum annuisset, agi sibi gratias passus est,

nec erubuit beneficii inuidia. salarium tamen proconsulare solitum

offerri et quibusdam a se ipso concessum Agricolae non dedit, siue offensus non petitum, siue ex conscientia, ne quod uetuerat uidere-

tur emisse.

Proprium

Domitiani

humani uero

inreuocabilior,

natura

ingenii

est

praeceps

moderatione

in

tamen

odisse iram

et,

quem quo

laeseris;

obscurior,

prudentiaque

eo

Agricolae

41.2 possessione Shaw-Smith 41.3 ceterorum Grofius: eorum Ε, post quod lacunam statuit Bach: aliorum Croll 41.4 in ipsa gloria Courtney: 1 ipsam gloriam £: ipsa gloria Madvig — 42.2 simulationi ed. Bipont.: -ne E (parata simulatione Ernest?): -nis E"

42

62

CORNELII

TACITI

leniebatur, quia non contumacia neque inani iactatione libertatis famam

fatumque

mirari,

posse

obsequiumque laudis

prouocabat.

etiam ac

excedere

sub

sciant,

malis

modestiam, quo

quibus

principibus si

plerique

industria

per

moris

ac

abrupta,

Finis

uitae

eius

nobis

luctuosus,

amicis

in

tristis,

inlicita

uiros

uigor

sed

publicae usum ambitiosa morte inclaruerunt.

43

est

magnos

esse,

adsint,

eo

nullum

rei

extraneis

etiam

ignotisque non sine cura fuit. uulgus quoque et hic aliud agens populus et uentitauere ad domum et per fora et circulos locuti sunt; nec quisquam audita morte Agricolae aut laetatus aut statim oblitus est. augebat miserationem constans rumor ueneno interceptum. nobis nihil comperti adfirmare ausim; ceterum per omnem

ualetudinem

eius crebrius quam ex more principatus per nuntios uisentis et libertorum

primi

et medicorum

intimi

uenere,

siue

cura

illud siue

inquisitio erat. supremo quidem die momenta ipsa deficientis per

dispositos cursores nuntiata constabat, nullo credente sic adceler-

ari quae tristis audiret. speciem tamen doloris animi uultu prae se tulit, securus

iam

odii et qui facilius dissimularet gaudium

quam

metum. satis constabat lecto testamento Agricolae, quo coheredem

optimae uxori et piissimae filiae Domitianum scripsit, laetatum eum uelut honore iudicioque. tam caeca et corrupta mens adsiduis

adulationibus erat ut nesciret a bono patre non scribi heredem nisi malum principem.

44

Natus

Iuniis;

erat

excessit

Agricola quarto

Gaio

Caesare

tertium

consule

idibus

anno,

decimum

kalen-

et quinquagesimo

das Septembris Collega Priscinoque consulibus. quod si habitum quoque

eius posteri noscere uelint, decentior quam

sublimior fuit;

nihil impetus in uultu: gratia oris supererat. bonum uirum facile crederes,

magnum

spatio integrae

aeuum

peregit.

libenter. et ipse quidem,

aetatis ereptus, quantum

quippe

et uera

bona,

quamquam

ad gloriam,

quae

medio

in

longissimum

in uirtutibus

sita

sunt, impleuerat, et consulari ac triumphalibus ornamentis praed-

ito

quid

aliud

gaudebat, 42.4 plerique Schomann

adstruere

speciosae

43.1

Maehly

laetatus

fortuna

contigerant.

est

E

poterat?

filia

enisi post abrupta (est

secl.

opibus

atque

uxore add.

nescioquis)

nimiis

Heumann, 43.2

non

superstitibus post usum

«quod»

adfir-

mare Acidalius al alia 43.3 animi uultu Baehrens: animo uultuque Ε: habitu uultuque Ernesti 44-1 tertium Ursinus: ter E quarto Petavius: sexto E Priscinoque Hirschfeld: Priscoque E 44.2 impetus E: metus £" 44-4 filia...superstitibus post fama transpos. Doederlein, post amicitiis Urlichs

DE

VITA

IVLII

AGRICOLAE

63

adfinitatibus et amicitiis futura effugisse. nam, sicuti «non licuit

Q

potest uideri etiam beatus incolumi dignitate, florente fama, saluis

ei- durare in hanc beatissimi saeculi lucem ac principem Traianum

uidere, quod augurio uotisque apud nostras aures ominabatur, ita

festinatae

mortis grande

solacium

tulit euasisse postremum

illud

tempus quo Domitianus non iam per interualla ac spiramenta tem-

porum sed continuo et uelut uno ictu rem publicam exhausit. non uidit Agricola obsessam curiam et clausum armis senatum et eadem

strage

tot consularium

caedes,

tot nobilissimarum

femi-

45

narum exilia et fugas. una adhuc uictoria Carus Mettius censebatur, et intra Albanam

Baebius

etiam

tum

arcem

sententia

reus erat. mox

Messalini

nostrae

strepebat,

duxere

et Massa

Heluidium

in

sanguine Senecio perfudit. Nero tamen subtraxit oculos suos iussitque scelera, non spectauit; praecipua sub Domitiano miseriarum pars erat uidere et aspici, cum suspiria nostra subscriberentur, cum

denotandis tot hominum palloribus sufficeret saeuus ille uultus et

rubor,

quo

se contra pudorem

muniebat.

tu uero

felix, Agricola,

non uitae tantum claritate sed etiam opportunitate mortis. Vt perhibent qui interfuere nouissimis sermonibus tuis, constans et libens fatum excepisti, tamquam pro uirili portione innocen-

parentis erepti auget maestitiam quod adsidere ualetudini, fouere deficientem, satiari uultu complexuque non contigit. excepissemus certe mandata uocesque, quas penitus animo figeremus. noster hic dolor, nostrum uulnus: nobis tam longae absentiae condicione ante quadriennium amissus est. omnia sine dubio, optime parentum, adsidente

amantissima

uxore

superfuere

honori

tuo; paucioribus

tamen lacrimis comploratus es, et nouissima in luce desiderauere aliquid oculi tui. $i quis piorum manibus locus, 51 (utsapientibus placet) non cum corpore extinguuntur magnae animae, placide quiescas, nosque domum tuam ab infirmo desiderio et muliebribus lamentis ad contemplationem

uirtutum

tuarum

uoces,

quas

Q

tiam principi donares. sed mihi filiaeque eius praeter acerbitatem

neque

lugeri

neque

44.5 non licuit ei post Dahl suppletum est grande solacium £: graue solE" exhausit £: hausit E" 45.1 uisus Reizenstem: alii alia 45.-5 amissus es Rhenanus comploratus E: compositus £" 46.1 nosque Urlichs

46

64

CORNELII

plangi

fas est. admiratione

suppeditet, simi

similitudine

cuiusque

pietas.

secum

reuoluant,

te potius

colamus:

id filiae

patris, sic mariti memoriam

TACITI

quoque

uenerari

formamque

et laudibus

is uerus

honos,

uxorique ut omnia

ac figuram

animi

et, si natura ea

coniunctis-

praeceperim, facta dictaque magis

sic eius

quam

cor-

poris complectantur, non quia intercedendum putem imaginibus quae

marmore

aut

aere

finguntur,

sed,

ut uultus

hominum,

ita

simulacra uultus imbecilla ac mortalia sunt, forma mentis aeterna,

quam tenere et exprimere non per alienam materiam et artem sed tuis ipse moribus possis. quidquid ex Agricola amauimus, mirati sumus, manet mansurumque

in aeternitate temporum

quidquid

est in animis hominum,

fama rerum: nam

multos ueterum

uelut

inglorios et ignobiles obliuio obruit; Agricola posteritati narratus et traditus superstes erit. 46.2 temporalibus

wt dittographiam post ἴῈ potius

Grotius, colamus Muretus (decoremus mamque T, Muretus: famamque E

secl. Muretus

Ursinus): militum decoramus E ut (del., ut uid.) tuis E 46.4

similitudine

40.3 for«ut» Wood-

man: cum post hominum tum et post temporum nterpungunt plerique, tricolon interpretati fama Halm obruit Haupt: obruet E

COMMENTARY

1-3 PREFACE Prefatorial statements are a feature of many genres of classical literature, both prose and verse, and theoretical writers developed various conventions for their composition. In the first century BC, for example, the author of the Rhetorica ad Herennium said that in the case of epideictic (or 'display’) oratory the preface should be drawn from the person of the speaker or that of his subject or that of his audience or from his topic (3.11). He then illustrates each of these in turn. Thus, in the case of a preface which is centred on the person of the speaker, he says that, if we set out to praise someone in our speech, we might explain that we shall be doing so out of duty (officium), since a bond of relationship

(causa necessitudinis) exists

between us. In the late republic Cornelius Nepos used a preface to introduce his biographies of Latin historians (fr. 2.1—2) and of foreign generals. In the latter case he begins by expressing a belief that many readers will find his type of writing to be trivial and unworthy of the personalities of great men (praef. 1 ‘qui hoc genus scripturae leue et non satis dignum summorum uirorum personis iudicent'); he then devotes most of his brief preface to forestalling such criticism, arguing that, because Greeks and Romans belong to different cultures, they practise different cultural habits. He concludes with a statement of his subject

(3): ‘in hoc

exponemus

libro

de uita excellentium imperatorum'. T. likewise concludes his preface with a statement of his subject: he says that he will be praising his father-in-law Agricola and that his biography is an expression of f:etas (3.9 ‘hic...liber honori Agricolae soceri mei destinatus, professione pietatis aut laudatus erit aut excusatus’). Yet, though the reference to fzetas seems merely to illustrate the advice given by the Rhetorica ad Herennium in cases where there is a special relationship between author and subject, it comes at the end ofa calculated and elaborate argument which has taken up the whole of the rest of the preface. The structure of the preface is deceptively simple. Opening and closing statements form a frame (1.1 — 3.3), within which the text 15 divided into two main parts: the first (À: 1.2—2.3) deals with the period before Domitian's death, the second

(B: 3.1—2) with the new dawn of Nerva and

Trajan. Each of these parts comprises two further chronological divisions which mirror each other chiastically: the first part is divided into the past (1.2-3) and T.'s own day (1.4-2.3), the second 15 divided into T.'s own day (3.1) and the past (3.2). This arrangement may be set out schematically thus: 65

66

COMMENTARY:

1

1.1 Opening statement A: 1.2—2.3 Before Domitian's death (a) 1.2-3 The past (1.2 apud priores) (b) 1.4-2.3 T.'s own day (1.4 at nunc)

Β: 3.1—2 The new dawn (b') 3.1 T.'s own day (3.1 Nunc)

(a') 3.2 The past 3.3 Closing statements

Since the closing statements effect a return to the topic introduced at the beginning, the function of 1.1 and 3.3 as a frame seems clear. Yet each statement is also closely integrated into its immediate context: the opening statement introduces the contrast between the past (antiquitus) and T.'s own day (nostris... temporibus) which will be elaborated in A, while the closing statements constitute the conclusion which results from the argument put forward in B. Moreover, whereas T.'s own day is to be equated with Domitian's reign in A, Domitian's reign has become the past in B: hence

there are, as it were,

two

'present'

times, each

signalled by nunc

(1.4, 3.1). Domitian's reign, which alone features in both main parts of the preface and which takes up more space than any other topic, emerges as being of central importance and is to be seen from the two different perspectives of the past (1.1, 1.2—3) and of the time of writing (3.1). Since T. begins by saying that in Domitian's day the biography of a deceased person was written only in cases of exceptional wertus (1.1), and since by definition T.’s own subject was a man of exceptional urrtus, the reader infers that T. is invoking those earlier biographers as necessary precedents for himself. The rarity of modern biography contrasts strongly with the old days, when the biographical habit was so common that even writing one's own life was regarded as a perfectly natural activity (1.2—3). But, when on the death of his anticipated subject T. came to embark on actually writing the man's biography, he was obliged to suspend his plan: the times had become so savage and so hostile to uzrtus that Arulenus Rusticus and Herennius Senecio, who are now revealed as his two precedents,

were executed for having written their biographies of Thrasea Paetus and Helvidius Priscus and copies of the biographies themselves were publicly burned (1.4-2.1). This bonfire marked a defining moment in a reign of terror

(2.2). Citizens did not have

the courage

to offer resistance:

they

were so cowed into servile submission that even the normal exchanges of speaking and listening were abandoned (2.3). But now, under the benevolent and restorative rule of Nerva and Tra-

jan, there has been a dramatic change for the better (3.1). Yet intellectual

COMMENTARY:

1.1

67

life cannot be revived overnight: on the one hand, people become accustomed to inactivity; on the other, Domitian's reign lasted so long that many potential authors had simply died, while the few who survived the fifteen years of living death were more used to silence than to eloquence and hence were out of practice (3.1—2). T. himself is one of those survivors, and, despite the inevitable effect which so long a silence has had on his powers of expression, he fully intends to write a history of Domitian's reign and its sequel. But that is for the future; in the meanwhile the biography of his father-in-law, which has now been resumed, will be either praised or excused by readers as work of devotion (3.3). This analysis, almost every element of which 15 controversial, attempts to clarify a story and an argument which T. has presented with considerable indirection. The most fundamental example of this lies in the relationship between the two main parts of the preface. The first part builds up to the universal silence which resulted from Domitian's reign of terror (2.3) and which had suppressed the Agricola even before T. started to write it (1.4). The second part likewise builds up to the same silence (3.2), which is made the excuse for any alleged inadequacy of style in — the Historzes (3.3). The inconsequence 15 striking. The Hzstorzes, an unwritten future work, has not even been hinted at in the whole of the preface hitherto; it was the Agricola,

now in progress at long last, which was at issue throughout 1.1-2.3, yet it here emerges only obliquely as a seeming afterthought (Ac intersm lLiber). T. has started his literary career the way he intended to go on. For studies of the preface and further bibliography see Sailor (2004); ΡΗ 25—90. 1.1 Clarorum uirorum facta moresque posteris tradere: tradere depends upon omzsit below (OLD omittere 8b, 'to neglect/fail to — *; cf. 5b). T.'s opening words, which are picked up at the end in ring composition (46.4 'Agricola posteritat? narratus et traditus superstes erit’), both define the present work and summarise the biographical genre: facta and mores, though regularly combined (TLL 8.1524.17ff.; note esp. Sall. C. 55.6), refer to two of the essential elements of biographical encomium, πράξεις or ‘deeds’ and ἐπιτηδεύματα or ‘behaviour/morality’

(the plural of mosis too

often rendered as ‘character’): see Menand. 372.2—13 (the text is uncertain in places: see Russell and Wilson ad loc.). Compare the ‘embedded’ mini-biographies of Catiline at Sall. C. 5.1-8 and of Eutherius at Amm. 16.7.4-10. clarorum uirorum is ‘an archaic and solemn form of words in a context of national praise' (Brink on Hor. Ep. 2.1.250). In the preface to his Orzgines, the first historical work certainly to have been written in Latin prose,

the elder Cato famously used similar words to defend the fact that he was spending his leisure time as a writer of history (fr. 2P — 2C 'clarorum

68

COMMENTARY:

1.1

oportere’; see further 40.4n. tranquillitatem). In another passage, which is closer in wording but which may or may not come from the preface, Cato referred to an ancient custom of singing the praises of famous men at banquets (fr. 118P — 113C 'clarorum uzrorum laudes atque uirtutes’), a custom clearly analogous to that of commemorating a famous man in a biography. Although neither passage constitutes Cato's actual first sentence, it is generally agreed that T. is alluding to one or both of them,' perhaps invoking Cato as a prototype of the uzrtus about to be personified bv A. within the biography (so O-R): for A. as clarus see 18.5n. Whatever T.'s motive, the allusiveness of his first two words leaves readers momentarily uncertain whether he is about to refer to himself or his subject, since clarorum hominum in the former passage is a reference to Cato as author, while clarorum wirorum in the latter passage is a reference to the subject matter of the commemorative songs. The uncertainty is appropriate and almost programmatic, since it is generally accepted that T. in the biography, at the same time as defending A., 15 also defending himself (Intro. p. g). This too was a feature of Cato, since he did not refrain from inserting his own speeches into the Orzgznes (Cic. Brut. 8g, De or. 1.227, Liv. 45.25.3, per. 49, Val. Max. 8.1 abs. 2, Gell. 6.3.7), even though the work was so depersonalised that he famously omitted to identify commanders by name (Nep. Cato 3.4, Plin. NH 8.11). The intertextual complexity of the opening words is further complicated by the fact that Cato's fr. 2 is itself compared by scholars with the opening of Xenophon's Sympostum, where Xenophon is making a different point again, namely, that the leisure time of famous men deserves consideration (1.1 'AAA' épol δοκεῖ τῶν καλῶν κἀγαθῶν ἀνδρῶν épya oU póvov τὰ μετὰ σπουδῆς πραττόμενα ἀξιομνημόνευτα εἶναι, ἀλλὰ Kal T& év Tals παιδιαῖς,

‘It seems to me, with reference to great and good men's deeds, that not only those done with serious intent are worthy of commemoration, but also those done in lighter moments’; cf. Plut. Alex. 1.2). Since ἀνδρῶν épya is paralleled by u?rorum facta here and ἀξιομνημόνευτα by digna memoratu

below (2), it of a ‘window ner in which 40.4; cf. also

seems likely that T. is also referring to Xenophon by means reference'.* One of the themes of T.'s biography is the manA. spent his leisure time, whether in or out of office (esp. 9.3, Galba at 6.gn. nertia).

! ‘If it seems far-fetched to suppose deliberate Catonian reminiscence in these two very ordinary words, ... the reader may consider the parallel case of the English words "famous men", which many people would doubtless link almost automatically with Ecclesiasticus 44' (Powell on Cic. Senec. 80).

?^ A ‘window reference' is one in which an author (C) alludes to a predecessor (B) in such a way as to reveal his awareness that B in turn is referring to a still earlier predecessor (A): see Thomas (1986) 188.

COMMENTARY:

1.1

69

Thus T., embarking on the biography of a man who was called 'Agricola' and whose distinction rested entirely upon his soldiering, has begun his work with a combined allusion to two writers who believed that farmers made the best soldiers (Xen. Oec. 5.14—16, Cato, Agr. cult. praef. 4 ‘ex agricolis et uiri fortissimi et milites strenuissimi gignuntur’). See also 46.4n. (Agricola).

antiquitus usitatum qualifies and is in apposition to the infinitival phrase Clarorum. .. tradere (‘ practised long ago’). The anagrammatical

assonance

(-2tus usit-) 15 typical of T. (TR 222—5)

and here

has the extra dimension of identical syllables at the end and beginning of adjacent words, a phenomenon criticised by ancient scholars on phonic grounds though not avoided in practice by Latin writers (e.g. again in T. at A. 1.10.7 ‘honora oratione’): see Austin on Virg. Aen. 2.27 'Dorica castra’, Godel (1984). It has been suggested that the frequency of assonantal effects implies oral delivery (CCRH 200 n. 27), and, although it is a serious (but common) mistake to think that silent reading was not normal in antiquity, it is clear from Pliny's letters that literarv recitations were a prominent feature of contemporary intellectual life; indeed one of Pliny's recitations (Ef. 9.27) is by an historian, whom some have thought to be T. himself (see 46.4n. manet). In general see Johnson and Parker (2009); for the audience of historiography see Marincola, CCRH 11—23. antiquitus 15 picked up and elaborated by apud priores below (2). Its combination with wusitatum has reminded scholars of antiquitus institutum at Plin. Ep. 8.14.4, which 15 the first of the allusions to T.'s preface which have been detected in Pliny's letter (see Whitton

(2010)

125).

ne nostris quidem temporibus...aetas omisit: T.'s reference to his own times, by which he means the principate of Domitian (AD 81—-96: see ΡΗ 260—2)

and

to which

he returns at 1.4 below

(at nunc), is conveyed

in

the two separate expressions nosiris. .. temporibus and aetas: this form of pleonasm is sometimes labelled 'disjunctiveness' (see Harrison on Virg. Aen. 10.103) and is regarded as a characteristic feature of Propertius (e.g. 1.4.7 'formosi temporis aetas’, 1.7.8 'aetatis tempora dura’), although juxtapositions of tempus and aetas are quite common elsewhere too (e.g. Cic. Brut. 8 ‘quo tempore aetas nostra...confugere deberet’). In many cases, such as Aen. 1.246 mare...pelago premit arua sonanti (‘the sea overwhelms the fields with the resounding sea’), scholars rationalise one of the two synonyms to eliminate or mitigate the oddity (hence pelago here can be taken — ‘sea’): 50 in T. aetasis a metonymy = ‘ the age' (cf. OLD 8c), as again at Cic. Fam. 5.15.3 'in ea tempora nostra aetas incidit'. The fact that T. could have written e.g. scriptores or auctores, but instead chose to write aetas, perhaps reflects an ambivalence on his part towards a period which had been notoriously cruel but during which

70

COMMENTARY:

1.1

he himself had prospered. For the converse dualism, portraying both past and present as ‘now’, see 1.4 nunc ~ 3.1 nunc (above, p. 66). T.'slanguage recalls Mart. 8.55.1, where the poet uses the same words to contrast Domitian's reign with an earlier age (' Temporibus nostris aetas cum cedat auorum’)

and laments the absence of literary (poetic) talent.

quamquam incuriosa suorum: though many scholars believe suorum is neuter, itis far more likely to be masculine: T. is drawing a pointed contrast between distinguished living contemporaries (suorum), whom it would have been much too dangerous for anyone to praise during Domitian's reign, and the magna...ac nobilis uzrtus which was represented by the long deceased Thrasea Paetus and Helvidius Priscus, whose biographies were risked by Arulenus Rusticus and Herennius Senecio respectively (see n. magna below). quamquam without a finite verb (OLD 4) is largelv post-Ciceronian and frequent in T. (again at 16.2, 22.1, 33.1, 30.1, 44.3); the gen. after zncuriosus, frequent in T., 15 said to be first here but cf. Sen. Ep. 95.50 ‘curiosi singulorum'. quotiens 'expresses only iteration, not frequency' (Long (1911) 21, after a discussion of this example). Here the implication is: ‘on the rare occasions that...’ magna aliqua ac nobilis uirtus uicit ac supergressa est uitium: the wording is generalised, but behind the generality there is a specific, though oblique,

reference

to Thrasea

Paetus and

Helvidius

Priscus which,

in a

manner characteristic of the preface, is revealed retrospectively at 2.1 below, where they are named and their biographers identified (see nn.): likewise uzrtutibus at 1.4 below. Since the two Stoics were dead when Arulenus Rusticus and Herennius Senecio wrote their biographies, the two authors constituted a precedent for T. when, on A.'s death in 93, he was about to start writing his fatherin-law's biography (1.4n. at nunc). Conversely, since Thrasea and Helvidius represented the opposite way of life from A.'s moderatio and quies, it 15 as if T. intends to establish the latter's type of uzrtus as a worthy alternative to theirs (see 42.4). The preface's preoccupation with uzrtus (also at 2 uzrtutzs and g u?rtutes) recalls esp. Sallust (on whom see Earl (1961)). nobilis, again qualifying uirtus at Vell. 79.1, here combines the notions of ‘noteworthy’ and 'heroic' (OLD g, 6); the co-ordination with magnus is originally Ciceronian (/nv. 2.7, Verr. 3.7, ND 1.4, Dom. 115), then e.g. Liv. 42.49.2, Sen. Suas. 2.12, Sen. Ira 2.32.3. uirtusis the subject of transitive uznco elsewhere (e.g. Cic. Sest. g2, Fin. 5.38, Liv. 25.14.1, 38.17.8, cf. CIL 1?.11 'nunquam uictus est uirtutei’), and likewise uztzum can be 'conquered' (Cic. Scaur. 44, [Quint.] Decl. 2.8, HA Alex. Sev. 9.5); but, in the light of the co-ordinated supergressa est here, uicit serves to introduce a series of metaphors drawn from roads and travel (see nn. below): cf. OLD uinco b ‘to prevail against, get the

COMMENTARY:

1.2

71

better of' conditions (cf. 7c 'to succeed in passing' a barrier). The association of virtue with road metaphors is common (e.g. Sall.J. 1.3; see N-R on Hor. Odes 3.24.44): normally virtue is the destination of the traveller, but in this case it is the traveller itself. On uzrtusin general see e.g. McDonnell (2006); in T. see Vielberg (1987)

199 (index s.v.).

uitium . .. ciuitatibus commune...inuidiam: T. seems to allude to Nepos, a predecessor in the biographical genre (Nep. Chabr. 3.3 ‘est enim hoc commune uitium in magnis liberisque ciuitatibus ut inuidia gloriae comes sit').

ignorantiam recti et inuidiam: the word order exemplifies conzunctio (the joining of two terms by the interposition of a third: cf. Rhet. Herenn. 4.38) and hence indicates that rect? 15 to be taken also with znuidiam (for the con-

cept cf. Rhet. Herenn. 4.36 ‘uirtutis comes, inuidia’). In other circumstances ignorantia and ?nuidia might be mutually exclusive (cf. Quint. 5.2.3 'aut de

inuidia aut de ignorantia’); here, however, it has been suggested that T. is

proposing the unusual notion that people's lack of understanding (2gnorantia) explains their /nuzdia (Leeman (1973) 200). Normally neut. sing. rectum — ‘moral conduct’ (OLD rectus 10b), but, since it can also mean

‘the straight road' (Vell. 2.3.4 ‘ubi semel recto deerratum est’), the road metaphor (see above) is sustained; for ignorantia recti cf. Quint. 12.1.12. For discussion of znuidia see Kaster anuidia utrtutis).

(2005)

84-103

(and

182

n. 15 for

1.2 sed apud priores...: sed contrasts the scarcity of modern commemorative writing, just described (1), with its abundance in times past (2—3); apud priores, which looks back to antiquitus (1), sets Ρ the further contrast with at nuncat 4 below. apud— ‘in the time of' (OLD 14) or possibly ‘in the case of' (OLD 15). The contrast between past and present also features at D. 1.1 priora saecula — nostra . . . aetas. digna memoratu is a Latin equivalent of the Greek &&óAoya, 'things worthy of recording' (e.g. Thuc. 1.1.1; cf. Xen. Symp. 1.1, quoted above); the expression, which first occurs in Livv (4.43.2, 25.2.5), is esp. common in the elder Pliny, Gellius and the H?storia Augusta and, like memoria digna etc., is often used to underline the selectivity of an author's work and to assure readers that only the most important topics are being described (M-W on A. 4.32.1; Oaklev on Liv. 7.10.5): see e.g. H. 2.24.1 'crebra magis quam digna memoratu proelia' (and H.), Serv. Aen. 1.373 'pontifex maximus...digna memoratu notare consueuerat' (of the Annales Maximz). T. seems to represent the ancients as acting with an eye on the recording of their achievements by historians. For the abl. memoratu see OLD memoratus? (a defective noun whose only other attested form is the dat. sing.). pronum magisque in aperto erat: pronum est -- inf. = ‘it 15 easy to — ' (OLD *b), while zn aperto — 'practicable' (H. 3.56.3; OLD apertum 1c);

72

COMMENTARY:

1.3

but the words also suggest a downhill and open road (OLD pronus 4a, apertus 3a): for their combination elsewhere cf. Virg. Aen. 5.212 ‘prona petit maria et pelago decurrit aperto' (literal), Tac. Agr. 33.4 'uirtus...in aperto, omniaque prona uictoribus' (again metaphorical); uertus 15 often associated with ‘openness’ (see e.g. Vell. 2.14.3, Sen. Ep. 43.3—5 and Summers' intro. n.). T. seems also to have in mind Sall.J 5.3 'inlustria magis magisque n aperto'. The uariatio of adj. ~ prepositional phrase 15 common in T. (Sorbom (1935) 92-3). celeberrimus quisque ingenio 'all those most celebrated for their talent', as Gell. g.2.1 ‘ingenio...amoeno...celebrem’ (cf. Ov. Met. 8.159). The contrast and complementarity between action and intellect (ut agere digna memoratu ~ ita...ingenio... ad prodendam ... memoriam) once again recalls Sallust (esp. ( 1.2-3.2). ad prodendam uirtutis memoriam: uzrtutis memoria (again at À. 14.51.2) appropriately suggests the language of the late republic (Caes. G. 2.21.2, 7.62.2, 77.4, Cic. Verr. 4.78); prodere memoriam seems similar (TLL 8.675.83-676.2: Cic., Cael., Caes., Nep., but also Gell.).

sine gratia aut ambitione 15 to be taken with ad prodendam...memoriam and relates to the impartiality conventionally claimed by historical writers (G. on A. 1.1.3); gratia regularlv features in such claims (see esp. the parodv at Sen. Apoc. 1.1). For gratia combined with ambitio elsewhere cf. A. 6.46.2, Vitr. 3 praef. 3, Vell. 126.5, Gell. 14.7.9; here the two words constitute the first element in an extended chiasmus: szne gratia aut ambitione 4 bonae. .. conscientiae pretio ~ fiduciam . . morum -- nec... citra fidem aut obtrectationa. bonae tantum conscientiae pretio ducebatur: the verb, ‘was led/influenced' (OLD metaphor (above); for

16), sustains and concludes the the combination with pretio (here —

travelling 'reward")

cf. Cic. Verr. 4.124, Off. 2.21. bona conscientia, exactly equivalent to our 'good conscience', seems first to become common in the vounger Seneca, Quintilian and pseudo-Quintilian, in all of whom it is very frequent. 1.3 ac plerique suam ipsi uitam narrare fiduciam potius morum... arbitrati sunt: sc. esse (cf. 4.1n. pater): 'and in fact many thought that the personal narration of their own life [lit. 'the fact that they themselves narrated their own life’] was tantamount rather to confidence in their behaviour...’. The Latinity is extremely unusual. Since an acc. + inf. may be treated as the subject of a sentence (e.g. Cic. Dezot. 1 'est...inusitatum regem reum capitis esse’, 'that a king should be on trial for his life 15 abnormal'; G-L $535), it is regular for an acc. 4+ inf. to be the subject of an indirect statement which itself requires the acc. - inf. construction. As arbitrati sunt is followed by an indirect statement, and the subject of the indirect statement is the infinitival phrase suam ... uitam narrare, there are

COMMENTARY:

1.3

73

(as it were) two reasons whv the demonstrative pronoun should be in the accusative case, zpsos. But the pronoun has been attracted into the case of plerique, presumably to make clear that suam...witam narrare refers neither to autobiographical writing in general nor to autobiographies written by third parties but to those written by 'the many' whose opinion T. is recording (plerique... arbitrati sunt). For a similar case cf. Luc. 9.1037-8 ‘tutumque putauit | iam bonus 6556 socer' ('and he thought that now being a good father-in-law was safe’) and note Sen. Suas. 6.15 'orationes. . .illis contrarias edere ac uel ipse palam pro contione recitare pollicebatur'. For esse— ‘to be tantamount to' see OLD sum 18a. uitam narrare, again at 4 below, — 'to write (a) biography' (similar 15 Nep. Pelop. 1.1 ‘ne non uitam eius enarrare sed historiam uidear scribere’; also Plut. Alex.

1.2 o0 γὰρ

ἱστορίας γράφομεν

ἀλλὰ

Pious,

‘for we

are not

writing histories but lives’). The transition from biography to autobiography, a genre in which the subject 15 by definition still alive when the work is being written, not only emphasises and confirms the lengths to which the ancients would go to commemorate achievement (ac— 'and in fact’, ‘and what 15 more': OLD 2a) but provides a foil for T.'s own abortive biography of a deceased individual at 4 below ('suam s? uitam narrare' — 4 ‘narraturo

mihi uitam

defuncti hominis’).

It may

also be relevant

(so

Leeman (1973) 201) that T.'s biography itself has an autobiographical aspect (Intro. p. 9). For the genre of autobiography see Smith and Powell (2009).

plerique, which occurs frequently in Agr., can mean ‘many’ or 'very many' (OLD 4) and also *most' or 'the majority' (OLD 3); in each case the meaning must be decided on its merits (see G. on A. 1.9.1).

fiduciam potius morum quam adrogantiam: although adrogantiam does not require qualification, this may perhaps be another example of con:unctio (see 1n. ignorantiam, above): 'confidence in their behaviour rather than arrogance arising from it' (for a similar gen. after adrogantia cf. Cic. De or. 1.246 ‘eius artis adrogantia’; the exact phrase morum adrogantiam occurs in the fourth-cent. AD /tzn. Alex. 40). There was naturally a narrow borderline between confidence and arrogance (Sen. Ep. 87.35, Quint. 4.1.33, 12.5.2): Quintilian discussed the issue of an orator's self-praise (11.1.15—28) and 'how to praise oneself inoffensively' was a rhetorical theme, on which Plutarch wrote an essay (De se ipsum citra inuidiam laudando); see further Marincola (1997) 175-9 (with the ancient sources listed at 176 n. 3). For fiducia morum cf. Val. Max. 3.3 ext. 2 (a gen. after fiducia, again at 3.1 and 14.3, 15 regular: see OLD). nec id Rutilio et Scauro citra fidem aut obtrectationi fuit: obtrectationz is not found elsewhere as a predicative dative. citra, when used of space (

acc.), means

‘on this side of' and hence, when

used in an extended

or metaphorical sense, as here and at 35.2 below, means 'stopping short

74

COMMENTARY:

1.4

of' or simply *without' (OLD 5; Austin on Quint. 12.2.1): 'and for Rutilius and Scaurus that [sc. writing autobiography] was not without credibility or a source of criticism'. T. is saying that the response to both men's works was the opposite of that which Cicero feared if he were himself to write an account of his own consulship (Fam. 5.12.8 ‘minor sit fides, minor auctoritas, multi denique reprehendant). The uariatio of prepositional phrase — predic. dat. seems unparalleled in T. (Sórbom (1935) 81—4). P. Rutilius Rufus (cos. 105 BC) was controversially convicted in 92 of extortion and went into exile at Smyrna, where he wrote 'de uita sua’; M. Aemilius Scaurus (cos. 115 BC), a prominent politician, also wrote 'de uita sua', perhaps the first work of its kind in Latin. The two men were political rivals and each took the other to court on charges of bribery: see esp. Cic. Brut. 110—160, where the two are compared as orators (Scaurus is commended for his fides and his de uita sua 15 described as self-praise, laudes; Rutilius is described as one of the 'Stoic orators’). See further BNP

1.215 s.v.

137 Ae. Scaurus, M., and 12.794-—5 s.v. I 5 R. Rufus, P.

adeo 'so true 15 it that... ' (OLD 5).

uirtutes iisdem temporibus optime aestimantur quibus facillime gignuntur 'uzrtutes are assessed best in the periods when they are produced most easily.” uzrtus aestimaturin the singular 15 common (Cic. Hor. Liv. Val. Max. younger Sen. al.); for u. gignitur cf. Cic. Fin. 5.38, Gell. 7.1.13; for bene aestimari cf. Sen. Ira 3.5.3, Ben. 2.29.5, Mart. 10.35.10. 1.4 atnunc narraturo mihi uitam defuncti hominis uenia opus fuit quam non petissem incusaturus: this difficult sentence has generated much controversy. The kev question 15 whether fuit refers to the past from the point of view of the author (i.e. is a genuine past and refers to Domitian's reign) or from that of the reader (i.e. is a kind of epistolary tense and refers to the time of writing in Nerva's reign). There are two main arguments in favour of the former. (a) The present sentence is explained by its sequel ('tam saeua...tempora’), which almost certainlv refers to Domitian'sreign (next n.). (b) Justas apud priores (2) picked up antzquitus (1), so

nunc here picks up nostris...temporibus (1), which referred to Domitian's reign. Thus nunc, though suggesting 'presentness' (above, p. 66), refers to an occasion in the (very recent) past in contrast with an earlier period (OLD 48, cf. 7). O-R, who also believe that uenia opus fuit refers to Domitian's reign, assume that the expression means 'I had to seek permission' and that T. ‘had sounded Domitian or his court for permission (venza) to compile Agricola's biography and had received a discouraging reply’. This is surely impossible. Since Domitian is portrayed as unremittingly hostile to A. and was rumoured to have been involved in his death, as T. tells us later (43.2), itisimplausible to think that T. would ever have sought permission to write the biography: the answer would have been clear in advance. It seems

COMMENTARY:

1.4

75

much more likely that the meaning 15 as follows: ‘But, when 7 was now about to describe the life of a deceased individual, I required a reprieve which I would not have sought had I been about to criticise' (see also 3 above, n. ac plerique). In other words — no doubt responding to a sense of moral obligation and fetas (mentioned at 3.3 and 46.2), and perhaps also to the general expectation of friends and family (A.'s wife and daughter are mentioned at 45.5—40.3) — T. had been fully intending to write the life of his father-in-law but, when in 93 he heard the dramatic news — reflected in the vivid use of nunc (above) - that Rusticus and Senecio had

been executed (2.1n. legimus), he felt such time as the climate became more uzrtus which Rusticus and Senecio had of Thrasea and Helvidius (1.1) was no

compelled to excuse himself until favourable. The magna...ac nobilis seen as justifying their biographies longer an adequate precedent for

T., since the times were now so hostile to uzrtutes (next n.). The addition of

the relative clause quam non petissem incusaturus is apologetic, underlining that it was only force of circumstances which obliged him to discontinue his original intention. Had circumstances been different, he would have embarked on the biography. uenia of various sorts is a frequent prefatorial motif (e.g. Postum. Albin. fr. 1P — 1C, Liv. praef. 7, Stat. Silv. 1 praef., Quint. 1 praef. 2, 6 praef 14, 12 praef. 4). ueniam petereis an extremely common expression but 115 meaning depends upon the context: it may mean 'to seek a favour' (as e.g. Catull. 15.2) or 'to seek pardon/forgiveness' (as e.g. RG 3.1 ‘uictorque omnibus ueniam petentibus ciuibus peperci’) or, as suggested for the present passage, 'to seeka reprieve' (ase.g. A. 1.75.9 'ueniam ordinis...petenti', Plin. Ep. 5.8.11 ‘ueniam ...aduocandi peto'; OLD uenia 5); with uenia we might understand

narrand? or narrandae uitae (i.e. 'reprieve from...’).

incusaturus 15 compressed for sz incusaturus fuissem; T. is fond of such pregnant participial constructions (G. on A. 1.31.1; cf. G-L $670.4 (1)). The participle 15 either being used absolutely (as Quint. 5.8.2) or has uitam as its understood object (as Quint. 5.13.39). tam saeua et infesta uirtutibus tempora: the sentence lacks an expressed verb: should we supply suntor erant? Many scholars prefer sunt, butitseems impossible that such brutal language as saeua and infesta would be used of a period which T. will soon describe altogether differently (3.1 ‘beatissimi saeculi', 'felicitatem temporum’). The elided verb is erant, referring to the reign of Domitian (for the ellipse of forms of esse, including tenses other than the present, see 4.1n. pater). This conclusion is supported by two considerations. (i) It 15 Domitian who is defined bv saeu:tia later in the preface (3.2 'saeuitia principis’) and by hostility to virtue elsewhere in the work (41.1 'infensus uirtutibus princeps’); in fact the biography's two other references to domestic saeuitia also apply to Domitian (39.3 'saeuae cogitationis’, 45.2 'saeuus ille uultus'), while for Juvenal the reign of

76

COMMENTARY:

2.1

Domitian was summed up in the expression 'tempora saeuitiae' (4.151). (ii) The present sentence 15 exemplified by what follows (2.1—2): the uzrtutes to which the times were hostile are Thrasea and Helvidius, while the

savagery of the times is illustrated by the fates of their biographers, Rusticus and Senecio, who were executed in 93 (45.1). Thus tempora here refers to the same period as, and picks up, nostris... temporibus at 1 above. It has been noted by scholars (e.g. Lausberg (1980) 424-5) that T.’s language here echoes that of Cicero, whose fear of his own times' hostility to uertus was a reason for his not writing his encomium of Cato (Orat. 35 ‘quem ipsum numquam attigissem, tempora timens inimica uirtuti’), although Cicero in fact succumbed to friendly pressure and wrote the work. For saeua tempora elsewhere cf. Stat. Silv. 2.1.190; for nfesta cf. Cic. Corn. 1.1 Crawford, Liv. 43.22.9. For infestus see 41.1n. (infensus). 2.1 legimus...capitale fuisse: since T. illustrates the savagery of the times (last n.) by the fates of the biographers who were invoked (albeit namelessly) as his models at 1.1, it is clear that this illustration has been carefully anticipated: the fates of Arulenus Rusticus and Herennius Senecio constitute the focal point of his argument, for which 1.1—4 has been an elaborate preparation. There is of course a certain grim disingenuousness in the late revelation here that the rare examples of Domitianic biographers to whom he alluded at 1.1 were each killed for their pains; but, if the interpretation

of legimus (below)

is correct, the revelation

in

the text mirrors the experience of T. himself. legimus has given rise to numerous questions: in particular, is it present tense (legimus) or past (legzmus),? and, if it 15 past, 15 it perfect (‘we have

read’) or aorist (‘we read’)? The executions of Rusticus and Senecio

(see

nn. below) were public knowledge:* thev followed public trials in the senate,? they achieved instant notoriety, and subsequently they became emblematic of Domitian's malevolence and cruelty. The burning of the pair's books 'in comitio ac foro' (below) was specifically intended to give the incineration as much publicity as possible. If T. is thought to be saying ‘we read'

(present

tense)

or ‘we have

read'

of such

events

(in the

acta senatus or senatorial records, as supposed by those who espouse this view), that constitutes a negation of the very notoriety and publicity which defined them. The most plausible explanation for the main verb is if T. was absent from Rome at the time of the outrages and, being ignorant of matters of such public knowledge, was obliged to read about them in some organ such as the gazette which evidently circulated in the provinces (cf. A. 16.22.3 'diurna populi Romani per prouincias, per exercitus...leguntur’) 3 This question would not have arisen in a recitatio.

^ Hence Watt (1987) emended

to scimus.

5 Hence Lipsius’ emendation uidimus.

COMMENTARY:

2.1

77

or possibly in personal letters from home. Since he was absent from Rome at the time of A.’s death on 23 August g3, as he later tells us himself (45.4—5); it therefore seems reasonable to agree with those scholars (e.g. Svme (1958) 25, 68, 131) who have suggested that he was still absent at the time of the executions and book-burning, which followed shortly after A.'s death (45.1 and nn.). On this interpretation legzmusis aorist (‘we read that...it had been a capital offence’), and the plural ‘we’ is not, as some have thought, a reference to a wider group (e.g. T.'s readers and/or fellow senators) but, like nobisat 43.2,is an authorial plural (though possibly including also T.'s wife, who we know from 45.4—5 shared his absence with him).

Interesting consequences flow from this interpretation. Since the exe-

cutions of Rusticus and Senecio were so well known, T. had no need to vol-

unteer the information that four vears earlier he had learned about them from a text; indeed the first readers of his preface, unless they were personally acquainted with him and with the movements of his career, are likely to have been almost as baffled as their modern counterparts by T.'s use of legimus. Only when such readers reached the end of the biography, and discovered both that T. had been abroad when A. died (45.4—5) and that his death was followed quickly by those of Rusticus and Senecio (45.1), would they have pieced together the items of evidence and inferred retrospectively from the verb the crucial information that T. had had no part in the executions (see further 45.1nn.). For the technique whereby a later passage explains an earlier (e sequentibus praecedentia) , characteristic of the preface (1.1n. magna), see Williams (1980) 23—4 and 292 (index). cum...laudati essent: in the ancient world the dividing-line between biography and encomium was so thin as to be almost non-existent (see Intro. p. 1): hence ‘had been praised (by)’ is virtually equivalent to ‘had had their biographies written (by)'. Aruleno Rustico and Herennio Senecioni are dat. of agent after a past passive verb (A-G §375, G-L §215.1, §354). Aruleno Rustico: Q. Junius Arulenus Rusticus, brother of Junius Mauricus (45.1n.), had been suffect consul as recently as g2. He was a Stoic and an admirer of Thrasea Paetus, and as tribune in 66 had offered to inter-

vene in the latter’s trial (A. 16.26.4—5). Suet. says that he was killed ‘quod Paeti Thraseae ...laudes edidisset’ (Dom. 10, cf. Plin. Ef. 1.5.2, 3.11.3); the date was almost certainly late 93 (cf. 45.1). For him see BNP 2.82 s.v. 2 Iunius A. Rusticus, Q.; for the connections between him and the other

three individuals mentioned here see Syme, RP 7.568—-87.

Paetus Thrasea: P. Clodius Thrasea Paetus, suffect consul in 56, was the

famous Stoic whose falling out with Nero 66 are narrated by T. in A.: the text of the is in mid-suicide (16.35.2). Along with his is described by T. as the personification

and eventual work breaks fellow Stoic, of virtue (A.

condemnation in off just as Thrasea Barea Soranus, he 16.21.1 'uirtutem

78

COMMENTARY:

2.1

ipsam’), and it 15 Thrasea who, along with Helvidius Priscus, lies behind T.’s references above to u?zrtus (1.1n.) and uirtutibus (1.4). On the basis of

T.'s presumed funerary inscription (CIL 6.41106 (—60.1574)), where the letters | CITO CA[ are preserved, it has been conjectured that T.'s full name was Cornelius Tacitus Caecina Paetus and that he was distantly related to Thrasea, whose

father-in-law was A. Caecina Paetus, suffect consul in 37

(see Birley (20002) 232). See also BNP 3.4609—70 s.v. Π 15 P. C. Thrasea Paetus. T., like other authors, commonly

inverts a person’s name

(W-M

on A.

3.21.93):in this sentence he inverts the names of the subjects but not those of their biographers, thereby forming two chiastic arrangements of nomen and cognomen. Herennio Senecioni: he had been quaestor (Plin. Ef. 7.33.5) butrefused a further political career because of hisantagonism towards Domitian, who was thereby enraged (Dio 67.13.2). He was accused ‘quod de uita Heluidi libros composuisset' (Plin. Ep. 7.19.5); again the date was almost certainly late 93 (cf. 45.1). For him see BNP 6.199 s.v. Π 11 H. Senecio. Priscus Heluidius: C. Helvidius Priscus the elder married Fannia, daughter of Thrasea Paetus; he became praetor in 70 but was executed in 74 (Suet. Vesp. 15, but the date is conjectural). Like Thrasea, to whose fortunes he was inextricably bound (A. 16.28.1, 29.2, 35.1), he was a Stoic. T. provides a memorable account of his personality and earlier career at H. 4.5—6. For him see BNP 6.123 s.v. 1 C.H. Priscus; for his son see 45.1n. capitale 'a capital offence' (OLD 1). saeuitum: sc. esse. The association of imperial actions with 'savagerv' is frequent in T, e.g. À. 1.53.3, 3.24.3, 4.1.1 (and M-W), 4.33.3, 4.57.1; also above, 1.4n. (tam saeua). delegato...ministerio ut...urerentur 'the task of burning...being delegated ...*: for mznisterium (OLD 2b) explained by an utclause see e.g. Plin. Ep. 9.28.1; the combination with delegare appears to be one of T.'s many borrowings from Livy (9.29.9). An abl. abs., such as delegato... ministerio here, 15 one of T.'s commonest methods of introducing the 'appendix' by which so many of his sentences are characterised (see Intro. PP- 32-3). In many cases, as here, the participle is timeless or ‘aoristic’, i.e. it does not strictlv refer to a time prior to the governing verb. ‘It was mainly Livy and Tacitus who...sacrificed accuracy by disregarding the sense of “prior completion" which is attached to the perfect participle passive in classical Latin' (NLS §103). For book-burning, ‘that Roman peculiarity’ (Thomas (1992) 169), see A. 4.35.4; Forbes (1936), esp. 122ff., Cramer (1945) (but his discussion does not extend to the cases mentioned by T. here).

triumuiris: sc. capitalibus: ‘a commission of three...responsible for prisons and executions' (OLD 1a): it was one of the four commissions

COMMENTARY:

2.2

79

comprising the vigintivirate, membership of which represented one of the first steps in a senatorial career (6.1n. ad capessendos). T. himself had been amember of another of these commissions, the decemun stlitibus tudicandis

(Birley (2000a) 237, cf. RGB 3-4). See further Oakley on Liv. 9.46.3 with refs. monumenta...ingeniorum is a regular phrase for ‘literary works’, e.g. A. 15.41.1 (of the books burned in the Neronian fire), Sen. Cons. Helv. 1.2 ‘clarissimorum ingeniorum monumenta'. clar(issim)a ingenia is a set phrase: T. had chosen the best precedents for the work which he had planned to start on A.'s death (1.4), their brilliance matching that of any famous subject (~ 1.1 Clarorum). See also next n.

in comitio ac foro: the comitium was an area north of the Forum Romanum and in the earlier republic was used for meetings of the popular assembly; the two locations are juxtaposed as early as the XII Tables (I.7 = Rhet. Herenn. 2.20). By the time of Augustus the comitzum ‘was no longer of any political significance and had virtually disappeared' (Patterson (2000) 15 (with map)); but in texts, as here in T., the two places continued to be referred to as separate entities (TLL 3.1801.50ff.). For the burning of books ‘in comitio' cf. Liv. 40.29.14; the forum was where the body of Julius Caesar was burned (Plut. Caes. 68.1—2 with Pelling), and, since it had become an area ‘filled with monuments of the most diverse types' (BNP 5.526 s.v. ΠῚ 8 Forum Romanum), T.’s choice of monumenta to

describe the books destroyed there (last n.) is somewhat pointed.

2.2 scilicet...arbitrabantur 'doubtless they thought that...’: the subject is Domitian and his supporters, who are too well known to need, and too criminal to be graced with, identification; their suppression here contrasts with T.’s elaborate listing (1 above) of the four names which they had wished to suppress. The tricolon structure underlines the almost apocalyptic significance with which the incinerations were invested: the authorities thought that their actions would prove a deterrent to those in the future who might be tempted to voice their opinion, to use their senatorial status as a justification for self-expression, or to appoint themselves keepers of the communal

conscience. But the sentence is ironical (OLD scilicet

4a): although the wox of writers such as T. was silenced for the duration of Domitian's reign, the existence of the present biography proves that the authorities did not effect the total destruction which they imagined. Cf. A. 4.35.5 'socordiam eorum inridere libet qui praesenti potentia credunt exstingui posse etiam sequentis aeui memoriam'. uocem populi Romani: the 'voice of the people' is a relatively common expr. in both singular (e.g. H. 1.40.1, Cic. Flac. 103, Mart. 10.6.8) and plural forms (e.g. H. 3.67.2, À. 3.14.4, Mart. 8.11.2). It was proverbial that the voice of the people was powerful (Aesch. Ag. 938) or even sacred (Sen. Contr. 1.1.10; Tosi $1). Cf. also 41.3 ‘poscebatur ore uulgi dux Agricola’.

80

COMMENTARY:

2.3

libertatem senatus: ‘Tacitus seems to have regarded the freedom of the Senate as freedom par excellence’ (Wirszubski (1950) 163; cf. Morford (1991), Oakley

(20092)).

conscientiam generis humani: it is ironical that elsewhere the phrase genus humanum can be used 'to suggest the emperor's (divine) superintendence

of the entire world'

(Hoffer

(2006)

77 n. 17), as at Plin. Ep.

10.1.2 'precor ergo ut tibi et per te generi humano prospera omnia, id est digna saeculo tuo, contingant’ (and earlier, in AD 48, at SEG 4.516 (col. ii, 14—15) 'the example of the excellent and truly most just leader, who, having taken all the human race [πᾶν 16 τῶν ἀνθρώπων yévos] into his personal care...’); here itis used of the emperor's potential opponents. expulsis...sapientiae professoribus: sapientiae here — 'philosophy' (OLD

ga), as often

in T.

(who,

outside D., avoids

the Greek philosophia

except at 4.3 and H. 3.81.1), and sapientiae professores 15 something of a set phrase — 'philosophers' (e.g. Cels. praef. 7, 14, Plin. NH 7.180, Quint. 1 praef. 15), though the 'standard and inoffensive term' professor is discarded by T. after this passage (Syme (1958) 553). Whether philosophers were banished once or twice by Domitian is debated, since the evidence is conflicting (see Jones (1992) 119-20); most scholars prefer a single expulsion in 93 (thus Sherwin-White (1966) App. VI), but a few think that there was also an earlier expulsion in 89 (see Birley (2000b) 10—13). omni bona arte in exilium acta: bonae artes (plural) are usually cultural pursuits or liberal studies (OLD ars 6a), but the phrase can also have a moral reference ('good qualities’), as at Cic. Cael. 77, and this meaning is frequent in Sallust (Earl (1961) 12) and T.; but the personified singular form here

(‘all morality driven into exile’) is remarkable.

The whole phrase forms a chiasmus with expulsis . . professoribus above. ne quid usquam honestum occurreret: honestum refers to moral worth or honourableness (OLD honestus 3a) and occurreret covers a range of meanings such as ‘meet the gaze’, ‘present itself’, ‘crop up' (OLD 8-10). For the tyrant's typical hostilitv to excellence and good men see 41.1n. (?nfensus). For the cynical negative purpose clause cf. Vell. 26.2 ‘ne quid usquam malis publicis deesset'; W. on Vell. 67.3. 2.9 dedimus profecto...'Without question we gave...' The adverb (OLD profecto 1) acknowledges the temporary truth of the authorities' intentions at 2 above (arbitrabantur) and the sentence as a whole repeats, in reverse (chiastic) order, the opposites of the three elements listed there: patzentiae — conscientiam, seruitute — libertatem, adempto . . . loquendi. . . commercio (cf. cum uoce) — uocem. The plural verb refers to T. and his contemporaries, especially his fellow-senators, but it disguises the fact that, when the philosophers were banished, T. was almost certainly absent from

COMMENTARY:

2.3

81

Rome and could not be held to share in the responsibility: see further 45.1n.

(mox).

grande patientiae documentum 'an extreme example of submissiveness' (OLD grandis 4a, patientia 3a): patientiain this sense ‘could be said to define the essence of slavery', implying 'a complete absence of will'; when in T. it does not refer to physical endurance (as it does at H. 4.60.1 ‘patientiae ...documentum’, cf. Sen. Prov. 3.9), it 'denotes only servility, above all the servility of the political elite in the face of the princeps’ power’ (Kaster (2002) 138-0, 143). documentum is strictly ‘an example serving as a warning, precedent, instruction' etc. (OLD 1); compare Curt. 10.3.4 'singulare certe ediderunt patientiae exemplum' (and see Bosworth (2004) 554—5). For grande see below, 3.2n. (grande). uetus aetas evidently refers to the late republic, to judge from ultzmum in libertate below: presumably, therefore, a reference to the first century Bc as a whole. The first paragraph of the preface thus ends on a similar contrast between past and present (s?cut uetus aetas — ita nos) as it began (antiquitus — nostris.. . temporibus . . aetas).

From Cic. Mil. 77 onwards, aetas is frequently subject of uidere (again at H. 1.43.1, À. 4.31.3); the phrase becomes a mannerism of the elder Pliny (e.g. NH 2.99 ‘trinos soles ... nostra aetas uidit Diuo Claudio principe, consulatu eius Cornelio Orfito collega’). quid ultimum in libertate esset: the adj. means 'extreme in degree’, ‘the ultimate' (OLD 7a); just conceivably T. was remembering lbertas ultzma at Luc. 7.580, though the sense is different. ita nos quid in seruitute 'so we

«saw/have

seen>

what

«the

ultimate

was/is> in servitude’: the elided words have to be supplied from the s?cutclause above. seruitus, not seruitium, is T.'s regular term for 'slavery' in Agr. (8:1) and

G. (2:0) but this pattern of usage is reversed in H. and A. (see

further 29.3n. ultionem). adempto ...loquendi audiendique commercio: the appended abl. abs. (above, 2n. delegato) explains seruitute, since the denial of free speech was the mark of the slave (Phaedr. 3 frol. 34—5 'seruitus obnoxia |...quae uolebat non audebat dicere' (cf. H. 1.1.4), Prop. 1.1.28 'sit modo libertas quae uelit ira loqui', Tib. 1.5.5—6 and Maltby's n., Sen. Ep. 47.3—4); it may also be relevant that speech conventionally distinguishes humans from animals (e.g. Cic. De or. 1.32 ‘hoc enim uno praestamus uel maxime feris quod conloquimur inter nos et quod exprimere dicendo sensa possumus', /nv. 1.5) and that, just as in Soviet Russia dissidents were incarcerated in psvchiatric ‘hospitals’, madmen were punished for speaking (Cels. 3.18.21). Compare esp. the picture of Tiberian Rome at A. 4.09.3 'congressus, conloquia, notae ignotaeque aures uitari; etiam muta atque inanima, tectum et parietes, circumspectabantur'. ?2nquisitiones

82

COMMENTARY:

3.1

(again at 43.2) is a reference to spying (OLD 2c); for commercio see OLD 4b ('exchange of speech' and the like). memoriam quoque...perdidissemus, si tam in nostra potestate esset obliuisci quam tacere: the imperfect tense in the protasis indicates an unreal circumstance in the past which continues into the present (G-L $597 R. 1): forgetting was not within T.'s power during Domitian's reign, nor is it within his power now (cf. Cic. Tusc. 3.35 ‘non est enim in nostra potestate . . obliuio'). The implication 15 that he would have forgotten if he had been able: the sentence thus amounts to a wish to forget (as was said famously of Themistocles: Cic. Acad. 2.2 ‘obliuisci se malle discere’, De or. 2.299, 351 'ut obliuionis artem quam memoriae malim’, cf. Plin. Ep. 8.14.7 'obliuiscenda magis quam tenenda’), the wish being due partly to the horrific nature of Domitian's reign and partly to guilt that T., like other senators, had not dared to oppose him (for 'the representation of guilt in enduring the harshness of tyranny’ see Ferri on [Sen.] Oct. 288-0). The contrast between obliuisci and tacere (for which cf. Cic. Flac. 61 ‘liceat haec nobis, si obliuisci non possumus, at tacere’) is pointed, in that silence lay

within the power of an author who had silence inscribed in his name. The inability to forget, as described here, is a prerequisite for T.'s promise at the end of the preface that his future work will include a record of their previous servitude (3.3 ‘memoriam prioris seruitutis’); but retaining the memory of Domitian's fifteen years is naturally incompatible with the notion, to be inferred from nostri superstites below (3.2n. pauci), that everyone had experienced a collective death during them. History, silence and memory are evocative and popular themes: see e.g. Hedrick (2000), Gowing (2005), Flower (2006), Gallia (2012). For memo-

riam perdere cf. Cic. Senec. 21. 3.1 redit animus is a very common phrase, esp. in Ovid (for some exs. see Gibson on Ov. AA 3.707), but the exact meaning of anzmus will depend upon the context. At 26.2 nonanis rediit animus the contrast is with znter somnum ac trepidationem earlier; ‘spirit’ or 'courage' 15 the expected contrast with trepidatio, but it is 'consciousness' which contrasts with the drowsiness of somnus. A somewhat similar ambiguity is perhaps present in our passage: the immediate contrast 15 with the fear implied at 2.3 above (esp. znquisl2ones) ; but T. is later to describe Domitian's reign in terms of suspended animation or living death (2n. pauc?): hence 'spirit' or ‘courage’ 15 the initial meaning, but ‘consciousness’ becomes active later as we read through the passage. For the common motif of ‘return’ or ‘restoration’ at the start of a new reign see W. on Vell. 89.2—6,

126

(intro. n.); also below

(n. res). redit 15

generally regarded as present tense, although redit would produce the fashionable clausula of cretic -- resolved spondee (demum redit ánimüs, cf. 1.1 rect(?) &t inuidiam), favoured by both Plinys, both Senecas and

COMMENTARY:

Quintilian (see Whitton

4.}

83

(2013), Intro. $5 ‘Rhythm’). Spengel suggested

rediit, ‘has returned'.

et 'and yet' (OLD 14a). quamquam...miscuerit...augeatque...suaserit: the subjunctive after quamquam occurs first in Livy and has become regular by the time of T. (NLS $245 Note, $249 (b); Carmody (1926) 119-21); and, whereas most first-century AD writers prefer quamuis to quamquam, T. generally prefers the latter (Adams

(1972)

362).

primo statim beatissimi saeculi ortu: similar language at 44.5 (‘hanc beatissimi saeculi lucem ac principem Traianum’) suggests that the image is that of a new dawn (OLD ortus 1b, 3). Though primus ortus is more commonly used of things other than dawn (thus e.g. A. 6.22.3), this 15 not always so: cf. Virg. Aen. 4.118, Sil. 13.132, and esp. Stat. Silv. 4.3.112 'primo ...ortu'. T. is contrasting the dawn of Nerva in the recent past (màscuerit) with the continuing daily efforts of his partner Trajan in the present (~ augeatque cotidie). The background to this imagery 15 the convention of comparing or identifving the ruler with the Sun (see further Du Quesnay (1995) 157). statim 15 regularly coupled with primus and similar words = ‘at/from the very start' etc., as again at 20.1 and 31.4 (OLD statzm 2b). ‘Saecular’ language was to be expected at the start of a new reign (cf. Calp. Sic. 1.44-5 'iuuenemque beata sequuntur | saecula’), and, just as Horace identified Augustus with his age (Odes 4.15.4 'tua, Caesar, aetas’: see Breed (2004)), so Pliny shortly after Nerva's death was writing to the new emperor Trajan with the phrase 'saeculo tuo' (Ep. 10.1.2 with Sherwin-White's n.), while the latter in his turn would later refer to ‘nostri

saeculi' (Ep. 10.97.2). See also below (n. felicitatem) , Courtnev on Juv. 4.68, Roche (2006) 199. Nerua Caesar: Nerva died on 27 January 98 and was deified. In the De aquaeductu, written by T.’s contemporary Frontinus, Nerva 15 Dzuus at 102.4 and 118.3 but not elsewhere (e.g. 1, 64.1); the inference seems to be that Nerva died during Frontinus' writing of the work, and, since Nerva is not called Dzuus by T. here (contrast H. 1.1.4) and Trajan is described as princeps at 44.5, some scholars believe that the same holds good for T. and the Agricola (see e.g. O-R 10-11, PH 259). See also below (n. Nerua Traianus).

res olim dissociabiles miscuerit — principatum ac libertatem: olim seems to combine the notions of 'formerly' and 'for a long time past' (OLD 1a, 2a). Since miscereis commonly used of mixing remedies (e.g. Cels. 3.6.16, 3.10.1), the verb perhaps anticipates the forthcoming medical imagery (see below): if so, Nerva is being seen as the caring doctor (an ancient and familiar metaphor: see e.g. Woodman (2010) 45 and n. g = PH 164 and n. 9). There is perhaps also a hint of the ‘mixed’ constitution (for which 566 A. 4.33.1, Plb. 6.3-10 with Walbank; von Fritz (1954), Lintott (1097)):

84

COMMENTARY:

3.1

cf. Cic. Rep. 2.42 ‘haec...ita mixta fuerunt...in hac ciuitate'. d2ssociabilis (for which see N-H on Hor. Odes 1.3.22) is extremely rare and its juxtaposition with m2scuerit is pointed, thereby emphasising Nerva's achievement: an analogous point 15 made by Pliny of Trajan (Pan. 24.1 'iunxisti enim ac miscuisti res diuersissimas, securitatem olim imperantis et incipientis pudorem’). An inscription dated on the day of Nerva's election (18 September 96) proclaimed Libertas Restituta (ILS 274, cf. Plin. Ep. 9.13.4 ‘primis diebus redditae libertatis’), Nero's coins displayed Liberlas Publica (BMC Emp. g, nos. 16, 46, 60, g1 etc.), and the relationship between principatus and . libertas occurs relatively often in earlv imperial texts (e.g. Mart. 5.19.6 'sub quo libertas principe tanta fuit?', Plin. Pan. 36.4 ‘eodem foro utuntur principatus et libertas', 44.6 ‘eadem quippe sub principe uirtutibus praemia quae in libertate', Plut. Galba 6.4 διαφυλάσσειν &pa τὴν ἡγεμονίαν kai τὴν ἐλευθερίαν (Galba writing to Verginius);

cf. Luc. 7.695-6 ‘par quod semper habemus, | libertas et Caesar', Tac. H. 4.64.2 ‘haud facile libertas et domini miscentur’); see also Lausberg (1980) 427-8. felicitatem temporum is the very phrase used by Nerva himself to describe his new reign (in an edict quoted by Plin. Ep. 10.58.7); T. repeats it of the reigns of Nerva and Trajan jointly (H. 1.1.4), Pliny applies it again to Trajan's reign (Ep. 10.12.2 'felicitas temporum, quae bonam conscientiam ciuium tuorum ad usum indulgentiae tuae prouocat et attollit’, cf. Pan. 74.1—5), and it was used on coins in the Antonine age to celebrate the birth of a new heir (Ferri on [Sen.] Oct. 280—1). Similar phraseology is found applied to Julius Julianus' governorship of Judaea in AD 125 (P.Yadin 15.10 τοῖς μακαριωτάτοις kaipols, ‘the happiest times': see Cotton

(1993) 104). Trajan would later refer to his own age as 'nostrorum temporum' (ap. Plin. Ep. 10.55). For political felicitas in general see W. on Vell. 100.3; for felicitatem augere cf. Cic. Inv. 1.94, Val. Max. 5.3.2a. It is often thought that T. cannot have been sincere in repeating political slogans such as this and the others in this sentence (see e.g. Leeman (1973) 203). Of course we cannot ever know whether T. as an individual really believed what he wrote, but, since the sentence as a whole is designed to defend Nerva and Trajan against possible criticism (see nn. below), it would be perverse to imagine that he undermined that defence by a use of language which was intended to be thought insincere. The context requires that he be thought perfectly sincere, and contemporaries for their part would not have been predisposed to question conformist terminology from an author whose political conformity had brought him such political success under Domitian (cf. Plin. Ef. 4.22.6). Itis too easy to read back into the Agricola the authorial attitudes which are commonly associated with the Histories and, especially, the Annals (being written about fifteen vears later, grande mortalis aeui spatium).

COMMENTARY:

4.}

85

Nerua Traianus: the former name indicates that T. is writing after Trajan's adoption by Nerva in the autumn (probably late October) of g7 (for the alternatives of mid-September or early November see Bennett (2001)

246 n. g1). Some scholars think that Trajan was already emperor when T. wrote the preface, as he clearly was when T. wrote 44.5 ('principem Traianum’); others believe that, even if the absence here of such a designation

as princeps or imperatoris regarded as without significance, the form Nerua Traianus — unparalleled except as part of Trajan's full title at Frontin. Aq. 03.4 — 'verv stronglv suggests that Trajan's adoption by Nerva was not only very recent but is also the most significant thing about him' (PH 259 n. 5). That Trajan nevertheless wished to remain his own man is suggested by the fact that, although he assumed the cognomen Nerva, he did not also take over Nerva's gentilicium Cocceius: ‘he remained Ulpius'

(Eck (2002)

225,

q.v.). nec spem modo ac uotum securitas publica, sed ipsius uoti fiduciam ac robur aetas suaserit: the transmitted reading robur adsumpserit is more difficult than is usually allowed. (a) The first clause lacks a verb, which is normally said to be supplied from the second clause by zeugma. Since the required sense would seem to be 'and the public security has not only encouraged our hopes and prayers', the verb to be supplied would be e.g. fouerit, which can be used with both ‘hope’ and 'prayer' asits objects (OLD 7). (b) adsumpseritin the second clause means ‘has taken on' (cf. OLD 10);

but it makes no sense to say that 'the public security ... has taken on confidence in its prayers’: we would expect adsumpserit to have as 115 subject a noun such as aetas, 'the age', to be supplied from the first clause bv a further zeugma. This seems one difficulty too far and it is possible that the text is corrupt. Perhaps adsumpserit hides an expression such as aetas suaserit (for which cf. Ov. F. 6.86, Plin. Ef. 4.24.3): 'and not onlv has the public security encouraged our hopes and prayers but the age has encouraged confidence in the effectiveness of [or perhaps 'strong confidence in’] those very prayers' (taking fiduciam ac robur as a kind of hendiadys). Reference

to 'hope(s)' and

'prayer(s)', a frequent combination,

is to

be expected at moments of political optimism (e.g. Vell. 103.4 'uota paene inserentium caelo manus spemque conceptam perpetuae securitatis', of Tiberius' adoption bv Augustus). securitas publica (again at Sen. Contr. 9.4.4, Sen. Ben. 6.15.8, Clem. 1.13.1, Ep. 73.2, [Quint.] Decl. 348.12, Flor. 2.19.1, Just. 18.3.18) recallsanother contemporary slogan: securitatem omn?um occurs in the same edict of Nerva as referred to above (n. felicitatem). For uoti fiducia cf. Plin. Pan. 94.5 ‘quo maiore fiducia isdem illis uotis...oro et obtestor'. natura tamen infirmitatis humanae: this is the first of the extenuating circumstances which T. adduces to explain why, despite the ministrations of Nerva and Trajan, intellectual life (ingenia studiaque below) has not made

86

COMMENTARY:

3.1

an immediate and complete recovery. The reference to human weakness 15 picked up by subit... amatur below, while tardzora . .. mala, which follows,

is explained by et ut...reuocaueris: the arrangement 15 thus chiastic. For infirmitas humana cf. D. 25.6, Liv. 30.31.6, [Quint.] Decl. 5.15 'humanae infirmitatis ista natura'. tardiora .. . reuocaueris: the proverbial expression fardiora... mala (next n.) is explained by an analogy (ut...sic...), both parts of which are further sub-divided into two: 'tardiora sunt remedia [a] quam mala [b]; et, ut

corpora nostra lente augescunt [a®], cito extinguuntur [b*], sic ingenia studiaque oppresseris [b'] facilius quam reuocaueris [a']’. The initial proverb

generates a chiasmus with the sicclause (a + b — b' - a'), which itself concludes another chiasmus with the utclause (a? + b? — b! - a'). See

further nn. below. tardiorà sunt rémédià quam mala: remedia indicates that mala has its common meaning of 'diseases' (OLD malum 7b, TLL 8.229.68ff.). The speed of disease (e.g. À. 15.38.3 'anteiit remedia uelocitate mali' (Nero's fire seen metaphorically as a fast-spreading epidemic), Sen. /ra 3.1.2 ‘lentum praecipitis mali remedium', Hippocr. Prognosticon init. ἐπειδὴ 8¢ oi ἄνθρωποι

ἀποθνήσκουσιν, ol p£v πρὶν f) καλέσαι TÓv inTpov, ὑπὸ τῆς ἰσχύος τῆς voucou, oi 6E Kai ἐσκαλεσάμενοι παραχρῆμα ἐτελεύτησαν, Ol μὲν ἡμέρην ζήσαντες, ol 66 dAiyw πλείονα χρόνον, πρὶν ἢ τὸν ἰητρὸν τῇ τέχνῃ πρὸς ἕκαστον νόσημα ἀνταγωνίσασθαι,) 'Some men, when they die, do so before calling the doctor, because

of the power of the disease, while others pass away immediately after calling him, some living for a day, others for a little longer but before the doctor by his skill can counteract each illness’), the slowness of medicine (Cic. Tusc. 3.35 ‘tarda...quidem medicina sed tamen magna’, Colum. 2.17.9 ‘pigriora...remedia’, Curt. 3.5.193 ‘lenta remedia’) and its untimeliness (Prop. 2.14.16, Ov. RA g1-2 with Pinotti, [Quint.] Decl. 12.23) are the related commonplaces lying behind T.’s expression, which, with its trochaic rhythm, reads like a proverb (see Tosi $1638; PH 271 n. 45) and perhaps deserves to be placed within quotation marks. et 'that is to say' (OLD

11): since the proverb

(last n.) is expressed in

metaphorical terms, its relevance requires explanation. ut...sic...:thespeed with which the body dies (ut corpora .. . cito extinguuntur) corresponds to the ease with which intellectual life is overwhelmed (szc ingenia studiaque oppresseris facilius); but the slowness with which the body develops (lente augescunt) corresponds only formally to the difficult resuscitation

of intellectual life (quam reuocaueris): there can be no

logical correspondence between the two because, whereas the development of the body is a precondition of its eventual death, exactly the

5 This reference is owed to Dr R. Flemming.

COMMENTARY:

3.1

87

opposite is the case with intellectual life: it is the overwhelming of intellectual life which is the precondition for its eventual resuscitation. The slow development of the human body stands out as the one anomalous element in this analogical sentence; but its presence is to be explained as a defence of Nerva and Trajan. Since the human body takes many years to reach maturity, readers must not expect Nerva and Trajan to have resuscitated ingenia studiaque fully within the few months that have passed since Domitian's death. The comparison with the body is designed to insinuate this temporal contrast and to deflect any blame away from Nerva and Trajan. cito extinguuntur ‘ are quickly extinguished’ (adversative asyndeton). For the idea cf. esp. Sen. NQ 3.27.2, Ep. 91.6. sic ingenia studiaque oppresseris facilius quam reuocaueris: ingenia studiaqueare a common combination (e.g. D. 14.2), especially in Cicero (e.g. Arch. 19, De or. 1.131, 3.16), to refer to intellectual or literary activity and the like. The complaint that literary (and esp. oratorical) standards were declining (e.g. Cic. Tusc. 2.5) became a commonplace in the first century AD (e.g. Sen. Contr. 1 praef. 6—10; Williams (1978) 6—51) and 15 of course the subject of T.'s Dialogus (see Mayer (2001) 12-16), whose dramatic date is 75. In the present sentence this notion is (as it were) applied to the particular circumstances of Domitian's principate and its aftermath, implying (but not expressly stating) that the prznceps had suppressed intellectual life, i.e. a much broader phenomenon than the biographical scarcity implied at 1.1 above.

(In fact major literary figures such as Martial, Silius, Statius,

Quintilian and perhaps Valerius Flaccus all published during the period; see also next n.) Since opprimere can be used of an overwhelming illness or

of 'instruments of death' (OLD 4b), and since reuocare means 'to recall to health or vigour, restore, revive' (OLD 11), both verbs continue the medi-

cal imagery of the passage. The plight of intellectual life is an illustration of the general truth that in the case of some

'diseases'

(mala) there is no

quick cure (compare also Lucr. 1.556—7 'quiduis citius dissolui posse uidemus | quam rursus refici’). oppresseris and reuocaueris are examples of the second-person singular potential subjunctive used to expressan indefinite ‘you’, an idiom in which ‘there is usually no observable distinction in sense between the present and the perfect tenses' (NLS§119): hence ‘you overwhelm talent and intellectual activity more easily than revive them'. The use of this subjunctive in narrative (as opposed to in speech) is more common in Latin historiography than biography, since it is largely absent from Nepos and Suetonius; yet, though T. is relativelv partial to the idiom in his major works (Carmody (1926) 53), itisdeploved with remarkable frequency in the Agrzcola, occurring on average once every 4.5 pages (11.3, 12.3, 22.4, 42.3, 44.2, 40.3: see nn.): this statistic, which excludes anv example in speech (g0.3, [?]

88

COMMENTARY:

3.2

32.2),isa far higher proportion than is found even in Velleius and Florus, the two historians to whom the idiom most appealed." In the present passage the usage constitutes a further exculpating device, inviting readers to reflect whether they could do any better than Nerva and Trajan in reviving literary activity (cf. refutaueris at A. 4.11.1, which ‘invites the reader's active thought', according to Gilmartin (1975) 119). subit quippe...amatur: T. continues his defence of Nerva and Trajan, offering two reasons why it has been difficult to revive intellectual life: the first (quippe), chiastically arranged, is that potential authors have been seduced

by idleness

(cf. ῥαθυμία

at [Long.]

Subl. 44.11);

the

second

is

given below (2 quid s?...). Pliny too comments about the desidia of the same period but restricts it to the audiences at recitations (Ε. 1.13.5 'auditorum ...desidia'): in his view, intellectual life 15 otherwise thriving (1.13.1 ‘uigent studia, proferunt se ingenia’, cf. 1.10.1 ‘urbs nostra liberalibus studiis...nunc maxime floret’); but T. wishes to play down the literary revival in order both to imply his own distinctiveness and to excuse his alleged authorial inexpertise (below, 3n. uel). The conceit of znuzsa primo — postremo amatur resembles Sen. Ep. 112.4 'homines uitia sua et amant simul et oderunt' (cf. 116.8 ‘uitia nostra...amamus’), but 566 esp. Auson. Ep. 23.29-31 'uitium fouet ipsa suum cessatio iugis. | dumque pudet tacuisse diu, placet officiorum | non seruare uices et amant longa otia culpam'. znertza and desidia are virtually synonymous (e.g. Cic. Leg. agr. 2.103, Plin. Ep. 2.10.8). subit = ‘comes over «one-": cf. Liv. 2.42.1 ‘dulcedo ...legis...subibat animos’. 3.2 quid si...uenimus?: T.'s second reason for the slowness of the literary revival is introduced by asyndeton and expressed by means of a difficult rhetorical question. quid 51 is a common idiom in which the reader supplies from the context the verb which has been elided after quid: see Oakley on Liv. 9.18.5, and cf. Williams (1957) 242-3 (the suggested meanings and elisions given at OLD quis' 13a are too restricted). Hence: ‘And what «can be done> if during fifteen years, a substantial period of mortal existence, many men fell to chance circumstances (all the most enthusiastic to the savagery of the princeps), and only a few of us are survivors — and, if I may say it thus, not merely of those others but also of ourselves, having had removed from the middle of our lives a total of years in which the young amongst us arrived at old age, and the elderly at almost the very boundaries of their completed span, but in silence?' This second reason, like the first (last n.), is arranged chiastically:

7 According to the standard discussion by Gilmartin (1975), Velleius uses the device once every 14.6 pages, Florus once every 12.6 pages (107 and n. 28). It is not clear why Gilmartin says that the Agricola has only the two examples at 22.4 and

44-2 (115 n. 33).

references spatium)

to the

length

and its deaths

COMMENTARY:

3.2

of Domitian's

reign

89

(per quindecim

annos...

(multe. .. interciderunt) are followed by references

to survivors of Domitian's reign (paucz...sumus) and 115 length (exemptus e media uita tot annis. .. uenimus). See also next nn.

Since mult:. . . interciderunt 15 a statement of indisputable fact, its appearance in a conditional clause must be merely a manner of speaking: with s2 we are obliged to supply mentally ‘as was in fact the case’. Yet it still seems very odd to be assured that during a fifteen-year period many people died from natural causes (fortuitis casibus). T.'s second explanation for the slowness of the literary revival appears to be the simple fact that there no longer exists a critical mass of potential writers: because fifteen years is so long a period of time ('grande mortalis aeui spatium’), it saw many people die off; as a result, only a few (paucti) remain, and, of those who remain, the once

vigorous have become elderly (zuuenes ad senectutem...), the once elderly are almost at death's door

(senes prope ad ipsos exactae aetatis terminos) , and

both groups are badly out of practice (per silentium). In other words, the few persons who could have constituted a literary revival lack the stamina and experience which are required for the task. There is only the rare exception such as T. himself, and, after the adverse circumstances of Domi-

tian's reign, he has only an unrefined and crude voice with which to implement his plan of writing a history of the vears 81 to the present (3.3 ‘incondita ac rudi uoce’): for a similar point η Plin. Ep. 8.14.2—3 (‘priorum temporum seruitus...aliarum optimarum artium ...obliuionem quandam et ignorantiam induxit...itaque reducta libertas rudes nos et imperitos deprehendit') see Whitton (2010) 123-4; also above, 1.1n. antzquitus and below, 3.3n. non tamen. per quindecim annos: 81—906, the reign of Domitian. grande mortalis aeui spatium: mortale aeuum is a Senecan expression (Tro. 52, Ep. 102.23, Cons. Marc. 25.1); the adj. underlines the fact that T. 15 talking about the death-rate (see above). The average life expectancv for a Roman was 25, but a Roman who had successfully reached 25 when Domitian became princeps, as was roughly the case with T. himself, could expect to live on average a further 32 years (Parkin (2003) 280, Cokayne (2003) 2—9): the principate of Domitian thus represented just over a quarter of such a person's life. For aeu? spatium (again at G. 37.3) cf. Ov. Met. 15.874 (the epilogue), Vell. 2.9.3, 89.6 (and W.), Sen. HF 749 (plur.), Vict. Caes. 35.193. grandis ‘probably had a freshness which magnus may have lacked’ (Brink on Hor. Ep. 2.2.178—9). promptissimus quisque saeuitia principis: promptizsszmus quisque occurs first in Cato, Orzg. 83P — 76C, then in Sall. H. 1.77.1 ‘rem publicam ...in periculis a promptissimo quoque defendi’, Liv. 23.18.5, 28.13.7; T. has it 4X elsewhere (H. 1.51.5, A. 2.81.2, 4.51.3, 12.38.3). The present example would be esp. pointed if T. had Sallust in mind. For the tyrant's tvpical

90

COMMENTARY:

3.2

elimination of good men see 41.1n. (?nfensus); for his saeuztia (s. principis again at Á. 1.74.2, 10.19.2) see 1.4n. (tam saeua), 2.1n. (saeuitum). A.’s own deliberate pursuit of otzum in Domitian's reign (40.4, cf. 42.1) precludes his being regarded as one of the prompt.. It is possible that promptissimus...principis is to be regarded as quasiparenthetic since it shares the same verb as multi and since multi — pauci is a regular distribution. pauci et...nostri superstites sumus 'a few of us are survivors and (if I may speak like this) not only of others but even/also of ourselves’: the sense of etis ‘and whatis more' (OLD 1a), emphasising the unusualness of the idea. (sumus cannot mean

'we were and still are survivors', since that

would require a longer period than the ‘quindecim annos' which T. specifies. Since the verb must therefore be a genuine present tense, it cannot be taken with per quindecim . . . aeui spatium, which refers to the past. Evidently the szclause is ‘punctuated’ by its two finite verbs, znterciderunt and sumus.) When T. says ‘we are survivors of ourselves', that *we' 15 divided into two elements, superstites sumus and nostrz, and, since those whom *we' have survived are also ‘we’ (nostr?), the dead *we' and the living ‘we’ are identical

with each other. T. 15 evidently deploying the notion of the 'divided self'. This notion is as old as Homer, whose protagonists are often represented as speaking to their θύμος or heart (see e.g. Rutherford (1989) 14ff.), but the earliest example involving a reflexive pronoun, as in T., appears to be Heraclitus, who said that ‘he went in search of himself’ (fr. 101 — Diog. Laer. 9.5 αὑτὸν... διζήσασθαι, cf. Sorabji (2006) g4, 52). Such reflexive expressions are associated particularly with the younger Seneca, in whom they are especially common (see Bartsch (2006) 247). Since the *we' (nostri) whom we have survived (superstites sumus) did not

literally die, T. is talking in terms of a metaphorical death and a metaphorical resurrection. Death is used frequently and variously as a metaphor, but it is more difficult to find models or parallels for the notion of resurrection. Cicero referred to his return from exile as the 'start of a second life' (Att. 4.1(73).8 'alterius uitae quoddam initium ordimur’) and a ‘birthday' (Att. 3.20(65).1). In other words there were two consecutive Ciceros: there was an old Cicero, who died and was succeeded by a second, new

Cicero. This is not quite the same as our passage, and the same is true of a cluster of Senecan texts too (Tranq. 11.12 ‘morti suae superstes’, Ef. 30.5 'Bassus noster uidebatur mihi prosequi se et componere et uiuere tamquam superstes sibi et sapienter ferre desiderium sui’, Ep. 32.5 ‘uiuit uita peracta’). Perhaps the most interesting passage comes from Lucretius, who in Book g discusses the man who resents the fact that after death he will be buried or cremated (870ff.). Such a man, says Lucretius, ‘does not remove or cast himself from life by the roots but unknowingly makes something of himself survive [esse...super]’; failing to make a distinction

COMMENTARY:

3.3

91

between himself and his dead body, he imagines himself to be that body ‘and, standing beside it, infects it with his own feeling... He does not see

that in real death there will be no other self [nullum. .. alzum se] to live and bewail his death’. The non-Epicurean’s notion that after his real death he will in some sense nevertheless survive is not unlike T.’s notion that he and others are survivors of their metaphorical death during Domitian's reign; but it is not identical and it will be clear from the preceding survey that nostri superstites sumus represents a highly unusual — and possibly unique — deployment of the 'divided self’. See further PH 277-81. The expression ut ?ta dixerim 15 found in Quint. 1.12.2, 9.4.01, Plin. Ep. 2.5.6, Front. Ant. elog. 2.2 (p. 135.17 vdH"), Apul. Flor. 2, Deo Socr. 19. Some scholars prefer to read Wolfflin's s?c, and it 15 true that ut sic dixerim occurs elsewhere in T. (D. 34.2, 40.3, G. 2.1, A. 14.53.4) and in Pliny (Pan. 42.3) and is much more common in Quint., ps.-Quint. and Apuleius; but the paradosis μ suggests that T. wrote 2£a here. The ut-clause 15 formally purposive, the 'polite' use of the perfect subjunctive (10.6n. addiderim) having replaced the present d?cam, which is regular in Cicero and other earlier authors. See further Mayer on D. 34.2. exemptis...annis: the exact expression seems unparalleled but is perhaps coined on the basis of the technical dies exzmere, ο pronounce days unfit for the transaction of business' (OLD eximo 3d, e.g. H. 3.81.2): in the case of Domitian's reign it was not individual days, but fifteen consecutive years, which were ‘pronounced’ unfit for the normal business of life. For demere annos cf. Hor. Odes 2.5.14 and N-H's note on related ideas. iuuenes ad senectutem, senes prope ad...aetatis terminos per silentium uenimus: though it is hard to define the age at which senectus was thought to start (it varied from 42 to 77: see Parkin (2003) 15-26, 279), T. himself, born between 56 and 58 (cf. Birley (2000a) 234-6), will not yet have become a senex when Domitian died. T. leaves unspoken (as needing no explanation) the fact that a third group, the adulescentes who have become tuuenes, were deterred by Domitian's saeuitia from ever embarking on literary training and so are in no position now to help revive literature. The idea of passing through life in silence is perhaps another Sallustian theme (cf. C. 1.1 ‘ne uitam silentio transeant', where silentio means ‘not speaking/writing' as well as ‘not being spoken about’). The repetition of senectutem ~ senes falls into the category of ‘climax’, viz. ‘a chain of cola in which each colon repeats a new element of the preceding colon' (Wills (1996) 320, remarking that it 15 relatively rare in prose). For exactae aetatis cf. OLD exigo 6b; for aetatis termini cf. H. 4.51.2, Sen. Ot. 5.7. aetatisis the last of three nouns — 'life' in this section: for such uariatio of vocabulary see Sorbom (1935) 16-29, esp. 28. 3.3 non tamen pigebit...memoriam prioris seruitutis ac testimonium praesentium bonorum composuisse: it was conventional for an author to

92

COMMENTARY:

3.3

state that he would deal with a given subject in a future work (W. on Vell. 89.1, 96.3): usually these statements are panegyrical in function (cf. H. 1.1.4 ‘quod si uita suppeditet, principatum diui Neruae et imperium Traiani, uberiorem securioremque materiam, senectuti seposui, rara temporum felicitate ubi sentire quae uelis et quae sentias dicere licet’) and their sincerity is often doubted; here T.’s future work will incorporate criticism (memoriam prioris seruitutis — Domitian's reign, picking up seruitute at 2.9) as well as panegyric (testzmontum praesentium bonorum — Nerva and Trajan, picking up beatissimz saeculi and felicitatem temporum at 3.1), and in the event he did write about Domitian's reign by including it in his H?storzes; but his promise that he would deal with Nerva and Trajan, which is repeated in the Histories (see above), remained unfulfilled. For the suggestion that przoris seruitutis is alluded to at Plin. Ep. 8.14.2 ‘priorum temporum seruitus’ see Whitton

(2010)

123.

Rather than being merely equivalent to componere, the perfect tense composuisse is usually explained as retrospective from the time of the future work's publication (as Liv. praef. 3 'iuuabit...consuluisse', Quint. 3.1.22 ‘non tamen ... pigebit meam quibusdam locis posuisse sententiam’). The verb shows that memoriam must here have its concrete sense of ‘memorial’ or ‘record’ (as Suet. Claud. 1.5 *uitae memoriam prosa oratione composuit’; OLD 8a); on the other occasions when the noun is juxtaposed with testimoniumit means 'memory' (Cic. Font. 23, Phil. 11.35). testimonzum here suggests praise, ‘testimonial’ (cf. OLD g); for its combination with componere cf. Cic. Att. 15.15.1 (where it means ‘testimony’). composuisse is the first of the so-called 'heroic clausulae' marking sentence-end in the Agricola. The others are spread more or less evenly throughout the work, except for a noticeable sparsity in the central military narrative and some evidence of minor clustering: cf. 4.1 nobilitas est; 10.9 cuneum tenuatur (n.), 10.4 nuenit domuitque, 18.5 periculum placuisset, 18.6 continuisse. Their total amounts to approx. 4% of the whole, significantly less than Sallust (10.6%) and Livy (11%), double that of Cicero (1.9%),

and

half the

‘norm’

for other

authors

(8.3%):

see Wilkinson

(1963) 161, Hellegouarc'h (1991) 2440. Quadrisyllabic endings are relatively rare in 'classical' hexameter poetry, and pentasyllabic endings even rarer (statistics in Bailey (1947) 1.115). uel incondita ac rudi uoce 'even if only [OLD uel 6] in an unrefined and crude voice’: uoce, which here refers to language/speech (OLD g), looks back to 2.3 uoce T. has now rediscovered, however inadequately, the voice he had ‘lost’ there. It was conventional for an author, esp. in a preface, to express his literary inadequacy and to state that his subject is bevond his powers of praise or criticism (e.g. Rhet. Herenn. 3.11; W. on Vell. 67.1, 104.4, 126.1). Here the statement, which is applied to the praise and criticism required in a future work (last n.) and whose relevance to the present

COMMENTARY:

4

93

work has to be inferred (below, n. professione), is made to arise as a natural consequence of the fifteen years' silence suggested by T.'s metaphorical death during Domitian's reign (2 per silentium). inconditus 15 often used of speech or language (OLD 1a; with uox again at Cic. Rep. §.3, Gell. 16.8.5); for its combination with rud?s cf. Curt. 7.8.10 'barbaris rudis et inconditus sensus’; oral delivery is the context at Quint. 11.3.32 ‘uox...rudis’ (as perhaps also at Cic. De or. 3.175 ‘quod ille rudis incondite fundit’). Scholars have again seen T.'s words as being alluded to in Plin. Ep. 8.14.3 ‘rudes nos et imperitos': see Whitton (2010) 124, and above, 3.2n. (quid st). hic interim liber: the ‘promise of a future work' topos (above) inevitably casts one's present text as a temporising production (see esp. Virg. G. 3.40 ‘interea Dryadum siluas saltusque sequamur’). Since adverbs of time and place are quite frequently used attributively as if they were adjs. (25.1n. motus), the word order may suggest that znterzm 15 to be taken with lber, ‘this interim book’; on the other hand, znterm is frequently interposed without any such suggestion, e.g. H. 1.80.1 'paruo interim initio', 3.1.2 ‘insessis interim Alpibus', 3.40.1 ‘Fabius interim Valens'. honori Agricolae soceri mei destinatus: on the first occasion that A. is named, T. states his relationship with his subject, as was expected in a preface (Rhet. Herenn. 3.11). This prepares for the justification of the biography: f?etas (next n.). professione pietatis aut laudatus erit aut excusatus: T.'s argumentis this: since he will be lacking literary expertise when he comes to write his future historical work (above, n. uel), there is an afortioriassumption - though he carefully avoids saying it — that he will be even more inexpert in the case of any work written prior to that. He thus has to produce some special justification for writing the present biography of A., viz. pietas (professioneis causal abl., ‘because of its declaration of devotion'). Compare how Propertius hopes that, if his literary strength is inadequate, he will be praised for his boldness (2.10.5—6 ‘quod si deficient uires, audacia certe | laus erit’).

4—9

EARLIER

YEARS

(AD 40-77)

Readers are introduced to T.’s father-in-law by an account of his earlier years which extends from his birth and origins (4.1-2) to his consulship and appointment to the governorship of Britain (9.6). The pattern of his life is established from the first: as a vouth he had a fiery mind (4.3 'incensum ac flagrantem animum") and he coveted gloria, but reason intervened (ratio) and his wisdom

led to moderation

('ex sapientia modum’).

As a

new recruit in Britain, though he did not show off (5.1 'nihil...in iactationem’), he displaved effort (zntentus) and a desire for glory (5.3 ‘gloriae cupido’); but subsequently in Rome he was a man of realistic quietism

94

COMMENTARY:

4.1

(6.3 ‘quiete et otio...inertia...silentium’), combining wisdom, moderation and reason (6.3—4 sapientia, medio, rationis). As a result he acquired a reputation despite himself (6.4 famae). In the aftermath of the civil war of 68—9 he again displayed effort (7.9 strenue), but, dispatched to Britain, initially he was obliged to curb his fiery spirit (8.1 'temperauit...uim suam ardoremque compescuit'); subsequently, though he did not show off to promote his own reputation (8.9 ‘nec...in suam famam’), his urrtus ensured that he emerged with unsought glory (‘uirtute...nec extra gloriam erat’, cf. 8.2 ‘uirtutes...gloriam’). As governor of Aquitania he neither showed off his uzrtus nor sought a reputation (9.4 ‘ne famam quidem ...ostentanda uirtute ...quaesiuit'), but it was fama which brought him the consulship and governorship of Britain (9.5—6). The consistent picture is that of an individual who represses his naturally ambitious instincts but, because of his sense of responsibility and determination to do his best, concludes each stage of his career with the gloria and fama which he desires but has tried to avoid. The characteristic balance of his life is revealed both in the very formulations which T. adopts to describe his behaviour (e.g. 6.4 utz...:ta, 8.3 uirtute ~ uerecundia, 9.3 ubi — ubi, nec... aut... aut) and in hisreflecting the complementary qualities of the Sallustian Cato and Caesar (esp. 9.3n. seuerus). His career is one of repeated preferment (5.1 electus, 6.5 electus, 7.3 praeposuit, electus, 8.2 praefecit, 9.1 praeposuit, 9.5 eligit, 9.6 praepositus est); his eventual appointment as governor of Britain seems inevitable (9.5—6). 4.1 Gnaeus Iulius Agricola...colonia ortus: it is standard for Nepos to begin his Lzves with the name (s) and origin of his subjects either as a form of title (e.g. Hann. 1.1 'Hannibal, Hamilcaris filius, Karthaginiensis') or as part of the opening sentence (e.g. Cato 1.1 'M. Cato, ortus municipio Tusculo, adulescentulus .. . uersatus est in Sabinis', Att. 1.1 "T. Pomponius

Atticus, ab origine ultima stirpis Romanae generatus...equestrem obtinuit dignitatem’). Suetonius' Lzves of the poets begin similarly, but, of his extant Lzves of the emperors, only those of Titus and Domitian begin with (a short form of) their name.

ortus 15 usually regarded as a synonym for natus, as often (e.g. Cic. Planc.

60; OLD orior b), but Birley prefers ‘came from', on the grounds that A.’s

parents, as members of the senatorial order, would have been required to live mainly at Rome (RGB 72 n. 40); it is, however, difficult to parallel this meaning of ortus - abl. of a place, unless ortum Corintho at Val. Max. 3.4.2 means not that Tarquinius Priscus was ‘born at Corinth' but that he 'came from Corinth’ (and, by implication, was born at Tarquinii). If Birley is correct, it means that T. has omitted all mention of A.'s actual birth apart from its date (13 June 40), which he provides towards the end of the work (44.1); for native land

(πατρίς) and birth

(yéveois) as standard topics 566

COMMENTARY:

4.1

95

Figure 2 Lead pipe from Chester (ILS 8704a — IRB 25) showing Agricola's full name Menander (369.18, 370.20—371.14). A.'s tria nomina are also found on a lead water pipe from Chester (/LS 8704a — IRB 25; see Figure 2) and an inscription from St Albans (see 21.1n. ut templa); his cognomen alone is found on a writing tablet from Carlisle (38.3n. 2n hibernis). For full details of his career see Raepsaet-Charlier (1991), RGB 71—95. uetere...colonia: Forum Julii in Gallia Narbonensis (mod. Fréjus on the south coast of France), which T. characteristically names bv a circumlocution, had been founded bv Julius Caesar and developed bv Augustus as a naval port, to which he had sent some of the ships captured from Mark Antonv at the Battle of Actium in 31 BC (A. 4.5.1): these associations explain uetere et inlustri here, although colonial status itself ‘was still usuallv counted an honour' (G. on A. 1.79.1, referring to Millar (1977) 400-1, 407-8). It was probablv the birthplace of the poet and politician Cornelius Gallus too. See Rivet (1988)

226-38.

utrumque auum procuratorem Caesarum: T. next moves on to A.'s familv background, yévos (Menand. 370.10-11). Procurator here describes a member of the equestrian order emploved in the service of the emperor (or of other members of the imperial familv), oftener called procurator Augusti. More widely, procurators (usually of freedman status) looked after

the emperor's estates, and in many provinces a procurator of equestrian

status was

also

responsible

for the

collection

of taxes

(cf.

15.2

below)

96

COMMENTARY:

4.1

and payment of troops.? On occasion conflict inevitably arose between the procurator and governor of a province, as T. elsewhere describes in Britain (A. 14.38.3); this was something which A. in his own career was determined to avoid (9.4). The plural Caesarum (cf. D. 7.1 ‘procuratores principum) is usually taken to indicate that A.'s grandfathers served more than one emperor (presumably Augustus and Tiberius), although some think the plural is simply generated because more than one grandfather is mentioned (so H., Raepsaet-Charlier (1991) 1820 n. 34). Nothing further 15 known of either man. 566 in general Brunt (1983) and (1990) 163—87, Demougin (1988) 712-43, OCD 1215-10 s.v. procurator. quae equestris nobilitas est: the antecedent of the rel. clause is procural20 (‘the office of procurator’), understood from procuratorem above. Since nobilitas in its technical sense could be conferred only by the holding of the consulship (which was of course open only to members of the senatorial, not equestrian, order: OCD 1018 s.v. nobilitas), the expr. is pointed, although the point is somewhat lessened by phrases such as eques nobilis in Livy (e.g. 40.47.10) and elsewhere (e.g. Frontin. Strat. 4.7.36): cf. the descriptions of Agrippa at Vell. 79.1 ‘uir uirtutis nobilissimae', 96.1 'qui nouitatem suam...nobilitauerat' (and W.). By equestris nobilitas 15 meant that rank of the equestrian order which was 'la plus prestigieuse, qui dominait toutes les autres' (Demougin (1988) 679): T. is perhaps being slightly defensive, as elsewhere in this account of A.'s background and early life. pater illi Iulius Graecinus: sc. fuit. In the interests of brevity T. regularly omits forms of esse where they can be understood from the context. Such omissions (or ‘ellipses’) can occur in main sentences (e.g. 1.4, 6.4, 9.3 (bis), 12.1—5

(multiple)), relative clauses

(e.g. 5.3; contrast 6.3

'tempo-

rum quibus...fuit'), the accusative and infinitive (e.g. 1.3, 9.4, 15.2) and elsewhere (e.g. 5.2n. trucidati). See in general Moore (1903). A.'s father's praenomen was Lucius, and he reached the praetorship ( CIL 6.41069, his funerary monument set up by his younger brother, Marcus): ‘uir egregius', according to Seneca (Ben. 2.21.5 (with a nice epitaph), Ep. 20.6), and author of a two-volume work on viticulture to which T. may allude at the very end of the biography (46.4n. Agrzcola). See RGB 72. senatorii ordinis ‘: for their numbers were greater than the Romans’’. For the topos of 'unequal numbers' see e.g. Liv. 6.13.1 and Oakley's n. (and Addenda (4.519)), esp. 6.18.5 'numerate ...quot ipsi sitis, quot aduersarios habeatis?' and K. T.'s regular word for 'for' is nam (13x in Agr.). enzm again occurs in (direct) speech at 31.2 neque enim, the other three occurrences being in narrative (16.1, and with neque/necat 6.4 and 33.5); in A., however, enzm 15 reserved mostly for speeches (Adams (19732) 132). For namque see 4.1n. et se: ‘The Roman army, being organised by legions, was easily calculable. If the Britons were to count themselves as well, thev would see how much more numerous thev were than the Romans' (O-R). The Britons are less unrealistic than Vercingetorix: 'consensui [sc. Galliae] ne orbis quidem terrarum possit obsistere' (Caes. G. 7.29.6). The correction sese seems the lectio facilior, produces less good logic, and fails to take account of the reciprocal counting implied by quantulum. sic Germanias excussisse iugum: examples drawn from history were particularly advisable in (the conclusion

to) a deliberative speech

(Rhet.

Herenn. 3.9 ‘maxime conducit quam plurima rerum ante gestarum exempla proferre’): here the example (introduced by sic as at H. 4.57.2, A. 4.39.5, 12.25.1) is Arminius' famous defeat of Quintilius Varus in the Teutoburg Forest in AD g (OCD 1446 s.v. Teutoburgiensis, saltus). As T. explains at A. 1.31.2, Germany was divided bv the Romans into Lower Germany (where the Varian disaster in fact took place) and Upper, but his frequent use of the plural Germaniae elsewhere in A. suggests that ‘Germany was thought of as a group of tribal areas’ (G. on A. 1.57.2): here in the mouths of Britons the plural implies a model of unified behaviour which thev themselves — equally diverse — should follow. The plural recurs at 28.1

(see n. cohors). For excutere tugum, used literally of horses at Curt.

4.15.16, cf. Sen. Ep. 51.8, Plin. Pan. 11.5, Flor. 1.30.2, Apul. Apol. 9o, Just. 15.4.12. flumine, non Oceano ‘by a river, not Ocean’: Latin commonly omits a word for ‘only’ (10.4n. quia). In view of the Romans' inferior numbers (quantulum), the Britons' point is that the Ocean, unlike the Rhine,

is a formidable obstacle to the sending of reinforcements. But there is a slight illogicality with the alleged ease with which the defeated legions are imagined as retreating (4 recessuros). 15.4 patriam, coniuges, parentes: standard topics of deliberative oratory (e.g. Rhet. Herenn. 3.5 ‘pro patria, parentibus...adire periculum et quemlibet suscipere laborem’) and esp. of the hortatio (e.g. H. 5.17.2

170

COMMENTARY:

15.4

'coniugum, parentum, patriae memores' and H., Thuc. 7.69.2 with Hornblower, Sall. C. 58.11, Liv. 8.10.4 with Oakley); see also 32.2 (n. aut nulla). The asyndeton between the three nouns in this first colon of an antithesis is varied by the co-ordination (et) of the two nouns in the second, a type of uariatio also found at the end of Calgacus' speech (32.4 'hic...poenae"); slightly different is the combination of asyndeton and co-ordination in a sequence, as at 38.1 (‘trahere...relinquere’): see Sorbom (1935) 60 and 55-

illis: the Romans. Both auaritia, justimputed to the Romans by T. himself above (12.6 and n.) and a familiar complaint on the lips of barbarians (M-W on A. 4.72.1), and luxuria were famously seen by Sallust (C. 5.8, 12.2) and Livy (fraef. 11) as contributing to the decline of Rome, though they disagree on when the decline set in. See esp. 30.4—5 and A. 14.32.3 ‘quam [sc. Britanniam] auaritia in bellum egerat’ (of the procurator of 60, Decianus Catus). For luxuria see 6.4n. (ut?).

causas belli ‘motives for war' (OLD causa 7a). Calgacus will later begin his famous speech with the words Quotiens causas belli . . intueor (30.1); in some respects the Boudican episode here is 'synecdochic' of the Mons Graupius narrative at the climax of A.’s governorship. ut diuus Iulius recessisset: the standard epithetis sarcastic in the mouths of Britons (cf. similar sarcasm from Arminius at A. 1.59.5 'illeinter numina dicatus Augustus’), and their comparison is somewhat tendentious: Caesar had withdrawn, but not in defeat.

modo uirtutem maiorum suorum aemularentur: modo + subjunc. = ‘only provided that' (OLD 4; A-G §528, G-L $573): the verb would be aemulemurin or. rect. The Britons speak as though they were quintessential Romans, having maiores who epitomise uzrtus (maorum uirtus is a regular expr.) and who invite emulation (for which see e.g. Oakley on Liv. 7.10.3); for ancestors as a standard motif of the hortatiosee 32.4n. (proinde). For uirtutem aemulari cf. Bell. Afr. 81.2, Liv. 7.7.3, Apul. Met. 1.23. neue...'and they should not...' For the thought cf. Nicias' speech to the Athenians at Thuc. 7.61.2. For pauescere+ abl. cf. Sall.J. 72.2. plus impetus felicibus, maiorem constantiam penes miseros esse: adversative asyndeton: ‘... greater resolution lay with the pitiable'. miseros is of course

an

allusion

to

the

Britons

(cf.

Britannorum...miserer:

below); plus impetus felicibus does not, however, allude but

is a general

statement

('the

fortunate

had

more

to the Romans drive’)

which

functions as a 'foil' to emphasise by contrast the greater resolution of the mzser? (constantia is another Roman characteristic, e.g. Liv. 30.7.6 ‘Romanae...constantiae’). The choice of the term felix heightens the persuasion,

since,

in addition

to 'successful',

it also

means

'lucky'

and

thus contains a hint that success both may not be deserved and may soon change. The Britons' sententia 15 pointed by double uariatio (plus +

COMMENTARY:

15.5

171

genit. ^ compar. adj.; dat. noun - penes - acc.). The contrast zmpetus — constantia seems not to be paralleled exactly (though cf. Val. Max. 3.8 ext. 3 'indignum iudicans tot et tam bene meritos indigna causa impetu inuidiae abripi, temeritati multitudinis constantiam suam obiecit’), but for

impetus — perseuerantia H. compares Liv. 5.6.8, 27.16.1, Just. 41.2.8. 15.5 etiam deos misereri: cf. Thuc. 7.77.4 (Nicias speaking) ‘we are now more deserving of the gods' pity than their jealousy’ and Hornblower's n. misereri picks up miserosabove (a device called 'resumption' by Wills (1996) 311—28, esp. 327—8). qui ... detinerent: the transmitted text 15 problematical since relegare is not otherwise found with 2n - abl.; the regular construction 15 in - accus. (as e.g. Liv. per. 23 ‘reliquiae Cannensis exercitus in Siciliam relegatae sunt’), for which the dat. is occasionally substituted in verse. It might be argued that in alia insula is to be taken with detznerent, but this disrupts what appears to be a chiastic arrangement: Romanum ducem + absentem — relegatum - exercitum. Lipsius proposed emending to religatum (‘moored’, ‘tied up’), but the transferred sense 15 not easy to parallel. S. P. Oaklev has suggested zn alia insula (‘who were detaining the absent Roman leader and his army, relegated as it was to another island’), but the emphasis on relegation to the island seems to imply, as indeed does the transmitted text, that the leader 15 not also on the island. Transposition of 19 alia insula and exercitum, on the other hand, preserves the chiasmus and

means that 221 alia insulais clearly to be taken with both ducem and exercitum. Whether absentem and relegatum are to be taken attributively (^who were detaining on another island the absent leader and his relegated army’) or predicatively (‘who were keeping the Roman leader absent, and his army relegated, on another island’)

is less clear; for the latter idiom

cf. OLD

detineo 4a. With alia T. attributes to the inhabitants the knowledge that Britain was itself an island, something the Romans did not know for certain until A.'s circumnavigation (10.4). quod difficillimum fuerit: the perf. subjunc. represents μ (‘has been’) in the imagined direct speech. The Britons are referring to the disunity which had characterised them hitherto (cf. 12.1—2 above and nn., esp. 22 commune non consulunt).

deliberare: Roman oratory was divided into three types: epideictic, deliberative, and judicial. ‘The epideictic kind is devoted to the praise or censure of some particular person. The deliberative consists in the discussion of policy and embraces persuasion and dissuasion. The judicial is based on legal controversy' (Rhet. Herenn. 1.2, trans. Caplan). Here the Britons deploy the technical verb for a deliberative speech, which is the kind they have been making: for this technique or similar elsewhere see e.g. 18.4, 34.1; Brink on Hor. AP 18, W. on Vell. 117.1. The clausula

172

COMMENTARY:

16.1

(= - — x) is Cicero's favourite ditrochee or 'dichoreus' (Wilkinson (1963) 161), the clausula which, when produced bv the tribune Carbo in a speech,

allegedly met with a shout of approval from his audience (Cic. Orat. 214 ‘hoc dichoreo tantus clamor contionis excitatus est’).

porro in eiusmodi consiliis periculosius esse deprehendi quam audere ‘, taking it further, ...’ (OLD porro 4): the adverb articulates the relationship between deliberare above and n ezusmod: consiliis: now they were deliberating, which was an improvement on their previous divisiveness; but such deliberation in its turn was dangerous (presumably because an assembly of Britons suggested trouble and invited a pre-emptive strike): they should therefore take bold action at once (cf. H. 1.21.2, 38.2, 40.1). Having begun their speech with the kev notion of 'advantage' (1n. nzhil profici), the Britons conclude with a sententza which refers to one of the two elements into which advantage was divided, viz. 'security' (in Greek 16 ἀκίνduvov): 'To consider Security is to provide some plan or other for ensuring the avoidance of a present or imminent danger' (Rhet. Herenn. 3.3 'tuta [sc. pars] est quae conficit instantis aut consequentis periculi uitationem qualibet ratione', trans. Caplan). 16.1 His atque talibus inuicem instincti: since inuicem seems to contrast with uniuersi below, the meaning probably is: ‘roused individually by these and other such words' (OLD imuicem 2). instincti (again at 35.2) isa regular term for the reaction to a speech (e.g. À. 2.46.3 with G.'s n., Liv. 6.14.9 with Oakley's); hzs atque talibus is a Sallustian formulation (J. 62.2) which T. resumes in A. (15x); see also G. on A. 1.5.1. Boudicca: the famous Boadicea of legend was the wife of Prasutagus, king of the Iceni, a tribe who inhabited what is now Norfolk. According to T.'s own later account (A. 14.31.1—2), the king had named Nero as one of his heirs in his will, hoping therebv to forestall any trouble when he died. But his plan failed, and on his death in 59 or 60 his people, including his wife and daughters, were subjected to the violence and depredations of which complaint was made in the Britons' speech above. As a result, the Iceni rebelled and were joined by their southern neighbours, the Trinovantes. The name is corrupt in H here; in the MSS of the Annals it is presented variously as boodicia (14.31.1), bouducca and boudicea (14.35.1), and Boudicca, boudicea and bouducca (14.37.2), of which the traditional ‘Boadicea’ is a corrupt derivation. Modern convention is to read Boudicca (M's reading at A. 14.37.2) in the Tacitean texts, although the actual spelling is thought to have been Boudica, pronounced 'Bowdee'kah' (Jackson (1979)). For her see Braund (1996) 132-46. neque enim sexum in imperiis discernunt: the truth of this statement has been disputed. T. later makes Boudica herself say 'solitum quidem Britannis feminarum ductu bellare' (A. 14.35.1), and we know that the

COMMENTARY:

16.2

173

Brigantes were ruled bv a queen, Cartimandua (31.4n., H. 3.45.1, A. 12.36.1); but the latter's rule is soon stigmatised as ignominious by her own people (A. 12.40.3 'stimulante ignominia ne feminae imperio subderentur’). Whatever the truth, it was standard practice for ancient writers to describe barbarians (as the Britons are designated just below) in terms which are the opposite to themselves. Boudica, being not onlv a leader (cf. G. 45.6 'femina dominatur’, of the Sithones, and Rives' n.) but also a warrior, is doubly ‘other’. It is this complex of ideas which leads T. elsewhere to describe various transgressive Roman women in 'barbarian' terms as military leaders: e.g. A. 1.69.1 'femina...munia ducis...induit’ (the elder Agrippina), 2.55.6 (Plancina), 12.37.4 'feminam signis Romanis praesidere' (the younger Agrippina); W-M on 3.33.2—3. The otherness of Boudica 15 underlined by the juxtaposition femzna duce, for which compare Virg. Aen. 1.364 ‘dux femina facti' (Dido). sumpsere ... bellum: T.'s extended account of the revolt will be found at A. 14.31—9; see also Dio 62.1-12. Both accounts are given under the year 61, but the revolt probably began in 6o (RGB 49), as indeed bienn?o (14.3n.) implies. See also Syme (1958) 762-—6. For the expr. cf. OLD sumo 14a (cf. 1b). ipsam coloniam: Camulodunum

(5.2n. trucidati).

ut sedem seruitutis ‘as being the seat of servitude' (OLD ut10). The temple to the deified Claudius at Camulodunum was seen ‘quasi arx aeternae dominationis' (A. 14.31.4). ullum in barbaris ingeniis saeuitiae genus 'any kind of savagerv «which there is> in barbarian temperaments' (see the account at A. 14.32.3, 33.2). The remark looks back to the Britons' allegation of Roman savagery at 15.2 (saeutret): it 15 the Britons themselves who are the true savages. Disparaging references to the ingenium of barbarians are conventional (e.g. H. 4.13.2, A. 4.46.1); barbara ingenia seems to be a unique variant on barbarorum ingenia, used by Livy (22.22.7, 29.23.6, cf. Val. Max. 5.1 ext. Ο); Hannibal).

for saeuitiae genus

cf. Sen.

Cons.

Marc.

6.2,

Gell.

2.6.7

(of

ira et uictoria: a striking personification. It is to be hoped that T. was

as familiar with the meaning of Boudica's name ('victory') as he seems to have been with that of the Ordovices (18.1n.). For the same two nouns

elsewhere cf. Sen. Tro. 280. 16.2 quod nisi Paulinus... 'and if Paulinus had not...' (OLD quod 1a): counterfactual statements are a standard historiographical method of heightening narrative vividness, though thev are perhaps more frequent (and more dramatic) in the inverted form

(i.e. when the conjunction fol-

lows the main sentence: see 33.1n. zamque, 377.1n. ni). T. uses quod nisi again at 26.2 and quod ni at 37.4 (neither recurs in any of his other works). For Paulinus' rescue journev from Anglesey via Londinium and

174

COMMENTARY:

16.2

Verulamium (St Albans) see A. 14.33.1—34.1; despite propereit *must have taken several weeks at least' (RGB 49). amissa Britannia foret: forms of the archaising forem are half as frequent in Agr. as those of essem, to which it is equivalent, and are dropped entirely in G. and D.; yet forem outnumbers essem by 3:1 in H. and by 2:1 in A. 1—12 but 15 itself dropped almost entirely in A. 13-16 (Syme (1958) 738—9; W-M on A. 3.14.4). Cf. 26.2 ‘debellatum...foret’, 37.4 'acceptum ...foret', 40.2. quam unius proelii fortuna ueteri patientiae restituit: this summary constitutes an ironic sequel to the barbarians' earlier complaints (15.1 patientza) and bravado (15.4 proelii unius); for a fuller account of the battle see A. 14.34—7. The idea of ‘only a single battle' 15 something of a commonplace, e.g. Nep. Epam. 8.4 ‘uno proelio non solum Thebas ab interitu retraxit sed etiam uniuersam Graeciam in libertatem uindicauit'; Oakley on Liv. 8.29.12. For froelii fortuna (also at H. 2.23.3, 3.18.2, 4.20.2) cf. Just. 1.7.5, 3.7.12; pugnae fortuna (also at G. 3.4, H. 4.33.2) is the expr. used exclusively by Livy (but note Liv. per. 27 ‘tot proeliorum aduersa fortuna’). It is not clear whether fortuna is nom. or abl. tenentibus arma plerisque: concessive; for lenere — retinere, as again at 3 below, see OLD 13b: T. has a general liking for simple rather than compound verbs. conscientia defectionis et proprius ex legato timor: though conscientia and temor are combined in various ways elsewhere (Cic. Fin. 2.53, Sall. ]. 35.4. Liv. 9.26.7, HA Aurel. 26.3), T. is here alluding to Liv. 23.15.7 ‘conscientia temptatae defectionis ac metus a praetore Romano' (a potential defector in the Second Punic War), but with added chiasmus. proprius . . timor— ‘their personal fear of the legate'. For consctentia . . . agitat cf. Sall. C. 5.7 'conscientia...agitabatur' (of Catiline; note also Sall. H. 1.77.7 ‘c. exagitati’); for tzmor cf. Hyg. Fab. 145.4 'timore exagitatam'. ne...adroganter in deditos et...durius consuleret: the clause depends upon /?mor : '(the fear) that...he might decide arrogantly and...too harshly against those who surrendered'. For the intrans. use of consulere see OLD 4b; for adroganter in deditos cf. Liv. 8.13.15 'crudeliter consulere in deditos' (the next sentence is alluded to later at 30.5: see n. Auferre), Curt. 5.6.16 ‘nec in deditos grauius consultum’; for durius cf. Caes. C. 1.22.6 ‘ut suae uitae durius consulere cogantur'. See also below, 3n. (massus).

egregius cetera is originally a Sallustian phrase (H. 4.70) which was taken over by Livy (1.32.2, 1.35.6 (ad cetera at 37.7.15)). cetera 15 adverbial acc. or acc. of respect (OLD ceterus 4; G-L $338); for quamquam see 1.1n. ut suae cuiusque iniuriae ultor: μ may be taken closely with suae and mean ‘as if' (OLD 8a): ‘ an avenger of each wrong as if done to

COMMENTARY:

16.3

175

himself [lit. as if his own]'. Alternatively u£ 15 to be taken with wltor and means 'acting as' (OLD g): 'acting as an avenger of each wrong done to himself'. Nipperdey preferred qu2sque to Wex's cuzusque, but the sense seems less good (‘as each man an avenger of the wrong done to him’) and ut quisque is normally used by T. to qualifv a plural (e.g. H. 1.57.1 'arma, pecuniam offerentes, ut quisque...opibus, ingenio ualidus’). 2nz uriae (-1arum) ultoris a Ciceronian expr.

16.3 Missus igitur Petronius Turpilianus: the logic seems to be that, since the Britons were retaining their arms because of their fear of Suetonius' harshness, therefore a more amenable governor was sent to succeed him

('tamquam exorabilior’, below): for discussion see Gambash

(2012).

The Britons' fears (tzmor, above) are represented in the Annals (14.38.3) as being encouraged, and perhaps even caused, by Julius Classicianus, the procurator who was feuding with Suetonius; according to T. his message was: ‘nouum legatum opperiendum esse, sine hostili ira et superbia uictoris clementer deditis consulturum'. P. Petronius Turpilianus, who was sent to Britain after resigning his consulship in 61 and stayed till 65, was the nephew or grand-nephew of the first governor of Britain, A. Plautius (14.1n. Consularium): ‘Doubtless a kinsman of Plautius would have seemed

a suitable person to restore confidence amongst the Britons' (RGB 50-2, quotation from 51). For zgztur, onlv here in second place in Agr., see 7.2n. tamquam ‘on the ground of his being...' (OLD *b). delictis hostium nouus 'new to the enemy's crimes' (OLD nouus 16a). paenitentiae mitior ‘more indulgent to the penitent’: mitis, which regularly takes a personal dative (e.g. A. 4.17.2 'domui Germanici mitis’, 'to the family of G.'), is constructed with the dative of a thing only at Colum. 10.21 manibus (so TLL8.1158.84—1159.1); but here paenitentiaeis abstract for concrete, as again at Plin. Ep. 10.96.2 'detur paenitentiae uenia’, and perhaps cf. ?ra et uictoria (for ?rat? uictores) at 1 above

(for the converse 566

4.2n. arcebat). compositis prioribus 'the previous situation having been settled' (OLD prior 3c).

nihil ultra ausus: more (critical) detail at A. 14.39.3 ‘non irritato hoste neque lacessitus honestum pacis nomen segni otio imposuit'. Trebellio Maximo: M. Trebellius Maximus, consul in 55, governed Britain from 63 to 69: see RGB 52-6. He has already been alluded to, though not by name, at 7.3 (n. legatis.. . consularibus). segnior et nullis castrorum experimentis: the unparalleled abl. phrase is usually rendered *with no experience of camp' (abl. of description or quality), i.e. Trebellius Maximus lacked militarv experience; but Trebellius is almost certainlv identical with the M. Trebellius who as legionary legate operated successfully in the Taurus mountains in 36 (A. 6.41.1),

176

COMMENTARY:

16.3

in which case he could not be described as lacking militarv experience. Birley (RGB 53—-5 and n. 134) therefore prefers ‘never ventured on a campaign’, the translation of Church and Brodribb (187*7),'? explaining that segnior...experimentis has to be seen in the light of compositis prioribus nihil ultra ausus above: Petronius did the minimum of campaigning but nothing further; Trebellius was even more unenterprising and did no campaigning at all. This interpretation is confirmed structurally, since the present sentence seems clearly elaborated by the next in the order abc & cab: thus segnior + nullis castrorum experimentis + comitate . . . brouinciam barbani .. . blandientibus + segnitiae excusationem -- adsuetus expeditionibus miles otio lasciuiret.'5 For experimentis perhaps compare Plin. NH 16.156 ‘belli pacisque experimentis', Oros. 5.5.15 ‘ad experimentum....belli'; for castra as typifying military operations and war see OLD 3. The uariatio of adj. ~ abl. phrase is common in T.: see Sórbom (1935) 89-91, though the present case should be classed as an abl. abs. comitate quadam curandi ‘with a certain affability in governing' (for the gen. gerund cf. Liv. 1.34.11 'comitate inuitandi'). The 'application of curare to military command is peculiarly Sallustian', esp. in its absolute use (e.g. C. 59.3, J. 46.7, 57.3, 60.1, 60.5); 'T. adopts Sallust’s usage and apparently extends it to the administration of provinces' (G. on A. 1.31.2). It is nevertheless possible that prouznciam (below) 15 to be taken with both curandz and tenuit (&nó κοινοῦ). For comitas see 22.4n.

(ut erat);

for its application to a governor see Damon on H. 1.13.4 'comiter administrata prouincia’. didicere ...blandientibus: it 15 typical of T.'s cynicism to suggest that indulgence in vice is an intellectual process (cf. A. 3.54.3 and W.) and that even (quoque) the formerly uneducated barbarians have mastered learning it. The comment anticipates 21.2 'discessum ad delenimenta uitiorum'. "The process could be called "Romanization"' (RGB 56). As evidence, O-R note that Roman-style housing at Lockleys (near Welwyn, in Hertfordshire) dates from around this time. ignoscere here — ‘to make allowances for, grant indulgence to' (OLD 2). interuentus ciuilium armorum praebuit iustam segnitiae excusationem: in contrast to the first years of his governorship, when he was culpable of spontaneous segnitia (above), the intervention of the civil wars in 68-9 presented Trebellius with ‘a justifiable excuse' for his idleness (OLD ?ustus

!** [n their edition (1869) of Agr., however, they render 'a man with no actual experience of campaigns'. '3 The correspondence between castrorum and expeditionibus suggests that castrorum experimentis is unlikely to be a reference to the 'practice camps' which archaeologists claim to have identified in Britain (esp. in Wales) and of which some may date to the governorship of Frontinus (17.2 below): see Davies (1968) 104 — (1989) 125—6, Davies and Jones (2006) 67—-90, esp. 89.

COMMENTARY:

16.4-5

177

5). ?usta . . excusatiois a regular expr. in Cicero, from whom itis taken over bv the younger Seneca and the younger Pliny. discordia laboratum: sc. est: ‘there was trouble from mutiny’, instead of the hard work to which the soldiers might have been put by their chief (e.g. Sall. J 44.3 ‘milites laborare coegisset’). laborareis regularly constructed + abl. (OLD 4); for discordia cf. Liv. 4.58.2 'discordia intestina laborarent Veientes'. For the mutiny see above, 7.3nn. (ub? and quippe). miles otio lasciuiret: T. talks elsewhere of soldiers 'running riot' (e.g. 5.1, Á. 1.10.2, 2.55.5, 4.2.1), and it was a commonplace that otzum was ‘res disciplinae inimicissima’ (Vell. 78.2, cf. e.g. G. 44.3 'otiosae...armatorum manus

facile lasciuiunt'), which

commanders

should do their utmost to

counteract (e.g. A. 11.18.2—3, 13.35.1, Frontin. Strat. 4.1.15; Ash on H. 2.34.1; also below, 22.2n. crebrae). T.'s expr. seems Livian (2.28.5 'otio lasciuire plebem’). The collective sing. mzles (for milites) occurs in prose from Quadrigarius but is common only in Livy, T. and Ammianus; for its use in verse see Austin on Virg. Aen. 2.20. 16.4 fuga ac latebris: each word (for their combination cf. Cic. Rab. Perd. 22, Sen. HF 1012, Herc. Oet. 1408) seems to be picked up by zndecorus atque humilis immediately below. uitata exercitus ira: "Irebellius Maximus per auaritiam ac sordes contemptus exercitui inuisusque' (H. 1.60); see also next n. precario mox praefuit 'subsequently he held command on sufferance’ (OLD precario 2b). Having initiallv fled from his angry and discontented soldiers (above), Trebellius evidently regained his position later as a result of the arrangement described at the end of the sentence (ac uelut... ducis salute. next n.). The arrangement can only have been temporarv, however, since Trebellius soon left Britain and fled to Vitellius in early 69 (H. 1.60 ‘ad Vitellium perfugerit’, 2.65.2 ‘profugerat Britannia ob iracundiam militum' with Ash's n.).

uelut pacta and the safety 1a). el, which posed before stetit 'came

exercitus licentia et ducis salute 'as if the licence of the army of its leader had been secured by negotiation' (OLD pacisco is wrongly transmitted after salute, is perhaps better transducis than deleted. to a standstill' (OLD 10c), as e.g. A. 3.72.3 'tanta uis unum

intra damnum

stetisset' (with W-M), and concluding a remarkably alliter-

ative sequence. 16.5 Vettius Bolanus: M. Vettius Bolanus, consul in 66 and the eighth governor of Britain, is the second of those whom T. has already named in his ‘biographical’ section (8.1 and n. Praeerat). ciuilibus bellis: uariatio after ciuilium armorum at g above, as often (Sórbom (1935) 23). According to T. elsewhere (H. 3.45.1—2), it was repeated rumours of the civil wars which encouraged Venutius, the former husband of Cartimandua, to lead the Brigantes (17.1 below) in revolt

178

COMMENTARY:

17.1

against Rome. Bolanus took over from Trebellius in the spring of 69; the revolt perhaps started in the autumn (Wellesley on H. §.45.1). agitauit is usually interpreted ironically (‘harassed’, ‘disturbed’, as in 2 above); but the sense 'controlled' or ‘managed’

(OLD 13) seems no less

apt (for the accompanving abl. see Sall. C. 9.5). eadem inertia erga hostes: i.e. the same as when Trebellius had been governor (so too with szmzlzsjust below). But quite the opposite impression of Bolanus' British activities is naturally given by Statius in his encomiastic poem to Bolanus' son (Szlv. 5.2.142-9 with Gibson's nn.: see also 5.1n. (Prima) above): see RGB 61. For erga see 5.3n. (temporibus).

similis petulantia castrorum 'there was a similar surliness in the camp’. nisi quod...loco auctoritatis 'except that Bolanus, harmless and not resented for any misconduct, had gained affection instead of influence’: if we are to infer from nullis delictis that Bolanus' predecessor was guiltv of misconduct, T. in the Haustories says that Trebellius was accused bv his rival, Roscius Caelius (7.3nn.), of defrauding the legions (1.60 'spoliatas et inopes legiones...obiectabat’). The antithesis carztatem — auctoritatis 15 Ciceronian

(Phil. 2.27).

17.1 sed, ubi cum cetero orbe Vespasianus et Britanniam recuperauit: scholars have seen this as exaggerated, on the grounds that it was Suetonius Paulinus who was responsible for the province's 'recovery' (from Boudica's revolt); but the ubzclause is clearly intended to contrast with the abl. abs. at 16.5 above (‘manentibus adhuc ciuilibus bellis') and refers

to Vespasian's role in ending the civil wars, something which had its effect as much

in Britain as in the rest of the Roman world

(of which Britain is

here described as a part: contrast 12.3 above). Hence the verb probably

has its less common sense of ‘restored’, 'revived' (OLD 2), the revival being

explained in the main clause. magni duces. .. minuta hostium spes: a tricolon with chiastic alliteration (m- e- e- m-) and polarised terms (magni — minuta). One of the main themes

of Agr. is 'greatness' (2n. uir magnus, 42.4). Petilius Cerialis: the third of the governors to have been mentioned by T. in the earlier ‘biographical’ section (8.2n.). The terror which he inflicted explains the enemy's diminished spes (above): for antitheses of the two terms see e.g. Cic. Sest. 38, Liv. 32.5.13, Sil. 17.62—3; for terrorem . .. intulit see 36.3n.(recentem). Brigantum ciuitatem...adgressus: the deferred participle is once again ‘timeless’, explaining the main verb (2.1n. delegato). The Brigantes inhabited a vast area from roughly the line of the future Hadrian's Wall as far south as the River Trent; they had been ruled over by Cartimandua (A. 12.936.1, 40.2), and their capital was Isurium (now Aldborough, a delightful village in North Yorkshire with surviving Roman mosaics). See Rivet and Smith (1979) 279-80, Hartlev and Fitts (1988).

COMMENTARY:

17.2

179

perhibetur: see 10.5n. proelia ...non incruenta 15 Livian (2.56.15). magnam...Brigantum partem: ‘There is general agreement that Cerialis moved

the Ninth, his old legion, forward from Lincoln

(Lzndum)

to

a new fortress at York (Eburacum) ... Archaeological evidence shows that he penetrated well into the northern Pennines; and the first Roman fort at Carlisle (Luguualium) turns out, on the evidence of dendrochronology, to have been built with timbers felled in the winter of 72—-3' (RGB 67). 17.2 et Cerialis quidem...obruisset 'And Cerialis for his part would have overshadowed [OLD gb] the administration [OLD 7] and reputation of any other successor'. quidem performs two functions here: as the presence of ef suggests, it both extends the reference to Cerialis in 1 above and also prepares for the otherwise unmarked contrast with Julius Frontinus below: for this double function (again at 38.1 and e.g. H. 1.63.1) see Solodow (1978) 62-3. cura (but not in this sense) is coupled with fama at Liv. 27.24.1, 35.41.2, Mart. 1.111.1; for alterzus see OLD alter' 2b. subiit . . . Iulius Frontinus: adversative asyndeton (3.1n. cito) after quidem above (as e.g. A. 5.6.2; Solodow (1978) 67—74), and completing a chiasmus. Sex. Julius Frontinus, the tenth governor of Britain (73/74—77), was highly distinguished: curator aquarum in 97, he achieved - exceptionally — three consulships (his last two are known to have been in g8 and 100) and wrote, among other works, the surviving Strategemata and De aquaeductu urbis Romae. See Eck (2002) 219-25, RGB 68—71. Although subire molem — ‘take on a burden' is readily paralleled (Curt. 10.5.37, Plin. Pan. 44.4 ‘quis [sc. successor] enim curae tuae molem sponte subeat?', Pan. Lat. 2.3.5), the reference in the preceding sentence to a successor strongly suggests that subat 15 intransitive — 'came next, succeeded'

(OLD 8a): hence it is sustznuit alone which means 'took on, shoul-

dered' (for which cf. Liv. 10.19.19, 36.7.10, Flor. 2.13.58, Pan. Lat. 7.5.2), although the two verbs can be coupled elsewhere with direct object(s) (e.g. Cic. Dom. 101, Virg. Aen. 10.798—-9, Sen. Brev. vit. 18.1). uir magnus (quantum licebat): the tense 15 important (contrast Val. Fl. 5.686 ‘magnis (quantum licet) .. .fatis'): whatever the realities of the situation, there is no suggestion that any limit is currently being placed upon Frontinus' greatness as there was in the past (he was, after all, curator aquarum and consul in the very vear that T. was writing). This is a tribute both to the current regime and to Frontinus himself (who indeed is called princeps uzr by Pliny, Ep. 4.8.3). Compare A. 14.47.1 'Regulus..., in quantum praeumbrante imperatoris fastigio datur, clarus'. Silurum gentem: the Silures (11.2n.) had had earlier conflicts with Rome under the leadership of Caratacus, who, in the aftermath of Osto-

rius Scapula's victory (14.1n. Consularium), was eventually handed over to

180

COMMENTARY:

18

the Romans by Cartimandua (A. 12.33.1—40.1). pugnacem . . gentemis originally Livian (22.37.8, then Curt. 3.9.3); for ualida cf. 12.2n. super uirtutem hostium locorum quoque difficultates eluctatus: the digression ends with chiasmus and an almost metatextual anagram. Normally uzrtus would be ascribed only to the Roman side (see e.g. Oakley on Liv. 6.13.1), but for its ascription to the enemy see e.g. 11.4, Caes. G. 1.1.4, 2.8.2, Liv. 4.32.1, Curt. 3.7.9, Just. 14.3.2 (and note 27.2 below); super= ‘in addition to' (OLD super? 7). The ideal general was expected to deal with difficult terrain (e.g. Onas. 7.1—2, cf. 18; below, 20.2n. praetemptare),

but T. here seems to be alluding to Velleius' panegyrical description of M. Lepidus' heroism with Legio XX in Dalmatia in AD 9 (115.2 'cum difficultate locorum et cum ui hostium luctatus’). Lepidus 15 much admired by T. and is the man generally agreed to display the closest resemblance to A. himself (see A. 4.20.2), while Legio XX had now been in Britain since 43. 18-28

THE

YEARS

AD 77-82

The narrative of A.'s proconsulship proceeds year by year from the summer of 77 (18.1 ‘media iam aestate’) and is marked by campaign seasons: 20.2 'aestas', 22.1 'tertius...annus', 23 'quarta aestas', 24.1 ‘quinto...anno’, 25.1 'aestate...sextum ...annum', 29.1 'initio aestatis’ (see further Hind (1985)). In this practice T. follows the historiographical tradition of Caesar (cf. G. 2.12.1 'ineunte aestate ... legatum misit’, the beginning of the second vear). Year 1 introduces a narrative pattern whereby the military action (18) is followed by an 'appended' section or 'coda' of a more thematic nature, in this case the elimination of abuses (19-20.1). This pattern is followed in each vear except 8o, the shortest year (23). Thus the campaigning season of 78 (20.2—3) 15 followed by the famous paragraph on the ‘Romanisation' of Britain (21); the advance to the Tay (22.1) develops into

a much

longer section on A. as the ideal general (22.2—4); the crossing of Clvde (24.1) is followed by a much longer excursus on the country Hibernia (24.2—3); and to the substantial account of campaigning in (25—27) is appended the bizarre and seemingly irrelevant episode of treacherous

Usipi

(28). Within

this framework,

however,

the of 82 the

it will be seen

that from vear to year T. varies the length of treatment and manner of focus. Throughout the narrative there is a marked absence of geographical information.

In Year

1 the only name

mentioned

(18.3)

is the island of

Mona (Anglesey). The campaigning season of Year 2 has been described as 'a triumph of applied waffle' (Woolliscroft and Hoffmann (2006) 176): ‘not a single place name is provided and we are given no real idea where these activities took place'. In Year 3 the only name is that of the Tay

COMMENTARY:

18.1

181

(22.1), though it is characteristic that the manuscripts offer an alternative (unknown)

name as well; and, although the Clyde and Forth are men-

tioned in Year 4 (23), the inferences which may be drawn from this reference for Year 5 (24.1) are heavily disputed. The Forth is seemingly a point of orientation in Year 6, but only to the extent of our being told that the action takes place 'across' or 'this side of' it (25.1 'trans Bodotriam’, 3 ‘citra Bodotriam). T. had no interest in such matters (above, pp. 13—15), being far more concerned to portray A. as a successful general extending the Roman empire. 16—20.1 The year 77 18.1 Hunc Britanniae statum, has bellorum uices... Agricola inuenit: after the double digression in 10-17, T. now returns to the main narrative which he left at 9.6 ' Britanniae praepositus est'. The transition is effected by the suggestion of ring composition (Britanniae — Britanniae), by the resumptive use of hic in anaphora

(OLD hic'

10), and by an allusion to the kind

of formulaic phraseologv which is frequent in Livy (e.g. 25.11.20 ‘hunc statum rerum Hannibal Tarenti relinquit') and used again by T. for transitional purposes in his later works (77. 1.11.3 ‘hic fuit rerum Romanarum status', 4.71.1, A. 1.16.1 ‘hic rerum urbanarum status erat’). ‘The adoption of this turn of phrase by Tacitus', it has been said, indicates 'the profound influence that Livy exercised on his style’ (C-L 48). inuenit self-evidently refers to the conditions which greeted A. on his arrival in Britain: hunc Britanniae statum therefore cannot refer to the picture of Britain presented in the digression as a whole, since the latter's ethno-geographical section (10-13.1) depends upon information acquired by A. only several years after his arrival (cf. 10.4) and is written from T.’s own perspective of AD g8 (Intro. p. 18). The phrase must be a reference to the immediately preceding summary of Britain under Frontinus at 17.2 and perhaps also under Cerialis at 17.1,just as Huncrerum cursumat 39.1 refers to the immediately preceding Mons Graupius: thus status (OLD status? Ob, cf. 7) means 'the circumstances affecting a...thing at a given time, state, condition

(with reference to the state, political situation etc.)'

and is explained more precisely by has bellorum uices, ‘the vicissitudes of warfare' (OLD uicis 7b). Both phrases together constitute an ex. of 'theme and variation'. belli uices is a favourite expr. of Statius (Theb. 4.620, 9.78, 566, 10.754) but 15 also at [Sen.] Oct. 480 and Sil. 3.13; the plural bellorum seems found only at Amm. 19.12.1 (where the meaning is different). For statum inuentre cf. Sen. NQ 2.41.2 (statum at [Quint.] Decl. 320.1 is a technical rhetorical term). media iam aestate transgressus: when used without an object (OLD 1b)

the verb takes 115 meaning from the context

(as 10.3, 24.1); here the

182

COMMENTARY:

18.1

reference is to crossing the English Channel (the fretum Ocean of 40.2). aestate, like aestas below (2), refers to the campaigning season of AD 77 (for the vear see RGB ηη--ὃ, Raepsaet-Charlier (1991) 1824-5, 1842—55); iam implies that A., like Julius Caesar before him (G. 4.20.1 'exigua parte aestatis reliqua’), arrived late in the season. For medzus used of periods of time see OLD medius' 6. See also next n. cum et milites...ad securitatem et hostes ad occasionem uerterentur: the clause explains or expands media iam aestate above (‘with the season alreadv at mid-point, when...'): for the technique see C-L 290-2 (although Walther, placing a full stop after znuenit, sees the clause as introducing Ordouicum. .. obtriuerat). The clause is problematic because it is not absolutely clear whether securitatem means ‘safety’ (OLD ga) or ‘carelessness’

(OLD

2). If the

former,

uerterentur will mean

‘were

turn-

ing their thoughts to’ (OLD 14b), as O-R and some others say: i.e. since there had been no campaign hitherto, the Roman soldiers had started to think that this was one season in which they had got away without risking their lives. On

the other hand, if securitas = 'carelessness', uerter-

entur will have a rather different nuance such as ‘were gravitating’: i.e. the lack of a campaign meant that discipline had broken down and the men were becoming sloppy. Which of these alternatives is more likely to be correct? It seems almost certain that the present clause is illustrated by the following sentence; and, if Ordouzcum ciuitas . . . obtriueratthere corresponds to hostes ad occasionem here, it could be argued from the parallelism and complementarity of the cum-lause that alam m finibus suis agentem corresponds to m2lites... ad securitatem. In that case securitatem will mean 'carelessness', which is also the meaning on the only other occasion

where

the

noun

is found

alongside

occasio,

namely,

Velleius'

description of the prelude to the Varzana clades in AD 9 (118.2): ‘iuuenis [i.e. Arminius]...segnitia ducis in occasionem sceleris usus est, haud imprudenter speculatus...frequentissimum initium esse calamitatis securitatem'. On the other hand, T.'s later reference to the soldiers' anticipating their rest-period for that vear (2 'praesumpta apud militem illius anni quies’) strongly supports the meaning ‘safety’. uelut omissa expeditione: uelut means either ‘on the grounds that’ (OLD 6), giving the soldiers' own reason for their reactions, or 'asif' (OLD 5c), providing an authorial comment from T.: while it is difficult to choose between these two, the latter is perhaps more pointed, in that it more readily invites the thought ‘but they reckoned without Agricola'. tamquam is sometimes similarly ambiguous (15.1n. nzhil). omissa expeditioneis Livian

(29.35.6).

Ordouicum

ciuitas...uniuersam

obtriuerat

the

'community

of the

Ordovices' inhabited the north-western corner of Wales; since their name

means ‘hammer-fighters’

(Rivet and Smith

(1979) 434), obtriuerat ('had

COMMENTARY:

18.2

183

crushed’) 15 evidently an etymological pun: these natives lived up to their name (see also 16.1n. ira). Their southern neighbours were the Silures (11.2, 17.2), and both tribes had been attacked by Ostorius Scapula (14.1n. Consularium) during his governorship of the province (A. 12.92.2—99). alam ...agentem: an ala was a cavalry unit of auxiliary troops, nominally 5oo-strong (quingenaria) but, even if at full strength (which Roman military units rarely were), more likely no more than 480 (comprising sixteen turmae, for which 566 36.9). agentem — 'operating' (OLD 23), as H. 1.70.1 ‘alam Silianam circa Padum agentem'. erecta 'was roused' (OLD7) or perhaps, though less likely, *was cheered’ (OLD 6); the literal meaning

of the verb

(‘was lifted up’) contrasts with

obtriuerat above: as the Romans are pounded down, 50 the Britons' hopes rise. For eo... initio (‘with that as the starting-point) cf. À. 1.16.2 ‘eo prinCipio'. 18.2 quibus bellum uolentibus erat: sc. & as antecedent, ‘those who desired war' or, rather more literally, ‘those to whom war came willingly’. This use of esse - a personal dat. part. (instead of e.g. gratumagreeing with bellum) is modelled on an idiom which was originally Greek (LS] βούλοuat II.4) and featured in Thuc. (e.g. 2.9.2 'the majority of the Plataeans did not wish to revolt from the Athenians', τῷ ... πλήθει... o0 βουλομένῳ ἦν), from whom it was taken over by Sall. twice ( 84.9 'neque plebi militia uolenti putabatur', 100.4 'uti militibus...labor uolentibus esset'); it is once in Livy (21.50.10 'quibusdam uolentibus nouas res fore’) and twice elsewhere in T. (H. 3.43.2 'ceterisque remanere...uolentibus fuit', A. 1.59.1 ‘ut quibusque bellum inuitis aut cupientibus erat’). probare exemplum ac recentis legati animum opperiri: the subject of the historic infinitives is the understood antecedent of quibus (last n.). Gudeman thinks that frobare exemplum mcans ‘they tested the conduct' of A., but other instances of the phrase (e.g. A. 4.43.5, Cic. Verr. 3.210, Liv. 42.1.12) show the meaning to be 'they approved of the precedent' of the Ordovices. Hence acis adversative ('and yet': OLD atque 9), and, as at 20.2 (n. laudare), the contrast between the two infinitives is underlined by their chiastic arrangement. antmum primarily = 'intentions' (OLD 7 Ὁ) but also suggests ‘disposition’ (OLD 14); recentis — 'newly arrived' (OLD recens' 1b), as 31.2. animum opperiri scems unparalleled. cum Agricola...ire obuiam discrimini statuit: thc new governor's first, decisive, action is conveyed by an inverted cum-construction (G-L. $581, NLS$237);alike unusual are its introduction by the historic infiniuve (C-L 5993-—4 compares only A. 1.11.9, 11.16.2, Liv. 5.46.1) and its interruption by a further clause (C-L 584-5), in this case concessive. The fronting of A.'s name ushers in a passage which - with its complex subordinations (analysed by C-L 659 n. 9), a quamquam-clause which refers to opposing

184

COMMENTARY:

18.2

voices, and the climax of A.'s resolute action — is significantly similar to the periodic sentence which immediately precedes the engagement at Mons Graupius (35.4 and nn.; also Intro. pp. 30-1). 'Agricola'sentry upon office is to be seen as a prelude to the decisive battle’ (Lausberg (1980) 416). On inverted cum-clauses see further e.g. 25.3n. (Ad); zre obuiam discrimini is surprisingly unparalleled. quamquam...: there are three possible ways of interpreting this clause. (1) Supply erantas the verb common to all three subjects (aestas, numeriand qutes), and take tarda et contraria as their (neut. plur.) predicates: 'although X Y Z were retarding and detrimental to someone about to embark on a war'. On this interpretation transuecta, sparsi and praesumpta are genuine participles. (ii) Regard transuecta, sparsiand praesumpta as finite verbs (supplving eratand erantas appropriate), and take tarda et contraria as neut. plur. adjectives in apposition: 'although the season had passed, the units were scattered across the province, and that year's recess had been anticipated by the soldiers, factors retarding and detrimental to ... ". (iii) Regard transuecta and sparsi as finite verbs, as in (ii), but take tarda et contraria as fem.

sing. adjectives agreeing with quzes. If we again supply erat, the meaning of the third element will be eher ' (although) that year's recess, retarding and detrimental to someone

about to embark on a war, had been antici-

pated by the soldiers' or '(although) the anticipation of that year's recess by the soldiers was retarding and detrimental to ... . It is difficult to know

which of (i)-(iii) is preferable.

(i) contains at least one statement of the

obvious, namely, that the passing of the campaigning season is detrimental to starting a war; (ii) offers a very natural series of ellipses of erat and erant but involves an apposition which is awkward in itself and produces an awkward transition to et plerisque. .. widebatur, (iii) retains the naturalness of (ii) without any of its awkwardness, although it is not mentioned as a possible interpretation by other commentators. Whichever interpretation is chosen, the quamquam-clause consists of an asyndetic tricolon crescendo followed bv a 'cap' (et... uidebatur). transuecta aestas: the verb in its literal sense is used of conveying or passing across (e.g. 4, 13.3, 37.2; OLD 1—2); its metaphorical application to time 15 seemingly paralleled only at H. 2.76.9 'abiit iam et transuectum est tempus'. If transuectais a genuine participle (interpretation (i) above), the phrase will be a nominative ex. of the ‘ab urbe condita construction' (10.4n. Romana): 'the passing of the season'. 50 too sparsz... numeri and

praesumpta . . . quies below.

numeri is a term for military units which was coming into use in the late first century AD (OLD 9). For sparsi cf. OLD 6a. praesumpta...illius anni quies: quies seems to be used in a quasitechnical sense to refer to the non-campaigning part of the year when the troops would retire to winter quarters (cf. Liv. 5.2.7 ‘ne hiemis quidem

COMMENTARY:

spatio, quae omnium

18.2

185

bellorum terra marique sit quies', quoted in OLD

2c; note also 2a, 4b). praesumo means

'to perform, enjoy, etc., before the

due time', i.e. ‘anticipate’ (OLD g, though our passage is listed under OLD 4 'to take for granted, presuppose, assume"); the expression is explained by the fact the soldiers had reached almost the end of the season without there being a campaign (1 'uelut omissa expeditione’). apud = ‘on the part of' (OLD 15a). tarda et contraria bellum incohaturo 'retarding and detrimental to someone about to start a war'; see above (n. quamquam) for the question whether the adjs. are neut. plur. or fem. sing. For the active sense of tardus see the exs. in OLD

1b (all poetic);

quzes tarda is found

(by implication)

at Serv. Aen. 8.30, where the meaning is ‘late sleep'. For bellum incohare cf. Liv. 35.15.6, 40.26.3, 41.23.13, Vell. 96.2, Suet. Aug. 16.1, Veg. 3.3.3, Just. g prol; the verb is a favourite of Livy, the vounger Pliny and T. plerisque custodiri suspecta potius uidebatur 'that the suspected areas should be guarded seemed preferable to many’: potius is a neut. comparative adj. (OLD potior* 5a), as Liv. 38.5.10 'potius uisum est Amphilochis...ferre opem’; for suspecta see OLD suspectus' 1a. custodiri implies being kept under close observation (OLD 7a). This 15 the first, but not the last, occasion on which A.'s excellence is enhanced by contrast with the preferences or advice of others (cf. 25.3, 35.4): it was a topos of panegyric to stress the uniqueness of the laudandus (Quint. 3.7.16). contractisque legionum uexillis 'and having mustered the legionary detachments' (OLD contraho 4a, uexillum g). Whether modica...manu (below)

is also to be taken with contractis or whether the phrase

(regular

in military narrative) is an 'abl. of military accompaniment' without cum (G-L $392 R. 1) is unclear. in aequum degredi is a rare variant (again at Liv. 22.12.7) on the common n aequum descendere. One must remember that North Wales is hilly and indeed boasts the highest mountain in the British Isles outside Scotland (Snowdon, 3500 ft). non audebant: barbarians are stereotypically cowardly when it comes to actual fighting (11.3n. in deposcendis), though their refusal to leave an easily defensible position such as high ground makes good strategic sense, and A.'s decision to charge uphill should seem the height of foolishness. ipse..., quo ceteris par animus simili periculo esset 'in order that the others should have equal courage in their common danger, he personally...’: this is the answer to the Ordovices' question about A.'s anzmus (above). T. is alluding to Sallust's description (of Catiline!) at C. 59.1 ‘quo militibus exaequato periculo animus amplior esset, ipse ...’, a passage also relevant to 35.4 (where see n. on the inadvisability of leading from the front). quoin a final clause without a comparative is an archaism which is

186

COMMENTARY:

18.3

found relatively infrequently in Cicero and other late-republican authors (Caes., Nep., Varro) except Sallust, who uses it relatively frequently. The usage becomes a favourite of T. (Adams (1972) 356). erexit aciem ‘he led the line uphill', a technical expr. (36.2n. erigere), but suggestive of a response to the provincials' earlier arousal (1 'erecta prouincia’). 18.3 caesaque prope uniuersa gente: this is A.'s response to the Ordovices' earlier opportunism (~ 1 'alam...prope uniuersam obtriuerat'). Battle narratives tend to generate universalising language (K. on Liv. 6.3.5 omnia) and claims to have annihilated the enemy are common, e.g. Caes. G. 2.28.1 ‘prope ad internecionem gente ac nomine Neruiorum redacto', Liv. 6.33.12 ‘in medio caesi ad unum omnes' with K. non ignarus instandum famae: sc. esse. 'not unaware that he should follow up his reputation', as H. 5.15.2 ‘instare fortunae' (and H.), Plin. Ep. 4.11.19 'non esse uerecundiae eius instandum’; OLD insto 7a. Awareness and strategic knowledge are further marks of the ideal general, as in T.’s presentation of Germanicus (e.g. A. 2.20.1 'nihil εχ his [enemy dispositions] Caesari incognitum: consilia, locos, prompta, occulta nouerat’); see

also K. on Liv. 6.30.7. And, whereas A. is anxious to suppress his military reputation amongst the audience back in Rome (below: 6 and nn.), he knows that he must capitalise on itin Britain if he is to continue successful on the battlefield. instare famaeis unparalleled (unless Haase's emendation ofA. 13.8.3 15 accepted). For T.’s variable use of ignarus (greatly preferred in H. and A. 6-14) and its archaising synonym nescius (preferred in A. 3—5 and 15—16 but completely absent from the minor works) see Adams

(1972) 367-8.

ac, prout prima cessissent, terrorem ceteris fore 'and, in accordance

with how his first actions turned out, terror would attend the rest’; cessissent (OLD 74) represents a fut. perf. in direct speech (NLS $272 (3) (d)); commentators are divided on whether ceteris is masc. or neut. ‘Max-

imising first impressions in warfare is a topos' (Ash on H. 2.20.2 'ut initia belli prouenissent, famam in cetera fore’; cf. also 3.70.3 ‘cetera secundum euentum proelii cessura', A. 12.31.2, 13.8.3, Thuc. 6.49.2, 7.42.3; Tosi $8811—12). For the importance of terror in Rome's relations with foreigners 566 Mattern (1999) 115-22, 221-2; here prout indicates proportion: the degree of terror felt by the rest would depend on how his first actions worked out. Monam insulam: the Isle of Anglesey (14.3n.). possessione 'occupation' (OLD 2a) or 'seizure' (OLD 2b, where this passage and A. 2.5.4 are listed). A plain abl. with reuocareis found at Virg. G. 4.88 ‘ubi ductores acie reuocaueris' and Liv. 25.36.2 (where some edd. read «e- proelio) ; F-A suggest that a has 'dropped out after znsula ', but H reads znsulam in full.

COMMENTARY:

18.4

187

supra memoraui: at 14.3—16.2. Cross-references are common in Latin prose authors (Starr (1981)), but, whereas supra is ‘by far the most common' referencing word in such cases, memorare (10.1n.) 15 correspondingly uncommon (ibid. 432). T. uses supra memoraui4x in A. and, in the plural form, 5x in H.; he is presumably imitating Sallust, in whom supra memoraui(mus) occurs 5x. Elsewhere onlv at Plin. NH 4.13 (plural; supra memorauit at Ascon. p. 7C.9-10). redigere in potestatem 'to bring under his control': administrative language (OLD redigo 10a). 18.4 ut in subitis consiliis ‘as in the case of improvised planning’: for this elliptical use of ut see 11.1n. The exact meaning of subita consilia will depend on context: Caesar, for ex., uses it to characterise the mercurial Gauls (G. 3.8.3); cf. also Liv. 2.28.1, 35.35.4, Plin. Ep. 3.16.9. Since preparation is a key quality of the ideal general (K. on Liv. 6.18.7), ut... consiliisis exculpatory, but the sentence of which it 15 part merely acts as a foil to underline A.'s resourcefulness (below). ratio et constantia ducis transuexit: adversative asyndeton: ‘ the leader's calculation and resolution conveyed across'. Though Julius Caesar did not lack ships when he proposed to cross the Rhine in 55 BC, he too resorted to an inventiveness (building a bridge) which would intimidate the locals: see G. 4.17-19 with Lausberg (1980) 416. The combination or juxtaposition of ratio and constantia seems exclusively Ciceronian

(Mil.

62,

Sull.

2,

Clu.

19, ND

2.56,

Div.

2.18,

Parad.

Stoic. 3.22). depositis...uenientibus: this complex sentence moves us from Roman stratagem to enemy reaction, as A. takes the island without striking a single blow. Its basic structure hinges around the consecutive clause and its introduction (?2ta repente. .. ut...): before that we have a participial phrase (depositis...sarcimis), a noun-object 4 relative clause (lectissimos ... quibus...), and finite verb (2znmisit); within the ut-clause we have a noun-subject + relative clause (Aostes quz...), finite verb (crediderint), and

a participial phrase (sic... uenzentibus). deponere sarcinas naturally recurs in militarv narrative (H. 4.34.3, Liv. 6.3.5, Curt. 4.9.9). quibus nota uada et patrius nandi usus quo ...regunt: many commentators, including O-R, explain nota uada as a reference to the Menai Strait (between the Welsh mainland and Anglesey), on the grounds that swimming featured in Paulinus' original attack on the island and that some of those earlier attackers were now pressed into a repeat performance in ‘the familiar shallows' in AD 77. Yet (a) according to T.'s own later account of the previous attack, swimming was restricted to (some of?) the cavalry and formed only a minor and incidental part of the action; the principal force crossed by ship (A. 14.29.3 ‘nauesque fabricatur plano alueo aduersus breue et incertum. sic pedes; equites uado secuti aut altiores inter

188

COMMENTARY:

undas adnantes equis tramisere’).

18.4

(b) Seventeen years had passed since

Paulinus' attack in AD 60, and, while there may have been a very few active survivors, it seems highly unlikely that they could now constitute an assault unit described as ‘lectissimos auxiliarium' and responsible for seizing the entire island. (c) patrius nandi usus suggests Batavians (36.1n.), who, since they were famous for their swimming (e.g. A. 2.8.3, CIL 3.3676; Ash on H. 2.17.2, H. on H. 4.12.3), would comprise a specialist force selected for this specific assault; and the present tense of regunt suggests that T. 15 talking in generalised terms. Hence, as F-A and Forni maintained, the quibusclause seems likelv to mean *who were used to shallows and had a native method of swimming whereby...’, a typically Tacitean circumlocution. It must be admitted, however, that the expr. nota uada elsewhere has a specific meaning, 'the familiar shallows' (H. 5.15.1, Caes. G. 4.26.2, Liv. 27.47.9, Ov. Met. 1.3770, Curt. 4.7.19, Val. Fl. 3.43, Frontin. Strat. 2.5.23). T.’s use of usus probablv comes under the heading 'a frequent or general practice' (OLD ga), like H. 5.6.3 ‘legendi usum', Virg. Aen. 12.396 'usumque medendi’; the very expr. patrius. .. ususoccurs at Val. Fl. 1.439 ‘nec patrio ... ignobilis usu’, ‘not unknown for his father's practice [sc. of herald]’. seque et arma et equos regunt: the formulaic tricolon u?r? arma equi (as e.g. H. 1.51.2: see Heubner (1963)) is here varied by the substitution of the reflexive pronoun. -que et ('both ...and...") 15 the archaic equivalent of the Greek τε kaí and 15 common in Sallust, Livy and T. but completely avoided by Cicero; usually -queis attached to a personal pronoun, as here. For instances of a second et, also as here, see e.g. A. 1.4.1,

Sall.] 76.6, Bell.

Afr. 93.3. regunt (‘manage’) combines the notions of control and direction. See 4.2n. (bonam),

10.6n.

(ac) for other combinations of conjunctions.

ut...nihil arduum aut inuictum crediderint sic ad bellum uenientibus: sc. esse or fore. 'that thev believed nothing was/would be steep or invincible for people who came to fight in such a manner'. Compare the German reaction to Germanicus at A. 2.25.3 (‘inuictos et nullis casibus superabiles Romanos praedicabant, qui perdita classe, amissis armis, post constrata equorum uirorumque corporibus litora eadem uirtute, pari ferocia et uelut aucti numero inrupissent’) or the remarks about the Athenians at Thuc. 7.28.3 (described by Hornblower ad loc. as ‘a classic piece of "presentation through negation"', comparing 1.105.4, 7.22.2, 7.42.2). The earlier British war-mongering (2 ‘quibus bellum uolentibus’) is here subtly answered by the effectiveness of the Roman war machine ('sic ad bellum uenientibus’).

obstupefacti hostes: in AD 82 Britons will be equally deployment of his fleet (25.2). Compare the boldness 2.6.10 'ipso miraculo audaciae obstupefecit hostes'. qui classem, qui nauis, qui mare expectabant: the qui, unusual in a narrative passage, marks three terms

astonished by A.'s of Horatius at Liv. triple anaphora of which are virtually

COMMENTARY:

18.5

189

synonymous: perhaps we are to imagine different groups of Britons, each voicing the same mistaken thought in a different way. expectabant is perhaps a metatextual allusion to the fact that it was standard in panegyric to say that the laudandus had performed 'supra exspectationem’ (Quint. 3.7.16). 18.5 Ita is to be taken closelv with the following ablative absolutes: because the inhabitants were so impressed by A.'s amphibious operation (‘nihil arduum...uenientibus’), they therefore sought peace and surrendered the island. clarus ac magnus haberi Agricola: though the combination of these adjs. is not uncommon, e.g. in the preface to Cato's Origines (fr. 2P — 2C 'Clarorum uirorum atque magnorum...': see above, 1.1n.), haberi shows that T. 15 alluding to Sallust: C. 53.1 'Cato clarus atque magnus habetur’, [ 92.1 ‘Marius... magnus et clarus antea, maior atque clarior haberi coepit’. Scholars prefer to privilege the first of these two passages, on the grounds that T. 15 presenting À. in partly Catonian terms (below, 6n. :psa; Lausberg (1980) 417-18). magnus, on the other hand, initiates the repeated theme of A.'s 'greatness' (cf. 44.2n. bonum), which at times is suggestive of Pompey the Great (42.4n. magnos). For clarusin historiographical contexts see K. on Liv. 6.1.3. quippe cui... placuisset: the clause 15 causal (G-L $626 N. 1, NLS $157 (1) (b) Note), explaining clarus ac magnus haberi Agricola. Τ. uses quippe qui only here for his usual ut qui. quod tempus alii...transigunt: the antecedent (fempus) has been attracted into the relative clause despite being in apposition to the notion of 'first steps' which is inherent in zngredienti prouinciam (‘a period which others...’).

per ostentationem et officiorum ambitum 'in ostentation and canvassing for compliments’ (OLD ambitus 7a, officium 2a); Gudeman seems to imply that the latter phrase — 'a round of official duties', but it is very doubtful whether ambitus can have this meaning. In his narrative of AD 62 T. puts into the mouth of Thrasea Paetus a speech condemning the way in which provincial governors tried to ingratiate themselves with the locals (A. 15.21.1—4 'at nunc colimus externos et adulamur...in modum candidatorum suffragia conquirimus’); Thrasea was talking about the end ofa term of office rather than the beginning, but the principle is the same. A. avoided both this vice and ostentatio, but, in an irony picked up later (6),

he nevertheless became clarus. labor et periculum corresponds to, and negates, ostentationem et...ambitum above. The combination of nouns is extremely common, esp. in Cicero and Livy; the former says that grounds for praising a man are 'si ab eo cum magno aliquid labore aut periculo aut utraque re,

cum

necesse

non

esset,

officii

causa...in

rem

publicam ...factum

190

COMMENTARY:

18.6

esse' (Inv. 2.35): not surprisingly /abor and periculum are conventionally associated with the ideal general (W. on Vell. 79.1). 18.6 nec is to be taken with both usus and uocabat. prosperitate rerum is an extremely rare variant (elsewhere only at Suet. Calig. 31.1) on the common prosperis rebus. in uanitatem 'for the sake of vanity' (OLD ?n 21a, uanitas 2b). (nec) expeditionem aut uictoriam uocabat uictos continuisse

‘nor did

he call it a campaign or a victory to have contained the conquered.' This reaffirmation of A.'s modesty is nevertheless consistent with the notion (above) that the island surrendered without a fight. Word play involving forms of wuictoria and uzctus is very common (e.g. Cic. Fam. g.2.2 ‘qui enim uictoria se efferunt quasi uictos nos intuentur’) and is here sharpened by the insertion of the alliterative uocabat. ne laureatis quidem gesta prosecutus est ‘and he did not follow up his achievements even with a laurel-wreathed despatch' (OLD prosequor 7a); see too 39.1 (n. epustulis). laureatae, originallv an adj. (e.g. Liv. 5.28.13 'litterae a Postumio laureatae sequuntur uictoriam populi Romani esse’), had evidently become a self-standing noun: see the Fast: Ostienses for February, AD 116 (= Smallwood §23) '... K. Mart. laureatae missae ad sen[atum ab Imp.] Traiano Aug. ob q[u]am causam Par[thicus appell.]'. For the practice of sending such a despatch to Rome to announce a victory see Plin. NH 15.139. ipsa dissimulatione famae famam auxit: 'Agricola through a positively evaluated dissimulatio paradoxically wins more fama than he otherwise would

have'

(Hardie

(2012)

276).

The

paradox,

which

is another

fea-

ture shared bv A. with the younger Cato (Sall. C. 54.6 ‘quo minus petebat

gloriam, eo magis illum sequebatur’; see above, 5n. clarus), is underlined

by the polyptoton (for this kind of 'compensatory' repetition see Wills (1996) 192-3). Additionally dissimulatio 'anticipates the dark narrative of events after Agricola's return to Rome from his success in Britain, where his own dissimulation is the onlv defence against a dissimulating emperor' (Hardie, ibid. 277). dissimulatio famaeseems unparalleled in classical Latin but famam augereis regular (H. on H. 2.1.2). aestimantibus is probably a type of compressed abl. abs. (for which see G. on A. 1.29.2; also below, 35.4n. promptior), appended to the sentence as often (Intro. p. 32), but dat. ‘of the person judging’ (as 10.3 transgresszs, 11.3 aestimanti) 15 also possible.

quanta futuri spe tam magna tacuisset: the abl. phrase (of cause or attendant circumstances) convevs the principal idea: 'reckoning how great must be his hopes for the future, given that he had kept silent about matters so important'. The perverse reading of A.'sactionsis shared, of course, by Domitian: see e.g. 39.2 and nn.

COMMENTARY:

19.1

Ceterum

effects the transition

19.1-2

from

191

the military narrative

(18),

which 15 essentially sequential, to the domestic 'coda' (19.1-20.1), which 15 essentially thematic: see above, p. 180. For ceterum see 11.1n. animorum prouinciae prudens: whereas at 18.2 the Britons waited to discover the new legate's anzmus, A.’s two earlier spells in the province as a subordinate (5, 7.3—8.3) mean that he is already knowledgeable about theirs, knowledge which can only have been improved by the most recent encounter (18.3—5). The structure of this sentence 15 chiastic up to sequerentur, though the last element (per... sequerentur) — which itself includes a chiasmus (parum...sequerentur) — is greatly elongated. The metonymy of prouincia for prouinciales s common (OLD 3d), as 15 prudens - gen. (OLD 3b). doctus ...parum profici armis 'having learned ...thatlittle was achieved by fighting’: for impersonal profici -- abl. see 15.1n. (nihil); the coupling of doctus with prudens (again at D. 32.2) occurs in a wide range of authors from Cicero and Lucretius onwards. per aliena experimenta: A.’s personal experience (above) 15 complemented by his familiarity with that of others; given the role played by exempla in Roman culture (46.2n. semilitudine), the importance of such familiarity was frequently stressed (e.g. D. 34.6 'eruditus...alienis experimentis', Sall. | 46.5, Frontin. Strat. 1.10.1 ‘Sertorius...experimento didicerat', Veg. 3.1.8, Just. 38.4.5) and could be learned from books as well as from life (see e.g. Sall. J. 85.13, Frontin. Strat. 1 praef. ‘probatis...experimentis').

There

is no need

to think

(with O-R)

that T. is

referring only to Boudica's revolt. For experimentum — 'experience' see OLD 2. causas bellorum statuit excidere: though we have just been given a stark account of A.’s military ruthlessness (18.9 'caesa...prope uniuersa gente’), he sees that insurrection is often prompted bv corruption and malpractice (iniuriae, cf. also 13.1), generating a vicious cycle which he proposes to interrupt by eradicating the causes: excidereis a metaphor from felling trees or pruning vines and the like (OLD 1b). T. returns to illustrate these causae/iniuriaein $4 below, using another horticultural metaphor for A.'s treatment of them

(cercumcisis: see n.), but before that there is an inter-

vening section (2—3), constituting a kind of 'foil', in which he describes,

with a further horticultural metaphor

(coercuit: see n.), A.’s methods

of

dealing with his own household from the points of view of employment and punishment. The clustering of horticultural images in this passage implies that on his arrival in the province as governor Agricola is ensuring that he is true to his name. See also 1.1n. (Clarorum).

19.2 ἃ 56 suisque orsus...: the 'foil' (see last n.) is introduced by adversative asyndeton: ‘ beginning with himself and his personnel [OLD

192

COMMENTARY:

19.2

suus 6a]...'. It was a commonplace that a leader should above all be able to control himself (e.g. Cic. Leg. Man. 38, Liv. 34.18.4; also above, 5.1n. nec).

domum suam coercuit: it is known that in the later Roman empire provincial governors had an elaborate headquarters staff or officium (RGB 11, with references; IRB 125-35), but there is very little literary and inscriptional evidence for 115 complement in the first century (Dise (1997) 273—5). The context (below, n. non studiis) shows that T. 15 thinking not of

A.'s personal am:ci but principally of those staff-members drawn from the army. For coercere in the context of restraining undesirable activities see Liv. 32.27.4 (the elder Cato as new provincial governor) 'Sardiniam M. Porcius Cato obtinebat, sanctus et innocens, asperior tamen in faenore coercendo habitus'. There is also an echo of A.'s youthful experience of his being restrained from philosophy (4.3), where, however, the metaphor 15 different (see n. studium). Here

coercere is a horticultural metaphor, for

which see OLD 3a—b; for domus — ‘household’ see OLD 6a, and note Saller

(1984) 342-9.

quod plerisque...'«something- which...': quod refers back to the whole of the main clause (G-L $614 R. 2); plerisqueis again used to emphasise by contrast the uniqueness of A. (see 18.2n. plerisque). The locus classicus for a description of the role and behaviour of a provincial governor is Cicero's famous letter of 60 Bc to his brother, Quintus (Q.Fr. 1.1). nihil per libertos seruosque publicae rei ‘ no public business through freedmen and slaves.” Whereas res publica = ‘the state’ or ‘the commonwealth', publica res — ‘public business', as G. 13.1 ‘nihil autem neque publicae neque priuatae rei nisi armati agunt’, Liv. 27.6.18 ‘hi censores neque senatum legerunt neque quicquam publicae rei egerunt'. It was T.'s view that 'keeping freedmen out of public service is a commendable policy' (Damon on H. 1.58.1 'Vitellius ministeria principatus per libertos agi solita in equites Romanos disponit’, quoting also 1.76.3, Plin. Pan. 88.1—2, Ep. 6.31.0), and acting through freedmen and slaves is so characteristic of the imperial regime in the first century (Osgood (2011) 38—40, 203—5) that Tiberius was praised for his restraint in this area (A. 4.6.4 ‘modesta seruitia, intra paucos libertos domus’) and Nero on becoming emperor assured senators that his household was separate from the state (A. 19.4.2 ‘discretam domum et rem publicam’). Indeed, while any form of delegation could be a sign of trust, it could also prove problematic, as was shown by the cases of Sejanus and, in the republic, Jugurtha (Sall. ]. 8.6 ‘igitur imperator omnis fere res asperas per Iugurtham agere’). In our passage a verb must be supplied, as at Cic. P. 1.6 ‘nihil per senatum, multa per populum', and the above quotations suggest that a form of agois

COMMENTARY:

19.3

193

the likeliest; in view of the remarkable sequence of nine historic infinitives which follows (the highest in Agr.: see 5.1n. noscere, 38.1n. Britann?), that form will be agere. Such ellipse of verbs of action or motion is common in T. (M-W on A. 4.12.4). non studiis priuatis nec ex commendatione aut precibus centurionem militesue adscire 'he did not engage a centurion or soldiers...' (sc. as members of his staff). Each of the rejected alternatives has a different reference: the ‘personal inclinations' are those of A. himself, the recommendations are those of third parties, and the pleas are those of the potential employees. Staff-members were called beneficiari (for some first-century examples in Britain see e.g. IRB 15 — ILS 2648, 129 — RIB 293 add.), since their release from routine military duties depended on the favour or beneficium of the governor (cf. e.g. H. 4.48.2): the beneficiary status of their employment serves to underline the propriety with which A. handled such matters

(below, n. sed optimum), as does his imperviousness

to letters of

recommendation, a standard feature of imperial life (e.g. Lendon (1997) 48—9, 61—4). adscireis a rare equivalent of the much commoner adsciscere and occurs earlier only at Virg. Aen. 12.38; but Puteolanus' emendation not only provides the appropriate term butalso a verb which T. likes (again at H. 4.24.1, 4.80.1, A. 1.3.5, then in later poetry). The uarzatio of abl. — prepositional phrase is common in T. (6.1n. per mutuam). centurionem: each Roman legion comprised sixty 'centuries', or companies of roughly 8o men; each century was commanded by a centurion, and six centuries constituted a cohort (cf. 28.1n. cohors). Centurions had

a reputation for violence (15.2n. alterius) and make frequent appearances in Roman historical narrative, esp. Caesarian (Brown (1999) 350). sed optimum quemque fidissimum putare: probably understand fore: ‘but he thought that all the best men would be the most loyal'. Loyalty was A.'s primary requirement in his staff, and he considered that the quality was most likely to be found in those of demonstrated sociomoral excellence. For the lapidary form of expression cf. Cic. Fin. 2.81 'optimum quidque rarissimum est', De or. 2.314 'firmissimum quodque sit primum’. 19.3 omnia scire, non omnia exsequi: whereas Julius Caesar paid no attention to some faults and did not punish fully those that he did see (Suet. D/ 67 ‘delicta neque obseruabat omnia neque pro modo exsequebatur’), A. knows every fault but omits to punish some altogether. Thus some faults are allowed to go unpunished bv each man (next n.), albeit for different reasons; but, whereas Caesar's observation is only partial, A.'s

comprehensive knowledge — emphasised here by the anaphoric jingle — marks him outas the superior leader (cf. Isoc. Evag. 42 αὐτὸς συνήδει!). Likewise Trajan's comprehensive knowledge of provincial matters is praised by

194

COMMENTARY:

19.4

Pliny (Pan. 70.3 ‘quidquid a quoque in prouinciis bene fieret, omnia te scire’), but the point there is of rewarding merit rather than punishing malpractice. For A. and Caesar see Intro. p. 29. For the sequence of compound verb (adscireabove) followed by simple

(sczre here) see W-M

on A.

3.29.1; for exsequi of punishment see OLD 3a. paruis peccatis...cum peccassent: the first two statements, whose chiastic arrangement underlines the complementary actions of granting pardon and accepting remorse (uenzam...seueritatem ~ poena....paenitentia), illustrate non omnia exsequi and its implication that some offences were indeed punished, while the third, itself chiastically arranged (non peccaturos praeponere ~ damnare cum peccassent), adds a new factor by way of climax. A.’s combination of clemency and severity recalls his civilian policy in Aquitania (9.3 'seuerus et saepius misericors') and is characteristic of the ideal general (see the exs. collected by W. on Vell. 114.3). ueniam...seueritatem commodare: the verb (‘he accorded': OLD 4) 5 not elsewhere combined with these nouns, which are likewise contrasted

at Sen. /[ra 2.10.4. seueritas was notionally a quality much prized in generals (see e.g. Phang (2008) 111-15), though in Flavian times it had been more honoured in its absence than in its presence

(Damon

on H. 1.5.2).

nec poena semper sed saepius paenitentia contentus esse: the (alliterative) chiasmus contrasts with the parallelism of paruss...uentam, magnis seueritatem above (cf. 20.2 and n. laudare). Some have considered poena to be nominative, but this interrupts the otherwise unbroken sequence of A.'s actions; others have suggested supplying u£ from contentus 6556 by zeugma, but poena contentusis a regular phrase (Cic. Lig. 13, Hirt. G. 8.21.2, 22.2, Sen. Clem. 1.9.11, [Quint.] Decl. 280.9, 301.14) and the meaning here (‘and he was not always satisfied with punishment but more often with remorse’) seems unexceptionable. For plays on poena ~ paenitentia elsewhere see e.g. H. 5.25.3, A. 1.45.2, Sen. Ira 3.26.2, Quint. 7.4.18. Kaster (2005: 82) maintains that the Romans 'almost never used the language of paenitentia’ to represent remorse but he does not mention this passage. officiis...praeponere ‘he would rather place likely non-offenders in charge of administrative duties’: officizs (see OLD 4-6) and administralonibus (OLD 3b) are almost synonymous and form a kind of hendiadys; nonis to be taken closely with peccaturos (a rare use of the fut. participle as a noun: see NLS$101). 19.4 frumenti et tributorum exactionem aequalitate munerum mollire: a very difficult passage. tributum is a general term for the imperial taxes, in many provinces collected in coin, but in some in kind; since frumentum is a tax of grain, the tributum in some provinces will have consisted of the frumentum. Whether that was the case in Britain is unclear. If frumenta in

COMMENTARY:

19.4

195

the next sentence refers back to (i.e. is synonymous with) :pso tributo in the appended abl. abs., as is certainly possible, frument: et tributorum here may be seen as a kind of hendiadys — 'grain taxes' (contrast 31.1 below). On this interpretation (and it is acknowledged that others are possible) aequalitate munerum would refer to the fair application of the grain tax amongst the local population (munus is short for munus publicum, a duty owed by an individual or community to the state: OLD 2a). exactiois a technical term for the ‘calling in of taxes, debts, etc.” (OLD 2a): since taxes

were the responsibility not of a province's governor but of its procurator (RGB 298), A. is evidently ‘pulling rank'. On taxation in general see e.g. Brunt (1990) chap. 15; the discussion of our passage by Mann (1985) is misleading. circumcisis quae...grauius tolerabantur 'lopping off the profitable schemes which were more painfully endured than the tax itself’: the schemes are described in the next two sentences (namque...). On the assumption that /rzbuto is being used to refer to the frumentum itself (so F.; see last n.), it perhaps gives additional point to the polyptoton with tributorum. in quaestum, again at G. 24.1, is a regular phrase — 'for profit’ (Cic. Flacc. 56, Quinct. 11, Manil. 4.132, Sen. Ben. 4.13.3, Quint. 7.2.24); circumcido, again at D. 32.4, is another horticultural metaphor (OLD 2c); for tolerare see 13.1n. per ludibrium...cogebantur ‘by way of mockery they were compelled to attend at closed store-houses and contrarily to buy grain and to provide compensation at a price.' Scholars have offered two possible interpretations, both of which assume that the Romans demanded their lew of grain at times when none was to be had. Either the provincials were compelled to fulfil their obligations by purchasing grain already stored in the Roman granaries (this seems to have been the scheme operated by Verres in Sicily: 566 Cic. Verr. 3.178) or— a variant on the former - the grain already in the Roman granaries was sold to the provincials at an exorbitant price and then bought back at an official (but of course much lower) price, the difference being pocketed by the Roman overseers (so e.g. Mann (1985) 22). In either case the grain never left the granaries. ultro indicates that a situation is the reverse of what it has been or should be (OLD 4); luere pretio is an obscure expression but the verb seems to refer to the discharging of the original demand for grain, while pretio seems to refer to the monetary cost incurred in discharging that demand. diuortia itinerum...lucrosum fieret: Cicero reports of some governors that 'they established the practice of ordering grain to the remotest places where transport was the most difficult’ (Verr. 3.190); their purpose was evidently to charge the provincials extortionately for the transportation, whereas in the case of Britain it is assumed that officials were accepting

196

COMMENTARY:

20.1

bribes from those who wished to avoid these long-distance journeys. Here the inherent contradictions of the situation are underlined bv the contrasts proximis — remota et auia and omnibus — paucis. diuortia itinerum et longinquitas regionum indicebatur: /ongznquitas regionum is Livian (7.29.1); diu. itinerum is also in Livy (44.2.7), where, however, it refers to a fork in the road (see Briscoe ad loc.), which clearly cannot be the meaning here. T.’s two elements are picked up chiastically by remota et auia below, which suggests that the whole expression here means ‘branch roads and distant regions would be prescribed’ (for this type of genitival phrase instead of adj. + noun see 21.2n. porticus). This extended meaning of diuortium is not recognised by OLD or TLL and is hard to parallel (Amm. 28.5.6 is not as clear as O-R imply). The singular verb takes its number from the nearest subject (G-L $285 E. 1). ut...deferrent: the verb — 'deliver', 'transport' (OLD 2a); the object (frumentum) has to be understood. Whether the clause is purposive or consecutive is unclear. proximis hibernis has been taken as a concessive abl. abs. ('despite the proximity of winter-quarters’), as separative with the verb (‘from’), orasan extended abl. of description or quality (‘communities with nearby winterquarters’, cf. NLS$83). The last of these perhaps provides the best contrast with n remota et auia. Watt (1987: 465) was anticipated by Ritter in proposing proximae. in remota et auia: the closest parallel for the combination seems to be Stat. Szlv. 4.5.290—30 'tene in remotis Syrtibus auia | Leptis creauit?’, but cf. also 37.5 'longinqua atque auia’. donec...fieret: ‘the subjunctive expresses the purpose in the minds of the officials’ (F-A): this is probably correct (‘until what was within every-

one's capability should become profitable for only a few’), although by T.’s time donec + subjunc. could be used without any idea of intention (NLS 8224 Note ii). For the meaning of in promptu see OLD promptus 3 (cf. 1b).

20.1 haec primo statim anno comprimendo ‘by suppressing these things ΪΠ his very first year’: for the idiomatic use of statim see 3.1n. (przmo). egregiam famam paci circumdedit: the metaphor, which envisages peace as being wrapped in clothing and is unique to T. (cf. D. 37.6, H. 4.11.2), 15 said to be formed on the analogy of the Greek περιβάλλω (e.g. [Long.] Subl. 1.3 ταῖς ἑαυτῶν περιέβαλον εὐκλείαις τὸν αἰῶνα, ‘they wrapped eternity around their fame") or περιτίθημι (e.g. Thuc. 4.87.6 τῇ πόλει τὸ κάλλιστον ὄνομα περιθεῖναι, ‘to wrap the fairest reputation around the city’);

very similar are the famous words of Anchises at Aen. 6.852 'pacique imponere morem' ('to deck peace with civilisation': OLD zmpono 6a). egregia fama is a combination exclusive to Sallust (H. 4.69.2, cf. Epist. Caes. 1.6.1) and T. (5x).

COMMENTARY:

20.2

197

quae...timebatur '(peace) which, thanks to the indifference or intolerance of his predecessors, had been feared no less than war' (sc. because it allowed the above abuses to flourish: cf. A. 12.33.1 'qui pacem nostram metuebant', also of Britain). The narrative of AD 77 concludes with Cicero’s second favourite clausula, cretic + spondee (Wilkinson (1963) 156); for the comparative form of the sententia see 1 2.6n. (ego).

20.2—21.2 The year 78 20.2 ubi aestas aduenit: that of 78. The verb is regular of periods, dates

and the like (OLD 6a) but is not elsewhere found with aestas.

contracto exercitu: this phrase, again at 32.1 and A. 14.38.1, is common in the historians and here presumably refers to the assembling of the troops from their various winter quarters at the start of the new campaigning season. multus in agmine ‘, were no source of alarm...' The personal use of annuus 15 elsewhere restricted to magistrates and the like (OLD 3) but a few slightly wider applications (Cic. Red. Sen. 4 ‘tutoribus annuis’, All. 5.19.3, 5.17.5) perhaps support the interpretation suggested here. moras obsidionis: military life was noted for its longueurs (e.g. Hor. Odes 2.7.6 *morantem ...diem") but ‘protracted sieges’ (for the form see 21.2n. porticus) were a special case (cf. e.g. Aesch. Ag. 558-66, Sall. H. 4.69.14 'exercitu in obsidio moranti frumentum defuit', Ov. F. 2.721 'patitur longas obsidione moras', Frontin. Straf. 1.3.4 ‘obsidionalis morae’, Flor. 1.40.17 ‘ex mora obsidi’). In T.'s minor works the archaising obszdzum is not used at all, and obsidio (the form used by Cic. and Liv.) only here; but he comes to prefer the former, esp. in A. 1-6 (W-M on A. 3.39.1). 22.9 sibi quisque praesidio: this must mean something like ‘each man was self-reliant', i.e. was not compelled to rely on outside help. The nonpersonal use of zntrepidus (above) 15 most unusual

(OLD 2).

inritis hostibus eoque desperantibus quia... ‘with the enemy thwarted and therefore in despair because ...’; the second appended abl. abs. of the paragraph is here continued, as often, by a subordinate clause. For eoque see 21.1n. soliti plerumque damna aestatis hibernis euentibus pensare: the characteristic uariatio of gen. noun - adj. 15 emphasised by the chiasmus. In classical Latin the pleonasm of solit? plerumque 15 extremely common in a wide variety of prose authors from Cicero (e.g. /nv. 2.58) and Caesar (C. 1.85.4) through to the second century (Gell. 12.1.17); the verse exs. are Ter. Phorm. 89—90 and Ov. Met. 12.277. For euentus— 'success' see OLD 2a; pensare (OLD 42a) = ‘to make up for — (with —)', as Sen. Oed. 490 ‘meliore damnum pensans marito', Suet. Aug. 25.4 'cuius abrupti damnum nulla captura pensari posset’: the verb is contrasted alliteratively with pellebantur at the end of the sentence. iuxta 'alike' (OLD 2), as Plin. NH 2.136 'iuxta hieme et aestate’.

22.4 nec...intercepit ‘and ...he did not usurp' (OLD 4b). On his first tour of duty in Britain, A. had seen his commanding officer take all the glory for himself (5.3); on his return, the then governor came to share the glorv with A., who responded with reciprocal deference (8.2-3). A.’s

COMMENTARY:

22.4

211

own governorship is framed by attempts at deflecting attention away from himself (19.6 —^ 39.1), a practice complemented by his refusal here to appropriate the praise due to others. seu...habebat ‘whether centurion or prefect, he would find an incorruptible witness to his deed.' For centurions see 19.2n. (centurionem); a praefectus was 'an officer of equestrian rank commanding an auxiliary infantrv (or cavalry) regiment of 500 men, which would form the garrison of a smaller castellum’ (O-R). For zncorruptus cf. Cic. Fin. 1.71 'incorruptis...testibus', Ov. χ P. 3.9.50, [Quint.] Decl. 3.10; amongst themselves the men might naturally exaggerate their own deeds (25.1), but formal recognition by A. consisted of nothing but the truth. apud quosdam...narrabatur 'on the part of some people he was described as being...' (OLD apud 15a, not 12). Ancient theories of praise recognised that the interspersing of 'gentle criticisms' would produce an impression of 'frankness' and so give greater credibility to the praise (Plut. Mor. 856D): here the first potential criticism in the panegyric is distanced by its attribution to other voices and will be turned to A.'s advantage in the next sentence (n.). ut erat comis bonis, ita aduersus malos iniucundus:

this looks like an

example of adversative asyndeton, acknowledging and justifving the criticism in the previous sentence, but placing it in a balanced perspective at the same time: ‘but, though he was affable to the bonz, nevertheless he was

indeed disagreeable to the mal:’. The balance 15 typical ofA. and symbolised by the use of ut...?ta (OLD ut 5b), its antitheses sharpened by chiasmus and uariatio (for dat. ~ aduersus cf. H. 1.35.2, À. 4.1.2). According to Onasander, the ideal general should be ‘affable... not so lenient as to be despised’

(2.2); affability (comitas)

‘is a leading trait of T.'s Ger-

manicus' (G. on A. 1.33.1). For the form aduersus see 9.4n. (aemulatione). ex iracundia nihil supererat secretum...offendere quam odisse: the final alliterative sententza, with 115 characteristic comparative formulation (5.3n. nec minus), relates to the ancient ethical theory that hatred was chronic anger (Cic. Tusc. 4.21 'odium ira inueterata’). A.’s anger was transitory and, like his well-intentioned outspokenness, means that he was the

antithesis of Domitian (39.3 'secreto suo ...reponere odium’, 42.3 ‘odisse quem laeseris...praeceps in iram ...inreuocabilior'). supererat — *was left over' (OLD 4). ut silentium eius non timeres ' with the result that you would not have feared his silence'. The present example and crederes at 44.2 (n. bonum) resemble those numerous cases in historiography where the generalising second-person potential subjunctive (see g.1n. 520) 15 used to engage readers in a description of character (see Gilmartin (1975), esp. 114-15 and n. 33). For the imperfect tense — 'vou would have — ' see NLS §121; for timeres again see H. 2.62.1 (cf. also Vell. 72.2 'Brutum amicum habere

212

malles, inimicum

COMMENTARY:

23

magis timeres Cassium’), and for exs. with ut, as here,

cf. H. 3.83.2, À. 1.32.3, 3.4.1.

23 The year 8o 23 Quarta aestas obtinendis quae percucurrerat insumpta: both the contrast with percucurrerat and the general context suggest that the primary sense of obtimendis 15 ‘securing’ (cf. OLD 8), though ‘retaining military control of' and ‘occupying’ (as 24.3 below) are obviously relevant (OLD 4—5); for the opposite cf. A. 15.8.2 ‘longinquis itineribus percursando quae obtineri nequibant'. For znsumo + dat., as usual in T., cf. G-G 658-9. The aestas 15 that of 80 (RGB 84). si uirtus...pateretur, inuentus in ipsa Britanniae terminus 'if the army's prowess and the glory of the Roman name had been destined to allow it, the boundarv of Britain would have been found within itself.' The syntax 15 difficult and the text disputed. The paradosis reads zn ipsa Britannia terminus, to which Shaw-Smith (1979) objected that termznus is curiously unspecified and unclear: he proposed n ipsa Britanniae terminus, noting that the same phrase recurs at 27.1 ‘penetrandam Caledoniam znueniendumque tandem Britanniae terminum' and 30.3 ‘nunc terminus Britanniae patet'. The proposal is strongly supported by the link between all three passages: in our passage the real terminus has not been found, at 27.1 the desire is to find the real terminus, and at 30.3 the real terminus is finally found. Whichever reading is preferred, a conceit 15 involved. The termznus of Britain would have been found within Britain itself only if the name is restricted to that part of the countrv this side of Caledonia, i.e. south of the Clvde/Forth line (see below); but, since itis clear from 27.1 and 30.3 that the real termznus of Britain lies at the far side of Caledonia, Britain in those

two passages is being used of the entire island. There is a similar slide in meaning at 10.9 ‘et est ea facies citra Caledoniam'. The imperfect subjunctive in a past unreal condition sometimes represents the continuation of the action into the present, as seems to be the case at 2.3 above (n. memoriam), sometimes represents the notions ‘was destined to' or *was likely to’ (NLS$199), which seems more appropriate here. With znuentus we should probably supply esset, though erat is preferred bv some edd. uwurrtus exercitusis a regular expr., Roman? nominis gloria more choice (H. on H. 5.16.2). Clota et Bodotria: the rivers Clyde and Forth (see Rivet and Smith (1979) 309-10 and 269-71 respectively). diuersi...reuectae 'carried back [i.e. inland] a great distance, by the tides of a different/opposite sea.' See also 10.6 above and, for the converse phenomenon, A. 6.33.3 'hibernus auster reuoluit fluctus pulsoque introrsus freto breuia litorum nudantur'.

COMMENTARY:

24.1

213

dirimuntur is ‘a word geographers like' (Horsfall on Virg. Aen. 7.227), elsewhere combined with spatio (-is) at Col. 4.53.2, Frontin. Strat. 2.2.8. The Antonine Wall, of which A.’s defences (next n.) seem to be a precur-

sor, was roughly 40 miles long. quod tum praesidiis firmabatur: praesidizs (-20) firmare (again at H. 2.83.2, 4.55.4) is a regular expr. (Cic. Sall. Liv. Vell. Curt. Suet.); see also 14.3n. (subactis). Whether the imperfect tense is continuative or inchoative is unclear. One of these praesidia has been excavated at Mollins, just south of the line of the Antonine Wall (see Hanson and Maxwell (1980)), another

one

(dated by a coin hoard

to 79/80)

near Dalkeith, 6 miles south of central Edinburgh (2007) and Hanson

at Elginhaugh

(see Hanson

et al.

(2007)).

omnis propior sinus: i.e. the swathe of territory south of the Clyde/Forth line. For fpropiorsee 14.1n. (proxima). uelut in aliam insulam: the Clyde/Forth line is so narrow and now so well defended (above) that the area north of it is as isolated as if it were

a real island such as Anglesey (15.5 ‘alia insula’): the comparison adds to the wildness and remoteness of Caledonia before it too succumbs to A.'s forces and is incorporated at the battle of Mons Graupius (see above, n. s uirtus, also Clarke (2001)

101). summouereis a regular term for driving off

hostile forces (OLD 3), e.g. HA Ant. Pius 5.4 ‘muro caespiticio summotis barbaris ducto' (of the Antonine Wall 60 years later). For the appended abl. abs. see 2.2n. (delegato).

24 The year 81 24.1 Quinto...anno: AD 81. naue primum transgressus: this passage has been called 'notorious', and the transmitted naue prima was obelised by Delz. (i) transgressus lacks an object and leaves the location of the crossing unspecified. (ii) Although prima seems to agree with naue (attempts to make it neut. plur. can be dismissed), the adj. customarily precedes its noun and does so on the rare other occasions when reference is made to 'the leading ship (s) : Liv. 37.23.8 ‘prima praetoria nauis Eudami erat’ (cf. Serv. Aen. 10.166 ‘Aeneae prima est nauis’), Caes. G. 4.23.2 'cum primis nauibus Britanniam attigit’, 25.6 (where the text is disputed). Exceptions to this rule relate to expressions of time and/or are clearly exceptional (as H. 4.43.4 ‘optimus est post malum principem dies primus’). Many scholars have tried to remove these two problems simultaneously by emending naue primainto a place-name or similar, e.g. Anauam (the River Annan) or maritima. None of these proposals is at all plausible, and the unqualified use of transgressus seems not very different from those at 10.3 and 18.1 above. O-R retained naue prima on

214

COMMENTARY:

24.1

the grounds that it ‘emphasises Agricola’s personal initiative’, but in fact it has the opposite effect, implying that A. would not normally have been in the leading ship, and in any case his initiative in this respect has already been noted at 20.2 above (‘aestuaria...ipse praetemptare"). By far the simplest solution is to assume that primu became prima either by misreading or by attraction (cf. 19.2): ‘having first crossed by ship, he tamed...’ It is naturally inferred from Clota et Bodotria (23, above) that A. moved north, crossing one of these estuaries, and from eam...partem Britanniae quae Hiberniam aspicit (below) that he was on the western side of the country: i.e. he crossed the Clyde

(see also Reed

(1971)).

ignotas ad id tempus gentes...domuit: for the panegyrical motif of conquering the unknown see 10.4 ?ncognitas ad id tempus insulas . . . domuitque and n. The people beyond the Clyde in mod. Argvll were called the Epidii by Ptolemy (Geog. 2.3.1, 8—9); domuit, often combined with gentes (as 13.3), fleetingly likens them to wild animals. crebris simul ac prosperis proeliis are clichés of military narrative (see H. on H. 2.24.1 and 2.93.2). eam...partem Britanniae quae Hiberniam aspicit copiis instruxit: if A. had

crossed

the Clyde

(see above),

it seems inevitable that this must

be a reference to the Mull of Kintyre (so too RGB 84), which is in fact the nearest point of Britain to Ireland. This idea has been called 'geographically and archaeologically absurd' (Haverfield (1913) 7 n. 10), and Reed (1971) 146-7 envisages a separate operation to the south in Galloway, from where Ireland may also be observed. In the first century AD aspicere becomes a regular term for places which command a view (OLD 7), whereas copais instruere seems to be used of a place only at Cic. Phil. 10.9 ‘Graecia...copiis non instructa solum sed etiam ornata’. in spem magis quam ob formidinem: for this antithesis cf. H. 2.46.2 ‘insistere spei...formidine properare', Sisenn. fr. 67P — 100C ‘uictoribus...spem,

uictis...formidinem'

(note

also

17.1n.

Petilius). For

ob

and the uariatio with n see 5.1n. (nihil). For in spem see H. on H. 4.42.1. si quidem Hibernia...ualentissimam imperii partem...miscuerit:

when 51 quidemis used causally, as here, the verb is normally in the indic., as

at G. 30.1 (itsonly other occurrence in T.): see OLD siquidem 3—4, Solodow (1978) 130-2. Butfut. perf. indic. would surelv make no sense here, so miscuerii must be perf. subjunc., perhaps analogous to its potential use in independent sentences

(NLS §120): ‘in as much as Hibernia would unite...'.

10.2

plain

The only earlier ex. of a similar subjunc. is said to be Cels. 1.2.6 'si quidem...corpus moueat’. It is unclear whether miscuerit here — 'unite, link’ or ‘produce (by joining) : neither seems paralleled when used of a singular entity such as imperii pars. For ualentissimam cf. the analogous H. 3.53.3 ‘Gallias Hispaniasque, ualidissimam terrarum partem’ (indir. speech). medio...opportuna: for the assumed orientation of Ireland see and

nn.

The

abl.

med?o is a favourite

of T.

(G-G

818a):

115

COMMENTARY:

24.2

combination with sita, instead of the common

215

n medi?o s., is unique

and

with znter almost so (only Vitr. 6.1.7 ‘nationes...inter axes...medio positae’). sita, having the same derivation as the noun sztus, prepares us for the brief ethno-geography which follows (see 10.1n. situm). Gallicum mare is an alternative name for Gallicus sinus (mentioned several times by Livy) and means roughly the Gulf of Lvon (Plin. NH 3.74, Apul. Mund. 6 (p. 141 Thomas) and in Oros.). magnis inuicem usibus 'to great mutual advantage' (abl. of attendant circumstances). For ;nu:cem see OLD 3. 24.2 spatium eius...nostri maris insulas superat: brachvlogical for 'exceeds «the extent of> the islands in our sea', a so-called ‘compendious comparison' (below).

(12.93n. dierum); 50 too Britanniae (next n.) and

a Britannia

in miniature

(21.1n. saluberrimis), so his account of 81 concludes with a

Just as T.'s account of 78 concluded with a formal ‘Kulturgeschichte’

brief ethno-geographical excursus on Hibernia (24.2-3), after which we are returned to the main narrative with Ceterum at 25.1

(n.). Islands were

intrinsically interesting (Gabba (1981) 55-8), and the excursus is parallel to the full-scale digression on Britain and its people which introduced A.'s conquests of the mainland (10-13.1); the difference 15 that Hibernia was destined to remain one of the 'ifs of history', the frequent subject of an old soldier's wishful thinking (24.3 'saepe ex eo audiui..."). The excursus begins with three comparisons (comparetur, superat (cf. OLD gc), and haud multum . . . differunt), asis to be expected in a geographical and ethnographical context (see e.g. 10.2,11.2—4, 12.3, 12.5; Hartog (1988) 225ff.). The area of Ireland is 31,520 sq. miles (= 81,638 sq. km.), whereas Sicily, the largest island in the Mediterranean (nostrz maris), is 9,927 sq. miles (25,711

sq. km.). See also 10.2n.

(maxima).

si Britanniae comparetur, angustius: though there are other possibilities, it is likely that Brztannzae is dat. after comparetur (as often) and that spatzum eius (above) is the subject of the conditional clause as well as of the main

sentence, thus producing another ‘compendious comparison’; angustius is the apodosis of the conditional clause: 'its area, smaller if compared to «that of> Britain,...'. The area of Britain

(10.2n. maxima) was thought

by Caesar to be twice that of Ireland (cf. Caes. G. 5.13.2 'Hibernia dimidio minor ut aestimatur quam Britannia’) and by Mela to be the same size (3.53 ‘paene par spatio’). Note also Plut. Caes. 23.3 with Pelling ad loc. solum caelumque: this regular doublet (again at G. 4, 29.2, H. 2.32.1, 5.7.2), found in a wide variety of early imperial authors (Vitr., Mela, Curt., Cels., Colum., both Plinys, Quint., Flor.), refers to two standard elements

of geographical descriptions (cf. 12.3 caelum, s solum). ingenia cultusque hominum haud multum a Britannia differunt: ?ngenium and cultus are likewise combined by Mela (2.9 ‘ingenia cultusque

216

COMMENTARY:

24.3

gentium differunt’, of Scythian tribes) but contrasted by Gell. (11.2 praef.). [in melius[ aditus portusque...cogniti: the obelised words make no sense. Rhenanus suggested either deleting zn or emending the whole to eius; but T. could scarcely have said that Ireland's approaches and harbours were better known than those of Britain, while ezus seems redundant after

its appearance at the start of the excursus. There isa popular belief that the words started life as a marginal correction of haud (above) by a patriotic Irish scribe which then found their way into the text: ‘... differ greatly from those of Britain, for the better' (a similar interpolation has been posited in the account of Ireland by Mela 3.53: see Romer (19g8) 117 n. 32). But there is no guarantee that Irish monks had anything to do with the transmission of the work. Wellesley (1960: 267) suggested that the transmitted differt (above) provoked the gloss £t melius (= runt melius, ‘plural better’), which was later incorporated in the text as 1271 melius. The two words were deleted by Wex. per commercia is found only in T. (again at 28.3, 39.1, G. 17.1, 24.2); the uariatio with the personal negotzatores is typical of him. 24.3 expulsum seditione domestica: barbarian leaders are frequently described as involved in internal rivalries and quarrels (e.g. À. 1.57.1—60.1, 2.1.1-4.%, 2.44.2—40.5, 2.62.1—3, 2.04.3—65.3, 2.88.1-2, 3.42.3 with W-M, 6.31—7); 1{15 part of the general picture of barbarian disunity (see 12.2n. nec). There was a well-established tradition of royal refugees seeking sanctuary with Rome, as 15 illustrated by the listat Aug. RG 32.1 (including two kings from Britain). sed:tio domestica, an expr. which T. is the last classical author to use (again at G. 29.1, H. 4.12.2), can however be applied equally to Roman affairs (e.g. Cic. Leg. agr. 2.90, Liv. 2.42.3). specie amicitiae is an expr. much used by T. elsewhere (7x), always with the implication that the appearance is false. Where imperialism is concerned, T. is nothing if not realistic. in occasionem retinebat *was holding him for the right opportunity’ (namely, an invasion); Juvenal in a well-known passage alludes to the presence of Roman arms in Ireland (2.159-61 'arma quidem ultra | litora Iuuernae promouimus et modo captas | Orcadas ac minima contentos nocte Britannos’), but in fact the Romans

never invaded.

:2n occa-

sionem, not a common phrase, sometimes takes a dependent genitive (as H. 1.80.2 'in occasionem praedarum’), sometimes not (H. 2.7.2, Flor. 2.19.84). ex eo: i.e. Agricola; see 4.3n. (memoria). idque etiam aduersus Britanniam profuturum si...'and it would also be a good thing in relation to Britain if...': 14 looks forward to the s?clause (cf. A. 3.6.1 and n.). prodesse aduersus + acc. is quite common in medical contexts — ‘be useful against' (TLL 10.2.2244.28-33), but here

COMMENTARY:

25.1

217

the preposition must have its much less common meaning of 'in relation to' (see W-M on A. 3.16.3). For the form aduersus see g.4n. (aemulatione). ubique: sc. essent. uelut e conspectu libertas tolleretur: ¢ conspectu (again at Sen. Ira 3.36.9 ‘sublatum e conspectu lumen est’), an alternative expr. to ex oculis tolli (Cic. Dom. 114, Hor. Odes 3.24.32, AP 182—-3), looks back to aspicit at 1 above: physical sight is transformed into the quasi-metaphorical. The motif will soon be taken up by Calgacus in his speech (30.2): 'nec ulla seruientium litora aspicientes, oculos quoque a contactu dominationis inuiolatos habebamus'. As usual, there is no suggestion that T. disagrees with his father-in-law's uncompromising attitude to the freedom of foreigners.

25—26 The year 82 25.1 Ceterum aestate qua sextum officii annum incohabat: AD 82; officium here — 'official position', 'commission' (OLD 6a). ceterum, as often (e.g. A. 4.12.1 and n.; G-G

167a), returns us to the main narrative after the excur-

sus on Hibernia. annum incohare (again at H. 1.11.3, À. 4.70.1, 13.10.1) seems to be a Tacitean mannerism found nowhere else. amplexus ciuitates trans Bodotriam sitas: some scholars have suggested understanding anzmo with amplexus (‘having designs upon’), but this is unlikelv if A. had already reached the Tav three years earlier (22.1). amplexus presumably looks back to that operation and means something like ‘having incorporated/assimilated’; this precise meaning, perhaps slightly different from 17.1 above, is not found in OLD but cf. Liv. 44.9.4 ‘Macedonum gens...Graeciam omnem et partem Thraeciae atque Illyrici amplexa’. quia...timebantur: this clause explains the main sentence portus classe explorauit; there is a contrast between previous operations south of the Tay (last n.) and those in the future beyond the Tay (next n.).

motus uniuersarum ultra gentium: adverbs of time and place are quite frequently used attributively as if they were adjs.: ‘all the farther peoples’. Cases where there is a possibility of the adverb's being taken predicatively (as 10.2 ‘nullis contra terris', ‘no lands being opposite’) seem markedly less striking than the present example; see also 3.3n. (hic interim). motus — ‘disturbances’, ‘uprisings’ (OLD ga-b), as H. 4.12.1 ‘gentium motu’. infesta hostili exercitu itinera ‘marches made dangerous by an enemy army.’ ?nfestus is the regular word for dangerous journeys (Lucr. 5.1124, Caes. C. 3.79.4, Hirt. G. 47.3, Cic. Phil. 12.25, Pollio ap. Cic. Fam. 10.39.5, Liv. 38.41.10, 39.1.6, Apul. Met. 9.10.5), and, though an abl. commonly expresses the source of the danger (TLL 7.1.1410.10-21), it is much less common when used personally (but cf. Frontin. Strat. 1.4 'de

218

COMMENTARY:

25.1

transducendo exercitu per loca hosti infesta', Suet. Aug. 8.1 'per infestas hostibus uias’). Some scholars, arguing that T. cannot have dignified the Britons by describing them as an ‘army’, prefer to read infesta hostibus exercitus itinera (*marches of his army made dangerous by the enemv'; H reads hostilis exercitus), but T. occasionally uses exercitus of other non-Romans (e.g. À. 12.14.1), asdo other authors (e.g. Vell. 112.2 'circumdatus hostili exercitu'). On infestus see also 41.1n. (?nfensus). portus classe explorauit: A. was following the exemplary prudence of Julius Caesar: ‘neque in Britanniam transuexit nisi ante per se portus et nauigationem ad insulam explorasset' (Suet. D/ 58.1). ab Agricola primum adsumpta in partem uirium: 7n partem means 'to be part (of)' (13.3n. adsumpto): the point is that this was the first joint land and naval expedition in British waters (for a similar point see Vell. 106.2 and W.), not that there was no previous fleet in Britain. T. elsewhere refers to the existence of a Britannica classis twelve vears earlier in AD 70 (H. 4.79.3), although a lack of ships caused problems for A. when he wished to invade Anglesey in 77 (18.4). Much about the fleet remains obscure: a praefectus is mentioned at 38.3 and a harbour at 38.4, but the former is not identified and the latter is a notorious passage (n. portum). For a survev of the evidence for the British fleet see Saddington (1990) 227—930. pars uirium is above all a Livian expr., though found elsewhere too (e.g. Val. Max. 1.1.15, Curt. 4.14.10, Frontin. Strat. 2.3.3, Just. 38.4.9 'Gallorum...nomen...in partem uirium suarum ipse numeret’). egregia specie is difficult to translate but perhaps *with exceptional pomp' (abl. of manner or of attendant circumstances), as A. 15.9.1 ‘hostiles turmae ... magna specie uolitabant’. Ancient shipping tended to attract appreciatory epithets, e.g. Sall. H. 4.69.13 'classe pulcherruma’ (also Val. Max. 8.1 abs. 4, Pan. Lat. 10.12.3), Vell. 79.2 'speciosissima classe'. So too in Greek, e.g. Hom. Od. 13.140, 175 περικαλλέα vija (‘very fine ship’), Sappho fr. 16.1-3. cum simul terra, simul mari bellum impellitur, ut saepe...: the paradosis of this passage reads zmpellitur ac saepe, which Rhenanus emended to zmpelleretur in order to make the verb co-ordinate with the two following subjunctives adtollerent and compararentur. This emendation has been almost universally accepted and would mean 'since the war was being pushed forward on both land and sea simultaneously, and often ...' (cf. Luc. 5.330-1'impulsi tantum quae praemia belli | auferat’); yet it is very hard to see why the imperf. subjunctive should have become corrupted into the present indicative. The sequence of the imperf. indicative sequebatur followed by cum -- pres. indicative zmpellitur reads like an inverted cum-construction (as at 3 below: see n. cum); if thatis so, acrequires changing to uf to generate the two following subjunctives (for the confusion of ac and ut see e.g. G. 16.2). The meaning will be: 'and the fleet was

COMMENTARY:

25.1

219

following along with exceptional pomp when war swept in on both land and sea, with the result that often...’; the image 15 that of a gale being driven along by its own force (as Sen. NQ 1.1.6 'nubes...maiore impetu impulsae', 3.10.2, 5.18.2, Phaedr. 955; Aetna 311). It may be asked how, since the Britons had no fleet, there can have been a war at sea. We must

assume a form of syllepsis and that bellum is to be taken literally with terra but metaphorically with mar: (for waging war with the elements see e.g. Lucr. 4.968, Curt. 9.4.14; OLD bellum 6b). It is clear from what follows that the two spheres of action, here signalled by semul...szmul, are treated in parallel (‘modo...modo...hinc...hinc’) and that the struggles at sea are presented in metaphorical terms as a military engagement (‘uzctus Oceanus’). For similar cases of syllepsis involving bellum see e.g. Liv. 7.22.4, Flor. 2.13.75 'acrius fuit cum ipso mari quam inter se nauibus bellum' (cf. 1.45.5 ‘maior cum Oceano quam cum ipsis nauibus rixa', Cic. Off. 3.99 ‘non modo cum hostibus uerum etiam cum fluctibus...dimicare melius fuit’). simul...simul (again at 36.1, 41.4) 15 esp. common in Livy (Oakley on 9.46.14) and Virgil (Tarrant on Aen. 12.268); itis absent from Cicero and first at Caes. G. 4.13.5. In T. 1 recurs at H. 1.50.1, A. 1.49.3, 14.40.1. pedes equesque et nauticus miles mixti copiis et laetitia: m?scer? can mean both ‘to join together' (as H. 3.18.2 ‘mixtus pedes equesque’) and ‘to join in — ' (as H. 1.9.3 ‘nec uitiis nec uiribus miscebantur', where there is the same typically Tacitean combination of abstract and concrete): here both meanings are said to be operative, the whole phrase being equivalent to ‘qui inter 56 mixti copias et laetitiam communicabant' (G-G 844b), or, as might be said in English, ‘having joined together in the provisions and happiness'. pedes equesque (see H. on H. 3.18.2) and miles (16.3n. miles) are collective singulars; nauticus milesis an unparalleled variantfor nautici, ‘sailors’. For the co-ordinating pattern A Bque et Csee 12.5n. (praeter) .

sua quisque facta, suos casus adtollerent: the verb may = 'praised' (OLD 10b) with facta but can only — 'exaggerated' (OLD 11c) with casus, which presumably = ‘perils’ (OLD 8). quisque normally takes a singular verb but the use of the plural

(a so-called

constructio ad sensum)

is common

in T.

(G-L $8211 R. 1 (a)). For the scene scholars quote Luc. 4.196—-202 'castris miles permixtus utrisque | errabat; duro concordes caespite mensas | instituunt...|extrahitinsomnes bellorum fabula noctes| ...| ... quae gesserunt fortia iactant’; cf. also Sall. J. 53.8 'milites...laeti...acta edocent atque T.'s mind), Tib. 1.10.29-32 'alius 511 fortis in armis, | sternat et aduersos Marte fauente duces, | ut mihi potanti possit sua dicere facta | miles'; Horsfall (10999)

116 (and below, 27.2n. eniquissema).

modo siluarum ac montium profunda, modo tempestatum ac fluctuum aduersa: szluarum profunda (again at Frontin. Strat. 1.3.10, Amm. 27.12.11)

220

COMMENTARY:

25.2

is a choice variant on sz/uae profundae (Lucr. 5.41, Virg. Aen. 7.515, Curt. 7.7.4) and means ‘deep woods’; presumably montzum profunda also refers to penetration deep into the mountains, but the notion of sheer drops (as Sil. 4.70 'sacros montes rupesque profundas') cannot be entirely ruled out. aduersa tempestas (again at H. 3.43.2) occurs in a wide range of prose authors from Cicero onwards, and aduersi fluctus 15 a regular expr. in verse as well as prose (Lucr. 6.725, Bell. Alex. 64.3, Liv. 30.24.8, Ov. Her. 7.40, Val. Fl. 8.327), but T. has varied these too by means of the neut. plur. + gen.,

for which

see

6.4n.

(/udos).

modo...modo,

refers respectively to the soldiers and sailors.

like

Various references to the silua Caledonia, of which

hinc...hinc below, the first is Plin. NH

4.102, have encouraged the belief that Scotland in Roman times was covered in forest, but modern scholarship suggests that such references to 'forests', a conventionallv evocative locale (e.g. Hardie on Virg. Aen. 9.381—-3; note also Schama (1995) 37-242, Schadee (2008) 178-9), are no more than a topos of ancient writing (see Breeze (1997), Smout (2000) 37—42), and indeed s:luaeare as frequently coupled with montes as are tempestates with fluctus. terra et hostis 'the land and enemy’: to get the full sense, the missing element has to be supplied from wuzctus below (next n.). uictus Oceanus 'the conquest of Ocean’: this and usa classis at 2 below are nominative examples of the 'ab urbe condita construction' (10.4n. tunc). Given the manifold dangers with which Ocean was associated, and which are amply described in Seneca's first Suasoria (see also next n.), it

is not surprising that A.'s troops should talk of ‘conquering’ Ocean (cf. Hor. Odes 3.24.41 ‘uincunt aequora nauitae’, Catalepton .48 ‘saepe mare audendo uincere’). When Julius Caesar celebrated his four triumphs in 40 BC, Ocean was depicted in gold as a captive (Flor. 2.13.88), and, when Claudius celebrated his invasion of Britain in AD 43, he affixed to his residence the corona naualis, ‘traiecti et quasi domiti Oceani insigne' (Suet. Claud. 17.3). iactantia: cf. A. 2.24.4 'ut quis ex longinquo reuenerat, miracula narrabant: uim turbinum et inauditas uolucres, monstra maris, ambiguas hominum et beluarum formas, uisa siue ex metu credita' (of the survivors of a storm off the northern coast of Germany in AD 16). ?actanta (again

at 39.1) first appears in T. and his contemporaries as a synonym for the older formation zactatio, which he uses at 5.1 and 42.3. T. uses zactatioonce again in G. (6.1) then drops it entirely in favour of zactantia (3 x in A.). 25.2 ut ex captiuis audiebatur: for the importance of prisoners of war as sources of information see Austin and Rankov (1995) 67—73, esp. 70, Bertrand (1997) 114. tamquam aperto ...clauderetur ‘on the grounds that, with the secret of their sea now opened up, the final refuge was closed to them if they

COMMENTARY: should be conquered' the sentence

25.3

29]

(OLD tamquam 7a): sui refers not to the subject of

(classzs) but to Britannos (see 7.1n. in praediis). The open but

unvoyaged 568 (cf. 10.2 ‘uasto atque aperto mari'; OLD apertus 3b) had previously been

the Britons' last resort; but, now

that the Romans

have

really opened it up (OLD aperio ga; cf. 22.1n. Tertius), it has become closed to them. 25.3 Ad manus et arma: the two nouns, which are regularly coupled (33.5; H. on H. 3.20.2), here perhaps constitute a hendiadys for 'armed combat'. The following episode (on which see further above, pp. 26—q) serves as a prelude to the final encounter at Mons Graupius (29-38). In ΔῊ inverted cum-lause A. receives intelligence which leads him to believe that the enemy will attack on several fronts (25.4 ‘cum interim cognoscit hostis pluribus agminibus inrupturos’); but, in a first peripeteia or reversal, the enemy learns of his response and changes tactics (26.1 ‘quod ubi cognitum host?, mutato repente consilio unzuers?.. . inrupere). But A. in his turn learns of the change

(edoctus) and, in a second inverted cum-clause, issues

the orders which result in the second peripeteia, namely, the defeat of the enemy (26.1-2). magno paratu, maiore fama (uti mos est de ignotis): since references to rumour and the unknown would be out of place if the action had already begun, it follows that paratu — ‘preparation’ (OLD 1a) rather than 'equipment' (OLD 2a). That the unknown gives rise to exaggeration is a commonplace (see e.g. G. on A. 2.82.1, W-M on 3.44.1); the word-play magnus -- mazor 15 frequent in a wide range of authors and genres, both verse and prose. oppugnare ultro castellum adorti: adorior (OLD 9b ‘to set to work (to do something), attempt, begin') + oppugnare is extremely frequent in Livy, having castellum as object twice (35.51.8, 43.21.4) and castra four times; cf. also Nep. Thras. 2.5, Curt. §.1.6. ultro is commonly used of ‘unprovoked military or other offensive action' (OLD 6a): cf. prouocantes below. metum ut prouocantes addiderant: the precise meaning of metum addere (again at H. 1.62.2, Ov. Ex P. 3.1.25, Curt. 8.3.5, Val. Fl. 3.428-0, Stat. Ach. 1.146) will vary according to context; here the likely meaning is that the Caledonians, ‘as being the challengers, had increased the dread’ of the Romans (OLD addo 11). For this common causal use of ut 4- adj./part., again at 26.1 below, see OLD 21a. regrediendumque... potius quam pellerentur: sc. esse with regrediendum and cedendum (whose combination may distantly echo Sall. H. fr. incert. 4 ‘regressi...cedebant’). In earlier periods it was regular for potius quam to be followed as well as preceded by an infinitive in indirect speech (e.g. Cic. Fam. 2.16.3 ‘nonne tibi adfirmaui quiduis me potius perpessurum quam ex Italia ad bellum ciuile exiturum?', 'Did I not tell vou that Iwould endure

222

COMMENTARY:

25.4-26.1

anything rather than leave Italy for civil war?'). From the time of Livy, however, the subjunctive came to be used in the quam-clause since in direct speech it was alreadv ‘required after quamin comparative clauses whenever the compared action is one that is to be, or was, purposely rejected' (NLS $253, Noteii: cf. the notes of Miller and G. respectivelv on A. 1.35.4; also G-L $644 R. 3 (a)). citrais used from the viewpoint of the author in Rome rather than that of the cowardlv warners in the text. ignaui specie prudentium admonebant: warning figures are common in ancient historiography and are esp. associated with Herodotus: most familiar is the 'tragic warner', whose advice is doomed to be dismissed or disbelieved until it is too late (see e.g. Oaklev on Liv. 9.3.4-13, pp. 68-9); but here T. is less interested in dramatic suspense than in deploying the advice of specious cowards who, reacting to the exaggerated rumour (mazore fama above) but in fact speaking from ignorance (cf. de ignotis above), enhance by contrast the genuine qualities of A. See also 18.2n. (plerisque). cum interim cognoscit...: sc. Agricola; in inverted cum-clauses the historic present is regular (NLS $237), as is ?nterim. We are presumably intended to assume that A.’s knowledge derived either from captives (mentioned at 2 above) or scouts (26.1 below). It was a general's responsibility to find out the enemy’s plans (e.g. A. 2.20.1; cf. Frontin. Strat. 1.2 'de explorandis consiliis hostium"). pluribus agminibus inrupturos: once again the expr. is Livian (1.14.11, 2.30.14); for the verb see further below

(26.1n. znter). It may be inferred

from et ?pse in tres partes below that A. was matching the enemy and hence that pluribus = ‘three’. 25.4 ne superante numero et peritia locorum circumiretur: the first abl. is instrumental, the second causal. The ideal general was naturally supposed to take counter-measures if he believed himself outnumbered by his opponents (Frontin. Strat. 2.3.17 ‘ne circumiretur ab hoste et peditum numero et maxime equitatu superante consecutus est [sc. Sulla]’). peritia locorumis originally a Sallustian expr. (/. 46.8, of Jugurtha) picked up by T. (again at A. 6.36.1) and then echoed by Amm. (19.8.5); we have already been informed of A.'s own skill in this respect (22.2).

incessit 'advanced' (OLD 3); the position of the subject within the abl. abs. is idiomatic, e.g. Liv. 45.10.5 'dimissis et ipse Attali nauibus...nauigare Aegyptum pergit’. 26.1 quod ubi cognitum hosti: whether host? 15 simple indir. object ('to the enemy’) or expresses agency (‘by the enemy’) is unclear; for the significance of the clause see above, 25.3n. (ad manus). nonam legionem ut maxime inualidam: Legio IX Hispana had arrived from Pannonia in the Claudian invasion of Britain in 43 (13.3n. transuectss) and, having suffered badly in the revolt of Boudica in 60, required supplementing thereafter (A. 14.32.3, 38.1). Its present weakness is

COMMENTARY:

26.2

223

usually explained with reference to an inscription (/LS 1025 — CIL 14.3612) which reveals that a detachment from the legion (wexillarz, cf. OLD 2) was sent to Germanv, presumably in the present vear, to help fight Domitian's war against the Chatti in 83 (39.1n. falsum). Syme, however, suggested alternatively that the legion's base at York, being the nearest out of the four legionary fortresses to the area being pacified, required garrisoning by a disproportionate number of troops (RP6.21 n. 17). See RGB 227-8, 281-2. The Ninth is the legion on whose supposed loss R. Sutcliff based her famous novel The eagle of the Ninth (1954). tnualidus 15 a regular military term for ‘below strength' (OLD 3), though not found elsewhere with legio. For ut see 25.3n. (metum). inter...inrupere: the verb is a favourite of Caesar and Livy (see Horsfall on Virg. Aen. 11.879); itisregularly used absolutely (OLD 2a), though here there seems a slight awkwardness in as much as znter. .. trepidationem refers to the unexpressed object of the verb. One might have expected T. to say e.g. tnter somnum et trebidationem suspensos. iamque...pugnabatur cum Agricola...iubet: the inverted cum-clause (25.4n. cum), here introduced bv the dramatic 'am d'ouverture' (33.1n. iamque), signals the second peripeteia of the episode: because of his superior knowledge

(edoctus) A. 15 able to surprise the Britons in their turn and

to force them into fighting on two fronts (cf. 5 'ancipiti malo"). iter hostium ab exploratoribus edoctus et uestigiis insecutus: in his use of scouts (for whose importance in Roman warfare see Austin and Rankov (1995) 42-54, esp. 43-4) A. is behaving exactly as a general should. edoctus is found with a retained acc. since Sall. C. 45.1, J 112.2 (TLL 5.2.106.39ff.). Since insequor regularly takes the acc., edd. assume that the word 15 here being used absolutely and that uestigzzs 15 abl. ('following in their tracks’).

adici

presumably

means

'to

be

added'

(as

Liv.

37.41.11

'clam-

ore...adiecto’) rather than 'to be intensified' (as Petron. 68.5 'adiectum ...clamorem’). Once A.'s forces are combined (ab unzuersis), they will

prove more than a match for the combined forces of the Britons (^ 1 untuersi).

propinqua luce 'in the approaching davlight’ (OLD propinquus 3a); the expr. is found in both verse (Luc. 5.678, Val. Fl. 3.213) and prose (Liv. 24.47.1, Plin. NH 2.195). fulsere signa: ‘a commonplace in the description of legions' (Wellesley on H. 3.18.1); noting that the eagles on the standards were normally of silver, N-H on Hor. Odes 1.7.19 remark that 'the Roman army made a glittering show’: see also e.g. H. 3.82.1, Liv. 28.14.10, Luc. 1.244; Ash on H. 2.49.1. 26.2 ancipiti malo is found in a sprinkling of authors from Sallust onwards (C. 29.1,J. 67.2) butisa favourite of Curtius.

224

COMMENTARY:

26.2

rediit animus: the contrast 15 with znter somnum

ac trepidationemat 1 above,

so animus combines the notions of ‘consciousness’ or 'courage' (~ trepidatio). See also 3.1n. (redit).

(~ somnus) and ‘spirit’

securi pro salute de gloria certabant 'careless [OLD 1, cf. 3] for their safety they contended about glory.' securus prois rare (edd. quote only H. 4.58.1, Sen. Const. sap. 2.1), whereas certare de is regular (5.2n. tum). The contrast between salus and gloria is exceptionally common, e.g. 31.9, Cic. Leg. Man. 19, Cael. 47, Sall. J. 94.5 'pro gloria...his, illis pro salute certantibus', 114.2 ‘pro salute, non pro gloria certare', Liv. 7.30.7, 22.60.13, Curt. 4.14.9. ultro quin etiam erupere: with the soldiers' break-out, the reversal is almost complete (~ 25.3 ultro, 26.1inrupere). ultro erumpere is originally a Livian expr. (23.18.6): here the adv. both underlines the Ninth's offensive action (25.3n. oppugnare) and implies that the situation is the reverse of what it has been (OLD 4). quin etiam (‘and furthermore': OLD quin 32) normally comes first in its word-group but is found in anastrophe in verse and post-Augustan prose (Ash on H. 2.17.2, with some statistics). atrox in ipsis portarum angustiis proelium is another expression comprised entirely of Livian elements: apart from Vell. 21.3 and 55.3, atrox proelium is exclusively and frequently Livian (19x) before T., who uses it elsewhere too (H. 2.66.2, 3.22.3); for in angustis portarum cf. Liv. 10.5.10, 30.5.10 (later in Just. 6.7.3, cf. Amm. 23.0.43). pulsi: sc. sunt. utroque exercitu: i.e. the uelocissimi equitum peditumqueand Legio IX. For the appended abl. abs. see 2.1n. (delegato). his ut tulisse opem, illis ne eguisse auxilio uiderentur: h5...:2//;s are in apposition to utroque exercitu and refer respectively to the uelocissimi and Legio IX. The wording alludes to the speech of Decius Mus to his troops at Liv. 7.35.4 'digni estis qui pauci pluribus opem tuleritis, ipsi nullius auxilio egueritis'. The uarzatio of nouns is of course also very characteristic ofT. (H. 5.3.1; Sorbom (1935) 16-29). quod nisi paludes et siluae fugientes texissent: nis: 15 a useful device for concluding battle-descriptions' (Ash on H. 2.14.3) and the whole is evidently another allusion to Livy: 3.22.9 'deletusque exercitus foret, ni fugientes siluae texissent'. paludes and siluae are combined from Catullus 115.5 onwards (e.g. Sall. H. 1.55.23, Liv. 5.53.9, 34.48.1) butare esp. common in Caesar (gx) to designate the landscape of northern Europe; T. follows suit (e.g. 31.1, G. 5.1, A. 1.67.2, 1.68.3, 2.5.3). Cf. also Mela 3.29 ‘magna ex parte siluis ac paludibus inuia' (of Germanv). For quod n?si see 16.2n. debellatum illa uictoria foret: Ptolemy (Geog. 2.3.7) says that in the territory of the Damnonii there was a fort named Victoria, which O-R tentatively identify with the fort at either Dalginross or Strageath (for which

COMMENTARY:

27.1

225

see Woolliscroft and Hoffmann (2006) 49-52 and 111-14 respectively, with excellent illustrations in pls. 6 and 14). It would certainly be neat if debellatum . . . uictoria (a unique variant on the Livian proelio debellare) were a reference to the name mentioned bv Ptolemy; on the other hand, if A. had already reached the Tay in 79 (22.1), itis more likely that he was operating much further north than the Gask Ridge line (RGB 87). It must be remembered that T. himself had neither any clear idea of Scottish topography nor any intention of providing his readers with a mental map (above,

pP- 13-15).

27.1 Cuius conscientia ac fama ferox 'defiant in their consciousness of this [victory] and in its fame’; the two nouns are often coupled or contrasted: to the exs. given by W. on Vell. 115.5 add e.g. Cic. Clu. 187, Sall. J 35.4, Sen. Ben. 6.42.2, Plin. Ep. 1.12.3. ferox shows that exercitus is singular; the verb (fremebant) is plural because the noun is collective. The sound-play introduced here by alliteration is continued in the soldiers' roar below: ‘wirtuti...inuium...inueniendumque fandem Brifanniae terminum continuo ...cursu'. m nihil uirtuti suae inuium et penetrandam Caledoniam: the troops are already in Caledonia (cf. 25.1 'trans Bodotriam', 3 ‘Caledoniam incolentes’): their point is that they want to penetrate even further, as is implied bv the following ?znueniendumque... Britanniae termimum (-que being quasi-epexegetic). penetrandam continues the notion of journeying latent in nzhzl...?nuzum, the literal/metaphorical status of which (cf. Ov. Met. 14.110 'inuia uirtuti nulla est uia’) is unclear; cf. Flor. 1.23.10 'inuios...montes...penetrauimus' and, for a comparable combination of ideas, Cic. Arch. 23 'si res eae quas gessimus orbis terrae regionibus definiuntur, cupere debemus, quo manuum nostrarum tela peruenerint, eodem gloriam famamque penetrare' and (see Bosworth (2004) 556) Curt. 9.2.9 ‘insatiabilis cupido famae nihil inuium, nihil remotum uideri sinebat'. For soldiers begging to attack or advance cf. e.g. A. 3.45.1 'certantibus inter se signiferis, fremente etiam gregario milite ne suetam requiem, ne spatia noctium opperiretur', Caes. C. 1.71.2, 2.34.4. illi . . erant: this is a further reversal: previously cowards had pretended to be sensible (25.3), now the cautious and wise were enthusiastic and boasting. For caut: ac sapientes cf. Cic. Att. 14.14.2, Calp. Decl. 18; for the antithesis cauti ~ prompti cf. Liv. 30.26.9. iniquissima . . imputantur: 7nzqua condicio 15 a frequent expr. in Cicero (though also occurring elsewhere), but the general sententiousness resembles Sall. J. 53.8 'quippe res humanae ita sese habent: in uictoria uel ignauis gloriari licet, aduorsae res etiam bonos detrectant’ (at 25.1 sua quisque facta Τ. had alluded to Sallust's previous sentence). For the sentiments themselves edd. compare H. 4.52.1, À. 3.53.3, 14.38.3; see also Hutchinson on Aesch. Sept. 4—9.

226

COMMENTARY:

27.2-28.1

27.2 non uirtute se uictos sed occasione et arte ducis rati: this is the Britons' response to the Romans' boast of uzrtus (1 above): compare the words of Vercingetorix at Caes. G. 7.29.2 ‘non uirtute neque in acie uicisse Romanos sed artificio quodam et scientia oppugnationis’; for the antithe515 uirtus — occas?o cf. Liv. 6.41.1, 7.12.4, Veg. Mil. 3.26. With the exception of the participle ratus, reoris elevated and poetic (Cic. De or. 3.153) and in its non-defective forms preferred by T. to puto (Adams (19732) 136). nihil ex adrogantia remittere quominus... ‘lost none of their arrogance but...’

(see OLD remittere 10b for its idiomatic construction with nzhil ex).

The rules and conventions for quom?inus and (esp.) quzn, which tends to be used only after a negative, are complicated and subtle (see A-G $$557—9, NLS §§184—7): sometimes quin - subjunc. can be translated as ‘but’, as in Sall. C. 53.6 ‘quos...silentio praeterire non fuit consilium quin utriusque naturam et mores...aperirem', 'and it was not my intention to pass over them in silence but I revealed the nature and behaviour of each’; although T. does not use this precise construction in Agr., he likes to use quominus as equivalent to quin, as here (see also 20.2n. nzhl). The clauses thereby introduced constitute a tricolon crescendo, the last two elements of which refer ominously to non-barbarian practices (see next n.).

coniuges...sancirent: two commonplaces of Latin historiography are that barbarians fight in the presence of their women (32.2n. nullae) and are irremediably disunited and quarrelsome (12.2n. nec, 15.5, 24.3n. expulsum); here the Britons are adopting the opposite practices in preparation for the final encounter at Mons Graupius. atque ita inritatis utrimque animis discessum: the last sentence of the year's campaign narrative combines closure with suspense (cf. A. 13.56.1 ‘atque ita infensis utrimque animis discessum’). znritatis animis is, fittingly, yet another expr. borrowed from Livy (15x); otherwise only Ter. Andr. 597 before T. 28.1 Eadem aestate: this linking phrase (again at A. 4.27.1 and 13.57.1), which otherwise is exclusively Livian (14x), assigns the following episode to AD 82; the fact that it 15 placed three vears earlier, in 79, by Dio (66.20.2: see O-R, Appendix 2; RGB 83) perhaps indicates the way in which such anecdotes can become 'free-floating' (for another ex. see PH 351). Since Martial in AD go/g1 refers to the Usipi as a standard example of treacherv (6.61.3 ‘leue flauorum ...genus Usiporum), their story had evidently become well known by the time that T. was writing his father-inlaw's biography. The episode itselfis quite extraordinary and defies satisfactory explanation. Though T. ends all but one of his other annual narratives with a similar 'appended' section or 'coda' (above, p. 180), the subject matter here 15 bizarre in the extreme: a cohort of Usipi, conscripted in Germany, murder their commanding officer and other soldiers, hijack three vessels, kill the

COMMENTARY:

28.1

227

helmsmen of two of them, and attempt to sail home. Despite raiding the British coast for food and water, they run out of supplies, resort to eating one another and, having unintentionally circumnavigated the island, they either get mistaken for pirates or are sold into slavery. Though introduced as a characteristic specimen of ancient historical writing (1 'magnum ac memorabile

facinus’), this is the stuff of the ancient novel

(Ash

(2010)

270 and n. 12). The story has no relevance to the biography of A., except in so far as he will send a fleet to circumnavigate Britain the following year (38.3—4); but it is Dio, not T., who makes the Usipi's 'success' the motivation for A. It is true that Calgacus in his speech before the battle of Mons Graupius uses the Usipi as an example to question the loyalty of Roman auxiliaries (32.3); but, although the example is not without its point (see n. Britann?), 'Calgacus' succinct retroactive reference might even make us suspicious that Tacitus has strategically incorporated the whole story of the Usipi shortly before this speech just to prepare the ground for the exemplum’

(Ash (2010) 278, who has a full, but different, discussion of the

whole episode). cohors Vsiporum per Germanias conscripta et in Britanniam transmissa: auxiliary cohorts, under the command of an equestrian praefectus (OLD 3a), were nominally 50oo-strong (cohortes quingenariae) but in practice, and if at full strength, seem to have consisted of no more than about 480 men

in six centuries of 8o, each led by a centurion; attached to some cohorts (called cohortes equitatae) were 120 cavalry in four squadrons (turmae) of

30, each led by a decurion. By the Flavian period cohortes miliariae had also been developed: though nominally double strength, they consisted of roughly 800 men. The Usipi, as they are usually called bv imperial writers including T. (who uses the alternative and seemingly earlier ‘Usipetes’ only at A. 1.51.2), are almost always closely associated with the Tencteri (e.g. Caes. G. 4.1.1) and had an extensive track record of attacking Roman interests; both peoples were known for their horsemanship, and indeed it has been suggested that the name Usipi is derived from the roots seen in Gaulish uesw- ‘good’ (cf. Bellovesus at Liv. 5.34.3) and epo- ‘horse’ (cf. Epona, eporediae). See also Rives on G. 32.1. At this time the Usipi were living in the lower Lahn valley, opposite the dividing-line between the Roman military districts of Germania Superior and Germania Inferior. this may explain the plural per Germanias (so O-R), although T.’s geographv is notoriously vague and he uses the plural on other occasions too (see 15.93n. sic). His straightforward phrasing here is punctuated by proper names, moving the group from their conscription to their deployment. magnum ac memorabile facinus ausa est: there 15 similar 'headlining' of a strange episode at Sall. . 79.1 ‘non indignum uidetur egregium atque mirabile facinus...memorare', but T. here leaves it unspecified precisely

228

COMMENTARY:

how we are to evaluate

the facinus

(Ash

28.1

(2010)

280-1):

historiography

deals naturally with items which are ‘great’ and ‘memorable’, like the Peloponnesian

War

(Thuc.

1.1.1

péyav τε... καὶ ἀξιολογώτατον,

cf. 1.2n.

digna memoratu), but itis clear from the other passages where the two adjs. are combined that the ethical quality of the actions in question can vary depending on the context (Ter. Haut. 314, Virg. Aen. 4.94, Liv. 39.51.10, Liv. ap. Sen. Suas. 6.22, Ov. Met. 10.608, Vell. 33.1, Luc. 4.496). Likewise facinus (often combined

with

audere)

can mean

either ‘deed’ or ‘crime,

outrage' (OLD 1—2): facinus memorabile 15 used to describe the Catilinarian conspiracy by Sallust (C. 4.4) but deeds of heroism by Livy (e.g. 38.24.2), in whom the phrase recurs (4x). Only with the first words of the next sentence does T. bring some clarification to the matter (occiso centurione...).

occiso centurione ac militibus: ‘Observe the singular predicate with two subjects, the verb usually preceding in such cases. This is particularly frequent in ablative absolutes, the predicate here agreeing with the nearer noun' (Gudeman): cf. e.g. H. 5.20.2 ‘occiso praefecto castrorum et quinque primoribus centurionum', Caes. G. 3.20.2 'auxiliis equitatuque comparato', Cic. Phil 4.16 'referente...hoc M. Seruilio collegisque eius'. qui...habebantur ‘who, mingling with the maniples, were being used as exemplary instructors to impart discipline.' Originally a maniple comprised two centuries and was equivalent to a thirtieth of a legion; the unit had long since become obsolete in practice, but T. retains the term throughout his works. exemplum et rectores, typically combining abstract and concrete (e.g. 20.2, 24.2), seems to bea hendiadys. For trado— 'to impart' 566 OLD 2c; for rector— 'instructor' and the like see OLD 5; for habeo — ‘to

employ' see OLD 23a. Liburnicas: this is an alternative form of Liburna and refers to a light type of ship which is frequently mentioned bv T. in the Histories but was most famously associated with Octavian's victory at Actium (e.g. Hor. Epod. 1.1 with Watson, Odes 1.377.30 with N-H). Liburn(?c) aewere double-banked

and had two rows of oars; thev are said to have required 60 rowers and been able to carry 30 troops (O-R; unfortunately the text of H. 5.23.1 15 badly corrupt). If the mutinous cohort was at anything like its full strength of 480 (see above, n. cohors), and even if the Usipi acted as their own oarsmen, there was still a serious discrepancy between the official capacity of the three vessels (270 in total) and the number of men actually on board. adactis per uim gubernatoribus presumably means something like ‘having threatened the helmsmen with violence' (lit. 'having been compelled by force’: OLD adigo 7). ascendere ‘boarded’

(OLD 4a) but whether historic inf. or the alterna-

tive form of ascenderunt 15 unclear.

COMMENTARY:

28.2

229

uno remig an achievement’. The desire for the unknown is conventional (e.g. Sen. Ot 5.1 'cupidinem...ignota noscendi’), but in his talk of boundaries (above), the unknown, and ‘beyond’ (below), Calga-

cus visualises the Romans as characterised by the insatiability of Alexander: see esp. Sen. Suas. 1.2 ‘quod noueram uici; nunc concupisco quod nescio’, 3 'orbem quem non noui quaero', Curt. 9.3.8 'in alium orbem paras ire et Indiam quaeris Indis quoque ignotam'. The successful conquest of previously unknown lands and peoples is a panegyrical topos (10.4n. incognitas, 24.1). sed nulla iam ultra gens, n:hil nisi fluctus ac saxa: the logic of Calgacus’ alliterative and assonantal statement is this: every every unknown quantity represents a challenge, but (sed) now there is no longer a challenge left but ourselves. For the motif see Hom. Il 15.737—40 οὐ pév 11 σχεδόν éc i πόλις πύργοις ἀραρυῖα, | ἦ K ἀπαμυναίμεσθ᾽ ἑτεραλκέα δῆμον éxovTes | &AN &v γὰρ Τρώων πεδίῳ πύκα θωρηκτάων | πόντῳ κεκλιμένοι (‘nearby there 15 no city

equipped with walls where with a decisive force we can defend ourselves; but we are encamped in the plain of the stoutly breast-plated Trojans, lying close to the sea’).

infestiores Romani: sc. sunt or perhaps sunt quam antea. Commentators suggest quam haec, but this seems too defeatist: Calgacus mentions waves and rocks not to point out the difficulties of escape but to emphasise that his men are the only remaining opposition (last n.). Cf. Liv. 21.41.15. For infestus see 41.1n. (nfensus). superbiam frustra...effugias: superbia is the typical characteristic of tyrants or those acting tyrannically and is a charge frequently made by Rome against her enemies (e.g. À. 14.26.2 'superbiam Parthorum’;

249

COMMENTARY:

30.4

Watson on Hor. Epod. 7.5—6) and, conversely, against Rome by her opponents (e.g. A. 2.15.3 ‘meminissent modo auaritiae, crudelitatis, superbiae', 3.40.3 'saeuitia ac superbia praesidentium’); for effugias cf. Liv. 40.4.14 'effugite superbiam regiam' (and for the indefinite second-person subjunctive see 3.1n. sc). obsequium 15 again coupled with modestza at 42.4: see above, Intro. p. 23. 30.4 raptores orbis: there is a similar slur in the hortatio which Silius gives to Hannibal before Cannae (9.200-1): 'raptor per saecula longa | Dardanus edomitum uobis spoliauerit orbem'. It had been a convention that Britain was 'another world' separate from ‘our Roman world' (12.3n. dierum), something of which Boudica makes much in the speech given to her by Dio (62.4.2); but the Roman advance to the north of Scotland gave the lie to this convention, as Calgacus seems to acknowledge by twice using ‘global’ terminology in his speech (again at 31.2 below); see also Clarke (2001)

100-4. Rather differently, the Samnite leader Telesinus called the

Romans 'raptores Italicae libertatis’ (Vell. 27.2). For the motif of robbery or plunder compare the anti-Roman letter which Sallust writes for Mithridates (H. 4.69.22 ‘latrones gentium', repeated against Alexander at Curt. 7.8.34): see further below. postquam...defuere terrae, mare scrutantur: the chiastic arrangement emphasises Calgacus' point, for which cf. Sen. NQ 3 praef. 10 'deficiente ad iniurias terra, errasse in Oceano ignota quaerentem'. The antithesis terrae ~ mare is the first of three others (locuples... auari ~ pauper ambitioss, Oriens — Occidens, opes ~ inopiam) and, together with the double anaphora (sz...szand mom...non) and totalising expression soli omnium, helps to generate a strong sense of closure at the end of the first paragraph of Calgacus' speech. mare scrutari (again at G. 45.4) belongs above all to the language of Senecan moralising (Clem. 1.3.5 *mare lucri causa scrutamur', Ep. 89.22 ‘profunda et insatiabilis gula hinc maria scrutatur', 119.7 'scrutatur maria ignota' (of Alexander), cf. Helv. 10.2) but note also Juv. 5.95-6 ‘scrutante macello | proxima'. For the motif terra deest in very different contexts cf. A. 13.50.1, Vell. 53.3. 51 locuples...ambitiosi: the alliterative pairing of auarus and ambitiosus, like that of the corresponding nouns, is esp. common in Seneca (see e.g. W-M on A. 3.13.1), though T. may perhaps be here thinking of Sallust (below, n. soli). The emphasis 15 on the second clause: wealthy peoples naturally provoke greed in others but an impoverished people, like the Caledonians, offers no such temptation — except military victories and territorial acquisitions to those in pursuit of glory (OLD ambitiosus 4). Compare the remarks attributed to Caratacus at Dio 61.33.3c and contrast those attributed to Boudica at Dio 62.3.5. quos non Oriens, non Occidens satiauerit: satiauerit 15 more pointed as a perfect subjunctive (causal/generic, cf. NLS $156) than a fut. pf. indic.:

COMMENTARY: 30.5

243

'whom neither east nor west could satisfy'. Calgacus speaks as if he knows that in Mithridates’ time the king accused the Romans of turning eastwards after they had exhausted the west (Sall. H. 4.69.17 'postquam ad occidentem pergentibus finem Occanus fecit, arma huc conuortisse’). Now, having exhausted the east too, they have turned north. For the topos of ruling 'from east to west' see W. on Vell. 126.9. soli omnium opes atque inopiam...concupiscunt: omnium is to be taken with soli (as G. 45.4 ‘mare scrutantur, ac soli omnium...’). opes and ?nofia are employed frequently in various types of word-play, which here gives point to the Romans' paradoxical craving. Some commentators compare Sallust's account of ambitio and auaritia (C. 11.3 ' concupiuit...insattabilis...neque copia neque inopia...'). For par! adfectu cf. Sen. Phoen. 383, 461. Ihe first paragraph of Calgacus' speech, like the last (32.4 cogitate), ends with Cicero's favourite ditrochee or ‘dichoreus’ (- « — x):see 15.5n. (delib-

erare). Calgacus' other exs. are g1.1 polluuntur, conteruntur, 31.2 dominis aluntur, 32.1 teneri putatis, 32.2 caritatis, tradiderunt. go.5 Auferre...appellant at the start of the second paragraph the change of subject (above, p. 236) is marked by three headlined outrages (‘auferre trucidare rapere’), each of which is elaborated in the sequel (auferre and rapere at 31.1—2 ‘liberos. .. pascit’, trucidare being reserved for last at g1.2 'nos...in excidium petimur’). The last five words of this opening sentence are extremely famous (see e.g. Birley (1999) xxxix, Benario (2007) 66) and are often used by modern politicians and the like as a free-

standing epigram, but this obscures the fact that, just as appellant governs imperium as well as pacem, so the plural falsis nominibus refers to pacem as well as tmperium. The sentence also represents a sequence of ideas: solitudo is the result of the actions auferre trucidare rapere. Whether falsis nominibus, a common combination (e.g. H. 1.37.4 and H., Sen. Thy. 446 and Tarrant), is abl. of instrument or of attendant circumstances is hard to determine. On the idea that reality is often different from what is said to be the case see also 21.2n.

(uocabatur).

The Romans believed both that peace was achieved through victory in war (memorably expressed by Augustus at RG 14 'cum per totum imperium populi Romani terra marique esset parta uictoriis pax', cf. 26.2—9; Weinstock (1960) 45-0, 49-50) and that, although Rome's imperial

mission

was

to

rule,

to

civilise

and

to

be

merciful

to

the

con-

quered, nevertheless recalcitrant peoples should be suppressed (most eloquently voiced by Virgil at Aen. 6.851-3: see above, 20.2n. (parcendo), and Intro. pp. 17-18). It is not difficult to see how these beliefs could be abused in practice and seized upon by enemies. Calgacus’ famous words about solitudo derive ultimately from the younger Camillus' speech at Liv. 8.13.14-15 (‘itaque pacem uobis...parare in perpetuum uel

244

COMMENTARY:

saeuiendo

uel ignoscendo

31.1

potestis. uoltis crudeliter consulere in dedi-

tos uictosque? licet delere omne

Latium, uastas inde solitudines facere’),

which also seems to have influenced Curt. 8.8.10 ‘ueni enim in Asiam non ut...dimidiam partem terrarum solitudinem facerem sed ut illos quos bello subegissem uictoriae meae non paeniteret’, 9.2.24 ‘postquam solitudinem in Asia uincendo

fecistis’ (both Alexander),

Plin. NH 6.182

‘nec tamen arma Romana ibi solitudinem fecerunt’, Plin. Pan. 48.5 ‘nec umquam ex solitudine sua prodeuntem nisi ut solitudinem faceret' (of the monstrous Domitian). See also Tosi $1209. 31.1 liberos cuique ...natura carissimos esse uoluit 'Nature has willed that one's own children and relations are each person's dearest': suosrefers not to the subject of the sentence but to cuique (see 7.1n. in praedis); for the sense of uoluitsee OLD 13 (b for the acc. 4 inf.). natura . . 6556 uoluitisa common form of words in prose (e.g. Cic. Sull. 8, Deor. 3.195, Curt. 9.3.13, Sen. Cons. Pol. 4.3, Ep. 41.6, Plin. Ep. 8.13.2), matching the commonplace nature of the idea (e.g. Cic. Off. 1.12). per dilectus: earlier T. had stated that in his day 'Britanni dilectum ...impigre obeunt'

(13.1). seruituri (below)

expresses purpose

(NLS

$92 (d)).

etiam si hostilem libidinem effugerunt, nomine... hospitum polluuntur: the contrast is pointed by the chiastic arrange&nt of the terms and the alliterative word play (hostilem — hospitum), which is common elsewhere (e.g. of rape at Liv. 1.12.8 'perfidos hospites, imbelles hostes', 1.58.8 ‘hostis pro hospite', Ov. F. 2787 'hostis ut hospes'; Oakley on Liv. 6.26.3). For hostilem libidinem cf. esp. Rhet. Herenn. 4.12 (quoted in an example of the 'grand style’); polluo is regular of sexual defilement (OLD 4). Compare Caratacus' reminiscences about the Britons who had fought Caesar (A. 12.34 ‘quorum uirtute uacui a securibus et tributis intemerata coniugum et liberorum corporaretinerent’), the experiences of Prasutagus (A. 14.31.1) and the speech of Boudica (A. 14.35.1). For a study of women and Roman soldiers in the context of conquest see Phang (2004); she notes that, in order to celebrate the Claudian conquest of Britain, 'the builders of the Sebasteion at Aphrodisias in Asia Minor represented a heroic Claudius as overpowering a cringing female personification of Britannia' (217). See Figure 4. bona fortunaeque...contumelias conteruntur: by a mild syllepsis (cf. OLD contero 4) the verb applies equally to the three paired subjects (-que...atque...ac) in asyndeton; the first two are constructed with purposive 2n - acc. (5.1n. mhil), the third 15 varied bv a paired gerundival expression (ac) which itself is varied by a paired prepositional phrase

(et).

ager atque annus is suggestive rather than explicit: the possible mean-

ings are (i) ‘our fields and our time’, (ii) ‘our fields each year’ (a form of

COMMENTARY:

4-]'.!:::.' L

31.1

)flz Μ

245

!᾽ἢ͵ ᾿Ἠἧ

Figure 4 Claudius subduing Britannia (relief from the Sebasteion at Aphrodisias) hendiadvs), (111) ‘our fields and their vearly produce' (for this extended meaning of annus cf. G. 14.3; OLD 4). in frumentum ‘for grain' (OLD frumentum 1b). siluis ac paludibus emuniendis is perhaps instrumental abl., a use of the gerundive which is not applied to completed actions but is appropriate 'onlv when the action is in progress or contemplated' (NLS $206 Note 1), exactlv as here: *bv the continual metalling of woods and marshes'. On

246

COMMENTARY:

31.2

the other hand, the parallelism with the purposive i frumentum above may suggest rather the dat. gerundive of purpose, an archaising construction of which T. is generally fond (Adams (1972) 372—-3; cf. NLS $67 Note, $207 (4)): 'for the metalling of woods and marshes'. The compound verb appears to be unique in the sense of building a road through an area (TLL 5.2.544.02-3) but the simple form munio is so used (Liv. 21.37.2 ‘ad rupem muniendam’, /LS 39 "Tempe muniuit’). "The force of the compound e- is to convey the method by which Roman roads were built on an embankment’ (O-R). For the combination of forests and marshes see 26.2n. (quod nis?).

inter uerbera et contumelias: this combination

found first in Cicero

(again at A. 1.20.1) is

(Phil. 11.5), then in the Senecas

(Suas. 6.10, Const.

sap. 5.1). 31.2 nata seruituti mancipia...aluntur: the point is that a master, having bought a slave, would be responsible for the slave's livelihood if he wanted the slave to work for him and to be an economical proposition. Commentators compare the speech which Dio puts into the mouth of Boudica (62.3.3): ‘How much better it would be to have been sold to masters

once

for all [ἅπαξ...

πεπρᾶσθαι]

than...to

have

to ransom

our-

selves everv year!' (Loeb trans.). For nata seruituti cf. Cic. Prov. cons. 10, Liv. 36.17.5. Britannia...emit...pascit: adversative asyndeton again. em?it (~ ueneunt above) refers to paying tribute and taxes, pascit (~ aluntur) to providing grain and produce. seruitutem is most likely abstract, ‘slavery’, but could possibly be an example of the rare concrete use, ‘a bodv of slaves' (OLD 1C).

sicut in familia recentissimus quisque seruorum etiam conseruis ludibrio est, sic in hoc orbis terrarum uetere fWIatu noui nos et u_Fsin excidium petimur: the centuries-old notion of the globe’s enslavement to Rome (Intro. pp. 16—1%7), which T. will again use later in the mouths of the Batavi (H. 5.25.1 'totius orbis seruitium’), is here rendered more vivid by Calgacus in that he compares it to an actual slave household (OLD familia 2), elaborating his comparison with balance, alliteration and assonance. famulatu, with which familiais etymologically connected (Maltby 222), probably has its normal sense of 'servitude' (so OLD), although itis taken (uniquely) in a concrete sense by TLL 6.1.261.32—4; either way, the word 15 much commoner in later than in classical Latin, and the metaphorical use is extremely rare (Cic. Off. 3.117, but different). peto + ἴη - acc. = ‘to demand someone for a given state or treatment’ (OLD 12), as Sen. Apoc. 15.2, [Quint.] Decl. 342.1 'in seruitutem’; in excidium, though found with other verbs (Manil. 3.5, Stat. Theb. 5.683, Flor. 1.12.1,

1.22.15), is not elsewhere

cf. Hor. Ep. 2.1.38.

combined with feto. For nouz. .. et uiles

COMMENTARY:

31.3

247

quibus exercendis reseruemur: exercereis regular of cultivating farmland (e.g. Manil. 1.86 ‘arua’; Virg. G. 1.99 'tellurem' with Mynors' n.) and of operating mines (e.g. Liv. 45.29.11, Plin. NH 6.98 ‘metalla’), butitis only bv an extension of meaning that it is applied to the excavation of harbours (as if effodere, as Virg. Aen. 1.427): perhaps Calgacus has misread Aen. 4.86—7 ‘non arma iuuentus | exercet portusue ... '. (The notion that Calgacus is referring to the Caledonians 'acting as dockers etc.', as mentioned by O-R, seems far less plausible.) reseruo is regularly found with dat. of purpose (OLD g), but the dat. gerundive (above, 1n. szluzs) seems paralleled only at Col. 9.15.10. For Britain's natural resources see above, 12.6 and nn.: Calgacus is dissembling. 31.3 uirtus porro ac ferocia: ferociais a standard characteristic of barbarians (8.1n. feroci): surprisingly, its co-ordination with uzrtus, though found variously elsewhere in T. (A. 2.25.3, 11.19.1, H. 5.11.2), 15 only at Liv. 9.6.13; but cf. Virg. Aen. 12.19—20 'quantum ipse feroci | uirtute exsuperas' (of Turnus), imitated at Sil. 5.555 'uirtute feroci'. porrohere perhaps means ‘on top of that' (OLD 5, cf. 6 ‘furthermore’, ‘at the same time/on the other hand).

quo tutius, eo suspectius: the two adjectives agree in gender with secretum both because it 15 the nearer subject and because neuter takes precedence over feminine when things without life are combined (G-L $286 and 1 (a)- (b)). Calgacus' point seems to be that, although their distant seclu-

sion has protected them hitherto, it has also made them unfamiliar to the Romans, who are therefore all the more suspicious of them and hence all the keener to eliminate their perceived threat. Cf. also 30.3 'omne ignotum pro magnifico est' (n.) and, for a different association of the Caledonians with the unknown, 25.3. sublata spe ueniae: hope or its absence is a regular feature of hortationes (e.g. Virg. Aen. 2.354 'una salus uictis nullam sperare salutem’); ueniais the leniencv of the conqueror (OLD 4). spes ueniae, though found elsewhere, is above all a Livian expression

(7x) of which T. is fond

(elsewhere at H.

4.50.1, 70.5, À. 14.29.1). tandem sumite animum ‘it 15 now that you must take courage’: tandem emphasises the command and perhaps indicates impatience (OLD 1c). animum (-os) sumere is first in Liv. 6.23.3 and in its various meanings is seemingly more common in prose than verse (H. on H. 1.27.2, adding Vell. 2.17.3). tam quibus salus quam quibus gloria carissima est: the position of tam and quamimmediately before the two relative pronouns indicates that Calgacus 15 addressing two distinct groups (corresponding to the zgnau? and fortes mentioned at 30.1 above): those for whom survival is their main interest, and those for whom it 15 glory. In both cases the antecedent is uos. For the antithesis salus — gloria see 26.2n.

(securi). car(issim)a salus is very

248

COMMENTARY:

31.4

common in, and almost exclusive to, Cicero (note esp. Sen. Suas. 7.5 ‘tuis utar, Cicero, uerbis: “cara est cuiquam

salus...?”’); for c. gloria cf. De or.

1.19. 31.4 Brigantes femina duce...potuere: Calgacus ends the second main section of his speech with an exemplum a minore ad maius ductum (Quint. 5.11.9; Lausberg (1998) 199 $420 (1bp)) which acts as a foil for the sentence which follows (*nos...quos sibi Caledonia u?zros seposuerit’); the awkwardness is that the Brigantes under their queen Cartimandua (17.1n.) are nowhere said to have engaged in the actions listed here, which are specifically those of the Iceni under Boudica in 60/61

(16.1 and nn.). It

does not seem plausible either that the text 15 corrupt or that T. has represented Calgacus as getting his facts wrong (although Chaplin (2000: 38) has shown that this is characteristic of the deplovment of exempla by foreigners in Livy). Welleslev (1969: 268) ingeniously suggested that T. may have become confused because Camulodunum was not only the name of the well-known colonzain East Anglia which Boudica and her Iceni attacked (next n.), but also the name ofa (possibly Agricolan) fort and settlement in the territory of the Brigantes (the modern Slack in West Yorkshire).?? For fem?na duce cf. 16.1

(n. neque).

exurere coloniam, expugnare castra: the colonia is Camulodunum (16.1 ‘ipsam coloniam inuasere’; 5.2n. trucidati); the camp 15 presumably that of Legio IX at Lincoln, although T.’s later account says that it survived (A. 14.32.3): perhaps Calgacus is capitalising on the ambiguity of expugnare,

which, like the English 'to storm', can mean both 'to make an assault on'

and ‘to take bv assault' (TLL 5.2.1810.54ff. and 1807.60ff. respectively); for exaggeration in claims to have extirpated a people or place see 18.93n. (caesaque) . The verbs are more vivid as historic infinitives, though they may be dependent on potuere below. ac, nisi felicitas in socordiam uertisset, exuere iugum potuere: an indicative may be found in the apodosis of an unfulfilled condition when a modal verb is constructed with an infinitive or when an action is attempted or begun

but not finished

(NLS $200

(i) and

(ii), cf. A-G

8517b-c,

G-L

$597 RR. 2-3): such 'mixed' conditions, of which T. 15 fond (esp. in A.), recur at 37.1 'coeperant, ni...opposuisset' and in the acc. - inf. at 4.3 ‘hausisse, ni...coercuisset' (n.), 13.2 'agitasse ... ni...fuissent'. Here the sentence may be translated either as 'they would have been able to cast off the yoke if success had not turned...' or as 'they could have cast off the yoke but success turned...’. (If potuere is to be taken with exurere and expugnare above, in those two cases it means simply ‘were able'.)

29 Fora modern example of this confusion see the map in Mattern (1999) xx.

COMMENTARY:

31.4

249

socordia (only here in Agr.) and socors are archaising terms liked by the historians (Oaklev on Liv. 6.37.8). socordia normally — ‘lethargy’ and in this meaning is a favourite word of Sallust; but it can also — 'stupidity', a meaning associated esp. with the elder Cato (OLD b), or perhaps 'carelessness' (M-W on A. 4.31.2). Either of these last two meanings would be appropriate here. Since the corruptive effects of felicitas conventionally included greed (e.g. Sen. Clem. 1.1.7 'facit quidem auidos nimia felicitas' and Braund ad loc.), Calgacus may be implying that plunder and its enjoyment had diverted the Britons from capitalising on their success, a topos of barbarian behaviour (A. 1.65.6 'iuuit hostium auiditas, omissa caede praedam sectantium’; W-M on A. 3.20.1); cf. also Liv. 7.35.5 ‘cum eo hoste res est qui hesterno die delendi omnis exercitus fortuna per socordiam usus non sit’ (Decius' (OLD 21a).

hortatio). The

intransitive use of uerto is common

nos integri et indomiti...: adversative asyndeton. indomiti continues the metaphor of the yoke (cf. Catull. 63.33 'iuuenca uitans onus indomita iugi', 68B.118); there seems to be no ex. of integer used of untamed animals. in libertatem non ?2 paenitentiam laturi: there are two problems with the transmitted text: (1) paenitentiam has seemed to some schol-

ars to be a so-called *monastic corruption', an error by a scribe with Christian repentance on his mind

(see Ogilvie

(1971)

for this term and some

exs. from Liw), (2) laturilacks an expressed object. For a survey of scholarly responses to these problems see Karlsson (2005). The (unparalleled) contrast between :n libertatem and in paenitentiam seems an unexceptionable wav of expressing the hope that the actions of the Caledonians will result in their continued freedom rather than in any regret for having fought against the Romans (cf. OLD n 20 'so as to produce or result in a given condition’). For paenitentia as entailing the ideas of 'shortfall' or ‘not measuring up' (‘what I am or have done falls short of some

notional, superior alternative’)

see Kaster

(2005)

68, 77,

(and cf. 19.3n. nec). Wolfflin proposed :n patientiam (‘submission’) but this phrase occurs nowhere else in Latin, whereas in paenitentiam is found (admittedly after verbs expressing change) at H. 2.63.1, 4.37.2 and Suet. Claud. 13.2; cf. also Sen. Contr. 7.8.1 'in paenitentiam mortis recipienda est’. Several scholars have wished to understand nos as object of laturi, but this seems too harsh an ellipse: to produce an expression of motion Meister supplied nos, while Breithaupt emended to 2tur? (but this weakly anticipates 32.4). Others have preferred an expression of fighting or attack. Weinkauff proposed «bel-laturi, which is accepted by Soverini (cf. A. 2.44.2 ‘pro libertate bellantem' and H. 1.89.2 ‘in unius sollicitudinem aut decus...bellauerat’). Mohr suggested inserting arma, but

250

COMMENTARY:

32.1

arma ferre is not used elsewhere by T., who much prefers arma nferre. this has the double advantage that the sequence arma:n could easily have been omitted after -tzam and that :nlaturi continues the alliteration of which Calgacus is so fond. But, while the general sense of the passage seems clear, in the nature of things it is impossible to know exactly what T. wrote. primo statim congressu: primo congressu is perhaps technical in military narrative, being found in Caesar (C. 1.46.4, 1.47.1), Sallust ([ 74.3), Livy (4.33.1, where statzmis also interposed, 10.1.9, 38.17.6, 42.16.9) and Frontinus (Strat. 2.1.8) besides T. (again at 36.1, A. 2.3.1). For the combi-

nation primo statim see 3.1n. 32.1 An eandem...creditis?: though the general meaning of this sentence is clear enough, its mechanics are rather less straightforward than the absence of scholarly comment would lead one to believe. (a) Normally zdem qui has, or refers to, the same noun in the main clause

as in the relative (as 1.3 'iisdem temporibus...quibus...', Vell. 121.1 'eadem...uirtus...subsequenti tempore ... Tiberii fuit quae initio fuerat ; G-G 540a-b); only here in classical Latin, it seems, is there a different noun in each clause (so TLL 7.1.196.14-16): eandem must therefore refer not to strict identity but, as in such cases as Cic. Har. resp. 22 ‘tua tum ...eadem uirtus fuit quae...quondam tuo proauo', to similarity of type, size or degree etc. (‘the same kind/amount/degree as’). (b) Since adesse-- dat. (OLD 17b) is little different from esse4+ dat., Romanis...adesse 15 equivalent to ‘the Romans have’; but creditis cannot govern the subordinate clause as well as the main clause, since Calgacus would not wish to imply that the Romans' lasciuza was a matter of mere belief rather than of hard fact. Why, then, 15 quam ?n pace lasciuiam (adesse) in the accusative and infinitive? Either we must understand e.g. uidetis by zeugma ('as the indiscipline which you see they have in peacetime’) or a clause which in full form in oratio recta would have been quae in pace lasciuia adest has been attracted into the accusative and infinitive (‘as the indiscipline which they have in peacetime’). On the latter interpretation one might translate: ‘Or do you believe that the Romans have the same degree of manliness in war as they have of indiscipline in peacetime?' uzrtutem (nowhere else contrasted with lasciuia) looks back to u?ros above, contrasting the true manly courage of the Caledonians with its false counterpart amongst the Romans, whose previous successes have been due entirely to the disunity of their enemy. This sentence thus effects the transition to the third main topic of Calgacus' speech (see above, p. 236). nostris...clari: Calgacus turns to the topic of Caledonian unity (above, p. 237) by admitting, and re-casting, T.'s own statements at 12.2 above (esp. ‘nec aliud aduersus ualidissimas gentes pro nobis utilius quam quod Ϊῃ commune non consulunt'). dissensio and discordia are an exclusively

COMMENTARY:

32.2

251

Ciceronian combination (e.g. Phil. 8.7) before T. (again at D. 40.4); later at Gell. 2.12.1. The word order is interlaced (synchesis); the ablatives are causal. uitia hostium in gloriam exercitus sui uertunt: thc main clausc repcats what Calgacus has just said, but in different words ('theme and variation’) and differently arranged (acc. - gen. — acc. - gen.). For the notion of capitalising on others' defects cf. H. 4.34.9 ‘non minus uitiis hostium quam uirtute suorum fretus’; also 41.4 below, Cic. Leg. Man. 67. For in gloriam uertunt cf. Curt. g.10.28 'fortuna ... hoc quoque militiae probrum uertit in gloriam', Plin. Ef. 8.18.10. contractum ... dissoluent: the diversification of an army was discussed by Plb. 1.67.9ff. and Veg. 9.4 (1 'interdum mouet tumultum ex diuersis locis collectus exercitus’), though there the problem is mutiny and not, as here, desertion. Cf. also Hannibal and Caesar respectively on their opponents (Liv. 21.43.18, Luc. 7.270—4). nisi si is no different from nist (OLD nisi 7), here said with irony: 'unless

of course you think that...’ (the verb 15 putatis). pudet dictu is a unique variant on the normal fudet dicere (Sall. H. 1.58 eLc.).

licet...seruos is parenthetic: ‘(who), though they put their blood at the disposal of a foreign masterdom, nevertheless (have been) enemies longer than slaves'. Since commodareis regularly used of parts of the body such as one's ears or hands (OLD 2b), its application to blood is bitterly ironic; alienae 1s also pointed, since the lives of real slaves were at least at the disposal of their own master, and in the antithesis hostes ~ seruos Calgacus affects to forget the proverbial quot serui, tot hostes (Otto 319). 32.2 metus ac terror est, infirma uincla caritatis ‘it is {car and terror [sc. by which they are held], weak bonds of affection.' Beroaldus emended est to sunt, which most edd. accept, but it seems unlikely that an original sunt would have been corrupted to est. The singular verb is explained by the fact that its two abstract subjects form a single unit (4.9n. mitigauit). metus ac terror, a frequent combination, contrasts with fide et adfectu above: commentators note that the theme goes back at least to the Mytilenians’ speech at Thuc. g.12.1 ‘we were held in the alliance by fear rather than friendship' (see 4150 Oakley on Liv. 8.14.16); the specific contrast of metus and fides, though found elsewhere (e.g. H. 1.76.1, Cic. Rosc. Am. g1), 15 a favourite of Liw (c.g. 7.25.7 'cum cerneret metu tenendos quos fides non tenuisset', 26.12.6 ‘metu magis eos quam fide continente', 30.20.5 ‘quae pauca metu magis quam fide continebantur’). The 'bonds of affection’ isafairly common metaphor, but its application to caritas, rather than to the regular amor (e.g. Sen. fra 3.94.2), recurs only at A. 1.55.9 ‘gener inuisus inimici soceri, quaeque apud concordes uincula caritatis, incitamenta irarum apud infensos erant'. The archaising uinclum, only here in

252

COMMENTARY:

32.2

Agr., preponderates in A.; uinculum, absent from Agr., is used exclusively in G. and preponderates in H.: see Adams (1974) 575. For ?nfirma cf. Liv. 7.30.2, Val. Max. 9.12. ext. 1. quae ubi remoueris: the verb is either a future perfect indicative or an example of the second-person perfect subjunctive in a present generalising clause ('and when(ever) $217 (2) (c) Notei.

you remove

them’): for this idiom

see NLS

qui timere desierint odisse incipient: the relationship between fear and hatred is analysed at e.g. Sen. Clem. 1.12.3-4 (with Braund), where the proverbial oderint dum metuant — the tyrant's mantra — is quoted. The tense of incipient suggests that desterintis fut. perf. indic. (on the analogy of future vivid conditional clauses, in which the latter is regular). uictoriae incitamenta: the expression does not occur elsewhere, though cf. H. 4.18.2 (quoted next n.). zncitamentum, often -- gen. (e.g. A. 1.55.3, quoted above, n. metus), is a noun which T. likes (15x): see 20.2n. (znuitamenta).

nullae Romanos coniuges accendunt; nulli parentes fugam exprobraturi sunt: not only are the Romans far from home and parents but Calgacus is perhaps being made by T. to know that Roman soldiers were forbidden by law from marrying (though the presence of women in or near military camps was a fact of life). At A. 14.34.2 the Britons fight in the presence of their women, and a similar practice is attributed to the Germans (G. 7.2—8.1 with Rives' n.), the Batavi (H. 4.18.2 'coniuges...hortamenta uictoriae uel pulsis pudorem’), the Thracians (A. 4.51.2), and the Gauls (Caes. G. 1.51.2—3); but on this occasion the Britons have already removed their wives from the front line to safe areas (27.2n. coniuges), so Calgacus must be referring to their proximity rather than to their presence. T.’s language perhaps recalls Liv. 7.11.6 'pugnatum...in conspectu parentum contugumque ...quae magna...hortamenta animi...militem accendebant', where the Romans are watching from the walls (tezchoskopia): for further exs. see Oaklev ad loc. and Addenda (4.552). The contrast between accendere and exprobrare (again at A. 19.13.1) is perhaps also Livian (23.18.7); for fugam exprobrare cf. Ov. Met. 13.69. aut nulla plerisque patria aut alia est ‘most have either no fatherland or another one': i.e. either they have forgotten their original patria altogether or, if not, it was a different one from Rome/Italy. This is the converse of

the regular hortatory topos that one's own troops are fighting 'pro patria' (see 15.4n. patriam). Since Roman legions occasionally and other military units often bore names which reflected their origins (as Legio III Gallica, Ala I Tungrorum, Cohors II Gallorum), it was natural for Calgacus to be made to deny that the troops were culturally assimilated (modern scholars, e.g. Havnes (1999), have argued the opposite); moreover, since it was proverbial that ‘omne solum forti patria est’ (Ov. F. 1.493: see Otto 268;

COMMENTARY:

32.3

253

also Oakley on Liv. 9.4.11-12), he implicitly repeats his claim (1 above) that the Roman army lacked wrrtus. paucos numero: to have the enemy despising the numerical inferiority of the Romans was a topos: see 15.3n. (quantulum) and reference there to Oakley. numero is abl. of respect. trepidos ignorantia: ‘in ancientasin modern times the unknown proverbially caused more fear than the known' (Oakley on Liv. 9.24.8, adding Onas. 14.4); the abl. is causal. Calgacus' point, again similar to one made by Boudica in the speech given to her by Dio (62.5.6 'this region ...to them is unknown and hostile’), is elaborated in what follows.

caelum ipsum presumably refers both to climatic conditions and to the climatic region or ‘clime’ itself (OLD 7a-b). Compare Liv. 30.43.7 ‘multum etiam adiuturam notitiam maris terrarumque et uentorum, quae omnia ignaros turbatura hostes essent'. clausos...ac uinctos: in a reversal of roles, Calgacus imagines the Romans as wild beasts chained in a cage (Liv. 36.7.13 'ferae bestiae uinctae aut clausae’, cf. 35.18.6 'ferarum modo, quae claustris aut uinculis teneantur’): the combination of the two verbs does not always generate such an image (cf. Cic. Verr. 5.168, Liv. 2.24.6, Hyg. Fab. 7.2), but A. will respond with animal imagery in his speech below (34.2—3 and nn.). 32.3 ne terreat uanus aspectus et auri fulgor atque argenti: another example of 'theme and variation': uanus aspectus — 'empty appearance' (OLD uanus 2a, aspectus 5a) and 15 explained by aur...argenti. The Roman army was notable for its brilliant display (26.1n. fulsere signa), but ‘pretty but useless decoration in silver and gold' is a topos of battle narratives and hortationes (see Oakley on Liv. 9.40.4—6, adding Just. 11.19.11). For terreat.. . aspectus (again at H. 1.40.2, 4.78.2) cf. Virg. Aen. 11.699, Curt. 3.9.11, 6.6.34, Plin. NH g.50, Sen. Clem. 1.26.3 (Liv. 30.18.6 ‘ad ...aspectum territi’). quod neque tegit neque uulnerat: sc. eos and nos respectively with the verbs; the antecedents of quodare auri...atque argenti regarded as a single unit (G-L $285.2).

nostras manus: the most apt meaning would seem to 6 ‘our agents/instruments' (OLD 23), as in the controversial passage at 15.2 (n.

alterius). The

clausula

(znuen? |emus nostras manus)

is a favourite

of

Cicero (11.4n. quod). Britanni...Galli... Germani: the repetition of names from 1 above indicates the conclusion of Calgacus' third main topic, while the chiastic arrangement of deserent ... Germani — Usipi reliquerunt concludes the main part of the speech as a whole; what follows is by way of peroration. For the Usipi see 28 above. 'What the chieftain does not see — and what we do is that the Usipi are a terrible example for his larger point, that driving off Rome's empire is better than enduring it. For, whatever hope of safety

254

COMMENTARY:

32.4

the Usipi had had, they lost when they threw off their preceptors: without them, the world is one of universal depredation' (Sailor, CT 33). nec quicquam ultra formidinis looks back to nullae ultra terrae (30.1) and nulla iam ultra gens (30.3): the forthcoming battle is at the end of Britain, since there are no further lands or peoples to the north; if the Britons win the battle, there will be no further grounds for fear (OLD formido? c), since everything to the south (ultra) 15 in disarray. senum coloniae is a slighting reference to the fact that veterans had been settled at Camulodunum: see 5.2, where (unlike 31.4 above) a rhetorical plural is again used of the single colony (n. trucidati). The same contemptuous reference to veterans is put into Civilis' mouth at H. 4.14.4 ‘nec aliud in hibernis quam praedam et senes' (cf. 2.57.1). See further n. on aegra below. inter male parentes et iniuste imperantes: :nter is causal, 'given the circumstances of reluctant subjects and unjust rulers’ (OLD 8c, though none of the exs. is personal). For the word play (similar to A. 1.64.4) see e.g. Sen. Ep. 90.5 ‘cum bene imperanti bene pareretur nihilque rex maius minari male parentibus posset quam ut abirent e regno’; it is exceptionally common (see e.g. K. on Liv. 6.6.18 or W. on Vell. 79.1). inzuste imperantes 15 the language used in discussions of tyranny, empire, and similar topics (Cic. Off. 2.26, 2.29, Rep. 3.28). aegra municipia et discordantia: another rhetorical plural, since only Verulamium (St Albans) was strictly a municipium; like Camulodunum and Londinium,

it had suffered badly in the rebellion of 60/61, and the loss

of life in all three places together allegedly came to 70,000 (A. 14.33.2). Since physical and mental harmony were regarded as essential for health and well-being, it is possible that the metaphorical aegra (which perhaps = "weak': OLD 4) is sustained by d?scordantia (elsewhere in T. only in A.: see G. on 1.9.4), despite the lack of parallels; either way, the disunity imputed here to the Romans and their allies contrasts with the previous disunity of the Britons themselves

(1 'dissensionibus ac discordiis’).

32.4 ibi tributa et metalla et ceterae seruientium poenae summarily repeats Calgacus' second main topic (31.1-2 ‘tributum... seruitutem ...metalla') at the expense of some illogicality (cf. neque... nobis . . metalla...sunt above). For seruzentium cf. 30.2 (n. nec); the coordination of the three nouns varies the asyndeton in the first half of the antithesis (cf. 15.4n. patriam). quas...in hoc campo est: although the general sense is clear ('enduring or avenging these punishments depends on this battle’), we might have expected 29 + abl. to be constructed with e.g. stat (OLD sto 21a) or positum est (OLD pono 23) rather than est, the precise usage of which is hard to parallel (perhaps see OLD sum 11b). perferre and ulcisci are the subjects of est; poenas perferreis uncommon (Cic. Leg. agr. 2.92 (cf. Sest. 145), Nep. Arzst.

COMMENTARY:

353.1

255

1.5), p. ulcisci unparalleled. The motif of 'everlasting slavery' is deployed three times in the speech which Caesar gives to Critognatus (C. 7.77.9, 15 'aeternam ...seruitutem’,

16); for the uariatio of in aeternum — statim (an

unparalleled antithesis) see Sorbom (1935) 95. For campus— ‘battlefield’ see OLD 4a; for the motif of ‘this field' etc. cf. e.g. Luc. 7.260 ‘haec acies’, Stat. Theb. 11.485—6 ‘noster | hic campus nosterque dies’, Claud. 26.572—3 ‘hic omnia campus | uindicat’ (for hicalone, as in hic. .. exercitus above, cf.

e.g. Liv. 21.41.15, 43.5). proinde...et maiores uestros et posteros cogitate: for the combination of 'ancestors', a motif of the hortatio (15.4 and n. modo, A. 12.34 'uocabat...nomina maiorum' (of Caratacus), Thuc. 7.69.2), and ‘posterity’, mentioned in Boudica's speech in Dio (62.4.3), cf. Caes. G. 7.77.12—13 ‘nostri maiores...posteris prodi pulcherrimum [sc. exemplum]’, Curt. 4.14.25 'ite alacres et spiritus pleni, ut quam gloriam accepistis a maioribus uestris, posteris relinquatis'. cogztate — ‘reflect on’, ‘bear in mind’, 'think about' (OLD 8): being the current word, it is used bv T. mainly in speeches and only here in Agr.; its only appearance in Sallust 15 also in a speech (C. 44.5); in the narratives of H. and A. T. prefers the archaising reputo (see Adams (19732) 136). proinde in the sense of ‘accordingly’ or 'therefore’ (OLD g) is 'particularly associated with speech: until Apuleius, the vast majority of instances introduce a hortative sentence, or some form of instruction' (Holmes (1997/08) 63). The word in this sense is used by T. only in direct or indirect speech (ibid. 67). See also 33.6n. (proinde). 33.1 Excepere orationem alacres...dissonis: since the Roman soldiers will also receive A.'s speech with alacritas (35.1), a motif regular in Livy (Oakley on 7.33.4), it follows that ut barbaris moris refers only to fremitu . . . dissonis; as is to be inferred from the clause, raucous clamour 15

an ethnographical commonplace of (esp. northern) barbarians (Oakley on Liv. 7.10.8). moris (sc. est) is an example of ‘hypercharacterisation’

(to

use the technical term of linguistics). In Latin a possessive or qualitative genitive can be used to denote a custom, as at Cic. De or. 2.169 'barbarorum est in diem uiuere' ('itis customary of barbarians to live for the day’); but of course authors could also express the same idea by using an actual word for 'custom' in the nominative followed by either a genitive (as at [Quint.] Decl. 13.11 ‘barbarorum mos est’, ‘it is a custom of barbarians’) or a possessive dative (as at A. 4.49.3 ‘ut mos barbaris’, ‘as is the barbarians' custom’). In our passage the two conventions have been combined, the former type of genitive being applied to the actual noun meaning 'custom' (= something like ‘it is customary of the custom of barbarians’): the construction, which seems first attested at Cic. // Verr. 1.66 and 15 quite regular (see G-L 234 $366 R. 2, OLD mos 3b), recurs at 39.1 and 42.5. frem:tus is found nowhere

else with cantus but is common

with clamor(es); the

combination of cantus and clamor(es), again at H. 5.15.2,is Livian (5.37.8,

256

COMMENTARY:

33

25.24.5); clamores dissoni, though found elsewhere, is a favourite of Livy (7 X ). excipiois the regular word for greeting a speech in a given manner (OLD 9a).

33-34 The hortatio of Agricola It was a convention of historiography that a hortatio should be delivered by each side before a forthcoming encounter. A.'s corresponding horta?0 is not quite half the length of that of Calgacus and makes three main points: 33.2—3 past victories, 33.4—6 fight not flight, 34 beasts at bay. To most readers it has seemed the inferior speech in almost every respect. The disappointments have been well summed up by Rutherford (2010: 315—16). 'Calgacus' speech is sharper and more memorable for its sententiae, its forceful, aggressive language, and its bold turns of phrase.' In particular, ‘the most challenging parts of his speech, the denunciation of Roman empire-building as motivated by greed and disastrous for their subjects, receive no response...In terms of the balance of arguments within Tacitus' text, the silence on this theme is striking. Agricola makes no use of the crucial terms /ibertas and servitus.' These comments need to be seen in context. While it is true that Calgacus' speech is noteworthy for 115 stylistic effects, A. does not want for anaphora (33.2 tot...tot, 33.4 quando . .. quando, 34.1 uestra. .. uestros, hi...hi), balance (33.2 seu...seu, 39.2—3 Me...UO0S...ego...uos, 33.5 ut...ia, 33.6 ut...?ta [text], 34.2 quomodo...sic), word play (33.3 ergo egress? ego, 34.3 . bella. . . rebelland?), and alliteration (33.4 flumina fatigarent, uota uirtusque, uictoribus. .. uictis, 33.5 periculosissima . . . prosperrima, 33.0 terga tuta, 34.2 siluas saltusque). He too deploys Ciceronian clausulae (33.3n. finem, 33.0nn. iam pridem and nec inglorium, 34.1n. hi) and intertextual allusions (e.g. 33.5n. sed manus); but it is the suggestion of epic which is perhaps the most noteworthy feature. The speech is introduced by an elaborate and suspenseful sentence of notably epic composition (33.1n. zamque), and, while it is conventional to compare barbarians to wild beasts, the simile of the hunt casts the comparison in an epic mould - as well as being transformed brilliantly into a gladiatorial metaphor at the very end (34.2—3nn.). Seen in these terms, A.'s speech is not so much inferior as different. On

the

matter

of seruztus and

Bbertas,

Rutherford

himself acknowl-

edges that it would be unrealistic to expect A. to address the larger issues of Roman expansion: 'his soldiers presumably have no need to consider these matters, and would have little interest in hearing high-minded Panaetian doctrine.' But there is also a sense that Calgacus' accusations do not merit answering, because they have been pre-empted and made redundant by T.’s own text. A.'s relentless pursuit of barbarians across Britain, a pursuit which would have given few contemporaries any moral qualms

COMMENTARY:

33

957

at all, has been matched by, and co-ordinated with, the continuous intro-

duction of a more civilised way of life. seruztus has been shown to be not toiling in fields and mines and harbours, as the Caledonians claim (g1.2), but building porticoes and baths and dining rooms (21.2), which in turn develop out of A.'s wholesale Romanisation of the province (21.1—2nn.). Calgacus' accusations are represented as twenty vears out of date, more appropriate to the days of Prasutagus and Boudica than to the governorship of A. The seruztus depicted earlier in the work is utterly different from, and considerably more attractive than, the allegations of the Caledonians;

and there is a noble defiance in A.'s leaving unspoken and implicit the view that 'templa, fora, domos' constitute a preferable way of life for people whom he had found to be 'dispersi ac rudes eoque in bella faciles' (21.1). Acenturv earlier Virgil had put into the mouth of Anchises a memorable statement of Rome's global mission (Aen. 6.851—3), and A. can be forgiven for thinking that, in implementing that mission to the letter, he had nothing for which to apologise. iamque...ita disseruit: 'Some of the finest lines of the Aeneid characteristically open with the adverb zam, which, given the evocativeness of the poet, adds an exceptional depth and movement. The brilliant image of an army on the march, deploying on the vastness of a plain...' (C-L 497, quoting Aen. 9.25 'Ilamque omnis campis exercitus ibat apertis’). The 'zam d'ouverture' occurs very often in the Aeneid and is regularly followed by ΔῊ inverted cum-lause (e.g. tamque...cum...at 2.730—1), a combination which has an 'epic flavour' (Horsfall on Aen. 7.25). Since an inverted cumclause is used by T. here to introduce A.'s answering speech (‘cum Agricola...disseruit’), and since T. is himself fond of the zamque...cum combination and its equivalents (e.g. 26.1, 29.4, H. 2.61), it seems very likely (1) that zamque here 15 part of the same combination,

(ii) that there should

be only light punctuation after procursu (as after gestantes at 29.4), and (iii) that szmul (which links the two opening sentences) means 'and at the

same time', as often in T. (G-G 1503b): compare Curt. 4.2.21 'Iamque paulum moles aqua eminebat, et simul aggeris latitudo crescebat urbique admouebatur, cum Tyrii... . A.’s speech 15 thus given an appropriately epic build-up. On all these various narrative devices see C-L 513—17, 561—95. agmina et armorum fulgores: with these words (which constitute four hexameter feet) we need to understand apparebant or something similar (the imperfect tense is the most frequent before an inverted cum-clause: C-L 594). armorum fulgor (again at A. 1.68.4) is first at Hor. Odes 2.1.19 (his description of Pollio's history), then Liv. 22.28.8 and Curt. 4.13.2, but only here is it plural, ‘flashes’; taken with the accompanying genitive, it comes very close to meaning ‘flashing weapons' (21.2n. porticus; and see 26.1n. (fulsere) for the glitter of a Roman army). The columns, weapons

258

COMMENTARY:

33.2

and battle-line (acies) are those of the Caledonians, since the Romans are

evidently still within their fortifications (cf. uix munzmentis coercitum militem

below).

audentissimi cuiusque procursu 'as all the boldest rushed forwards'. The superlative form recurs only at Gell. 6.2.10. quamquam: 566 1.1n. For munzmentis coercitum cf. Plin. NH 12.5. accendendum adhuc ratus: adhuc— ‘still further' (OLD 7a). A. 15 taking a leaf out of Caesar's book (cf. C. 3.92.4 'est quaedam animi incitatio atque alacritas naturaliter innata omnibus, quae studio pugnae incenditur; hanc non reprimere, sed augere imperatores debent’). ita disseruit is exclusively Livian (8x) before T. (again at H. 1.83.1); later at Gell. 12.5.6. The expression, when contrasted with locutus fertur at 20.4, suggests that A.'s speech has ‘a greater ontological stature' than that of his enemy (so Laird (1999) 126). For the collective malitem see 16.gn. 33.2 Septimus annus est...ex quo...Britanniam uicistis: where Calgacus began with himself, A. leads with his men and (see below) the populus Romanus. A. cannot, however, mean

that his men

'have been

[sc. and

still are] conquering Britain', for which Latin requires the present tense (as Luc. 1.300 'decimo iam uincitis anno’); he must be referring to his first victory in AD 77 (18.5-6) and Britanniam uicistis must be equivalent to something like 2n Britannia uictores fuistis ('It is the seventh year since you were victorious in Britain’). Some comm. speak as though this were a conventional way to open a speech but, although the form of words is quite normal (e.g. Hor. Sat. 2.6.40—2 'septimus octauo propior iam fugerit annus | ex quo Maecenas me coepit habere suorum | in numero’), in fact the only two oratorical parallels are in T. himself (H. 1.29.2, A. 14.53.2). Perhaps Dr Johnson opens his famous address to Lord Chesterfield, his patron, with a contaminatio of Horace and T.: 'Seven vears, my lord, have now past since I waited in your outward rooms or was repulsed from your door...' commilitones: the use of this ‘comradely address' (H.) was associated esp. with the first Roman invader of Britain, Julius Caesar (Suet. D] 67.2 ‘nec milites eos pro contione sed blandiore nomine commilitones appellabat', Sen. Ben. 5.14.1—2), but it was eschewed by the more aloof Augustus (Suet. Aug. 25.1 with Louis' n.); see also Ash on H. 2.66.2. The hints that A. has been surpassing Caesar through imitation are reprised here (see Intro. p. 29). uirtute et auspiciis populi Romani... Britanniam uicistis: there are two difficulties with the transmitted zmperi Romani. (a) There is no parallel anywhere for the expr. uertus imperiz Romani, which several scholars have declared to be impossible here; they have inserted wuestra either before or after uzrtute. (b) Nor 15 there any parallel for the expr. auspicia impera Roman:. In public life (as opposed to contexts where the word is being used

COMMENTARY:

gg.2

259

in an extended or metaphorical sense) the auspices can be held only by a person (see the many exs. collected at TLL 2.1547.15-1548.28): during the republic this person would be a magistrate (thus Scipio in Liv. 21.40.93 ‘ille exercitus... mcis auspiciis rem gerit’), while under the empire it was the fprinceps (thus À. 2.40.1 'recepta signa cum Varo amissa ductu Germanici, auspiciis Tiberii'); see Oakley on Liv. 8.91.1. The only exception seems to be Luc. 9.22 'auspiciis raptus patriae' (of Cato). Both these difficulties can be avoided simultaneously if we assume that the problem lies with :mperit. In many ways the easiest hypothesis is that an abbreviated ‘imp.’ was intended to mean 'imperatoris' but instead was wrongly interpreted as 'imperii'; but it must be very doubtful whether T. would want À. to make a positive (or indeed any) reference to Domitian. Here it is assumed that a scribe has simply substituted one set phrase for another, as happens often, and has written 'the Roman empire' when he should have written 'the Roman people'. Both utrtus popult Romanti (as H. 4.73.1 ‘neque ego umquam...populi Romani uirtutem armis adfirmaui', Cic. Sest. 188, Liv. 3.61.5, 21.20.2, 20.41.12, Val. Max. 7.2.93) and auspicia populi Romani (e.g. Cic. Dom. 48 and esp. Liv. 30.14.8 'Syphax populi Romani auspiciis uictus captusque est') are regular phrases, although it is true that the latter, where the auspices are represented as being held by the people by whom the individual auspice-holder is invested with power, 15 restricted to republican texts. On the other hand, unless auspiciisis itself corrupt, our options are somewhat limited. The term cannot refer to A. himself, since he was fighting under Domitian's auspices, not his own. The only parallel for the combination of uzrtus and auspicia (-um) 15 Plaut. Epid. 381 'uirtute atque auspicio Epidici cum praeda in castra redeo'. fide atque opera nostra: because uicistis 15 second-person, these words too have caused problems: is nostra used for mea? or is it a genuine plural? or should we emend to uestra? The first of these can be dismissed: the general's loyalty (fides) is taken for granted and does not need stressing, while it seems arrogant and tactless to say that the soldiers' success was due to the general's effort (opera). On the whole, the second alternative seems the least problematical and is perhaps supported by me — uos and ego — uosin the next two sentences. Δ. 15 stressing both the army's mutual loyalty (nostra here almost — Σ εΥ̓ nos), thus countering immediately Calgacus’ prophecy of disaffection (32.1—3), and the combined hard work of himself and his men; but with uicistts he 15 made ‘to emphasize the achievement of the troops’ (O-R). For fides and opera elsewhere cf. Cic. Fam. 6.10a.1, 6.19.4, 19.09.1, Nep. Cimon 4.9, Liv. 44.19.14 and (but rather different) Val. Max. 9.8 ext. 4. tot expeditionibus, tot proeliis: abl. of time: Π so many campaigns, 50 many battles’. For the combination of nouns cf. Sall. H. 2.98.6, Liv. 37.53.18.

260

COMMENTARY:

33.3-4

neque me militum neque uos ducis paenituit: for the close intertwining of leader and soldiers cf. Liv. 6.12.11 'nec dux legiones nec fortuna fefellit ducem', 8.30.5 ‘non dux militi, non miles duci defuit' and see Keitel

(1987b) 74.

33.3 ergo egressi, ego..., uos...terminos: T. makes Seneca employ a similar division in his speech to Nero at A. 14.54.1: 'sed uterque mensuram impleuimus, et tu...et ego...'. terminos egredi (again at H. 4.51.2) is first at Cic. Quznct. 35, then ILS 212 (Claudius' speech) 2.30, Colum. 1.8.7, 11.1.23 (bis). Compare the significance which Velleius attaches to Tiberius' record-breaking operations in the North Sea in AD 5 (106.2-3), and, on ‘firsts’ in general, see 13.1n. (przmus). finem Britanniae ... tenemus: this statement may simply mean ‘our hold on Britain is nota matter of rumour but of hard fact’, but, noting that tenere has a similar range to the English 'to grasp', commentators prefer to see a mild syllepsis here: the verb is used of an intellectual grasp with fama and rumore (i.e. OLD 23) but of a more physical grasp with castrzs et armis (cf. OLD 5a, 9ga-b): ‘our grasp of the end of Britain depends not on report and rumour but on camps and arms' (these two elements are then picked up by znuenta and subacta respectively just below). That is: A. 15 making a similar point to that which

T. himself had

made

at 10.1, namely,

that

earlier assumptions about Britain (‘nondum comperta’) had been superseded by the genuine knowledge ('fide rerum’) which came with conquest (‘tum primum perdomita est’): for the relationship between knowledge and conquest see e.g. Schadee (2008) 164. It may seem a little strange that A. introduces an intellectual dimension into his pre-battle speech but his reference to fama and rumor serves to correct Calgacus' claim that Caledonia's very seclusion had protected it against fama (30.3 and n. nos). fama and rumor are fairly commonly combined; for arma and castra (again at G. 41.1) cf. e.g. Cic. Marc. 30, Mil. 74, Off. 2.84, Liv. 22.60.26, Sil. 2.286, Plin. Ep. 9.2.4. armis tenere (see H. on H. 2.17.1) is considerably more common than castris t. (Caes. C. 1.68.1, Liv. 32.11.2). The clausula (- - — x) is

Cicero's favourite (again at 34.1 ?nterrogate, 34.9 ederetis): see 15.5n. (deliberare) , 30.4n.

(soli).

inuenta Britannia et subacta: the brevity effectively means 'no sooner

discovered than tamed’: cf. 10.4 'inuenit domuitque'

(n.). subacta (cf. OLD

4a) hints at the animal imagery which will come to the fore below. 33-4 saepe in agmine 'often while on the march' (as Liv. 10.45.12 ‘saepe in agmine', Vell. 47.1). Quando animus ? ‘when do they intend to come?' (sc. eis est). The passage has been much emended; uen?zre (producing Cicero's favourite clausula) could easily have been omitted before the following ueniunt, for which it forms an effective preparation. anzmus4- inf. 15 mostly poetic (OLD *b) but cf. Suet. D/82.4 ‘fuerat animus...trahere'. With the

COMMENTARY:

33.5

261

soldiers' quoted talk compare the more extended direct speech at Liv. 0.3.1—3, esp. 'cum pro ingenio quisque fremerent “per obices uiarum", alius “per aduersa montium, per siluas, qua ferri arma poterunt, eamus; modo ad hostem peruenire liceat..."'. e latebris suis extrusi: for the wild animal imagery see esp. Curt. 9.3.8 'inter feras serpentesque degentes eruere ex latebris...expetis’. Barbarians, esp. northern, are often compared to beasts (W-M on A. 3.47.4, Oakley on Liv. 7.10.3): cf. 34.2—3 and nn. uota uirtusque in aperto: sc. sunt or erunt; A.’s meaning is that there is now, or soon will be, no obstacle to our prayers and prowess; in aperto not only contrasts with the lairs from which the beasts have been forced out (above), but also introduces the travel imagery of prona... aduersa below: see 1.2 ‘pronum magisque in aperto' and n. For the motif cf. Virg. Aen. 10.279 'quod uotis optastis adest, perfringere dextra' (where Harrison quotes Hom. Il 16.207-8), Liv. 34.13.5 'tempus...quod saepe optastis uenit, quo uobis potestas fieret uirtutem uestram ostendendi', Curt. 9.9.4 'adesse finem laboris omnibus uotis expetitum: iam nihil gloriae deesse, nihil obstare uirtuti', Luc. 7.251—2 'adest totiens optatae copia pugnae. | nil opus est uotis’; for uota — wuirtus elsewhere cf. Liv. 22.5.2 'nec enim inde uotis aut inploratione deum sed ui ac uirtute euadendum esse’, Vell. 45.3 *uotisque Italiae ac decretis senatus, uirtute atque actione Annii Milonis...dignitati patriaeque restitutus est’. omniaque prona uictoribus atque eadem uicüis aduersa: again sc. sunt or erunt. In view of what precedes (‘ueniunt...in aperto’),itlooksat first sight as though wictoribus refers to the Romans and wictis to the Caledonians; but the next sentence explains ο

aduersa (cf. nam) and clearly refers to the

Romans and to their plight if thev are beaten. Alternatively both uwuzctoribus and uictis could be taken as entirely general references to ‘conquerors’ and 'conquered'. The travel imagerv of in aperto (above) 15 continued by the chiastic prona — aduersa (cf. Virg. G. 1.201—3 'aduerso ...flumine' — 'prono...amni', Plin. NH 2.179 Π alia aduerso, in alia prono mari’); contrast Caes. G. 2.28.1 'uictoribus nihil impeditum, uictis nihil tutum'. omnia ...prona uictoribus (again at H. 3.64.1) may perhaps derive from Sall. . 114.2 ‘omnia uirtuti suae prona esse' (see further nn. below), but cf. Curt. 6.3.16 ‘non mare illud, quod exaestuans iter fluctibus occupat, euntes nos moratur ...; plana omnia et prona sunt', Plin. Pan. 16.5 'omnia haec tam prona tamque cedentia uirtutibus tuis' (Val. Fl. 1.6go-1 15 perhaps slightly different). 33.5 superasse tantum itineris...: compare Hannibal's speech to his men at Liv. 21.43.9 'tantum itineris per tot montes fluminaque et tot armatas gentes emensos'. superare 15 the mot juste for overcoming difficult journeys and the like (OLD 1a-b, 5a), but 115 combination with Ζ 6 seems strangely unparalleled. For pulchrum ac decorum cf. Flor. 1.47.4.

969

COMMENTARY:

33.6

in frontem: a participle such as progredientibus has to be understood from the following fugientibus: 'although to have prevailed over such a journey, to have negotiated woods and crossed estuaries is fine and splendid forwards, nevertheless the most perilous things for fugitives are those which today are most propitious'. The ellipse is harsh; ShawSmith (1979) proposed zn frontem « nitenti- , ita .. . ita is Rhenanus' emendation of the transmitted ztem. Although ut...2tem is a recognised correlation (TLL 7.2.533.7-17), it does not have the concessive sense of ‘although ...nevertheless' which 15 commonly found with ut...ita (OLD ut 5b). item (nowhere else in T.) was almost certainly written

under the influence of the immediately preceding frontem. neque enim nobis...eadem abundantia: sc. erit quae hostibus (‘we shall not have the same...as the enemy’). 'The speaker here admits what Calgacus

had

asserted'

at 32.2

(Gudeman),

vet the admission

(in effect a

threat, like Hannibal's at Liv. 21.44.8) involves a slight illogicality: if the Romans were to retrace their footsteps southwards, as A. implies (above), neque...locorum eadem notitia would not strictly apply. locorum... notitia, again at H. §.59.2, 1s 4x in the Caesarian corpus, though also at Liv. 4.19.6, Plin. NH 5.38, Sen. Ep. 71.10, Quint. 6.4.8. The parallel arrangement of A.'s phrases contrasts with the quasi-chiastic structure of the previous sentence. sed manus et arma et in his omnia: there is perhaps a 'slide' in the argument here, since, although the words contrast with meque enim . .. abundantia above, it is more

natural

(but not, admittedly,

essen-

tial) to supply sunt rather than a future tense here. By stripping the soldiers down to their essentials - weapons and the hands that wield them — A. offers them an image of self-sufficiency to counter the exhaustion and potential adversity of the landscape and bodily needs. See further Phang (2008) 105—7 on armour and the body. The sentence may be another Sallustian allusion ( 51.4 ‘in armis omnia sita’). For manus et arma see 25.3n. 33.6 iam pridem mihi decretum est neque exercitus neque ducis terga tuta esse: the prospect of battle is terrifying and some soldiers will naturally want to flee instead of fight; A. endeavours to counteract this desire by giving them the reassuring benefit of his extensive experience: ‘I decided long ago that flight is unsafe for both an army and its leader'. This is his version of the point made by Calgacus at 30.1 (n. ?ta). terga tutais otherwise poetical (Ov. Met. 5.161, Luc. 3.49-509); the clausula is Cicero's second favourite (again at 34.3 deprehensi sunt and causas rebellandi): see 30.1n. (?mminente).

proinde ut honesta mors turpi uita eodem loco sita sunt: the transmitted problematic. et honesta mors turpi uita preceded; it is et zncolumitas ac decus

potior, ita incolumitas ac decus text reads proinde et...et...but is potior is illogical after what has eodem loco sita sunt which 15 the

COMMENTARY:

g34.1

263

logical sequel to A.'s argument. Sleeman suggested translating et...et as "while ...yet', thereby making the first statement subordinate to the second; but it must be very doubtful whether this meaning is possible. Nipperdey transposed the two statements, but this seems excessively radical, esp. since he felt obliged also to delete the first et. There 15 in fact no easy solution to the problem, but perhaps it is worth suggesting that et...et be emended to ut...?ta. On this suggestion proinde will not mean 'accordingly' or 'therefore', as it does at 32.4 (n.), but will have its comparative sense

(‘in the same

way

(as)’:

OLD

1-2),

of which

T. is fond

(Holmes

(1997/98) 67—9; see also 10.5n. ne uentis quidem). Thus: ‘just as an honourable death is preferable to a disgraceful life, so safety and glory are situated side by side’. honesta mors turpi uita potior: an honourable death was better than a disgraceful 6 (Nep. Chabr. 4.3 ‘praestare honestam mortem...turpi uitae’) and a disgraceful life could be redeemed by an honourable death (Cic. Quiznct. 49 ‘mors honesta saepe uitam quoque turpem exornat'; W. on Vell. 87.1): the most famous example of such antitheses was Achilles, who preferred a glorious death to a humdrum life (Hom. 7l g.410-16; for another Homeric motif see 45.4n. excepissemus). These and similar sentiments are proverbial (see Oakley on Liv. 9.3.3), but the language specifically recalls (and reverses) Sall.J. 67.3 ‘turpis uita...potior'. incolumitas ac decus eodem loco sita sunt: since A. has argued that flight is dangerous, the conclusion must be that it is safer to fight, which is also the nobler option. nec inglorium fuerit...cecidisse ‘nor would it be inglorious to fall...': almost as an afterthought A. adds that, if the worst came to the worst and (despite his arguments) soldiers had to die, it would be glorious to have died at the world's end. The potential subjunctive (10.6n. addiderim) is greatly preferable to the bluntness of a fut. perfect here. terrarum ac naturae, which picks up and generalises finem Britanniae (3), is a further ex. of concrete and abstract (20.2n. laudare); again (cf. 30.3—4nn.) the talk of boundaries invokes the exploits of Alexander the Great (e.g. Sen. Suas. 1.1, 1.3), though naturae finis is common in other contexts too. The clausula is that which Cicero made

notorious with his esse uideatur (cf. D.

23.1 with Maver's n.). 34.1 constitisset ‘had taken up position’, a technical meaning (OLD consisto 6b). The tense realistically supposes that the Caledonians have now completed the movements described at 33.1 'iamque...acies'. For nouae . .. atque ignota cf. Cic. Rep. 1.1, 1.25, Sen. Ag. 269. hortarer: speeches often include an explicit reference to their formal category (15.5n. deliberare); for some other exs. in a hortatio cf. Sall. C. 20.15, Liv. 21.40.2. nunc ‘, asitis’ (OLD nunc11a).

264

COMMENTARY:

34.2

uestra decóra recensete: the noun almost certainly has the more abstract (and rarer) meaning of 'achievements' here (OLD 2b), picking up decus at 35.6, rather than 'decorations' (as at 29.4): although dona could be awarded immediately after a battle as well as at the completion of a campaign (Plb. 6.39.2; Maxfield (1981) 132-3), it would have been invidious and counter-productive to single out award-holders. uestros oculos interrogate: whereas Calgacus was gratified that his men's eyes had had no contact with Roman oppression (30.2), A. urges his men now to fix their eves on the Caledonians. His expression, which is unparalleled, seems to envisage the eves as witnesses in court. hi sunt quos...clamore debellastis: the reference is to 26.1—2, esp. 'ab uniuersis adici clamorem'. For furto noctis, 'night-time stealth' (OLD furtum 2a), cf. Curt. 4.13.9, Sen. Ag. 626, Sil. 6.70-1; winning with '(only) a shout' (as again at e.g. Cic. Caec. 43 'saepe clamore ipso militum aut instructione aspectuque signorum magnas copias pulsas esse et uidimus et audiuimus’) has its roots in Homeric epic (see Lendon (2001) 170-1): the theme is resumed with sono (2) below. For the hortatory topos ‘we are

fighting the same enemy as before' see e.g. Liv. 21.40.5, Dio 62.11.3; W-M on A. 3.45.2, K. and Oakley on Liv. 6.7.6. The point is that ‘quae ex usu nota sunt non timentur' (Veg. 3.12.7). hi...fugacissimi ideoque tam diu superstites: a similar point 15 made by Arminius at A. 2.15.1 ‘hos esse Romanos Variani exercitus fugacissimos'. 34.2 quomodo siluas saltusque penetrantibus fortissimum quodque animal contra ruere, pauida et inertia ipso agminis sono pelli «pro-bantur: the transmitted text reads pellebantur and for this reason editors have been concerned to explain that the simile cannot be general but is ‘intended to refer to the campaigns which thev had gone through' (O-R); nobis or uobis is to be understood with penetrantibus, and the only remaining question is whether ruere is historic infinitive (H., Soverini)

or the alternative form

of ruerunt

(O-R).

It follows from

this

explanation, however, that the Britons are being compared to themselves ('sic acerrimi Britannorum...’), which seems, if not quite impossible, at least highly improbable. Rather, the simile seems clearly to be generalised, like that in the Iliad to which it 15 said to allude (21.573ff., cf. Curt. 3.8.10), and the text requires emendation. Puteolanus' pelluntur leaves undisturbed the question of ruere, Wex's pelli solent produces both the right sense and an auxiliary verb on which ruere can depend, but seems rather far from the paradosis. A simpler emendation would be pelli probantur (an abbreviated pro could easily have been omitted in error): ‘Just as it is proved repeatedly that all the bravest animals attack those penetrating their woods and denes, while the frightened weaklings cower at the very sound of the column...’. For the construction cf. e.g. À. 15.51.4 ‘suspectante Nerone haud falsa esse etiam quae uera non probabantur' (and

COMMENTARY:

34.3

265

note 15.20.9 'usu probatum est...leges egregias, exempla honesta...ex delictis aliorum gigni'; G-G 11972), Ov. Met. 2.92 ‘patrio pater esse metu probor', Vitr. 7.9.5 'sic refrigeratum 51 restituatur in pristinum colorem, sine uitio esse probabitur'. agmen 15 the proper term for a ‘body of people moving or acting together' (OLD 4a), such as hunters (Ov. Hal. 53—4 ‘impiger ecce leo uenantum sternere pergit | agmina’), but suggests also the column of troops with whom the hunters are implicitly compared; such

transpositions

are

common

in similes

(many

further

exs. below).

For the two categories of hunted animals, brave and timid (lion, bear and boar ~ hares and deer), see Ov. Hal. 49-65, where the point, as here, 15 that all the bravest ones invite their own death by their aggression. For the notion that Domitian's reign had recently given an impetus to the visual depiction of hunting in art see Tuck (2005). siluas saltusque penetrantibus: the combination of nouns (again at A. 2.14.2) is found often but appealed especially to poets. For penetrare cf. Plin. NH 8.50, Frontin. Strat. 1.2.2 (both of woods); Sil. 2.442 saltus. contra ruere: cf. Val. Fl. 3.706 'aspera nunc pauidos contra ruit agna leones?' (an adynaton); H. aptly compares Ov. Met. 8.343—4 ‘ruit spargitque canes, ut quisque ruenti | obstat’ (of a charging boar); cf. also Hal. 60-1 ‘actus aper ...| 56 ruit'. pauida et inertia: the only parallel for this combination seems to be a controversial line of Horace (AP 172 'iners,

auidusque futuri’); for iners — 'spiritless, cowardly' see e.g. Virg. Aen. 4.158 'pecora inter inertia' (with DServius), Sen. Ep. 55.5 ‘qui uelut timidum atque iners animal metu oblituit’; OLD 3. acerrimi: the adj. has a wide range of meanings ('brave, enthusiastic, fierce’); esp. relevant is its application to wild animals (OLD ob). It was a topos that all the bravest men die in battle (see Oakley on Liv. 9.32.8 'anceps proelium...fortisimum quemque absumit’). reliquus est numerus ignauorum et timentium: numerus 15 said to have its contemptuous meaning of ‘mob’ here (G-G g8gb, OLD 8b); reliquus perhaps has a similar connotation (cf. Liv. 21.40.10 'reliquias extremas hostis, non hostem habetis’). ignauus, regular of cowardlv persons (25.3; OLD 3), can also be used of animals (e.g. H. 3.36.1, Hor. Epod. 6.2). 34.3 quos quod tandem inuenistis ‘and, as to the fact that [OLD quod6a] you have tracked them down at last' (quos quod — et quod eos); like deprehens: below

(OLD

1b), znuen?stis maintains the animal language

(cf. e.g. Lucr.

4.99%, Gratt. Cyn. 205—6 'premit inuentas, non inuentura latentes | illa feras’): discovery of their prey's habitat (33.3 'inuenta Britannia’, cf. 23) has now brought A.’s men face to face with the prev itself. nouissimae

res

'their

extreme

circumstances’

(OLD

nouissimus

8),

responding to necessitatem in the introduction to Calgacus' speech (30.1),

266

COMMENTARY:

355.1

where causas belli and hodiernum diem are answered by magnum diem and causas rebellandi below. Cf. Liv. 21.40.6 ‘nec nunc illi quia audent sed quia necesse est pugnaturi sunt'. extremo metu: causal abl.: ‘paralysis on account of their extreme fear'. in quibus pulchram et spectabilem uictoriam ederetis: now that the wild beasts have been stopped in their tracks (OLD uestigium 1a), the hunt 15 replaced by the arena (~ 37.2): spectaculum edereis the regular expression for putting on a show in the arena (G-G 334b) and pulchrum is a regular word for describing such a show (e.g. Cic. Verr. 5.77, Plin. Pan. 34.1). The arena was of course the focal point of an amphitheatre, in which tiered seating would rise to a considerable height; in fact it is the Britons whose arrangement on a hillside resembles the rows of spectators (35.3 below): is À. suggesting that they will be witnessing their own destruction? ederetis is purposive. transigite cum expeditionibus ‘have done with campaigns!' (OLD tran52g0 3C): the first element in A.'s concluding tricolon crescendo looks back to the beginning of his speech, esp. 33.2 'tot expeditionibus'. imponite quadraginta annis magnum diem 'crown these forty years with a great dav' (lit. ‘place a great day upon...’): the expr. is analogous to ultimam manum imponere or finem imponere (OLD impono 13a) and is A.'s response to Calgacus' ‘this day' topos (30.1n. hodiernum). As at the start of his speech, A. adopts an historical perspective, here longer than the seven years mentioned there (33.2). The paradosis reads quznquaginta but AD 83 15 40 years from Claudius' invasion in 43 (13.3) and A. has no reason to exaggerate. Numerals are easily corrupted in transmission. adprobate ...causas rebellandi ‘prove to the commonwealth that neither delaying the war nor reasons to rebel could ever have been imputed to the army.” O-R suggest that moras belli is an allusion to the lack of initiative shown by Trebellius and Bolanus (16.3—5) and causas rebellandi to Paulinus' severity (16.2); but this seems contradicted by their prior assertion that exercituzis ‘emphatic’: the point 15 not that A. himself but that his army should not show weakness (‘fight now and make sure that our victory is so complete that the enemy is never tempted to resist us again’). The tense of potuisseis explained by the fact that, by the time news of the engagement is received and assessed in Rome (cf. 39.1), the battle will be long over. Word-play of the type belli — rebellandi is a particular feature of Livy (e.g. 7.27.7, 34.13.0). 35.1 alacritas: see 33.1n. (Excepere). ad arma discursum is above all a Livian expression (6x) but also at Sen. Suas. 6.5, Curt. 9.7.8, Sen. Ira §.2.3, Brev. vit. 3.1. The motif of rushing to arms 15 as old as Homer (Od. 24.466).

COMMENTARY:

35.2

267

35.2-36.2 The battle The battle itself is constructed on a 'ring' principle, with two outer sections of introduction and conclusion (35.2-4 Deployment — 38.1—-2 Aftermath) framing three inner sections of fighting which pivot around the short,

central,

section

(36 The

first attack, 37.1-2

Introduction

of the

alae, 37.3—6 The climax). In each of these three sections À. 15 named and, appropriately for his role as the ideal general, intervenes decisively to bring about a feripeteia or dramatic change in the action (36.2 ‘donec Agricola...', 37.1 ‘ni...Agricola...’, 37.4 ‘ni...Agricola...’). The first section ends with terrified, riderless horses symbolising the plight of the Caledonians (36.3); the third section ends with a defiant horse failing to rescue his young Roman master from danger, a vivid image of the cost of victory (37.6). The narrative is a sustained and classic example of dramatic historiography. Vividness or enargeia is achieved by repeated appeals to sight (36.9 ‘pugnae facies’, 37.2 'grande et atrox spectaculum', 38.2 'faciem uictoriae’), and realism is suggested by the constant resort to technical military language (e.g. 36.2n. erigere) and to many of the topoi which in other military narratives characterise fighting (e.g. 37.4n. et sicubi [ad fin.]) and the enemy (e.g. 36.1n. ingentibus). Throughout the engagement, as in the biography as a whole, the qualities of the ideal general are attributed to A. himself (35.2n. dzsposuit, 37.1n. ad subita). More or less marked references to Sallust (e.g. 35.3, 36.3) and Liw (e.g. 36.3, 38.1) are combined with particular allusion to Sallust's description of the battle of Cirta (37.2n. Tum), and it has also been suggested that the whole battle is modelled on Caesar's account of Pharsalus (Intro. p. 29 n. 97). In addition there are allusions to the epic poetry of Virgil (37.3nn. /acer? and etzam) and Silius (37.5n. longinqua), as well as an actual quotation from Lucan (38.2n. zncerta). Echoes of the leaders' pre-battle speeches provide a grim context for the victory of the Romans and the defeat of the Britons (35.93n. ?n speciem, 36.1n. primo, 37.4n. ualidas) , and itis especially dramatic that the hunting and slaughter of wild animals, foretold by A. in his hortatio (34.2—3), are realised at the climax of the battle (37.2 ‘spectaculum’, 37.4 ‘indaginis modo’). Nothing could be more human, however, than T.’s picture of the aftermath, where

a remarkable series of historic infinitives itemises the miseries of the conquered (38.1n. Britann?) and in an eerie silence smoke is seen rising from the abandoned homesteads (38.2n. fumantia). 35.2—4 Deployment 35.2 Instinctos ruentesque: for this combination of emotion and reaction cf. Liv. 4.33.2 'ignibus armata ingens multitudo ... uelut fanatico instincta

268

COMMENTARY:

furore cursu in under 'To rush swoop' in OLD likely meanings

35.2

hostem ruit' (of the people of Fidenae). or descend (upon) in hostile or aggressive (ruo 5), but this seems inconsistent with are ‘To move swiftly or impetuously' (1a)

with haste or impatience'

(2). For instinctos see 16.1n.

ruentes is listed fashion, charge, dZsposuit, more or “To proceed

(His).

disposuit: the appropriate deployment of troops was naturally a mark of the ideal general (see e.g. Frontin. Strat. 2.5; Oaklev on Liv. 9.17.15) and it is noteworthy that A. can execute his reponsibility in their current state of inspired excitement. ut peditum auxilia...mediam aciem firmarent: auxzla = ‘auxiliary contingents’ (OLD auxilium 5); octo milium is a genitive of description (NLS §84). The 8ooo auxiliaries will have included the four Batavian and two Tungrian cohorts mentioned below (36.1); whether any of these cohorts was already double strength (milaria: see 28.1n. cohors) by this time 15 unknown.

mediam aciem firmare, ‘to make

a strong middle

of the line’, 15

Livian (21.56.2, 22.46.3, 23.29.4, 27.14.4). equitum tria milia cornibus adfunderentur: these mounted troops are evidently different from the four alee introduced into the battle later (37.1). The use of adfundi— 'to be spread' is not common but evidently liked by Seneca (cf. OLD 5); cornibus is presumably abl. of place ‘where’ (NLS 853 Note); cf. Sen. NQ 1 praef. 10 ‘equitem...a lateribus adfusum' (‘cavalry...spread out on the flanks’). legiones pro uallo stetere: ‘in front of the rampart’ (as Liv. 37.38.8 'pro uallo stetere’; OLD pro 1a), not 'along' (for which see G. on A. 2.13.4): the reference is to the Roman marching camp, previously constructed. We do not know which of A.'s four legions (II Augusta, II Adiutrix, IX Hispana, XX Valeria Victrix: see Appendix 1) were present. ingens uictoriae decus...si pellerentur: the appositional expression, difficult in construction, is also difficult in meaning: ' «intended to be> a great distinction in victory, of fighting without Roman bloodshed, and a support, if «the auxiliaries> were being beaten back'. decus is followed by a defining genitive (though a gerund seems unparalleled), which makes it likely that (despite H. 2.78.2 'Iudaicae uictoriae decus', Val. Max. 1.5.1 'inclitae...uictoriae decus’, Flor. 1.45.26 'maximum uictoriae decus') uctoriae is dative rather than another genitive: either way, in English decus has to be taken twice. auxiltum is parallel to decus but is followed bv a conditional clause instead of a noun: for this type of uariatio see A. 4.51.2 ‘his partae uictoriae spes et, si cedant, insignitius flagitium', 6.34.3 ‘plus decoris uictores aut, 51 terga darent, flagitii atque periculi laturos', 12.20.1 *modicam uictoribus laudem ac multum infamiae, si pellerentur’; Sorbom (1935) 118 (who takes our uictoriae as gen.). For citra = ‘without’ see 1.9n. (nec 2d); c. sanguinem recurs at Sen. Clem. 1.25.1, Ep. 108.18, Liv. per. 126 and cf. Luc. 6.211

COMMENTARY: g5.3

269

‘citra...cruorem’. ingens. .. decusis originally Livian (1.99.3, 5.44.93), then Sen. Med. 226. O-R maintained that withholding the legions was a ‘new method' of fighüng which 15 illustrated on Trajan’s Column and which T. here and at H. 5.16.1 is 'the first literary author to describe'. It was, however, argued by Rainbird (1969) and more lengthily by Gilliver (1996) that A. at Mons Graupius was simply adopting regular tactics, auxiliaries being more versatile than the heavily armed legions and better suited to hghting on unfavourable terrain; by referring not to these reasons but to the preservation of citizen lives T. enhances the ease and nature of A.'s victory. Certainly at Pharsalus in 48 Bc Pompey 15 alleged to have promised his men that 'sine periculo legionum et paene sine uulnere bellum conficiemus' (Caes. C. 3.86.4); at Idisiaviso in AD 16 the legions were not in the front line (Tac. A. 2.16.9), there is no reference to their

actually fighting, and the result was ‘magna...uictoria neque cruenta nobis' (2.18.1); in 28 against the Frisii a legion had to be summoned while the battle was in progress (A. 4.73.2); and in 60 'Corbulo inmissis Hiberis . .. hostilem...audaciam externo sanguine ultus est' (A. 14.23.3). See also, for the ‘bloodless’ motif elsewhere, Caes. G. 7.20.12 'sine uestro sanguine' (Vercingetorix 15 speaking), Tac. A. 3.39.2 'sine nostro sanguine', 12.17.2. 35.3 in speciem simul ac terrorem: 1515 purposive (OLD 21a) and the words (again at A. 2.6.2 'in speciem ac terrorem’) are almost a hendiadys (‘to create a terrifying sight’). 'Observe that what Calgacus had warned against and ridiculed...is here asserted to have been resorted to by the Caledonians themselves'

(Gudeman, with reference to 32.9); but see also

next. n. editioribus locis: the Caledonians are not all show (last n.): they have also taken the position of advantage (cf. 18.2n. non). editusis frequent with locus, esp. in military narrative, but the comparative form is a speciality of Sallust (/. 58.3, 98.7, H. 1.140; also Caes. C. 1.7.5). in aequo: sc. essef. ceteri per adcliue iugum conexi uelut insurgerent: ?nsurgere, already choice when applied to ‘motionless things' such as a wood (A. 2.16.1 ‘insurgebat silua’ and G.; OLD g), is here applied to an army in a transferred usage (hence the extenuating uelut); per 2 ‘all over’ (OLD 4a) and is to be taken with both conext (presumably a reference to rows or ranks) and ?nsurgerent: ‘while the others soared (as it were) up the sloping ridge across which they were linked'. For his description of the hillside T. has perhaps remembered Ov. F. 5.154 ‘leniter adcliui constituere iugo' (leniter adcliuis 15 1tself Caesarian: see G. 2.29.9, 7.19.1). media campi: i.e. the area between the Caledonians and Romans. For the neut. plur. + gen. see 6.4n. (/udos).

270

COMMENTARY:

35.4

couinnarius eques: 'a strange circumlocution’ (Sleeman), if is right (Rhenanus inserted ef). Presumably couinnarius is here val (‘the charioted cavalry’) but a noun at 36.93 below: compare of c.g. sagittarius, likewise adjectival (c.g. Liv. 37.38.9 'sagittarii OLD sagittartus!

1) and substantival

(OLD sagittarius®

the text adjectithe use equites;

1); equesis of course

a collective singular (25.1n. pedes). The mid-first-century geographer Mela explains how the Britons fight (3.52): 'dimicant...curribus Gallice armatis: couinnos uocant, quorum falcatis axibus utuntur' (cf. Sil. 17.416—17 ‘caerulus haud aliter, cum dimicat, incola Thules | agmina falcigero circumuenit arta couinno). Also used allegedly by the Belgae (Luc. 1.426), these scythed chariots are elsewhere called quadngae falcatae they were said to have originated in Asia Minor and are attested most recently as used against Julius Caesar by Pharnaces at the battle of Zela in 47 BC ( Bell. Alex. 75.2). Roman generals were supposed to be able to deal with them (caltrops, tribulti, are mentioned by Veg. 3.24.1—4; see the replicas illustrated in Davies and Jones (2006) colour fig. 54). See also 12.1n. (et curru).

strepitu ac discursu complebat: the two nouns are combined elsewhere only in a letter of Pliny which exhibits other phraseological similarities to T.'s oeuvre (Ef. 1.9.7 'tu quoque strepitum istum inanemque discursum...relinque'): see CCT 34. When describing the Britons' method of fighting from chariots Caesar specifies both their movement and the noise of the wheels (G. 4.393.1 ‘genus hoc est ex essedis pugnae: primo per omnes partes perequitant et tela coniciunt atque ipso terrore equorum et strepitu rotarum ordines plerumque perturbant). For strepitu . . . complebat cf. Liv. 24.21.9 ‘omnia...strepitu...complentur’; for discursu c. cf. Apul. Met. 11.7.2 (Plin. NH 14.13 15 slightly different). In view of the parallels (Lucr. 2.323-4 'praeterea magnae legiones cum loca cursu |camporum complent belli simulacra cientes’, Virg. Aen. 7.641—3 'pandite nunc Helicona, deae, cantusque mouete, [qui bello exciti reges, quae quemque secutae | complerint campos acies’; cf. Liv 25.12.6) itistempting tosuggest that campos complere 15 originally Ennian. 35.4 tum Agricola.. .constitit: for the arrangement of this sentence 566 Intro. pp. 30-1. Its orderliness matches A.’s military dispositions. diductis ordinibus: the same phrase is used of Cleandrias and Hannibal by Frontinus (Strat. 2.3.12, 2.6.4); 1t 15 strikingly absent from Caesar and Livy. porrectior: comparative adjs. can denote ‘a considerable or excessive degree of a quality’ (A-G $291a, cf. G-L $297.2): here 'rather extended’ is perhaps more likely than 'too extended'. The combination with acies 15 regular but infrequent in military narratives (Sall. /. 52.6, Bell. Afr. 17.1, Liv. 25.21.6).

COMMENTARY:

36.1

27]

arcessendas plerique legiones admonebant: sc. esse. The independent nature of A.'s generalship was likewise underlined at the start and in the centre of the British narrative (cf. 18.2n. plerisque, 25.9 ‘ignaui specie prudentium admoncbant' and n.).

promptior in spem et firmus aduersis: it is not clear whether aduersis 15 dat. (TLL 6.1.817.39 and evidently Gudeman ad loc.), abl. (O-R) or even ‘a form of abbreviated ablative absolute T. gladly affects, for brevity, variety, or otherwise' (G. on Λ. 2.14.3, comparing e.g. H. 1.10.1 ‘uir secundis aduersisque iuxta famosus ; 566 also above, 18.6n. aesttmantibus). For simi-

lar uariatio of abl. or dat. with prepositional phrases see 6.1n. (per mutuam) and 6.4n. (ut) respectively. pedes ante uexilla constitit: the banners are those of the auxiliaries (2 above). pedes 1s a noun used adjectivally: it is usually interpreted here as meaning 'on foot’ (OLD 1), but, since the ideal general also fulfilled the role of the soldier (W. on Vell. 85.5), there 15 perhaps also the connotation of ‘as a foot-soldier’

(cf. OLD 2a). Others of whom

this 15 said included

Catiline (Sall. C. 59.1 ‘ipse pedes’) and Caesar (G. 1.25.1, cf. Plut. Caes. 18.3). According to Onasander, however, the ideal general should not fight in person but should ride around on horseback to encourage and direct as appropriate (393.1, 6): see Oakley on Liv. 6.8.1-4 with Addenda (4.515). Yet A. nevertheless manages to play this role too (37.4 'frequens ubique’). 36 The first attack 36.1 primo congressu: this is the moment when Calgacus had wanted his men to show their mettle (31.4 'primo statim congressu' and n.). simulque...uitare uel excutere: the structure of the sentence has cncouraged some comm. to understand the two ablatives nearer the verbs as instrumental (gladüs...caetris) and to take the two verbs chiastically with them: 'to avoid our men's missiles with their shields and to shake them off with their huge swords'. But excutere is more naturally applied to shields (Virg. Aen. 10.777 'illa [1.e. hasta] uolans clipeo est excussa’, cf. Veg. 1.4.8 'tractare scutum et obliquis ictibus uenientia tela deflectere’; OLD excutio 1b), which leaves gladits and uitare high and dry. It 15 perhaps better, therefore, to take the infinitives chiastically with constantia and arte, and to understand the two other ablatives as descriptive: ‘with their huge swords and short shields, the Britons avoided our men's missiles skilfully or shook them ΟΗ resolutely'. For simul...simul see 25.1n. (cum).

ingentibus gladiis: it was conventional to describe barbarians as using huge (but ineffective) swords to match their huge (but vulnerable)

272

COMMENTARY:

36.1

bodies (Oakley on Liv. 7.10.5-9). In fact Roman swords were themselves getting longer in the late first century AD, as evidenced by a collection found at Newstead in Scotland, dated to the Agricolan and immediately post-Agricolan years (James (2011) 151). breuibus caetris: small leather shields; ‘a useful ethnographic ornament' (Horsfall on Virg. Aen. 7.732, q.v.). magnam uim telorum superfundere: us telorum, with or without magna (or zngens), 15 fairly common in battle descriptions but 15 usually combined with effundere (Liv. 27.18.11, Curt. 8.10.31 'ingens uis telorum in propugnatores effusa est’, Sil. 15.764—5 'tum uero effunditur ingens | telorum uis’); here, however, the Britons are throwing from higher ground

(35-3)-

donec Agricola...cohortes...cohortatus est: A.’s first decisive intervention in the battle is signalled by an ex. of ‘le donec de rupture’, which, introduced into historical narrative by Livy, becomes a favourite device of T.; both the use of preceding historic infinitives and the considerable build-up are regular: see C-L 617-40, esp. 635. cohors and cohortari were thought to be etymologically connected (Maltby 139): itis as if the Batavi and Tungri are being urged to live up to their units (the same assonance recurs at A. 3.20.1—2). Batauorum:

the Batavi came from the modern

Netherlands, where the

name survives as the area known as Betuwe. Thev are mentioned frequently by T. (see e.g. G. 29.1 and Rives' n., A. 2.8.3 and G.'s n.) and were famous for their aquatic skills (cf. 18.4n. quibus). In AD 69 eight Batavian cohorts on the Rhine under their leader C. Julius Civilis took advantage of the ongoing civil war to revolt against Rome; after their defeat in 70 they were

reconstituted

as I, II, III and

IX

(or VIIII)

and

it is assumed

that

these are the quattuor Batauorum cohortes to which T. here refers. Though all four cohorts seem to have been withdrawn from Britain over the next few years, evidence that shortly after Mons Graupius Cohors VIIII Batauorum was stationed at Vindolanda, just south of the future Hadrian's Wall, is to be found

in the famous writing tablets: see Birlev

(2002)

27

(plate

17), 41—5, 62-9 (esp. 65), 188 (index s.v. cohorts); also Hassall (1970) 135 and James (2011) 145-7. Tungrorum: the Tungri originated from around modern Tongres/ Tongeren in eastern Belgium (see Rives on G. 2.3). Shortly after the battle of Mons Graupius (c. AD 85-92) the Cohors I Tungrorum under the command of Julius Verecundus was stationed at Vindolanda, where one of the writing tablets lists - most evocatively - the strength of the cohort as it was on 18 May of the vear in question (7ab. Vind. 2.154 — IRB 273);a funerary inscription probably of early Hadrianic date records the death of a centurion serving with the same cohort (Birley (2002) 74 (plate 50) — IRB 280), while numerous inscriptions record the two Tungrian cohorts in

COMMENTARY:

36.2

273

northern Britain during the second and third centuries: see Birlev (2002) 41—2, 45, 60-2, 70-6, 188 (index s.v. cohorts). ut rem ad mucrones ac manus adducerent: perhaps surprisingly this alliterative combination is found elsewhere only at Petron. Sat. 80.4 ‘conuertite huc manus, imprimite mucrones'. quod...ipsis uetustate militiae exercitatum ‘which was familiar to them from their long experience of soldiering’: the general sense is more mundane than the details, since the meaning of uetustasseems uncommon outside T. (OLD 4b, cf. G-G 17632) and this ex. of exercitare/-tatus 15 not registered under any of the meanings in OLD s.vv. or TLL 5.2.1387.8—13809.60. quod refers to fighting at close quarters, i.e. rem... adducerent. hostibus...gerentibus: the words parua...gerentibus were deleted by Wex as an unwanted repetition of zngentibus gladiis et breuibus caetris at 1 above; Nipperdey proposed extending the deletion to /olerabant below, while Haase preferred to delete only nam...tolerabant. Good arguments can be found for each of these suggestions, but it seems paradoxical to consider deleting perfectly intelligible passages (esp. if they contain possible allusions to Livy: see next n.), when $1 itselfis intrinsically problematic (see 1n. szmulque).

gladii sine mucrone...in arto pugnam non tolerabant: cf. Liv. 22.46.5 ‘gladii Gallis praelongi ac sine mucronibus' and 28.33.9 'in arto pugna Romano aptior quam Hispano militi futura uidebatur'. non tolerabant transfers the action of the fighters to that of their weapons and seems much bolder than the use of the same verb at A. 2.6.2. An enemy's deficiency in swordsmanship, esp. in the case of Celts, was a topos (Oakley on Liv. 7.10.5—9, esp. p. 136). For toleraresee 13.1n. complexum armorum: i.e. fighting at close quarters (OLD complexus 3c). For the uariatio with the prepositional zn arto bugnam below see Sorbom

(1935) 79-80.

36.2 miscere ictus ‘to exchange blows': the verb is regular in descrip-

tions of combat (OLD 10, cf. 3b) butis not elsewhere combined with :ctus.

Gudeman compares Aesch. Pers. 1052—3 μεμίξεται | ... πλαγά, but the meaning there is ‘mingle’, not ‘exchange’. ferire umbonibus: the aggressive use of shields in combat is regular (e.g. Liv. 5.47.4 'umbone ictum', 9.41.18 'umbonibus...sternuntur hostes’; 566 also Oakley on 8.38.11) and perhaps illustrated on the Trajanic tropaeum at Adamklissi (metope XVIII); ferire seems nevertheless unusual, but cf. Claud. Raft. 2.224 'umbone ferit’. ora fodere: Caesar's order at Pharsalus was to strike at the faces of the Pompeians (N-H on Hor. Odes 2.1.20) and Germanicus' at Idisiaviso was the same (A. 2.14.3 ‘denserentictus, ora mucronibus quaererent’); for the expr. (also at A. 2.21.1) see Oakley on Liv. 8.10.6.

adstiterant ‘had taken up position/made a stand' (OLD assisto 3a).

974

COMMENTARY:

36.3

erigere in colles aciem 'to march the line up into the hills’: the expr. (again at 18.2, H. 3.71.1, 4.71.5) is technical (OLD erigo 2a), e.g. Liv. 10.26.8 ‘in collem...aciem erexit', 28.15.10. aemulatione et impetu conisae seems a hendiadys, 'straining with a rivalrous impulse’: the two nouns are combined elsewhere only at A. 15.34.4 ‘quasi impetus antiqui et bona aemulatione’; for nitor and compounds in descriptions of battle see Oakley on Liv. 6.12.8. proximos quosque: normally this idiom (superlative 4- quzsque) requires the singular but there are exceptions, e.g. Cic. Off. 2.75. plerique ... relinquebantur: 'the Roman attack was 50 effective that in their eagerness to clinch victory they pressed on leaving many of the British unscathed or only wounded' (O-R); for semzneces aut integri cf. H. 3.28. uictoriaeis objective genitive, ‘in their haste for victory' (TLL6.1.614.23-32): cf. Cic. Phil. 3.2 'festinatio ... uictoriae ... auida'. 36.3 equitum turmae: a (urma was a division of an ala (18.1n.) and consisted of 30 horsemen. [fugere couinnarii[: some scholars regard these two words as a parenthesis, but this would be very odd without some explanatory particle such as enzm. Others insert enzm, transposing it from where it is wrongly transmitted after ea below; but fugereremains a problem, since one would surely expect either a pluperfect instead of a perfect or (if fugere be regarded as an infinitive) the presence of a modal verb. Both problems would be met by the insertion of ut *when', but not only has this construction appeared in the immediately preceding sentence but its insertion after the subject (‘equitum

turmae’)

seems odd. Given the uncertain transmission in this

section of the principal MS (see further below), the corruption may be more extensive, e.g. znterim equitum turmae fugare couinnarios et peditum se proelio miscuere, but the truth seems irrecoverable. peditum se proelio miscuere: an example of so-called ‘mimetic syntax’, where the construction mirrors the action which is described. recentem terrorem intulerant: recens terror (again at A. 14.23.1) is a Livian expr. (36.9.13, 36.20.5). terrorem inferre (again at 17.1) is extremely common in Livy but is also found elsewhere, in verse as well as prose. densis tamen hostium agminibus et inaequalibus locis haerebant: cavalry action is ‘essentially an affair of movement and fluidity’ (O-R) and an encounter between cavalry and infantrv could be difficult for men on horseback (Keegan (1976) 153-9). Vegetius recommends ‘loca...inaequalia’ if a general wants to fight successfully against cavalry with infantry, whereas here cavalry have been brought in to fight infantry, a tactic for which 'loca...plana' are recommended (3.13.3). This is perhaps an argument in favour of the view, adopted by some scholars, that the equitum turmae are those of the Caledonians and that Ahostzum here refers to the Romans. Against this, however, are the facts that equitum turmae, a

COMMENTARY:

36.3

275

very common phrase, is almost always used of the Romans and that the expression here seems naturally to look back to equitum at 35.2.?! densum agmen 15 absent from extant Livy, though common elsewhere, esp. in Virgil (see H. on H. 2.22.1). haereo can be constructed with dat. or abl.

minime enim equestris ea pugnae facies erat: the transmitted text reads minimeque and has enim after ea, where it is clearly wrong: edd. either delete enzm or transpose it to follow fugereabove. It seems more likely, however, that -equeis a mistaken anticipation of equestris, as Till (1943: 67 and n. 5) suggested, and has ousted enzm from its correct place. In the margin of H &i has been proposed for ea, but the latter exemplifies the type of transference known as enallage (4.2n. omnem): hence 'the appearance of that battle’; for pugnae facies see H. 2.42.2 (and H. for other exs.). For other battle descriptions where the fighting was different from what it was supposed or expected to be see e.g. A. 1.49.1 'diuersa omnium ...ciuilium armorum facies', Thuc. 7.62.2—4, Sall. J. 59.1 ‘non, uti equestri proelio solet', Liv. 22.47.1 ‘minime equestris more pugnae'. aegre cliuo stantes: as Till (1043:67) remarks, this whole passage is ‘not well transmitted' in H, which here offers aegradiu aut stante. A very great number of conjectures have been made, and there seems no way of recapturing what T. actually wrote. Triller's suggestion, printed here, is closer to the paradosis than those of most other scholars, though T. uses clzuus only once elsewhere (H. 3.71.1, of the Capitol); for stare + abl. cf. A. 12.36.2 'stetere ...cohortes campo'. From the transmitted aut, however, and perhaps also from the presence of szmulbelow, it may be feared that something has been omitted from the text and that the corruption is therefore considerably more extensive. It is not immediately clear whether T. is talking about Britons or Romans; but, since simul equorum corporibus impellerentur seems to be a natural consequence of the involvement of the cavalry in the previous sentence ('densis tamen hosttum agminibus .. . haerebant"), it is perhaps more likely that the Britons are the subject. uagi currus, exterriti sine rectoribus equi: scholars have debated whether exterriti...equi refers to Romans or Britons, but the juxtaposition of chariots, hitherto mentioned only in connection with the Britons, clearly implies the latter. ‘They are loose horses from the Caledonian chariots' (O-R). For the scene cf. Curt. 8.14.9 'ut dissipatos tota acie ?! [nterestingly the same phrase, equitum turmas, is likewise disputed at Caes. G. 4.33.1. O-R repeat F.’s statement that turmae (unqualified) is ‘rarely used of other than Roman cavalry' but T. regularly uses the term of foreigners (e.g. G. 7.2, H. 1.79.2, A. 2.52.1, 6.35.1), including the Britons (A. 14.34.2). It is sometimes said that the first evidence for the Britons' use of cavalry rather than horse-drawn chariots is found in the famous Vindolanda inscription (/RB 274 — Tab. Vindol. 2.164) which refers to the 'Brittunculi' and is dated to the turn of the first/second century AD; but Caesar refers to British cavalry at G. 4.24.1.

276

COMMENTARY:

37.1

currus uagari sine rectoribus uidit' and perhaps Philostr. /mag. 2.5.1 ἵπποι Te ἀτακτοῦντες

ueT'

ἐκπλήξεως

(‘horses

running

wildlv

in

terror',

Loeb

trans.). exterreo, a favourite word of Virgil and having an 'artificial flavour at the time of Tacitus', is used only here in the minor works (in imitation of Sall. H. 1.139 ‘equi sine rectoribus exterriti'); but he uses it increasinglv often and by A. 11-16 it is more than 3 times as frequent as the commonplace terreo (Adams (1972) 362-3). ut quemque formido tulerat explains í(ransuersos aut obuios below: depending on where their terror took them, the horses sometimes knocked into men from the side, sometimes head-on.

37.1—2 Introduction of the alae 47.1 summa collium: see 6.4n. (/udos). uacui 'idly' (OLD 112). ni...Agricola...opposuisset: a ‘nz de rupture' marks A.'s second decisive intervention in the battle (cf. 36.1n. donec). In this construction, for which the archaising ? (4.3n.) 15 considerably more popular than ns? in the major historians, the combination of indicative in the apodosis and subjunctive in the protasis (for which see 3 1.4n. ac nis?) 15 standard; for the particular type where an incipient action is described in the main clause, as here (coeperant), see C-L 613—14 (and in general 597—617,637-8). The passage can be translated either as 'began to surround the rear of the winners «and would have succeeded> had not Agricola...’ oras ‘began to surround the rear of the winners but Agricola...'. id ipsum ueritus: not 'fearing' but 'regarding asa possible source of danger' (cf. OLD uereor 2), a caution that marks the ideal general (e.g. Caes. G. 2.11.2 'ac re statim Caesar per speculatores cognita insidias ueritus’, C. 3.112.5 'hoc tum ueritus Caesar...Pharum prehendit atque ibi praesidium posuit’, Liv. 33.37.2 '[consul] ueritus ne intercluderetur simul a Bois Liguribusque exercitum eadem uia qua adduxerat reduxit’ and cf. Vell. 118.2 with W.).

ad subita belli retentas: the ideal general was expected to keep troops in reserve for emergencies and the like: see Onas. 22.1—3, Frontin. Strat. 2.3.21—2, Veg. 3.17; K. on Liv. 6.23.12, Oakley on Liv. 9.17.15. subita belli (again at H. 5.13.4) is Livian (6.32.5 with Oakley's n., 25.15.20, cf. 4.27.1) but also at Flor. 1.1.15. For alaesee 18.1n. adcucurrerant: the sense 'charge, attack', as here, is considerably rarer

than 'run up (to help)': see OLD 2. tanto acrius pulsos in fugam disiecisset 'all the more harshly repelled them and scattered them in flight': for acriuscf. A. 4.73.3 ‘acri pugna hoste pulso’; for zn fugam disicere cf. Amm. 24.4.4 'disiectis in fugam’.

COMMENTARY:

37.2-3

277

37.2 consilium...uersum: the turning-point of the action is marked by the word for ‘turn’ and re-emphasised by the attack on the ‘turned away’ part of the enemy (next n.). auersam hostium aciem seems to mean 'the rear of the enemy's line' (as Liv. 3.70.4, 22.48.4, 30.35.1; OLD auersus 4a) rather than 'the rear line of the enemv'

(OLD 4).

37.2—6 The climax Tum uero patentibus locis grande et atrox spectaculum: sequi, uulnerare, capere: now starts the spectacle which A. had promised his troops earlier (34.3); tum is regular at 'the climax of a battle when one side begins to retreat' (Oakley on Liv. 7.33.15) and patens locus (sing. or plur.) is found frequently in numerous authors, esp. in battle scenes; but T. is alluding in particular to Sallust’s description of the battle of Cirta (f. 101.11): 'tum spectaculum horribile in campis patentibus: sequi, fugere, occidi, capi' (see further below).

atrox is used as an enticement to readers

(e.g. Sall.

J 5.1 ‘bellum scripturus sum ... primum quia magnum et atrox...fuit’), though its combination with spectaculum recurs only at H. 2.70.1; for the adj. grandis see 3.2n.; for the asyndetic historic infinitives see A. 4.25.2 ‘trahi, occidi, capi’.

atque eosdem oblatis aliis trucidare ‘and they butchered them likewise when others were presented’: eosdem refers to the implied object of capere, for oblatis see OLD 1a—b, though the exact meaning is hard to pin down. trucidare implies resistless slaughter, often of animals (W. on Vell. 119.2, adding K. on Liv. 6.37.2), and see n. below (cateruae). 37.3 hostium...offerre: the 'fronted' genitive applies both to cateruae armatorum and to quidam inermes: ‘now, depending on each man's temperament, companies of the enemy who were armed presented their backs to smaller numbers, while some who were unarmed rushed up of their own accord and offered themselves for death.' For prout . .. ingentum erat see Liv. 8.21.1 and Oakley ad loc. cateruae ... terga praestare: since each noun can be used of animals as well as of persons, thev allow the continuation of the metaphor implied by spectaculum above and prepare for the simile of the hunt below (4). cateruae armatorum (again at H. 2.88.9, A. 12.33) is originally Livian (25.39.4); ¢ praestare (instead of praebere) is paralleled only at Juv. 15.75. ultro ruere ac se morti offerre: the latter expr. is used to describe brave gladiators by Cicero (M:il. g2 ‘in gladiatoriis pugnis...fortes et animosos et se acriter ipsos morti offerentes seruari cupimus’), of whose own brave death it was also used (Sen. Suas. 6.18 'fortiter se morti non praebuerit tantum

sed obtulerit'); Caesar puts it into the mouth

of Critognatus

(G.

7.77.5 ‘qui se ultro morti offerant facilius reperiuntur quam qui dolorem

278

COMMENTARY:

37.4

patienter ferant’). For ultro ruere cf. Val. Fl. 6.242 ‘ultroque ruunt in funera Colchi'. laceri artus et cruenta humus: the latter expr. is as rare (only Virg. Aen. 11.668-9) as the former is common (Ov., Manil., Sen. trag., Luc., Sil., Mart., Stat., Plin., [Quint.]); scholars also compare Sall. . 101.11 'humus

infecta sanguine' (again from the battle at Cirta). aliquando etiam uictis ira uirtusque: although ?ra and uzrtus are found together quite often (e.g. Liv. 9.14.9 ‘nec uirtus modo insita sed ira etiam exulceratos ignominia stimularet animos’), T. is alluding simultaneously to Virg. Aen. 9.794—5 'neque terga | ?ra dare aut uzrtus patitur’ (of the l4s redit...uzrtus (of the defeated Trojans). For anger as a motivating factor in battle see Oakley on Liv. 9.13.3 and Intro. n. to g.1.1-16.19 (p. 21).

adpropinquauerunt: the subject is the Caledonians. 47.4 primos sequentium incautos 'the first, incautious, pursuers'. Pursuit themes the remainder of the paragraph (5 sequi, fim?s sequendz). locorum gnari is a favourite expr. of T. (H. 2.13.1, 2.85.2, 5.6.4) but seems otherwise regular (Plaut., Sall., Liv., Val.Fl., Amm.).

quod ni...Agricola...iussisset: this is A.'s third decisive intervention

and

it results

in the

final

dénouement

(5—6

below).

The

niclause

(see

also 26.2n.) is divided into the two principal groups to whom A. issues orders, the ‘strong light-armed cohorts' (ualidas... modo) and the cavalry (et sicubi . . . persultare) , of whom the latter are themselves divided (bv szmul)

into two sub-groups (those who dismountand those who evidently do not): see further nn. below. For quod ni see also 16.2n., and for nzsee 4.3n. frequens ubique 'in constant attendance everywhere' (OLD frequens 5). Cf. 20.2 ‘multus in agmine'. ualidas et expeditas cohortes indaginis modo: an :ndago was a ring of huntsmen or nets for the trapping of prey (OLD 1a), and the comparison (like Liv. 7.37.14 'uelut indagine dissipatos...ageret’: see Oakley's n.) here constitutes another actualisation of a passage in A.’s speech (34.2nn.). There is, however, no expressed infinitive saying what the cohorts have been ordered to do: is persultareat the very end of the clause to be applied to the cohorts as well as to the cavalry? or are we to understand a different verb by zeugma? Ifpersultareis applied to the cohorts, the verb is being used intransitively (OLD 2 ‘range freely’), since the complex word order of sicubi...persultare below makes it almost impossible to extract an appropriate object; it follows that there would be a slight syllepsis, since persultare is being used transitively (OLD 2 ‘scour’) of the cavalry. On the whole it is probably easier to understand a different verb altogether, e.g. agere ('he ordered the strong and light-armed cohorts to act [or perhaps to drive] in the manner of a cordon’), butitis nota straightforward decision.

COMMENTARY:

37.5

279

ualidae cohortes (again at H. 3.78.2) 15 originally a Livian expr. (6.24.2, 21.60.4) found also in Luc. 1.305 and Frontin. Strat. 2.3.5; expeditae cohortes (7x in T.) isaltogether more common (Caes., Sall., Liv.). etsicubi artiora erant.. . persultare iussisset: although Rhenanus' emendation equitem persultare has been generally accepted, nevertheless this second part of the clause is even more difficult than the first. A. orders part of the cavalry (partem equitum) to dismount and to scour the denser areas «of the woods> on foot; at the same time, according to the paradosis as emended by Rhenanus, he also orders the cavalry (equitem) to scour the thinner areas of the woods «on horseback>. This paraphrase ought to make clear that, despite the virtual unanimity of editors, it is simply not possible to make any sort of meaningful contrast between partem equitum and equitem. Τ 6 distributive expression partem equitum requires to be followed by a matching expression: evidently a scribe repeated a form of eques instead of the correct partem. sicubi artiora erant = ‘any denser areas that there were' (lit. ‘if anywhere there were denser areas’): the clause provides the first object of persultare (1.e. as if artiora sicubi erant) and 15 parallel to, and varied by, rariores siluas.

The clause as a whole may be translated as follows: 'And, had not Agricola (who was in constant attendance everywhere) ordered the strong lightarmed cohorts to act in the manner of a cordon and some of the cavalry to dismount and scour any denser areas that there were and others «to scour> the sparser woods...' Cavalry fighting on foot, and often restoring the situation, is a regular military topos (Oakley on Liv. 6.24.10). acceptum...foret: the sentence ends on the same theme as it began (per nimiam fiduciam — incautos), emphasising the disaster which would have taken place if A. had not intervened; excessive confidence is always militarily dangerous (out of several exs. see Liv. 33.14.9 'cum id effusius hostes et, ut fit ab nimia fiducia, neglegentius etiam

facerent’; also

Oakley on 6.3.5). In the description of Cirta of which there have been clear echoes in this paragraph (see above), Sallust had referred to literal wounds (/. 101.11 'uolneribus acceptis’), a reference to which an allusion is perhaps made by T.'s metaphorical wound here (OLD uulnus 4). Elsewhere T. uses the metaphor of the wound for mental or emotional blows of a domestic nature (7.1, 29.1, 45.5; OLD 4). For foret see 16.2n. (amissa).

37.5 ubi compositos firmis ordinibus sequi rursus uidere: the subject 15 the Caledonians. compositos . . . ordinibus — ‘arranged in solid ranks’ (OLD compono 6, firmus 3b); the contrast is with primos sequentium incautos above (4), underlining the effect of A.'s intervention. rursus is probably to be taken with udere = ‘conversely’ (OLD 6). uersi: sc. sunt (though some edd., interpreting uers? as a participle, continue the sentence to petiere).

280

COMMENTARY:

37.6

rari et uitabundi inuicem ‘ scattered and avoiding one another' (OLD rarus 2b): the adjectives correspond to non agminibus and mec alius alium respectantes respectively. uztabundus (again at H. 3.37.2) was a favourite word of Sallust (4x, including/. 101.9), then once in extant Livy (25.13.4) and Ammianus (30.1.13). Adjectives in -bundus are almost always used predicatively with verbs of motion (as here), they often indicate the visibility of the action in question, and they are favoured by historians in descriptive and dramatic passages. See G. on A. 1.17.1. longinqua atque auia petiere combines Sil. 13.92 ‘auia tunc longinqua placent' with 4.177-8 ‘auia...| dum petit'; T. repeats both laterat A. 2.68.1 'auia...petiit', A. 15.11.2 ‘longinqua et auia...repetiuere’. finis sequendi nox et satietas fuit: pursuit being ended by nightfall is a variation on the topos of night ending a battle (for both see Oakley on Liv. 7.33.15 with Addenda (4.564)); for night and satiety combined cf. Amm. 19.2.14 'tandemque nox finem caedibus fecit, et satias aerumnarum indutias partibus dederat longiores' (rather different); for satiety and closure cf. Cic. Phil. 5.23 'cum autem semel gladium scelere imbuisset, nulla res ei finem

caedendi

nisi .. .satietas attulisset', Sil. 4.110

'nec finis satiasue,

noui sed sanguinis ardor'. 37.9 caesa...cecidere: ancient battle descriptions are often formally rounded off with casualty figures, frequently — as here — including both named and unnamed dead. 10,000 is a typical round number and is presumably no more genuine than e.g. Liv. 10.45.11 'capta aut caesa ad decem milia hominum’, while 360 is a multiple of both 12 and g, each of them a rhetorical number (Fehling (1989) 221-4, 226-32); whether the totals are invented bv T. or derived from an estimate bv A. is impossible to know. During the republic, only generals who had killed at least 5,000 of the enemy in a single battle were permitted to hold a triumph (Val. Max.

2.8.1). A.'s figure 15 exactly twice the minimum.

‘This under-

lines his standing as the true heir of the great Republican commanders' (Birley (1999) 90). ‘The quasi-adverbial use of ad with numerals to mean “up to" or “almost” is a development of the regular prepositional use. The usage is most striking when found with a declinable numeral in the nominative' (Oakley on Liv. 9.42.8, q.v.). Here decem milia is in the nominative but is of course indistinguishable in form from the accusative; see also OLD ad 20a. According to the fifth-century AD historian Orosius (7.10.4), T. η the H?stories was reticent about casualty numbers, invoking the authority of Sallust and others: 'nam quanta fuerint Diurpanei Dacorum regis cum Fusco duce proelia quantaeque Romanorum clades longo textu euoluerem, nisi Cornelius Tacitus, qui hanc historiam diligentissime contexuit, de reticendo interfectorum numero et Sallustium Crispum et alios auctores quamplurimos sanxisse, et se ipsum idem potissimum elegisse dixisset'.

COMMENTARY:

38.1

281

The only ex. in H. as extant is 2.17.2; in A. cf. 4.73.4, 14.37.2 (Boudica's revolt).

in quis: the archaising equivalent of quibus occurs only here in T.'s minor works; in A. 1—12 itis preferred marginally to quibus, which is the preferred form in H. and A. 13-16. See Adams (19732) 135. Aulus Atticus is the only member of A.'s forces to be named in the entire biography. He was perhaps called A. Julius Atticus and came from the same family in Gallia Narbonensis as the Julius Atticus who wrote on viticulture and of whom A.'s father was 'uelut discipulus' (Colum. 1.1.14): see RGB 93 (referring to Syme, RP 1.375). iuuenili ardore et ferocia equi: ferocia usually denotes defiance or, in animals, a refusal to be managed (Ov. Her. 4.79 'ferocis equi luctantia colla’;

OLD ferocia 1a; also above, 11.4n. ut quos), 50 this chiastic and varied

epitaph poignantly suggests a voung man carried away by his own initial enthusiasm (zuuenlis ardor also at D. 2.1, [Sen.] Oct. 189) and then, when he realised he had advanced too far, tragically unable to wheel his horse round. 38.1—2 Aftermath 38.1 Et nox quidem...laeta uictoribus: quidem both picks up and extends the reference to nox at 37.5 above and also prepares for the otherwise unmarked contrast with proximus dies at 2 below: for this double function see 17.2 (n. Cerialis quidem). For gaudio. .. laeta cf. Curt. 8.4.27, Plin. Pan. 02.4 ‘diem...gaudio laetum’; for praeda...laeta (again at A. 14.93.2) cf. Liv. 27.31.3, Apul. Met. 8.31.1; for the combination of abstract and concrete see e.g. 20.2, 24.2, 28.1, 29.4. Britanni...concitari: the Caledonians' actions are described in a series of ten historic infinitives, the highest number in T.; there are six at 20.2,

seven at 5.1, nine at 19.2—4 and A. 4.51.1, eleven at Sall. J 66.1. See also 5.1n. (noscere). For the uariatio of asyndeton and coordination see 15.4n. (patriam).

mixto ... ploratu: the expr. 15 originally Lucretian (2.579-80, of babies) but T. is no doubt thinking of Livy (5.21.11 'mixto mulierum ac puerorum ploratu', cf. 2.33.8 'clamor...oppidanorum mixtus muliebri puerilique ploratu’). Such lamentation is a regular feature of military narratives, esp. (but not exclusively) in sieges (Quint. 8.3.68; Oakley on Liv. 6.3.4).

miscere . ..consilia aliqua, dein separare: aliqua is pathetic: the barbarians can now share only a few, indeterminate plans but, even so, they soon revert to type (separare): see 15.5 and 32.1. For miscere. .. consilia (again at H. 2.7.2, À. 15.68.2) cf. [Quint.] Decl. 14.9; for c. separare cf. Caes. G. 7.63.8, Liv. 23.20.4; the two verbs are natural opposites (Col. 3.19.3, Plin.

282

COMMENTARY:

38.2

NH 17.187, Quint. 4.2.101, 6.5.1, [Quint.] Decl. 365.5). For night as the time for planning see Oaklev on Liv. 9.3.1. frangi aspectu...concitari: aspectu (abl.) 15 quite commonly found with similar verbs (TLL 2.803.57ff.), but neither of these seems paralleled elsewhere. pignorum = 'their loved ones' (OLD 4b), whose presence near the battlefield had been supposed to spur them to victory (32.2). satisque constabat: the formulaic expr. (13.2n.) is designed both to distance and to vouch for the horror of mercy killing. For the action against wives and children see Liv. 21.14.4 (Saguntum), 31.18.7 (Abydus). 38.2 proximus dies faciem uictoriae latius aperuit: the verb suggests not simply visibility after the darkness of night (cf. H. 4.29.3, Sen. Ira 2.22.9 'ueritatem diesaperit'; OLD aperio 11a, Horsfall on Virg. Aen. 3.206) but also, as the presence of faciem and latius implies, the opening up of the landscape itself as the shadows disappear (as Sil. 17.603 'prostratis facies aperitur dira maniplis'; OLD aperio 11b): for facies of landscape cf. A. 4.67.2, 14.10.2; OLD 2a. uastum...silentium: the adj. combines the notions of desolation and almost limitless extent: ‘no writer exploited the resonances of uastus as effectively as Tacitus' (Oakley on Liv. 10.34.6). The expr., which here calls to mind the solitudo of Calgacus' indictment (30.5), recurs at H. 3.13.2, A. 4.50.4 and cf. also Mela 3.95, Luc. 5.508, Sil. 1.67. See also 41.2n. (sleri). secreti 15 assumed to mean ‘deserted’: if so, it 15 a highly unusual sense (OLD 4b).

fumantia...tecta: compare the teichoskopia conjured up by Quinctius in his speech in Livy (3.68.2 ‘ex muris uisite agros uestros ferro ignique uastatos...fumare incensa passim tecta') or Ovid's image of Troy destroyed (Met. 13.421 "Iroades et patriae fumantia tecta relinquunt’). On smoking ruins see further Horsfall on Virg. Aen. 3.3. incerta fugae uestigia: it is simplest to take these words as dependent on compertum (est) below and to understand esse (‘when it had been discov-

ered that the tracks...’); butitis not certain whether fugaeis to be taken with uestigia, in which case fugae virtually — fugzentzum (cf. 16.3n. paenitentiae), or whether with ?ncerta, in which case fugae has an extended sense ('uncertain of the means of escape' (OLD 7a), though that is not what the

same phrase means at Sil. 4.112). In the line of Lucan which T. is quoting, incerta is used predicatively (8.4 'incerta fugae uestigia turbat', Pompey ‘confused the tracks of his flight «and made them> unclear’). exacta iam aestate ‘with the season already completed', as Caes. G. 3.28.1 ‘prope exacta iam aestas erat’; OLD exigo Ob. aestate combines with spargi (below) to produce a metaphor of sowing (see A. 3.21.4 and W-M's n.). In his first year in Britain, A. had disregarded the lateness of the season, unexpectedly initiating a campaign ‘quamquam transuecta aestas' (18.2).

COMMENTARY:

38.3

283

in fines boreos totum exercitum deducit: the paradosis reads borestorum, conventionally taken to refer to an otherwise unattested people, the Boresti (see Rivet and Smith (1979) 272—3). Wolfson (2008: 65—74) has argued that this be emended to boreos totum. Though boreus is frequent only in the obscure astronomical author Hyginus (note Astron. 4.11 ‘ad boreum

finem’), it occurs once each in Ov.

(Tr. 4.8.41

'sub axe boreo’)

and Amm. (23.0.74 ‘per boreum latus’) and frequently in the late grammarian Priscian (e.g. Perieg. 315 ‘boreisin finibus’), and this evidence (esp. Ammianus, an imitator of T.) is perhaps sufficient to justify its use here. Hence: ‘he withdrew his entire army to the northern territories’ (deducere is technical for the withdrawing of troops: OLD 1a). Wolfson suggests that T. is referring to the very north-eastern tip of Scotland (Caithness).

38.3—4 Coda: circumnavigation 38.3 circumuehi Britanniam: this is the expedition to which reference was made in the digression at 10.4 'tunc primum Romana classis circumuecta insulam esse Britanniam adfirmauit'. Given the ancients’ fear of unseasonal sailing, it seems extraordinarv that A. would have issued such an order in northern

and

uncharted

waters in the autumn

(2 'exacta iam

aestate'). The use of an acc. + inf. with praecipio to express an order seems to start with early first-century authors: for the passive, as here, cf. e.g. Val. Max. 5.5.9 'praecepit...praetorium ei statui’ (TLL 10.2.449.11ff.). See also 46.4 ‘praeceperim...uenerari’. praecesserat terror will refer either to the sailings of the previous year (25.2) or to the news of Mons Graupius or both.

nouarum gentium: since we do not know the location of Mons of these destinations.

it is impossible to know whom T. means, especiallv either A.'s point of departure (fines at 2 above) or Graupius or the routes by which he reached either If we assume that he was now marching south down

the eastern side of Scotland, the tribes in question were the Cornovii, Lugi, Smertae, Decantae, Vacomagi and Taexali, most of them little more than

names. One difficulty with this assumption, however, is that A. may well have encountered the same tribes on his march north to Mons Graupius: if so, itis hard to see how they can now be called ‘new’. Perhaps T. means ‘the new tribes encountered in this year’, as opposed to the nouae gentes of an earlier campaigning season (see 22.1 and nn.). Itis unlikelv that A. led his troops down the west coast of Scotland, since the terrain there is much less congenial to a marching army. ipsa transitus mora ‘by the very slowness of his passage' (OLD mora 3). in hibernis: unknown. It is shown by a writing tablet (Tomlin (1992) 150 with Plate 111) that men of Legio XX were at Carlisle (Luguvalium)

984

COMMENTARY:

38.4

on 7 November of this year, 83, and another writing tablet from there (Tomlin (1998) 74-5 with Plate XVIII = IRB 286) records the presence of a singularis Agricolae ('bodyguard of Agricola’), suggesting that that may well have been A.'s own winter quarters. 38.4 secunda tempestate et fama: there is a slight syllepsis here, ‘because of the favourable weather and to favourable report’. portum tenuit is the regular Latin expr. for 'reached harbour' and is extremely common in, for ex., Livy; cf. OLD teneo 5a. The name of the harbour is otherwise unknown. Rhenanus and Lipsius wanted a reference to Rutupiae (mod. Richborough) on the Kent coast, since that came to be a base for the Roman fleet, but this seems not to square with other inferences from the text (next n.). Wolfson

(2008: 35—46) wanted a reference

to Shetland (‘Tulensem portum’) and hence preferred to base his argument on the marginal reading of H; but d?specta est et Thuleat 10.4 implies that there was never a landing on the island (see n.); he was also required to emend

(ru- to trux (‘ruthless because of the favourable weather and its

reputation’), which seems quite implausible. For further discussion see Rivet and Smith (1979) 478-80. unde proximo Britanniae latere praelecto omnis redierat: we might have expected redierat to be preceded by quo (‘to where it had returned") but unde has to be taken with praelecto ('from where it had skirted the nearest flank of Britain and returned in its entirety’): cf. Liv. 36.21.5 ‘a Patris... Acarnaniae litora legit’. The technique whereby a relative pronoun

or equivalent

(here

unde)

takes its form

not from

the main

verb

of the relative clause but from some subordinate expression is regular, e.g. Cic. Off. 2.89 ‘Catonis senis, a quo cum quaereretur quid . .. expediret, respondit’: see Mayer (2008) 342-3. For the quasi-adverbial use of omnis cf. Liv. 32.8.3 'alteri consuli nouus omnis exercitus decretus' ('to the other consul a new army was decreed in its entirety’; Horsfall on Virg. Aen. 7.598—9). Many commentators understand proximo to mean ‘nearest to Rome', as at 14.1 (n.); but the juxtaposition of unde suggests that the adj. refers to the flank of Britain nearest to the harbour from which the fleet set out. We do not of course know where the harbour was, but it may be inferred from praefecto classis . . praecipit at $ above that it was in the north of Scotland: thus the point of unde proximo Britanniae latere praelecto seems to be that, on first leaving the harbour, the fleet did not make

the

short voyage round the northern tip of Scotland but sailed directly south down the coast of Britain on which the harbour was situated. We do not know whether this was the east or west coast, but, if Thule at 10.4 is to be

identified with Shetland, the reference there to the onset of winter perhaps indicates that the fleet started down the east coast of Britain in the autumn and, having then sailed back up the west coast, rounded the tip of

COMMENTARY:

99

285

Scotland and finally sighted Shetland as winter approached; but all this 15 mere speculaton.

39-49

THE

FINAL

DECADE

(Ap 84-939)

The narrative of A.'s final decade is dominated by the person of Domitian. Though we last heard of him as an irresponsible young man fourteen years previously (7.2), his introduction in the preface was such that his principate has cast a long and ominous shadow over the intervening narrative (3.2 'per quindecim annos...sacuitia principis’). As the penultimate section of the work opens, Domitian is feigning delight, and hiding his anxiety, at A.’s success in Britain (39.1 'fronte laetus, pectore anxius'). He sees himself as ‘the good leader' (39.2 'ducis boni'), and decides to store away his hatred of A. until the impact of the latter’s reputation has worn ΟἹ (39.93 ‘reponere odium’). As the section closes, the reverse 15 the case: Domitian 15 feigning grief, and hiding his joy, at A.'simpending death (43.9 'speciem...doloris', 'qui .. . dissimularet gaudium’). Though he is still in the grip of hatred (49.93 'securus iam odii'), he also experiences genuine delight when he learns the terms of A.’s hinal testament (43.4 ‘laetatum eum’); but he has misinterpreted the will, not realising that 115 terms indict him as an evil ruler (43.4 ‘malum principem’). Domitian emerges from these framing scenes as the deceiver whose first victim 15 himself. While A. 15 preparing to depart from Britain and leave his army behind, the prospect of his governing Syria is deliberately raised by Domitian (40.1); but, when the triumphant general has actually left his province and 115 army, his return to Rome 15 stage-managed in advance (40.3 'ita ut pracceptum crat’): he is obliged to behave as the opposite of returning victor and his interview with the princeps is cut short; Syria is never mentioned again. When, some years later, it was A.’s turn to cast lots for the governorship of Africa or Asia, he is advised obliquely to forget all about it (42.1) and 15 compelled to attend an interview with Domitian at which, instead of being formally dispatched to a prestigious province, he embarks on a farcical 'dialogue of refusal' which is prescribed for him in advance (42.2 and n. offensus). Domitian's confrontations with his general over the matter of provinces thus form, as it were, an inner frame, and the reasons why À. received such

treatment are illustrated in the centre of the narrative. À. was a man of uirtus and gloria (41.1 — 41.4), whose tried and tested excellence in war contrasted comprehensively with the incompetents responsible for successive military defeats (41.2—9). In spite of his precautions, he, not Domitian,

286

COMMENTARY:

39.1

was the leader whose name was ΟἹ evervone's lips (41.93 ‘poscebatur ore uulgi dux Agricola’). 39.1 Hunc rerum cursum...excepit the expr. is elliptical: ‘received «news of> this course of events'. The subject is Domitian, inferred from Domitiano in the ut-clause. Hunc sums up what precedes (cf. 18.1); cursus rerum 15 a fairly common phrase but the order of words 15 almost invariable (as at H. 4.34.1); another exception is Amm. 17.9.6. quamquam...auctum: for the importance and role of such letters see Osgood (2009), esp. 349—51, and above, 18.6n. (ne laureatis); augere is a technical rhetorical term for magnifving or emphasising (OLD 11). uerborum iactantia is the contemporary equivalent of the older w. zactatio (Cic. Sull. 24, Fam. 11.20.2 (D. Brutus)): see 25.1n. (zactantia). For quamquam see 1.1n. moris: cf. 33.1n. (Excepere). fronte laetus, pectore

anxius:

T. has varied,

but thereby

normalised,

the exclusively poetic expressions /aeta frons (e.g. Virg. Aen. 6.862, Luc. 9.1106—7 'pectora laeta |fronte tegunt’) and anxia pectora (e.g. Ov. Her. 20.198, Sil. 2.285—6). derisui fuisse nuper: this predicative dat. is found elsewhere only at Phaedr. 1.11.1-2 'Virtutis expers, uerbis iactans gloriam, | ignotos fallit, notis est derisui', but znrisu? is at Caes. C. 2.15.1 and risu in Livy (OLD 1b). nuper qualifies fuusse, for its position cf. 42.1 below. falsum...triumphum: Cicero had once complained of the fictitious triumphs (fals: triumphi) which were inserted in some family records (Brut. 62); Domitian's actual triumph over the German tribe the Chatti in 83 was represented as false (Plin. Pan. 16.3 ‘mimicos currus...falsae simulacra uictoriae', Dio 67.4.1): the enemy were 'triumphati magis quam uicti'

(G. 37.5).

emptis agrees with the understood antecedent (e.g. 225 or hominibus) of quorum; for the appended abl. abs. and 'aoristic' participle see 2.1n. (delegato). For per commercia see 28.3n. quorum...formarentur: habitus has so many different, but closely related, shades of meaning that it is not absolutely clear what is meant. In the 'triumph' which Caligula celebrated over Germany it was height and hair that were the key factors (Suet. Calzg. 47 ‘Galliarum quoque procerissimum quemque ... legitac seposuit ad pompam coegitque ... rutilare et submittere comam'; cf. Pers. 6.43—7; Beard (2007) 185-6, Ostenberg (2009) 152—6); buta person's height cannot be 'got up' (OLD formo 2a) to resemble that of a German captive. Perhaps 'style of dress' (OLD habitus® 3) is meant. It is also not clear whether the subjunctive is final/jussive (‘were to be fashioned': NLS8134) or consecutive/generic ('could be fashioned': NLS 88155—-8). For crinis see 11.2n. (tortz); O-R are mistaken in

COMMENTARY:

39.2

287

saying that the collective singular form ‘is only used of a single individual’: see e.g. Sen. NQ 1.17.7. at nunc

ueram...:

the acc. - inf. after conscientia

(OLD

2a)

resumes

until uzrtutem 6556 (2). uera uictoria15 orig. Livian (29.14.7, then Flor. 1.6.6, 1.34.12). 39.2 formidolosum: sc. esse. For the danger posed to Domitian in this passage see Lendon (1997) 108-13. priuati hominis nomen supra principem adtolli: the expr. is again strictly elliptical (‘that the name/reputation of a private individual was being elevated above «that of> the princeps’) but serves to heighten Domitian's alarm and indignation: it is as if a mere name is being elevated above the realitv of the princeps. For such compendious comparisons 566 12.3n. (dierum). nomen adtolli derives from Luc. 7.11, where Pompeius Magnus, a man famouslv associated with his own name and now regarded as a military conqueror par excellence (Grenade (1950) 57-8), 15 dreaming of one of his triumphs; if Domitian envisages A. as a second Pompev (see also below, 41.4n. szmul), it 15 relevant that Augustus and Tiberius regarded their respect for Pompev's name as an indication of their own moderation (see W-M on A. 3.72.2 manente tamen nomine Pompei and ref. there to Aug. RG 20.1). To emphasise still further the gap between himself and A., Domitian 15 capitalising on the ambiguity of priuatus (see OLD priuatus! 2a—b): though A. was a priuatus in the 1655 technical sense of ‘subject’, the suggestion is that he was a priuatus in the sense that he was holding no public office, which was of course false. studia fori et ciuilium artium decus: the first element of this formally chiastic phrase refers to the processes of law; other exs. of the phrase ciuiles artesindicate nothing more specific than the pursuits of civilian as opposed to military life (Cic. Brut. 155, Liv. 10.15.12, Vell. 97.2, HA Alex. Sev. 3.1). in . silentzum acta suggests the 'banishment' mentioned at 2.2 ‘omni bona arte in exilium acta' (n.); for szlentium see also g.2.

51 militarem gloriam alius occuparet: gloria militarzs 15 a common expr., but g. occupare is more choice (Val. Max. 4.3.8, Plin. NH 36.37, Frontin. Strat. 2.4.1); the verb suggests the kind of military action involved in winning glorza. cetera utcumque facilius dissimulari ‘other qualities were somehow more easily ignored’: for this sense of dzssimulo see OLD g; for utcumque see OLD 2. For the tyrant's dissimulatio, which in his paranoia Domitian here uses of his reactions to a third party, see above, 1n. (fronte). ducis boni imperatoriam uirtutem: adversative asyndeton, and essehas to be understood. Most scholars seem to think that the meaning is ‘ the excellence of a good leader was an imperial «excellence-', but the sense is at least as compelling, if not more so, if one takes it the other way round: ‘ a commander's prowess belonged to the Good Leader'

288

COMMENTARY:

(for this sense

of zmperatoria(e)

on

289),

39.3

utrtus (-tutes)

cf. Cic. Leg. Man.

20, Liv.

9.18.5, Quint. 7.10.13, 8.2.11). The fact 15 that the sentence makes sense both ways, its ‘palindromic’ nature emphasising its message: however the matter was looked at, military distinction was the exclusive possession of the emperor. See esp. Plato, Leg. 832c 'the ruler, through fear of the subject, will never voluntarily allow him to become ...in any way warlike'. ‘The title dux undergoes a revival with Domitian and Trajan’ (Ferri [Sen.]

Oct.

and,

on

the

second

of the

above

interpretations,

Domitian is describing himself in the same reverential terms as Augustus was addressed by Horace (Odes 4.5.5 and g7 'dux bone’): ‘The term dux...especially suits the return of the zmperator rebus prospere gestis. .. However, the epithet bone underlines Augustus' untiring service to the state and anticipates his role as a benefactor in a civil rather than a military capacity’ (Du Quesnay (1995) 156). On the present occasion it was A. who was the returning victor and, in Domitian's eyes, perhaps a potential zmperator. For some other exs. of dux bonus cf. Cic. Off. 3.100, Sall. H. 1.77.21, Liv. 7.40.15. 39.3 curis exercitus is a common phrase (esp. in poetry), but the participle is perhaps the first of the medical terms which cluster in this sentence: exercere — ‘to trouble' is very commonly used of illness and the like by Celsus (e.g. 1.3.13, 1.6.1, 2.1.9, 2.1.18, 3.0.16, 4.24.2). quodque saeuae cogitationis indicium erat: the antecedent of the rel. clause is secreto suo satiatus below. indicium 15 another medical term, ‘symptom' (OLD 4a); for Domitian's 'savage thinking' compare the enraged Dido at Serv. Aen. 4.642 'furiata saeuis cogitationibus': his saeuztia was a recurrent theme of the preface (1.4, 2.1, 3.2). secreto suo satiatus 'satisfving himself with his usual [OLD suus 10—11] seclusion'. Secretiveness was a characteristic of the typical tyrant (Plat. Rep. 579B—C); that of Domitian (cf. Dio 67.1.1) was described thus at Plin. Pan. 48.5: ‘non adire quisquam, non adloqui audebat, tenebras semper secretumque captantem nec umquam ex solitudine sua prodeuntem nisi ut solitudinem faceret'. Since seclusion was also a natural resort of the ill (e.g. A. 14.50.3), the phrase may also contribute to the metaphor of illness. optimum: sc. esse or fore. ‘he decided it would be best for the present to store away his hatred' (OLD reponere g). T. returns to this tyrannical trait with Tiberius (A. 1.69.5 'odia...quae reconderet auctaque promeret') and Nero (16.5.9 ‘dissimulatum ad praesens et mox redditum odium’).

donec impetus famae et fauor exercitus languesceret: since mpetus (OLD gb) and languesco (OLD 1, 3b) can be used of illness (cf. A. 4.21.1 'si impetus offensionis languerat, memoria ualebat' and M-W), the implication seems to be that Domitian is suffering from an 'attack' of A.'s fame and intends to wait until it weakens.

COMMENTARY:

40.1

289

nam etiam tum...obtinebat: 'Agricola, with four legions at his back, might have imitated the example of previous provincial governors, like Vitellius and Vespasian, and aspired to become emperor' (Sleeman). 40.1 triumphalia ornamenta: ‘After 10 BC no senator outside the imperial family was permitted to celebrate a triumph. Instead deserving commanders were on imperial initiative awarded trzumphalia ornamenta’ (OCD 1048—9 s.v. ornamenta). Itis not clear what the ornamentaincluded: perhaps no more than 'the usual toga praetexta of a magistrate’ (Beard (2007) 70); at any rate, 'Agricola with the ornamenta triumphalia earned a distinction not to be achieved

by other generals in Domitian's wars, who,

so far as

known, had only consular dona' (Syme, RP 3.184). Though the ornamenta are a commonplace of modern scholarship, the term is relatively uncom-

mon in classical Latin: 5 x in Velleius (in the order o.t.), once elsewhere in

T. (44.3, where the order 15 again /.o.) and 10x in Suetonius (who prefers t.o. by 7:3). inlustris statuae honorem: the ornamenta ‘seem to have been accompanied by the grant of a statue of the honorand in that most triumphal of monuments, the Forum of Augustus' (Beard (2007) 70, referring to Dio 55.10.3 and Swan's n.). See also OLD triumphalis 1c, adding A. 4.23.1. quidquid pro triumpho datur: given the 'triumphal' context, it is natural to translate this as 'evervthing that is given in place of a triumph', esp. since the possible cvnicism seems typical of the authorial persona often associated with T.; but, since the reference

to the tri-

umphal ornaments (above) has already made this point implicitly, such an interpretation perhaps seems rather flat, and it may be wondered how much literarv capital could be derived from drawing explicit attention to a substitution which was now more than a century old. An alternative 15 to assume that trzumphus is here being used in its non-technical sense of 'victory' (OLD 4): 'everything thatis given as reward for a victory'. multo uerborum honore cumulata: the neut. plur. participle agrees with all three preceding elements, ornamenta, honorem, quidquid (see A-G $287.3), and means perhaps ‘enhanced’ or ‘crowned’ (cf. OLD cumulo 5—6); in the sense ‘piled up’ it also constitutes metalinguistic comment on the repetitions triumphalia — triumpho and honorem — honore. The set phrase uerborum honoris to be expected in this context (cf. esp. H. 4.4.2 ‘multo cum honore uerborum Muciano triumphalia...data' and H.). decerni in senatu iubet: the convention was that the senate decreed such honours on the emperor's initiative (Domitian being the subject of tubet): cf. e.g. Plin. Ep. 2.7.1 'asenatu Vestricio Spurinnae principe auctore triumphalis statua decreta est'. According to the transmitted text of Dio 66.20.3, À. was given triumphal ornaments ‘by Titus’, but the text presumably requires emendation (Birley (2009) 54 and n. 19).

290

COMMENTARY:

40.2

additque insuper opinionem Syriam . . destinari: if one retains the paradosis addique, the Latin may appear to say that Domitian ordered the opinio to be appended to the decretum, which is impossible. This possibility is excluded bv Muretus' simple change to additque. Domitian 15 now responsible for two clearly separate actions, ordering the decree and (in addition) spreading the kind of rumour which attended an earlier stage of A.'s career (9.5 'comitante opinione Britanniam ei prouinciam dari’). In fact, however, ‘it was anomalous for a “vir triumphalis" to proceed to another governorship' (Svme, RP3.1384). For the expr. cf. Caes. C. 2.40.3 ‘addita...opinione’. Atili Rufi consularis: epigraphic evidence shows that T. Atilius Rufus was legate of Pannonia in 8o and governor of Syria in 85; neither the date of his consulship nor the identity of his Syrian successor is known. He is one of the 'governors dving in Syria' discussed by Syme (RP 3.1376—92), and 'the last of the four legates whom the pages of Tacitus happen to disclose dying in office' (ibid. 1384). maioribus refers to persons 'greater in power or reputation’ (OLD maior 6a, cf. 7); a governor of Syria was in command of a powerful army (four legions before AD 770, three thereafter). uacuam...reseruatamis chiastic. 40.2 credidere plerique...compositum est: the sentence 15 somewhat parenthetic and seems designed to support the notion (which otherwise might have seemed incredible) that the princeps himself was spreading rumours about A.'s further advancement. libertum...codicillos...tulisse:

cod?cilli is the technical term, of which

]. is our first evidence, for a 'letter of appointment' (Millar (1977) 288—90, 305-6, 310-11; cf. OLD ga): see also D. 7.2, À. 13.20.2. From inscriptional evidence we know of lbert: a codicillis (e.g. ILS 1529). ministeriis: abstract for concrete, 'attendants', ‘agents’ (OLD 5a).

cum eo praecepto ut, si in Britannia foret, traderentur: the point is that, if A. were still in Britain, he would still be in command of his legions and hence would be in a position to ensure that Domitian acted on what he was rumoured to have promised. in ipso freto Oceani: i.e. the English Channel. siue uerum istud siue ex ingenio principis fictum ac compositum est ‘whether

that was true or whether

it was fabricated

and

compiled/

concocted in accordance with the princeps' temperament': for this elliptical use of szue...stue (again at 42.2, 43.2) see OLD 5a; for ex see OLD 20a (though 18a 'as a result of' is perhaps also possible). The use of iste to refer to something just mentioned (i.e. as opposed to its secondperson sense — 'that of yours') seems to be colloquial and is almost always restricted by T. to speeches; the one other exception appears to be A. 16.16.2, another authorial comment (a type of passage where he

COMMENTARY:

40.3

291

sometimes admits speech usages): see Adams (1973a) 131 and n. 68. The exact meaning of compositum (OLD 8 or 10a) is not really clarified by other cases where these two verbs are co-ordinated (Liv. 40.12.7, Val. Max. 7.4.2, Sen. Cons. Helv. 17.1, Quint. 8 praef. 23, Tac. H. 3.78.2). 40.3 successori suo: ‘This unnamed successor cannot be securely identified, but might conceivably be Sallustius Lucullus' (RGB g4): for speculation about this man, the date of whose consulship is unknown, see RGB 05—9, where it is argued that he governed Britain between 84 and 87. quietam tutamque: for sentence-final -que 566 10.4n. (?nuenit). ne notabilis...introitus esset: "The entry of a great man into the city, with the welcome he receives, is a traditional topic’ (Pearce (1970) 314, with a wealth of illustrative material). A. proposes to avoid or reverse all the usual manifestations of an aduentus, one of which is 'the crowding and density of the welcomers'

('celebritate et frequentia occurrentium’):

see

e.g. A. 3.9.2—3 'ripa frequenti ...celebritate loci' (of Cn. Calpurnius Piso), Cic. Mur. 68—g. He behaves instead as if he were returning defeated (cf. e.g. Pind. Pyth. 8.81—7). See also O'Sullivan (2011) 67-8. For occurrentium see below (n. turbae).

officio here means 'a gathering of persons to pay their respects' (OLD 2b): again contrast Calpurnius Piso at A. 3.9.2 ‘magno clientium agmine' and compare the endangered Seneca at A. 14.56.3 ‘prohibet coetus salutantium'. noctu...noctu...uenit: it was of course usual to arrive by day for maximum publicity (so Piso again at A. 3.9.2 d?e), and Cicero was scathing about an earlier Piso for arriving bv night (PZs. 53 ‘nonne tibi nox erat pro die, solitudo pro frequentia?’, cf. Liv. 9.7.10-11 with Oakley on 9.6.7 (pp. 101—-2)); but Augustus preferred to arrive by night precisely to avoid being greeted by crowds (Suet. Aug. 53.2, Dio 54.10.4, 54.25.4). Palatium here, as often, denotes the complex of imperial buildings ('the palace’), not the Palatine Hill on which it was situated. Domitian was responsible for developing the site by building the Domus Flavia and the Domus Augustana (Jones (1992) 95-6). ita ut praeceptum erat: the meaning is not clear, but on the most cvnical interpretation the phrase is equivalent to ‘as the rules had dictated’: i.e. behaviour under a tyrant is, as it were, prescribed in advance. The only plausible alternative is to understand a Domitiano, and to assume that the princeps himself or one of his aides had instructed A. how to arrive. exceptusque breui osculo: the normal phrase osculo excipereis sometimes used of the formal greeting extended bv the emperor to important persons (Plin. Pan. 23.1, Suet. Otho 6.2); A. 15 dignified by no more than the breue osculum with which Nero used to greet his mother (A. 13.18.3). nullo sermone contrasts strongly with the many words of honour with which Domitian praised A. publicly before he left his command (1 above).

292

COMMENTARY:

40.4

turbae seruientium inmixtus est: the verb is usually taken as a genuine passive here: ‘he was made to mingle' (as 28.1 above) rather than ‘he mingled'. The combination with turbae (again at H. 3.74.1) is orig. Livian (3.50.10, 8.24.15, 22.60.2, 23.23.8). seruzentium (see 30.2n. nec) 15 the second in a sequence of substantival present participles (cf. occurrentzum above and laudantes at 41.1 below): for the phenomenon in general, of which T. is fond, see Adams (1973b). 40.4 uti militare nomen...aliis uirtutibus temperaret 'in order to moderate his military reputation by means of other urrtutes': this is (as it were) A.'s way of responding to Domitian's reaction to his nomen at 39.2 above (see also next n.). militare nomen had occurred at Sall. H. 3.48.6 (where, however, nomen — 'pretext'). Itis of course ironical to describe tranquillitas and ofzum (below) in terms of uzrtutes (6.3n. quiete). graue inter otiosos: in view of the earlier contrast between Domitian's pretensions and A.'s real military achievements (39.1-2), it seems likely that otiosos — ‘the unmilitary' and includes the princeps; the meaning of graue will depend upon its focalisation, whether Domitian's ('unpleasant’, ‘a serious matter':

OLD

14a)

or A.'s (‘dangerous’).

Some, however,

believe that otiosos — ‘idlers’, who have time and inclination to praise a hero and thus endanger him. T. perhaps alludes to Liv. 7.4.3 'cognomen ...Imperiosi, graue liberae ciuitati’, although graue nomen (again at A. 3.24.2) is not uncommon. tranquillitatem atque otium penitus duxit: the paradosis reads auxit, which Wex emended to hausit. The emendation is generally accepted because

(i) penitus hausisseis at D. 30.3, (11) the verb 15 said to sustain the

metaphor in temperaret (‘dilute’) above. It seems likely, however, that T. 15 alluding to, and reversing, the peroration of Cicero's Pro Archia (30 ‘cum usque ad extremum spatium nullum tranquillum atque otiosum spiritum duxerimus'), a passage which he may well have in mind again at 46.1 and 46.3 below (nn.). duxitis closer to the paradosis, and for the meaning (‘he drew

deep breaths of tranquillitas and otzum') see OLD duco 25a. For the expr. note also Ov. Ex P. 1.5.43—4 'qui segnia ducam | otia’, where, however, Gaertner prefers the meaning ‘spend’. The combination of the two nouns tranquillitas and otium 15 above all Ciceronian: he uses it several times to describe either civic and political stability (P?s. g, Sest. 110) or disengagement from political life (Mur. 55, Marc. 27): T. usesitin the former sense at D. 8.2 and in the latter sense here. In view of the allusion to the elder Cato at the start of Agr. (1.1n. Clarorum), it

is relevant that Cicero says of Cato that even as an older man he preferred not to live ‘in illa tranquillitate atque otio' (Rep. 1.1): his reason was that the existence of u?rtus depended upon its use, and ‘usus...eius est maximus ciuitatis gubernatio' (Rep. 1.2). T. is making the paradoxical point that, under Domitian, uzrtus for A. consisted in precisely the opposite of

COMMENTARY:

40.4

293

what it meant for Cato (see also above, 6.95n. quiete). Moreover, although otium here has a different shade of meaning from ot2osos above, the very identity of the terminology suggests that, by following otzum, À. is coming to resemble the ofzoszamongst whom Domitian himself was included: 566 below, 41.1n. (znfensus), for the motif that tyrants are surrounded by men like themselves. cultu modicus...uno aut altero amicorum comitatus: just as ostentatious dress and the like bespoke majesty and importance, so a modest appearance denoted the opposite: so e.g. the future Ptolemy I (Curt. 9.8.23 ‘modico ciuilique cultu, liberalis imprimis adituque facili nihil ex fastu regiae adsumpserat') and esp. Mark Antony (App. BC 5.76 ‘he spent the winter at Athens...exchanging the display of a commander for the simplicity of private life, wearing the square-cut pallium and the Attic shoe, and with no crowd at his doors. He went out, in like manner, without the

insignia of office, accompanied by two friends and two attendants...At the end of the winter he was like another man. He changed his clothing, and with his clothing his whole appearance. There was straightway a crowd around hisdoors...and all things thatinspire terror and awe' (Loeb trans.)). Note also Agesilaus at Xen. Ages. 11.11. See in general WallaceHadrill (1982) 33—5. Note the partitive genitive amicorum: it was not that A. did not have many friends but that he thought it prudent to appear in public with only one or two (again cf. Seneca at A. 14.50.93 ‘uitat comitantis'). For his characteristic moderation and modesty see 4.3n. (retinuitque). For cultu modicus cf. Plin. Pan. 88.7. comitatus is regularly constructed with the simple abl. rather than by a(5), e.g. A. 14.8.4, Cic. Cael. 34, Virg. Aen. 1.312, Ov. F. 3.603, Curt. 8.14.15, 10.8.3: ‘the *accompanying" is not regarded as an act done to another person, but as a simple act of association' (Austin on Virg. loc. cit.).

adeo ut...interpretarentur: since the point depends on those features of A. which people could see (uiso aspectoquebelow), adeoappears to qualify the whole of cultu. .. comitatus: for this (common)

use see M-W on A. 4.1.2.

per ambitionem ‘by their display' (OLD ambitio 6). aestimare (5.1n. electus) + per (instead of an abl.) is most unusual but cf. Suet. Calig. 60 ‘per haec aestimare'. For A. and ambitio see 29.1n. uiso aspectoque: the two verbs, which seem to be used without distinction (cf. A. 3.45.1, Cic. Phil. 5.29, T?m. 13, Val. Max. 2.1.5), are picked up at 45.2: everyone was watching everyone else; the question was whether they were taken in by appearances (below, n. pauct). quaererent famam: normally famam quaerere = ‘to find/seek a reputation' (see 9.4n. ne famam, 14.2), but here the meaning is 'inquired about’ (OLD 8a).

pauci interpretarentur: ‘Do not supply famam' (Sleeman); ‘It is natural to supply famam' (O-R). The seeming parallelism of plerique. .. quaererent

294

COMMENTARY:

41.1

famam ~ pauct interpretarentur suggests the latter; others wish to supplv ‘his modest behaviour’ from earlier in the sentence, but this seems extremely difficult; Kritz plausibly understood eum (= Agricola) from the preceding abl. abs.: at H. 1.14.2 ?nterpretari is again used with reference to a person but without an expressed object ('Piso ...deterius interpretantibus tristior habebatur’). The 'discerning few' (again at A. 4.33.2) reflect poorly on the majority, who, judging only by appearance, failed to realise the explanation for A.'s reversals of behaviour. 41.1 accusatus...absolutus est: the verbs are to be understood informally; they do not refer to formal prosecutions and acquittals. querela laesi cuiusquam 'complaint from any wronged person’. infensus uirtutibus princeps 'the princeps hostility to uzrtutes': this 15 a nominative ex. of the so-called 'ab urbe condita construction' (10.4n. Romana), except with an adj. instead of a participle. T. here picks up the phrase infesta uirtutibus tempora at the very start of the biography (1.4) and assigns the responsibility to Domitian in person. It was a commonplace that tyrants hate excellence and/or eliminate good men (e.g. A. 16.21.1 ‘uirtutem ipsam exscindere concupiuit' (Nero), Hdt. 5.92, Eur. Suppl. 444—0, Plato, Leg. 832c, Arist. Pol. 1284a26—-33, 1311a20-2, Sall. C. 7.2 ‘regibus boni quam mali suspectiores sunt, semperque iis aliena uirtus formidolosa est’, Plin. Ep. 2.1.9 'suspectus atque etiam inuisus uirtutibus', Pan. 14.5 'iners ipse alienisque uirtutibus tunc quoque inuidus imperator' (of Domitian), 68.3 'cum a malo principe tamquam successor timeatur quisquis est dignior’), of which the converse is that thev gravitate towards the base and/or are surrounded by men like themselves (e.g. Eur.

Ion 62*7—8); very often these two ideasare combined (e.g. Hdt. 3.80.4, Xen. Hiero 5.1—2 (cf. 6.15), Plato, Rep. 5675-5684, Arist. Pol. 1314a2-10): note esp. Tiberius at Tac. A. 1.80.2 'neque enim eminentis uirtutes sectabatur et rursum uitia oderat: ex optimis periculum sibi, a pessimis dedecus publicum metuebat’. See also W. on Vell. 88.2. Alone of prose authors T. has a general preference for infensus over infestus, though the latter predominates in the minor works (3:1 in Agr.); in verse Virgil is the only author who shows the same preference, albeit marginally: see W-M on A. 3.15.2. gloria uiri: this pregnant use of uzr, sometimes instead of a pronoun (again at H. 2.68.4, 3.80.2, A. 2.03.3 ‘magnitudinem uiri’, 12.19.1 'claritudine uiri’), connotes impressiveness and 15 common in epic (e.g. Virg. Aen. 6.174; the oblique cases of ?s were generally avoided in verse). It may be relevant that some addressed Domitian as 'Dominus et Deus' (Dio 67.13.3—4; Gibson on Stat. Silv. 5.1.37), whereas A.'s glory was merely that of a urr. pessimum inimicorum genus, laudantes: the immediately following context (2—3) suggests that T. 15 referring to genuine eulogists who, by

COMMENTARY:

41.2

295

endangering A. unwittingly, were tantamount to enemies; but at the very end he introduces genuine enemies who deplov false praise (4 'pessimi...extimulabant): pessemum...laudantes 15 applicable to both groups. The eulogists are the third item in a list, the various elements of which (princeps, uirtutibus, gloria, laudantes), are picked up in ring composition at the end of the paragraph (~ 4 sermonibus, principem, uirtutibus, gloria). For the substantival participle laudantes see above, 40.3n. (turbae). 41.2 et ea insecuta sunt rei publicae tempora quae...non sinerent: ef confirms and illustrates the preceding: ‘And in fact there followed such conditions of the res publica as did not allow...' (the rel. clause 15 consecutive or ‘of characteristic’: NLS8158, A-G 8$534—5). Scholars have detected an allusion to Cic. De or. 3.8 'sed ei tamen rem publicam casus secuti sunt...', a famous passage to which T. returns at 43.1 (n. Finzs) and 45.1 (n. non u1dit).

sileri Agricolam non sinerent 'did notallow A. to go unmentioned' (OLD sileo 3b); to describe the praising of A., T. alludes to the formula non (te) silebo, common in works of praise (W-M on A. 3.65.1). The clamour at the end of the biography (note esp. 4 below) contrasts strikingly with the silence which has marked both Domitian's reign (2.3, 3.2, 39.2) and A.’s own career (6.4); even the battle of Mons Graupius had ended in an eerie silence (38.2). On silence in T. see Strocchio (1992). tot exercitus...amissi, tot militares uiri...capti: these participial expressions explain, and are in apposition to, fempora above. The Dacians had crossed the Danube, invaded the province of Moesia and killed its governor, Oppius Sabinus; subsequently the praetorian prefect, Cornelius Fuscus, attempted a counter-attack across the Danube into

Dacia but was killed and his army wiped out. The precise date of each event is disputed, but the general period was 84—86. Towards the end of the decade Domitian attacked the German tribes of the Marcomanni and Quadi and was defeated; and in 92 a legion and its commander were lost when the Sarmatians invaded Pannonia. See Suet. Dom. 6 with Jones' nn., Dio 67.7-8, 10; also Jones (1992) 126—-59. For the co-ordination of the places (A Bque et C et D) see 12.5n.

(praeter).

temeritate aut per ignauiam ducum: femeritas was one of the factors which Cicero expected to be singled out in a historical narrative (De or. 2.63), and temeritas ducum is a regular feature: see e.g. Sall. ( 85.47, Suet. Tib. 18.1; Oakley on Liv. 6.22.6. For the contrast with zgnauza, found in numerous authors, see e.g. Liv. 2.65.4, 6.24.6, 9.5.6. For the uaratio of abl. — per+ acc. 566 6.1n. (per mutuam). militares uiri is a phrase used several times by T. to designate 'fighting men' (Wellesley on H. 3.73.2): they may be as important as Vespasian (H. 2.75.1) or Corbulo (A. 15.26.3) or anonymous (A. 15.10.1): see also A. 4.42.2, 15.67.3 (tribunes). The phrase is absent from Caesar,

296

COMMENTARY:

41.3

Sallust and Curtius but used by Livy (10.24.4, 30.15.13, 30.37.9, 35.26.10), from whom T. perhaps derived it. In modern scholarship the phrase has sometimes been used to designate individuals from the Flavian period onwards whose careers had certain distinguishing features such as an early consulship prior to the holding of an important military command: A. himself is the classic example (for discussion see Campbell (1975), Saddington (1991)). expugnati et capti: the application of these verbs to one and the same subject/object 15 Livian (26.47.9 'naues...expugnatae captaeque’, 309.31.17, 40.14.11), though he never applies the combination to persons; expugnare of persons is first in Caes. G. 7.10.1 (Cic. Verr. 4.94 deos, by metonymy for lemplum). de limite imperii et ripa: the bank is that of the Danube; since the Danube formed the frontier (/zmite), etis probably epexegetic. In the same way the Rhine 15 described as a limes at Pan. Lat. 10.7.3. In a discussion of the meaning of the term /zmes, however, Isaac (1988) maintains that T. 15 distinguishing between a land- and river-boundary (128). possessione: sc. emperiz; scholars have compared Cic. Acad. 2.132 ‘non de terminis sed de tota possessione contentio' (philosophical metaphor). Shaw-Smith (1995: 14) would insert prouinciarum before possessione. 41.3 cum damna damnis continuarentur: such polvptoton is common (e.g. Pan. Lat. 12.22.2 ‘bellis bella continuas’), and lends itself esp. to tragic circumstances (see e.g. Wills (1996) 220 on Seneca) or cynical comment (e.g. Val. Max. 7.4 ext. 1 'ut metum metu, uim ui discuteret’; H. on H. 2.68.3). omnis annus funeribus et cladibus insigniretur: this seems to be a Tacitean variant on the Livian 2znsign?s annus (e.g. 2.21.5 'insignis hicannus est nuntio Tarquini mortis’, 4.12.6 'annus...multiplici clade ac periculo insignis'). The coupling of funeraand cladesis also Livian (9.22.5, 28.41.3). It is as if the speakers have been reading 'Livian' reports of the army disasters. omnis annus — ‘every year'. poscebatur...dux Agricola: compare Vell. 111.2 ‘ut praesidium militum res publica ab Augusto ducem in bellum poposcit Tiberium' (of the Pannonian war of 6). For ore uulgi cf. 2.2n. (uocem).

comparantibus cunctis: contemporaries now perform the kind of syncri525 which T. himself had performed at 13.1—17.2 (above, pp. 158-9). For uigorem, constantiam (below) cf. Plin. NH 7.91 on Julius Caesar; bellis (-0) expertus is a common phrase (e.g. Virg. Aen. 7.235, Pollio, fr. 8P = 11C). ceterorum: the paradosis eorum seems intolerably flat and obscure. Some scholars have proposed a lacuna, for which Halm suggested quzbus exercitus committi solerent. But the apparent contrasts uzgorem — ?nertia and constantiam ~ formidine suggest that expertum bellis anzmum requires a third contrasting element such as qui potius tirones quam duces agerent (‘who were

COMMENTARY:

41.4

997

acting more like novices than leaders'). Most readers, however, are likely

to find this too bold; and, of the more economical proposals which have

been made, ceterorum seems the best.

41.4 aures uerberatas: T. likes to focus on ears ous extended senses (e.g. D. 9.2, 10.8, H. 3.10.3, 4.69.1, 3); ‘assailing the ears’ occurs as earlv as uerbero' 4). For Domitian's ears cf. Juv. 4.86 ‘quid

and/or use them in vari3.50.3, À. 1.31.5, 4.29.3, Plaut. Amph. 333 (OLD uiolentius aure tyranni?’

For satis constat (above) see 13.2n.

dum...exstimulabant: better sense seems to result if dumis regarded as

causal (

as much as...’),a use of which T. is fond, rather than temporal,

but there is admittedly no parallel for its combination with a tense other than the present (G-G 322b). exstzmulare can be used with a desirable or undesirable subject (OLD b): here it has both.

libertorum: for Domitian's freedmen see Jones (1992) 61-9. amore et fide...malignitate et liuore: sc. erga Domitianum with the first pair of nouns, and erga Agricolam with the second. For malicious praise see esp. Pan. Lat. 3.4.5—5.1 ('accusandi genus quod nullus refelleret’). amor and fides are quite commonly combined; for malignitas and liuor cf. Plin. Pan. 58.4, Suet. Calig. 34.1. deterioribus: masculine or neuter? O-R repeat the statement of earlier edd. that T. constructs pronus with 171 - acc. for persons, the implication being that he uses the dative for things; but this is not so: for the dat. pers. Cf. 33.4, H. 3.64.1 (both wictoribus); for zn of things cf. 6.2 ‘in omnem auiditatem', H.

1.69 'in misericordiam', 4.76.2

‘in arma’; for the dat. of

things cf. A. 4.29.3 offensioni. Grammar alone cannot therefore answer the question. deterioribus cannot be a varied repetition of pessimi, which seems very weak and implies wrongly that Domitian has seen through the pessima. The term must refer either to such persons as the inferior generals or to the failings by which inferior persons are characterised. Since T. in the next sentence

mentions

u2tza as a decisive factor

(next n.), deterzoribus 15

perhaps neuter; but he may have wanted to capitalise on the ambiguity of the grammatical form. simul

suis

uirtutibus,

simul

uitiis

aliorum:

scholars

are

divided

on

whether the witia are the 'flaws' of the deficient generals (2—3) or the ‘vices’ of the lbertorum pessimz; the contrast with suzs uzrtutibus suggests the former. T. is alluding (with added chiasmus) to Cic. Leg. Man. 67 'quasi uero Cn. Pompeium non cum suis uirtutibus, tum etiam alienis uitiis magnum esse uideamus', where Cicero is urging supreme command to be given to Pompey (68 'quare nolite dubitare quin huic uni credatis omnia’). A. seems to be developing into the Pompey-like figure whom Domitian dreaded (39.2n. priuati); indeed it has been argued (by Cova (1999) 137-9) that the four cardinal military qualities of scientia rei militaris, uzrtus, auctoritas and felicitas which Cicero claimed for Pompey

298

COMMENTARY:

42.1

(Leg. Man. 28—48) are displaved above all by A. in the course of T.'s biography. See further 42.4n. (magnos). For simul...sitmulsee 25.1n. (cum). in ipsa gloria praeceps agebatur: the paradosis reads 1271 ?psam gloriam, which seems contradicted by the fact that A. has been endangered by gloria for some considerable time (esp. 1 ‘gloria uiri'). Emendation seems required, and :fsa gloria was suggested by Madvig and znter ipsam gloriam by Baehrens. The former results in an awkward double abl., the latter seems

questionable Latin and is certainly unparalleled. E. Courtney's zn zpsa gloria (‘in the midst of his very glory’) avoids these difficulties and is preferable to either. It would be natural for a scribe to write in - acc. with praeceps agi, but no destination 15 required to be expressed (see e.g. Sall. C. 31.9). praeceps, which 15 commonly combined with agz is similar in meaning to, if perhaps rather stronger than, pronus (e.g. Sen. Ep. 97.10 ‘non pronum est tantum ad uitia sed praeceps’): A. and Domitian are each on a downward path. See further 42.3n. (Domitiani). 42.1 Aderat iam annus: :am(que) aderatis a common expr. in both prose and verse to signal the next stage in a story, but only here (it seems) are adv. and verb transposed. For adsum of time see OLD 10. The year was probably 9o; T. marks time differently within the Britain narrative (where each year is narrated, however brieflv) and without. quo...sortiretur ‘in which he was to cast lots for' (the subject 15 A. and not, despite Soverini, Domitian): this is a regular but infrequent use of the imperf. subjunc. to express futurity from a past point of view (Handford (1947) 84-5). Africa and Asia were the two most important and illustrious ‘public’ or senatorial provinces, and they were awarded to the two most senior ex-consuls bv lot. At this time the normal interval between consulship and governorship was about a dozen years: T.'s own governorship of Asia in 112/13 involved a slightly longer interval, and this would also have been the case with A. too, on the assumptions that his consulship had been in 76 and that annus here refers to 9o. If A.'s name had gone forwards to sortition, he had a fifty-per-cent chance of being appointed to Africa, the only senatorial province with a legionary force (Legio III Augusta); this may well have been an important consideration in the charade which follows. occiso Ciuica nuper: C. Vettulenus Civica Cerealis was put to death in 87/88 while actually proconsul of Asia: Suetonius uses his case to illustrate Domitian's saeuitia and says that the man had been plotting revolution (Dom. 10.1—2 and Jones' n.). nec...consilium deerat nec...exemplum: this reads like the scenario with which T. himself was confronted at the time when he proposed to write A.'s biography (1.4n. at nunc): with the killing of Arulenus Rusticus and Herennius Senecio, Domitian had created a precedent (exemplum) which acted for T. as a warning (consilium). This is one of several passages

COMMENTARY:

42.2

299

where there is a parallel between T. and A. and also where the end of the biography resonates with the beginning. qui...ultro...interrogarent ‘to ask A. of their own accord...’. ultro (OLD ultro 5a, cf. 6b) implies that the peritz approached A. without Domitian's knowledge, but later it seems that Domitian knew what was happening (1 pertraxere ad Domitium, 2 . paratus... compositus); whether this means that initially the perit had read the princeps mind (cf. A. 3.22.2 and W-M), or whether they were aware of his intentions and simply devised their own way of fulfilling them, 15 unclear. Though the perit (primo...mox...postremo...) and Domitian (paratus — compositus, et audüt ~ et passus est) each go through their choreographed parts, the picture is impressionistic and suggestive rather than precisely detailed. The introduction of a simple indir. question by -ne (OLD -ne 4a) is only here in Agr. and once in H. (2.76.1, in a speech), though there are a few more exs. in A. (G-G g11b).

primo occultius quietem et otium laudare: dramatic irony: Domitian's agents use the very terms which had guided A.’s life at the very start of his career (6.3 and n. quiete). occultius (~ non iam obscuri suadentes below) refers

to the agents' indirect manner of speech (OLD occulte 2): A. has to read between the lines, and, since he was already leading a life of tranquillitas and otum (40.4), we are invited to imagine that he was completelv baffled by this initial approach. in adprobanda excusatione 'in winning approval for...': excusatio here — ‘a plea to be excused from carrying out...a public duty’ (OLD 2); for adprobare see 5.1n. operam offerre — ‘to offer support’ (as e.g. Liv. 40.23.1, Curt. 6.2.20; cf. OLD opera 6): just as A.'s agreement ο the plan is taken for granted (adprobanda), so the agents, adding insult to injury, do not credit him with the abilitv to carry it out successfully. Yet this in its turn isa tribute to his straightforwardness, in contrast to the deceitfulness of the princeps (2 ‘paratus simulationi’).

non iam obscuri suadentes simul terrentesque ‘no longer cryptic urging and terrorising simultaneously’: for the word order cf. H. 4.77.1 ‘pugnari simul uincique suos’, A. 5.10.2 ‘fingebant simul credebantque’; for obscurus of speakers or writers see OLD 8b. pertraxere: on the assumption that A. was not literally dragged by force, the verb must mean ‘constrained him to come' (OLD 2b); vet it perhaps still retains the implication that the result would be ‘punishment or some other unpleasant fate' (OLD 2a).

42.2 paratus simulationi, in adrogantiam compositus: the transmitted paratus simulatione can mean only ‘prepared by simulation"; but this invites the question 'prepared for what?' and does not correspond to the story, since Domitian's role in the pretence is still to come. We expect something like ‘ready with his simulation' (hence Ernesti's parata stmulatione) or

300

COMMENTARY:

42.2

‘prepared for simulation' (hence the Bipontine ed.’s paratus simulation?). The latter is the easier change, and, for the uariatio of dat. ~ prep. phrase (here arranged chiastically), cf. 6.4n. (utz?). There is perhaps also a distant

allusion to Sall. C. 31.7 'paratus ad dissimulanda omnia'. componere in + acc., ‘to modify one's appearance (so as to give a particular impression)’, is peculiarly Tacitean (W-M on A. 3.44.4; OLD 11); for paratus — compositus cf. Cic. I Verr. 32, HA Ant. Diad. 3.2. preces excusantis: A. carries out the plan exactly (— 1 excusatzone); for the absolute use of the verb

('to excuse oneself’ from

something)

cf. D.

5.1, Suet. Aug. 35.2, Apul. Met. 1.26.1, 2.18.1 (OLD 4a). The combination of preces and adnuisset (below) again suggests Domitian as the 'Dominus et Deus' (41.1n. gloria). agi sibi gratias passus est: ‘The way to live long at court was to return thanks for injuries received' (Sleeman): cf. H. 2.71.2, À. 14.50.3 'Seneca (qui finis omnium cum dominante sermonum) grates agit', 14.64.3, 15.71.1, Sen. Ira 2.33.2, Tranq. 14.5 'agebant enim gratias et quorum liberi occisi et quorum bona ablata erant', Ben. 4.17.3 'gratias agi sibi ab iis quos adflixere patiuntur' (a passage which T. perhaps remembered). audiüt . . . passus estis again chiastic. nec erubuit beneficii inuidia 'and he did not blush at the invidiousness of his good deed’: not onlv beneficzz but also the whole sentence 15 ironical, since Domitian's naturally ruddy complexion (45.2n. cum) meant that he could not be seen blushing at even his most outrageous actions. Contrast Tiberiusat A. 6.23.1 ‘nec erubuit’. For the connection between znuzdiaand shame 566 Kaster (2005) 92-9 (esp. 95 for this episode). For the causal abl. cf. G. 28.4; OLD erubesco 1b. This sentence is itself chiastic with the preceding, so that audzt...2nuidia produces the arrangementABBAAB (A= main verbs): for such arrangements see Kraus (1994) 22. salarium: this term, first attested in the elder Pliny (see OLD), refers to the pay which was introduced by Augustus for provincial governors (see Dio 52.23.1, 53.15.4 with Rich's n.). solitum offerri et quibusdam a se ipso concessum: if the salary was usually offered but actually granted in only a few cases, the implication is that the beneficiary was expected to decline it. The offer was a mere ritual (see also next n.). The phrase is yet another chiasmus. offensus non petitum: sc. ?d and esse, for the rare acc. + inf. see OLD offendo 7b (adding Suet. Tib. 34.2, Serv. Aen. 4.484). Since the salary was ‘usually offered' (above), we must infer that it was not offered on this occasion and that A.'s subsequent failure to ask for it constituted a possible reason for Domitian's not awarding it to him. The whole court scene is an elaborate game. For the uariatio of participle —^ prep. phrase (ex conscientia) see 5.2n. (non sane); conscientia here — 'a sense of guilt' (OLD gd).

COMMENTARY:

42.3

301

ne quod uetuerat uideretur emisse: what Domitian had forbidden was A.'s sortition for the proconsulship of Africa or Asia (1); if he had nevertheless given him a proconsular salary, he might be said to have bought him off. If there was no salary, there was no public evidence to connect A. with either province: the real reason for his failure to participate in sortition (viz. the threat of violence: 1 terrentes) remained secret. For a different

interpretation see Courtnev (1971). 42.3 Proprium humani ingenii est odisse quem laeseris: this famous sententia functions primarily as a foil for the following statement about Domitian, but it is also the first sign of closure as the penultimate section of the biography nears its end (for epigrams as closural devices see Smith (1968) 196-210). For the sentiment (with which A. 1.33.1 15 compared by G. ad loc.) see Sen. /ra 2.33.1 ‘quos laeserunt, et oderunt’. For proprium see OLD 4b; the constr. with the inf. is common. /aeseris 15 an ex. of the second-person (perfect) subjunctive (3.1n. sic), common in 'gnomic generalizations' (Gilmartin (1975) 103). Domitiani uero natura praeceps in iram...inreuocabilior: this general statement, in which another of Domitian's natural proclivities (cf. 41.4 ‘pronum deterioribus’) contrasts with the headlong descent forced upon A. (41.4 'praeceps agebatur’), is adversative in relation to the preceding sententia

(cf.

uero)

but concessive

in relation

to what

follows

(cf.

tamen

below). praeceps in is quite common (H. on H. 1.24.1); for zn ?ram cf. Liv. 293.7.12 ‘quamquam praeceps ingenio in iram erat' (of Hannibal); natura praeceps is Sallustian ( 6.3, g2.5). It is unclear whether the subject of quo . . . inreuocabilior 15 natura or Domitian, but the latter 15 perhaps more likely; the comparative ?nreuocabilior occurs only here. moderatione ... prudentiaque: for this combination cf. Cic. Tusc. 5.101, Cels. 7.29.1, Val. Max. 9.3.2. non contumacia neque inani iactatione libertatis: contumacia and 115 cognates are regular in T. to describe those who challenge the status quo: 566 Damon on Z. 1.3.1 and 1.51.4 (‘T.’s warmest praises go to those who show how to avoid it’). This is the final occasion on which T. refers to A.'s

avoidance of self-advertisement. znan?is picked up by zn nullum...usumat 4 below; its combination with zactatione is Senecan

(Ben. 2.11.6, Brev. vit.

13.7; tnanis iactantia at Quint. 1.8.18). neque...famam fatumque prouocabat 'he would not...encourage fame and «therefore- his fate’ (OLD prouoco 2): A. maintains to the end of his career the attempt to avoid fama which has characterised his whole life (see also Hardie

(2012)

281). The

combination

of the two nouns is

Virgilian (Aen. 7.79, 8.731, the latter closing a prominent passage already alluded to by T.: see 6.5n. tum); for their etymological connection (both were derived from far?) see Maltby 222.

302

COMMENTARY: 42.4

42.4 sciant...: for the sententious jussive subjunctive cf. Sen. Suas. 6.1 'sciant posteri potuisse Antonio seruire rem publicam, non potuisse Ciceronem’, Sen. Ep. 95.28 'sciant protinus hi qui iactationem ex istis peti et gloriam aiebant non ostendi ista sed conscientiae dari'. Though the acc. - inf. which 15 thereby introduced is a generalisation, its terms have an obvious relevance to A. quibus moris est inlicita mirari ‘whose custom is to admire unlawful conduct’: the attractions of the forbidden are proverbial; here inlicita means ‘not approved by the Emperor rather than contrary to the constitution' (O-R): the implication is that the behaviour of conformists like A. should

be the object of admiration instead (cf. 46.2 'admiratione te ...colamus"). For morzs cf. 33.1n. (Excepere). sub malis principibus: the malus princeps (again at 43.4, H. 2.37.2, 4.42.0) is a major theme of Pliny’s Panegyricus (e.g. 44.1—40.3, 53.3—06, 67.3—68.3) and also in the Historia Augusta. magnos uiros esse: if there have been earlier hints that A. is to be seen as a second Pompeius Magnus (39.2n. priuati, 41.4n. simul), the choice of adjective here seems significant (see further 44.2n. bonum). Another uzr magnus was Frontinus (17.2n.). obsequiumque .. . adsint: the balance required by the times was achieved by A. in his life and career (4.3n. retznuitque, 6.3n. quietre, 22.4n. ut). For obsequium . .. ac modestiam see 30.93n. (superbiam); for industria ac uigor cf. Vitr. 5.5.6, Plin. Ep. 1.14.7 (a letter of recommendation); for adesse— ‘to be present’

(of conditions, characteristics etc.) see OLD

17a.

(obsequiumque ac modestiam...) eo laudis excedere quo plerique per abrupta, sed...inclaruerunt: this passage has caused considerable difficulty: quo cannot be taken with inclaruerunt but requires a verb of motion which needs to be either understood or supplied. Some have supplied enzs? (‘having struggled up’) after either abrupta or usum, but neither seems at all plausible. Others have favoured plerigue « qui>, a verb of motion being understood with plerique: this detaches inclarueruntfrom quobut also places per abrupta in a different clause from the understood verb of motion with which it requires to be taken. O-R retain the paradosis, understanding (from excedere) a verb of motion with plerique and regarding sed as introducing a new main verb: ‘(compliance and modesty) reach the same peak of praise as many men Nero nevertheless did not gaze upon his victims. For the comparison between Nero and Domitian cf. Plin. Pan. 53.4 (also suggesting Domitian's sympathy for Nero), Juv. 4.38 and Courtney; for the sensitivity of the murderer cf. A. 12.47.5 ‘uisui tamen consuluit, ne coram interficeret’; also Sen. Contr. 5.8 ‘0 hominem

in sua crudelitate fastidiosum, qui,

cum uellet torquere, tamen nolebat audire!' uidere et aspici: these two elements look back to the immediately preceding sentences: our sight of the victims (uzsus) was painful enough but, whereas Nero could not bear to look upon his victims (‘subtraxit oculos suos...non spectauit’), a particularly distressing feature of Domitian's rule was that not only were we compelled to watch the victims (uzdere — uisus)

but the victims were gloated over by the princeps (aspici — subtraxit oculos suos . .. non spectauit). These visual references are framed by those to notseeing and seeing with which the paragraph began (non uzdit, the blind Catullus Messalinus) and will end (denotand?s); see also below, 5n. (desider-

auere). For discussion of the emperor as observer see Bartsch (19094) 1ff., esp. 31—5. According to Suetonius, Domitian had large eves but dull vision (Dom.

18.1

'grandibus oculis, uerum

acie hebetiore’). For the combina-

tion of these two verbs see 40.4n.(uso);

for active — passive cf. 5.1n.

(nosci).

cum suspiria nostra subscriberentur: compare the treatment received under Tiberius by Nero Caesar (A. 4.61.2 ‘cum uxor...suspiria... patefaceret') and Drusus Caesar (A. 6.24.1 'adstitisse...qui uultum, gemitus, occultum etiam murmur exciperent’). subscribo here — 'record, note down' (OLD 4): for this practice cf. e.g. A. 4.60.1, 4.67.4. cum...sufficeret saeuus ille uultus et rubor: the last two nouns are evidently a hendiadys: 'that savage red face'. The ideal ruler was supposed to look with benevolence on his people (Hor. Odes 4.5.6—7 'uultus ubi tuus | adfulsit populo’) and to have a countenance which matched his greatness (W. on Vell. g4.2); but the gaze and appearance of the tyrant were in every way the opposite of these (e.g. Cic. Off. 1.112 ‘moriendum potius quam tyranni uultus aspiciendus fuit', Hor. Odes 3.3.3 'uultus instantis tyranni’ and N-R, Sen. Contr. 2.5.4 'intentum tyranni uultum ...oculos minaces’, [Sen.] Oct. 108-10 ‘poena nam grauior nece est | uidere tumidos et truces...| uultus tyranni’). Domitian's face, like that of Pompey the Great (next n.), was famously flushed (Suet. Dom. 18.1

‘uultu...ruboris...pleno’, Plin. Pan. 48.4 'ipse...uisu...terribilis: super-

bia in fronte, ira in oculis,...in

ore impudentia

multo

rubore

suffusa',

390

COMMENTARY: 45.3

cf. Philostr. Vit. Apoll. 7.28; above, 42.2n. (mec erubuit), Hulls (2007) 202—4) and is in striking contrast to the pallor of his terrified subjects: the abstract hominum palloribus instead of hominibus pallentibus (cf. 21.2n. ingenia) is esp. graphic. Cf. Juv. 4.74—5 Π quorum facie miserae magnaeque sedebat | pallor amicitiae'. denotare recurs in T. only at A. 3.53.1 'denotantibus uobis ora ac metum singulorum', where it may mean either ‘point out' (OLD 2a) or simply ‘observe’ (OLD 3): see W-M's n. The same question arises here; but aspic: above supports the latter, which also seems more sinister (cf. Cic. Cat. 1.2 ‘notat et designat oculis ad caedem unum quemque nostrum’). sufficeret appears to be ironic, 'had sufficient strength/capacity for ~’, 'stood up to' (OLD 4a).

quo se contra pudorem muniebat: an anonymous verse about Pompey the Great (Vers. Pop. 2 (in Courtney, FLP) 'quem non pudet et rubet non est homo sed sopio’) jokes obscenely on his combination of ruddiness and shamelessness: whereas blushing was normally a sign of shame (e.g. Ter. Andr. 878 ‘color pudoris signum', Cic. Tusc. 4.19 ‘fit ut pudorem rubor...consequatur’), a naturally red complexion hid blushing and 50 was regarded as a sign of shamelessness: see La Penna (1975). On blushing see further Sen. Ep. 11.1—7; Kaster (2005) 20-2 and n. 2. 45.3 tu uero felix ... opportunitate mortis: uero (OLD ga) confirms the preceding paragraph, whose two topics of a distinguished life (44.1—4) and an opportune death (44.5—45.2) are summed up by uitae... claritate (an unparalleled expr.) and opportunitate mortis respectively. The apostrophe, evidently a standard feature of funerary literature (van Dam on Stat. Silv. 2.1.227—30; and cf. the Laudatio Turiae, with Horsfall (1983) 89), raises the emotional tone (cf. Rhet. Herenn. 4.22 'exclamatio est quae conficit significationem doloris’), and the whole sentence is another allusion to Cic. De or. 3.12: 'ego uero te, Crasse, cum uitae flore tum mortis opportunitate ...exstinctum

6556 arbitror', which

T. has made

chiastic

(mortzs

opportunitas recurs at Cic. Brut. 4); cf. also Xen. Ages. 10.3 θάνατος óopoíos, Virg. Aen. 11.159 'felix morte tua neque in hunc seruata dolorem' (with Horsfall), Sen. Cons. Marc. 20.4 'cogita quantum boni opportuna mors habeat’, 5 ‘felix mori potuit’; Gibson on Stat. Silv. 5.1.221. In the minor works T. follows Cicero in using claritas to the exclusion of claritudo; in H. he uses both equally, but in A. the archaising claritudo has almost ousted its synonym (Syme (1958) 716, H. on H. 2.78.2). For A.’s ‘brilliance’ cf. also 1.1 Clarorum, 18.5 clarus... Agricola.

qui interfuere nouissimis sermonibus tuis: the nouussima uerba of a dving person are a standard feature of death literature: see e.g. A. 2.71.3 (of the dving Germanicus), 15.62-3 (Seneca), Prop. 3.7.55, 57-64, El. Maec. 2.1-34, Cons. Liv. 307-8, Stat. Silv. 2.1.148-53 (and van Dam on 152), 5.1.176-93 (and Gibson on 17560). The reference here prepares us for the

COMMENTARY: 45.4 information

321

(below: 4—5 sed...) that T. and his wife were absent at the

time of A.’s death and thus missed his last words; but it also continues the

allusion to Cic. De Or. 3.16 ('nos enim, qui ipsi sermoni non interfuissemus... ). For perhibent sec 10.5n. constans et libens fatum excepisti, tamquam...innocentiam principi donares ‘you accepted your fate resolutely and willingly, as if you were doing everything within a man's power to bestow innocence on the princeps . Given the contemporary rumour that A. had been poisoned by Domitian (49.2), I.'s point seems to be that, so far from voicing accusations against the frinceps in his ‘final conversations' in the way that Germanicus on his death-bed alleged that he had been poisoned by Piso and Plancina (A. 2.70.1, 71.1-72.1), À. by his demeanour suggested that nothing was untoward (constans perhaps looks back to constans rumor at 49.2). The sentence does not settle the question whether or not À. was actually poisoned by Domitian; nor is it clear whether tamguam. .. donares is T.'s own comment or is imagined as spoken by A.'s interlocutors, although the latter seems more likely in the context. The language of the main clause is Senecan (Ep. 76.29 ‘excipere fatum', 95.5 'ne constanter quidem aut libenter'); for innocentiam donare cf. H. 4.58.6, where Vocula prays to the gods to save the troops from commitüng the crime of mutiny ('militibus Romanis...innocentiam detis'). pro utrili portione is a rare variant on pro uinli parte ( OLD portio 1a). 45.4 sed mihi filiaeque eius...non contigit: here T. begins to explain the one respect in which the timing of A.’s death was inopportune: he and A.'s daughter could not be present at his death-bed since they happened to be abroad. The fronting of the datives mhi filiaeque both contrasts with qui interfuere. . . tuis above and suggests that they are to be taken alike with erepti, augeat and non contigil. acerbitatem parentis erepti: acerbitas presumably has the same associations with premature death as the adj. acerbus (OLD acerbus 4): cf. CIL 14.9977 'soror et mater acerbitate accepta fecerunt [sc. monumentum]'. eripiois commonly used of death (OLD 3a) and is esp. emotive when used of close relatives (e.g. Aug. in Suet. 71b. 29.1 'quoniam atrox fortuna Gaium et Lucium filios mihi eripuit’); the image presumably derives from 'the hands of Death' (Tib. 1.3.4; McKeown on Ov. Am. 2.6.39—40, van Dam on Stat. Szlv. 2.1.197—9). For the substantival use of the participle ('the scizing of one’s parent') sce 7.2n. (nuntio). auget maestitiam quod ...non contigit: in his advice for the conquestio (lament), Cicero's eighth point is that one should mention something which did not happen but which ought to have done: 'hoc modo “non adfui, non uidi, non postremam uocem eius audiui, non extremum spiritum eius excepi"' (/nv. 1.108). Cf. also Cons. Liv. 95-9, Sen. Cons. Marc. 9.2, Plut. Cons. Apoll. 117B.

392

COMMENTARY: 45.5

The apparent echo here (auget maestitiam . . . deficientem) of 43.2—3 (augebat miserationem . . . deficientis) underlines the fact that T. has described A.'s death twice, once as the conclusion to the strictly biographical narrative (43.1—4) and again as part of the conquestio here (45.3—5); this 15 another example, albeit on a smaller scale, of the duality evident in his accounts of

Britain (above, p. 159). T. and his wife had already been absent from Rome for four years when A. died (5 below): the nature of his absence is unknown

ualetudini...deficientem:

(Intro. p. 6).

for the uariatio of abstract — concrete see

20.2n. (laudare). Both adsidere (OLD 1c) and fouere (OLD 5b) are esp. used

of attending and caring for the sick. satiari uultu complexuque: some take uultu as referring to the couple's (T.'s and his wife's) gazing at A., but this is not what the noun means. uultu here properly means 'the distinctive appearance of an individual's face' (OLD 4), with perhaps the strong additionalimplication of its owner's gaze

(cf. OLD ga; Fitch on Sen. HF640-1);

complexu, on the other hand,

presumably refers to the couple's embracing of A. For satiari cf. OLD 3. For the last embraces and/or sight of a dying person see Eleg. Maec. 2.0--10, Prop. 4.11.64, Cons. Lzv. 89—94, Stat. Silv. 5.1.173—5, 194 (and Gibson's n.), 5.8—9, Quint. 6 pr. 12 (also below, n. desiderauere); for the expr. cf. Liv.

37.37.8 'satiatus...complexu filii’; also Sen. Tro. 761—2 ‘amplexu ultimo | auidos dolores satio'. excepissemus certe mandata uocesque: sc. ‘if we had been present'. The verb, repeated from g above in a different sense (the figure traductio), implies both that A.'s final words were precious (‘we would have seized on’) and that they were difficult to hear (‘we would have caught, picked up ): see OLD excipio 14a. A dying person's mandata or last wishes (e.g. A. 2.71.9 ‘quae uoluerit meminisse, quae mandauerit exsequi’, of the dying Germanicus) are naturally a special feature of their nouissima uerba (above,

3n. qui) and it was a particular misfortune to be deprived of such consolation: the classic ex. 15 Hom. Il. 24.743—5 (see Macleod's and Richardson's nn.), and Lipsius in his Curae Secundae (1588), referring to these lines, saw the whole scene as Homeric

(‘Plane Homericum

est’).

quas penitus animo figeremus: quas — ut eas. For anzmo figere cf. OLD figo 8a, adding Petron. 115.15. 45.5 noster hic dolor, nostrum uulnus 'this is our pain, our wound’: hic (= hoc, attracted into the gender of dolor) is anticipatory of nobis... amissus est; the arrangement of first-person terms (noster... nostrum...: nobis...) is 2 + 1, the converse of 45.1 nostrae...; nos...nos...(1 4 2). moster... nostrum are emphatic: our particular pain, as opposed to that felt by others. For metaphorical uulnus see 7.1n. (Sequens). nobis...amissus est: Rhenanus' es 15 very tempting (cf. tuo and comploratus esin the succeeding sentences); but the previous sentence has been

COMMENTARY:

46.1

323

cast in the third person (cf. ezus) and there is no justification for changing the transmitted reading. condicione — 'on account of, owing to the circumstance of' (4 gen.): the idiom 15 not recognised bv OLD but cf. Val. Max. 3.8 ext. 4, 5.3.3 'cladis condicione'. ante quadriennium (‘four years previously’) is an odd equivalent of quadriennio ante (‘previously by four years'): see OLD quadrienntum b. omnia...superfuere honori tuo 'evervthing...was available for your honour' (OLD supersum 7). The presence of one's nearest and dearest at the death-bed was a standard motif, e.g. Ov. Am. §.9.49-50, Prop. 4.11.63—4, Cons. Liv. 809—94. amantissima uxoresuggests formulaic phraseology, as CIL 3.11022 'coniugi amantissimae'. szne dubio contrasts with tamen below. comploratus es is the reading of E, with compositus written in the margin. In a context of death the natural meaning of the latter is ‘you were laid out for burial’, ‘laid to rest' (OLD 4c), as H. 1.47.2 ‘Pisonem Verania uxor..., T. Vinium

Crispina filia composuere’;

but this would

be very odd, since

the following reference to A.'s final gaze indicates that he is still alive. nouissima in luce is exceptionally effective here, combining the notions of 'the light of day’, ‘life’ and 'vision', and evoking the memorable formula for death, /uce careo (OLD careo 1b). The adj., which is used esp. of death (OLD nouissimus 3), looks back to nouwssimis sermonibus at 3 above: his last words have given way to his last gaze; and his last gaze was neither upon his daughter and her husband nor upon ‘hanc beatissimi saeculi lucem’ of Trajan's reign (44.5). For the expr. cf. Ov. Her. 9.167, Met. 1.772, Luc. 7.90. desiderauere aliquid oculi tui: aliqu:d ‘pathetically expresses the condition of the sick man, who, already in the shadow of death, is vet dimly conscious of the absence of his only child' (Gudeman). The lastlook ofa dying person is a regular motif (e.g. Tib. 1.1.59, Ov. Met. 4.145-6, 7.860, Cons. Liv. g2, Stat. Silv. 2.1.148—9g with van Dam): cf. esp. Stat. Silv. 5.1.174—5 ‘nec sole supremo | lumina, sed dulci mauult satiare marito' (and Gibson ad loc.), Theb. 8.040—50 'illam unam neglecto lumine caeli | aspicit et uultu non exsatiatur amato'. Here the sight which A. missed on his deathbed closes the section which began with other missed sights (1 non uidit). For oculi desiderant cf. Cic. Planc. 13, Curt. 10.2.23, [Quint.] Decl. 16.11. 46.1 51 quis piorum manibus locus: for the widely held notion that 'the blessed dead' have their own abode see e.g. Cic. Phil. 14.32 'piorum...sedem et locum', Hor. Odes 2.13.23 'sedes... piorum' with N-H's nn., Mart. 12.52.11 'accipient olim cum te loca laeta piorum’, Sil. 13.703; Austin on Virg. Aen. 6.637—78; also below, n. s For the expr. piorum manibus cf. Val. Fl. 1.650 ‘Elysium manesque piorum’, Serv. Aen. 3.302. Austin (1939) pointed out the lyric rhythms in this last paragraph: here the first three words form the first half of an Alcaic line (which T. could

324

COMMENTARY:

46.1

have completed, had he inserted est before locus); see also nn. below and 10.3n. (Btore).

si (ut sápientibus placet) non...animae: the opening of the clause recalls Cic. Arch. 30 'siue, ut sapientissimi homines putauerunt,...’ (but T.'s exact wording is Sallustian: H. 2.28 'sapientibus ut placet’), the centre recalls 3.1 above: ‘ut corpora nostra ... extinguuntur' (one of the many elements of ring composition framing the biography). Conditional clauses in such contexts are conventional: e.g. Cic. Tusc. 1.98 'sin uera sunt quae dicuntur, migrationem esse mortem in eas oras quae qui e uita excesserunt incolunt...’, Ov. Am. 3.9.59—60, 65, Cons. Liv. 320, Sen. Ep. 63.16 'si modo uera sapientium fama est recipitque nos locus aliquis ... ', Plut. Cons. Apoll. 120B 'if the account of the ancient poets and philosophers is true, as it most likely is, and there is for the blessed departed ...a place set apart in which their souls spend their time...'. anzmae here means ‘souls’, as at Ov. TY. 3.3.59 'utinam pereant animae cum corpore nostrae' (OLD ra); the word in this sense is not elsewhere qualified by magnae, but T. seems to be alluding to the notion of the 3rd-cent. Bc Stoic philosopher, Chrysippus, that the souls of ‘the great and the good’ will live on (e.g. Cic. Rep. 6.13 and Powell, Sest. 143 and Kaster). See esp. Cic. Senec. 79-81 (and Powell ad loc.) and also the related s2 post fata manent sensus (Courtney (1995) 403 on 1994.39). For exstinguuntur cf. Ter. Ad. 314, Plin. NH 20.98, Serv. Aen. 1.93. The sapientesare philosophers (OLD sapiens? b); the rhythm is again notable

(last half of an Alcaic line).

placide quiescas: this seems to be a funerary formula (Tib. 2.4.49, CLE 541.12 ‘hic ego sepultus iaceo placidusque quiesco’, cf. 467.8), but notat Front. Ep. Ant. 1.1.3 (p. 86.22 vdH?) ‘nox quae sequitur fac iam placide quiescas’. Cf. the formula b(ene) g(uzescas) in epitaphs (Lattimore (1942) 72). When used of the dead, quiescoseems largely poetical (OLD 3), as does quies (OLD 93). magn(ae) animae plácide quiescas forms the complete last line of an Alcaic stanza. nosque domum tüàm 15 the last half of an Alcaic line; domum tuam 15 in apposition to nos. (Gudeman favoured Urlichs' nosque domum, on the grounds that two separate entities, T. and the women, are required to correspond to :nfirmo desiderio and muliebribus lamentis below. But this seems unnecessarilv fussy and destroys the rhythm.) desiderio looks back to deszderauere above (45.5): T. and his wife now miss A. in the same way as he on his deathbed missed them. et muliebribus lamentis: grief and lament were conventionally associated with women (see 29.1n.), and A.'s daughter and wife are about to be mentioned (3): for the contrast with uzrtutum below cf. Hor. Epod. 16.39 ‘uos, quibus est uirtus, muliebrem

tollite luctum’

(and Watson

ad loc.).

The injunction to refrain from, or to place a limit upon, grief is conventional in death-bed scenes, consolations and the like: e.g. A. 2.71.3, 3.6.1—3

COMMENTARY:

(and W-M

ad loc.),15.62.2; N-H

46.2

325

on Hor. Odes 2.20.23. uoces imitates the

convention of epitaphs and epitaphic literature that it is the dead person who issues the injunction (Lattimore (1942) 217ff.). The metre is again Alcaic (see nn. above).

ad contemplationem uirtutum is an allusion to Sen. Cons. Marc. 24.4 ‘harum contemplatione uirtutum’. quas néqué lugeri néqué plangi fas est: cf. Pliny's letter on the death of Verginius Rufus, 'si tamen fas est aut flere aut omnino mortem uocare...' (Ep. 2.1.10: see also below, 4n. manet), which in turn alludes to the epitaph attributed to Naevius: 'immortales mortales si foret fas flere' (fr. 2 Courtney, FLP). The first five words close a dactylic hexameter. 46.2 admiratione te potius et laudibus is the text rightly printed bv O-R and Soverini. te is needed as object of the main verb and, though potius strictly qualifies admaratione, it is normal for the unemphatic pronoun to be placed in second position. That the transmitted temporalibus is a form of dittography after te potzus was seen by Muretus. The combination adm:ratione...et laudibus, which here presumably includes a reference to the present biography (Intro. p. 1), is esp. Ciceronian (De or. 2.347, 3.52, 3.101, Brut. 159). si natura suppeditet 'if nature should permit' or perhaps ‘with nature's help' (OLD suppedito 1): the point is that adm?ratio and laudes are within our own control, but any effort to resemble A. (szmilitudzne) will require

something extra. similitudine colamus: the paradosis reads mlitum decoramus. Grotius’ similitudine, which 15 generally accepted, is verv plausible: it prepares for the main theme of $5 below and 15 a natural counterpart of admiratione above (cf. Sen. Ep. 84.8 'si cuius in te comparebit similitudo, quem admiratio tibi altius finxerit'). It is less easy to choose between Muretus' colamus and Ursinus'

decoremus. When

colo means

‘honour’,

there is usu-

ally some suggestion of divine worship, but T. could be hinting at the Roman practice of assimilating the deceased to a divinity (see Gibson on Stat. Silv. 5.1.231—4); the verb is combined with /aude at Stat. Theb. 2.295 but the text has been questioned. decoro also means 'honour' (OLD 2) and

the combination

with

/aude is Ciceronian

(Arch.

22, Lig. 6; also

clausula

(esse uideatur).

Front. p. 216.17 vdH?); on the other hand, the corruption is rather more difficult to explain

and we

lose the

Ciceronian

For exemplarity as a central feature of Roman culture see e.g. W. on Vell. 126.5, W-M on A. 3.55.5; M. Roller, CCRH 214-30, Kraus (2010) 409-15. is uerus honos, ea...pietas: cf. Stat. Theb. 10.711 'haec pietas, hic uerus honos’; the two nouns are also combined at Curt. 5.2.18, Plin. Ep. 7.24.8. pietas, which is naturally found in contexts of bereavement (see Gibson on Stat. Szlv. 5.1, Ep. 2, quoting Plin. Ep. 5.16.8), looks back to 3.3 professione

326

COMMENTARY:

46.3

pietatis, one of those frequent links between the beginning and end of the biography; for uerus honos cf. Sen. Ep. 115.10, Plin. Pan. 33.4, 84.8. contunctissimi cutusque — ‘of all those closest to you'. 46.3 id filiae quoque uxorique praeceperim: whereas s and ea (above) were retrospective, 2d looks forward to wenerars, which is in apposition: hence it seems that praeceperim is an additional injunction, aimed specifically at the two women. Yet, since the sentence concludes (exprimere...tuis...moribus) by repeating the same notion of moral imitation as already expressed by similitudine, there seems to be a sense in which :d is retrospective too, forming the third member ofa tricolon: if so, then

the women

have been

excluded

On the other hand, although the women tence as far as complectantur, the sentence person in the masculine (?pse...posszs). would instruct’) see 10.6n. (addiderim).

from

the exhortation

colamus.

are the main subject of the senends with a generalised secondFor the subjunc. praecepersm (‘1 For the inf. uenerari see 38.3n.

(citrcumuehi).

reuoluant ‘go back over' (OLD 2c, cf. d; W-M on A. 3.18.4). formamque ac figuram animi...complectantur: the combination of forma and figura is almost exclusively Ciceronian: for the association with complecti cf. Cic. Tusc. 1.37 'animos enim per se ipsos uiuentes non poterant mente complecti, formam aliquam figuramque quaerebant’. Pliny in his turn alludes to T.: ‘non ergo perpetua principi fama, quae inuitum manet, sed bona concupiscenda est: ea porro non ?magznibus et statuis sed uirtute ac meritis prorogatur. quin etiam leuiora haec, formam principis figuramque, non aurum melius uel argentum quam fauor hominum exprzmat teneatque (Pan. 55.10—11). Though complecti refers to mental retention and is almost equivalent to 'remember' (OLD 4b, cf. a), there is perhaps also the suggestion that embracing A.'s anzmus will continue, or make up for, the physical embrace at his death (— 45.4 complexu). non quia...putem...sed...: a rejected reason expressed bv the subjunc. 4 non quia, rather than non quod ΟΥ non quo, ‘is commoner in early Latin, and comes into favour again with Livy' (NLS$243); here the sequel (as at D. 37.0, H. 1.15.2) is expressed by sedand a main clause: cf. e.g. Cic. Tusc. 1.1. intercedendum (sc. esse) — ‘that a veto should be placed (on)’: see

OLD 5. imaginibus quae .. . finguntur: Menand. prescribes that paintings should be made of the deceased (414.26—7), and of course it was normal to have portraits or busts in the home (van Dam on Stat. Szlv. 2.1.191—3); but statues and the like are often used as foils for more permanent or immaterial memorials (e.g. Pind. Nem. 5.1—5, Isoc. Evag. 73, Sall.J. 4.6, Hor. Odes 3.30.1ff., 4.2.19—20, 4.8.13—-15, Ep. 2.1.248 and Brink ad loc.), and this is the case here (see below, n. forma). For T.'s expr. cf. Cic. Mil. 79, De or. 2.350, Sen. Ef. 13.12, Quint. 11.2.21.

COMMENTARY: 46.4

327

ut uultus hominum, ita simulacra uultus...: the comparison between human and material 15 emphasised by the repetition of uultus (for which see Wills (1996) 254-68, esp. 260) and strengthened by the fact that T. has described the latter in terms primarily used of the former: for similar techniques see N-H on Hor. Odes 2.9 (intro. n., p. 135); Gibson on Stat. Silv. 5.1.16; and nn. below. forma mentis aeterna: sc. est (and adversative asyndeton). The preceding contrast between human and material gives way to one between material and immaterial, for various forms of which cf. Isoc. Evag. 73-7, Cic. Arch. go, Sen. Cons. Marc. 24.5, Cons. Pol. 18.2, Stat. Silv. 5.1.1-15. The contrast 15 once again tightened because forma, which here seems to be used almost in a Platonic sense, can also be used of statuary (OLD 14b): see also next n. For forma...aeterna cf. Manil. 1.211 'haec aeterna manet...forma'. quam tenere et exprimere...possis: the contrast between statuary and the immaterial continues but in a compressed manner: in the case of the latter, a mental image has to be grasped (tenere OLD 23a) before it can be expressed (exprimere. OLD 6a) 1n one's own behaviour in order to produce a likeness (2 above: similitudo); but, in the case of statuary, the image has to

be sculpted (exprimere. OLD 6b) before it can be physically grasped (tenere OLD 1—2),so tenere et exprimereis here hysteron proteron. The contrast is all the more marked because the ancients believed that a person's character could indeed be represented in statuary (e.g. Plut. Mar. 2.1, cf. Xen. Mem. 3.10). Note the double uanriatio of per -- acc. — abl. (6.1n. per mutuam) and of the antithesis alienam ~ tuis ipse. For materiam et artem ct. e.g. Liv. 45.27.11 'omni genere et materiae et artium insignia', Quint. 2.19.3. The generalising second-person subjunc. possis (3.1n. sic), here perhaps analogous to its usc in philosophical dialogues (for which sce Gilmartin (1975) 101-2), explains the singular (uis. 46.4 quidquid ex Agricola...fama rerum: considerable problems are posed by the traditional presentation of this sentence: quidquid ex Agricola amauimus, quidquid mirati sumus, manet mansurumque est in animis hominum, in aeternitate temporum, fama rerum.

(1) fama rerum 1s an cxtremely common phrase in prosc (esp. Livy) and verse but it seems always to refer to the specific deeds of a person and not to ‘history’ in a general sense. But in that case fama rerum is of an entirely different order of expression from the two preceding phrases and therefore, regardless of whether we understand or (with Halm) supply in, it is unlikely to constitute the third member of a potential tricolon. It also makes poor, if any, sense to say 'everything we loved and admired in

328

COMMENTARY:

46.4

A. will live on....in the fame of his deeds'. Other scholars suppose that fama is a causal abl. (‘everything we loved and admired in A. will live on...because of the fame of his deeds’); but the juxtaposition of a different type of abl. 15 extremely awkward, the phrase scems just tagged on at the end of the sentence, and the expression itself is redundant after quidquid . .. mirati sumus. (11) The twin clauses quidquid ex... mirati sumus implicitly refer to personal qualities (esp. amauimus) and, after the argument in 2—3 above, we expect T. to say 'whatever we loved and admired in A. will live on because we shall indeed imitate him in our own behaviour ;

instead, however, we are told that these qualities will live on because (nam below) T. has written the man's biography. And this apparent non sequitur is compounded because the biography says very little about A.’s personal qualities but, as is underlined by fama rerum and (below) inglorios et ignobiles, concentrates almost exclusively on A.'s public life as a successful military general. These problems may be solved if we punctuate the passage in such a way as to separate the personal qualities implied by quidquid ex . . . mirati sumus from the public life implied by fama rerum. There is more than one way of achieving this separation but the suggestion offered here is as follows: quidquid ex Agricola amauimus, quidquid mirati sumus, manet mansurumque est in animis hominum, «ut- in aeternitate temporum famá rerum: nam... (‘whatever we loved and admired

in À. endures and will endure

in the

hearts of men, just as the fame of his deeds in the eternity of time: for, unlike those whose glory and renown are destroyed by oblivion, he will survive, having been described and handed down to posterity’). ut, mistakenly inserted in the previous sentence and easily lost between hominü and in, gencrates a chiasmus with quidquid . . . hominum which serves to introduce the two final sentences; it also takes it for granted that the fame of A.'s deeds will endure in the eternity of time: the reason for this (nam) is that T. has written his biography. fama manet 15 originally a poetic expr. (Catull. 61.222—9, Ov. F. 4.150, [Sen.] Epigr. 447.3, Luc. 8.74) which occurs first in prose in Pliny (Ε. 9.93.1, cf. Pan. 55.10) and T. (H. 1.30.2, 2.68.4, where H.'s n. should be corrected). quidquid ex Agricola amauimus ‘whatever we loved in A.’: for this odd use of ex cf. e.g. Prop. 1.4.28 ‘nec quicquam ex illa quod querar inueniam' and Shackleton Bailey (1956) 18. For the idea cf. Sen. Ep. 99.4 'magna pars ex iis quos amauimus, licet ipsos casus abstulerit, apud nos manet'. manet mansurumque est: such repetitions (for which see Wills (1996) 302—9) are frequent in consolatory and similar contexts (e.g. Vell. 66.5

COMMENTARY:

46.4

399

‘uiuit uiuetque per omnem saeculorum memoriam' with W., Sen. Cons. Marc. 1.3 'cuius [sc. Cremuti Cordi] uiget uigebitque memoria', Plin. Ep. 2.1.11 ‘uiuit enim uiuetque semper, atque etiam latius in memoria hominum et sermone uersabitur’, the letter on the dead Verginius Rufus). Though exs. with manere are esp. common (e.g. Cic. Leg. 1.1, Plin. Pan. 93.1), that at Plin. Ep. g.27.2 ('liber...manet manebit legeturque semper') has prompted scholars to ask whether, in view of the verbal similarity to our passage, T. may be the anonymous historian described in the letter as bowing to friendly pressure and refraining from continuing his recitation of his work (see Whitton

(2012) 363—-4).

in animis hominum is an expr. much favoured by Cicero, though found also in some other authors. in aeternitate temporum: the expr. is unparalleled but Cicero twice has perpetuitas temporis ( Tusc. 1.977, Fin. 2.87) and diuturnitas temporis (Fin. 2.87, ND 2.5), the latter evidently litde different in meaning when the gen. is plural (Acad. 119 'diuturnitas temporum), as in T. 2n here is not local (as above) but temporal (OLD 35), as Quint. 12.1.21 'in hac quae superest aeternitate’. nam multos uéterum uélut inglorios et ignobiles obliuio obruit: T. draws to his close with a rhythmicafilssonantal and ägrzänatic aphorism which acts as a foil for the final sentence of all. The MSS read obruet, which Haupt changed to obruit (perfect). If the verb is future, the exact parallelism of the two final sentences (cf. erit) would seem to imply that A. 15 being contrasted with his contemporaries. But (a) itis extremely odd to refer to one's

contemporaries bv the term ueteres, even if seen from the perspective of the distant future; (b) there is no reason why these contemporaries should not

be memorialised in biographies of their own; (c) there is an unwelcome implication that T.'s biography will survive and those of other memorialists will not. Haupt's emendation avoids these problems. The point 15 then the same as Horace's famous lines: ‘uixere fortes ante Agamemnona | multi, sed omnes inlacrimabiles | urgentur ignotique longa | nocte, carent quia uate sacro' (Odes 4.9.25—8). obliuzo obruit is exclusively Ciceronian (Brult. 6o, Fin. 1.57, 2.105); for the two adjs. cf. Tusc. 3.57 'quod inglorius 511 atque ignobilis ad supremum diem peruenturus’. Agricola posteritati narratus et traditus superstes erit: adversative asyndeton. Ring composition with the opening words (1.1 'Clarorum uirorum facta moresque posteris tradere’, 1.4 ‘narraturo mihi’) indicates both closure and fulfilment, with the name of Agricola being a substitute for, and illustration of, ‘distinguished men’. This is also the final comparison between biographer and his subject: just as T. was a superstes after his metaphorical death in Domitian's reign (3.2), so A. will also be a superstes thanks to the power of literature. (superstes erit is an Ovidian tag, used for

330

COMMENTARY: 46.4

predictions of his own poetic immortality at Am. 1.15.42, Tr. 3.7.50; cf. also s. erisat Mart. 10.2.8, in a similar context.) posteritati.. . traditus may itselfbe a compliment, being a likely allusion to the work of A.'s father: the phrase recurs elsewhere at Colum. RR 1.1.14 ‘duo uolumina...praeceptorum de uineis Iulius Graecinus composita facetius et eruditius posteritati tradenda curauit’, where Columella may very well be quoting from the preface or epilogue to Julius Graecinus' volumes on viticulture (the only other ex. of the phrase seems to be Curt. 6.3.17).

APPENDIXES APPENDIX

1

Roman legions in Britain during Agricola's career II Augusta IX Hispana XIV Gemina XX Valeria Victrix II Adiutrix

¢. 55—75 Exeter; c. 75 onwards Caerleon 55—71 (or 43—-71) Lincoln; 71 onwards York 55—66 (c. 66—69 in Italy) and 69—70 Wroxeter; then withdrawn — 57—67 Usk; 67-75 Gloucester; 75-83/84 Wroxeter; 83/84 onwards Chester . 71—76 Lincoln; ¢. 78-83/84 Chester; 83/84—87 Inchtuthil

This table 15 based on RGB 227-8.

331

332

APPENDIXES

APPENDIX

2

Chronological tables for Agricola and Tacitus Agricola

77-84

Born (13 June) Military tribune in Britain Marries Domitia Decidiana Quaestor in Asia Tribunus Plebis Praetor Legate of Legio XX in Britain Governor of Aquitania Suffect consul Made pontifex Governor of Britain

83

Battle of Mons Graupius

93

Dies (23 August)

40

?58-60 61 63/64

66 68

70-72

73-76 76

77

Tacitus Born

76

Marries Agricola's daughter

?77-19 81

Military tribune (in Britain?) Quaestor

88

Praetor

97 97/98

Writes Agricola

Suffect consul

WORKS

CITED

References to the fragmentary Roman historians are given according to the edition of Peter (1914, 1906), to which is usually added the numeration in T. J. Cornell (ed.), The Fragments of the Roman Histonans (2013). Abbreviated references to scholarly literature usually take the form of name and date (and page-number), although dates are sometimes omitted in the case of certain works of reference. Apart from various standard commentaries, all abbreviations may be traced via the lists below. At regular intervals since 1964-65 H. W. Benario published comprehensive bibliographies on Tacitean scholarship, of which the most recent, and final, was in CW 98 (2005) 251—9336; for coverage since that date one has torely on the notices in the periodical Gnomon. A good bibliography of work on the Agricolawill be found in Soverini 27-44. There is no dedicated bibliography of Roman Britain, but a Bibliography of British and Irish History 1s available on the Brepolis website (www.brepolis.net), and the periodical Britannia appears annually. Roman officials who served in Britain are listed and discussed in RGB; personal names are listed by Kakoschke (2011).

AGRICOLA:

EDITIONS

AND

COMMENTARIES

This list provides details of the editions and commentaries which have been consulted; a much more comprehensive list will be found in Soverini 26-7. Church, A. J. and Brodribb, W. J. (1869). London Delz,J. (1983). Stuttgart (Teubner) Forni, G. (1962). Rome Furneaux, H. (1898). Oxford

Furneaux, H. and Anderson,J. G. C. (1922). Oxtford Gudeman, A. (1928). Boston Heubner, H. (1984). Gottingen Koestermann, E. (1962). Leipzig (Teubner) Lenchantn de Gubernatis, M. (1949). Turin Ogilvie, R. M. (1975). Oxford (OCT) Ogilvie, R. M. and Richmond, I. (1967). Oxford Pearce,J. W. E. (1901). London Ritter, F. (1848). Cambridge and London

Ruperti, G. A. (1832). Hanover Saint-Denis, E. de (1942). Paris

334

WORKS

CITED

Sleeman,J. H. (1914, repr. 1939). Cambndge Soverini, P. (2004). Alessandria Stadele, A. (* 2001). Düsseldorf and Zurich

Stuart, D. R. (1909). New York Walther, G. H. (1833). Halle

OTHER

WORKS

Adams, . N. (1972). 'Ihe language of the later books of Tacitus' Annals’,

CQ 22:350-73

(19732). 'The vocabulary of the speeches in Tacitus' historical works’, BICS 20:124-44 (1973b). 'The substantival present participle in Latin', Glotta 51:116—96 (1974). ‘The vocabulary of the later decades of Livy', Antichthon 8:54—062 (20092). Bilingualism and the Latin language. Cambridge (20093b). '* Romanitas" and the Latin language', CQ 53:184-205

(2007). The regional diversification of Latin 200 BC-AD 600. Cambridge

Adler, E. (2011). Valorizing the barbarians. Enemy speeches in Roman historiography. Austin Aili,

H.

(2009).

'Caesar's

elks:

interpolation,

myth,

or

fact?',

Eranos

105:4—17 Ando, C. (2005). 'Interpretatio romana’, CP 100:41—51 Ash, R. (20072). “Tacitus and the Battle of Mons Graupius: a historiographical route map?', CGRH 2:434—40 (2007b). Tacitus: Histories Book II. Cambridge (2009). ‘Fission and fusion: shifting Roman i1dentites in the Histores’, in CCT 85—99 (2010). ‘The great escape: Tacitus on the mutiny of the Usipi (Agricola 28)', in Kraus, Marincola and Pelling (2010) 275-93 (2012). 'Introduction', in ORT 1—-35 Austin, N. J. E. and Rankov, N. B. (1995). Exploratio: military and political intelligence in the Roman world from the Second Punic War to the battle of Adrianople. London and New York Austin, R. G. (1939). 'The epilogue to the Agricola’, CR 53:116—17 Bailey, C. (1947). 1 Lucreti Cari De rerum natura libri sex. Vols. 1—9. Oxford Baldi, P. (1979). ‘A structual ambiguity in Latin', CP74:49-52 Balsdon,J. P. V. D. (1979). Romans and aliens. London Barrett, A. A. (1989). Caligula: the corruption of power. London (1991). 'Claudius' British victory arch in Rome', Britannia 22:1-1¢9 Bartsch, S. (1994). Actors in the audience: theatricality and doublespeak from Nero to Hadrian. Cambridge, MA (2006). The mirror of the self. Chicago

WORKS

CITED

335

Beacham, R. C. (1999). Spectacle entertainments of early imperial Rome. New Haven Beard, M. (2007). The Roman triumph. Cambridge, MA and London Bebbington, D. W. (2008). 'Gladstone and the classics’, in L. Hardwick and C. Stray (edd.), À companton to classical receptions 86—97. Oxford Benario,

H.

W.

(2007).

‘Tacitus

Bili¢, T.

(2012).

'Crates

of Mallos

in

America’,

in

R.

Bedon

and

M.

Polfer (edd.), Etre romain: Hommages in memoriam C. M. Ternes 57—67. Remshalden Bennett,J. (2001). Trajan: Optimus Princeps. 2nd edn. Bloomington Bertrand, A. C. (1997). 'Stumbling through Gaul: maps, intelligence, and Caesar’s Bellum Gallicum' , Ancient History Bulletin 11:107—22 Bews, J. P. (1987). 'Language and style in Tacitus’ Agricola, GR 34: 201—11 Bickerman, E. J. (1952). 'Origines gentium’, CP 47:65-81 and

Pytheas

of Massalia:

examples

of Homeric exegesis in terms of mathematical geography', TAPA 142:295-328 Birley, A. R. (1999). Tacitus: Agricola and Germany. Oxford (2000a). 'The life and death of Cornelius Tacitus’, Historia 49:2930—47 (2000b). Onomasticon to the Younger Pliny. Munich and Leipzig (2002). Garrison life at Vindolanda: a band of brothers. Stroud and

Charleston, SC (2008). 'Britain: the Caesarian and Claudian invasions’, in I. Piso (ed.),

Die rómischen Provinzen. Begriff und Gründung 179-205. Cluj-Napoca (2009). 'The Agricola’, in CCT 47—58 Birley, E. (1946). ‘Britain under the Flavians: Agricola and his predecessors', DU/ 7:79-84 (repr. in Roman Bntain and the Roman army (Kendal

1953) 10-19)

Bosworth, A. B. (2004).

'Mountain and molchill? Cornclius Tacitus and

Quintus Curtius', CQ 54:551-67 Bowersock, G. W. (2005). 'The East-West orientation of Mediterranean studies and the meaning of North and South in antiquity', in W. V. Harris (ed.), Rethinking the Mediterranean 167—78. Oxford Bowman, A. K. (2003). Life and letters on the Roman frontier: Vindolanda and 115 people. ατὰ edn. London Bramble,J. C. (1974). Persius and the programmatic satire. Cambridge Braund, D. C. (1984). Rome and the friendly king. London (1996). Ruling Roman Bntain. Kings, queens, governors and emperors from Julius Caesar to Agricola. London Breed, B. W. (2004). ' Tua, Caesar, aetas: Horace Ode4.15 and the Augustan Age', AJP125:245-53 Breeze, D. J. (1997). ‘The great myth of Caledon', in T. C. Smout (ed.), Scottish woodland history 47—51. Edinburgh

336

WORKS

CITED

Brodersen, K. (1995). Terra cognita: Studien zur rómischen Raumfassung. Hildesheim (2004). 'Mapping (in) the ancient world’, /RS 94:189-90 (2012). ‘Cartography’, in Ducck (2012) 99-110 Brown, A. G. εἰ al. (2001). 'Roman vineyards in Britain: stratigraphic and palynological data from Wollaston in the Nene Valley, England', Antiq-

uity 75:745-57

Brown, R. D. (1999). 'Iwo Caesarian battle-descriptions: a study in con-

trast’, C/ 94:329-57

Brunt, P. A. (1983). 'Princeps and Equites’, /RS 73:42—75 (1990). Roman imperial themes. Oxford Burridge, R. A. (2004). What are the Gospels? A comparison with Graeco-Roman biography. 2nd edn. Grand Rapids and Cambridge Cairns, F. (1972). Generic composition in Greek and Roman poetry. Edinburgh Campbell, B. (1975). Who were the “Vin Militares"?', /R$ 65:11—31 (2012). Rivers and the power of ancient Rome. Chapel Hill Campbell, D. B. (2010). Mons GraupiusAD δ3. Oxford Carmody, W. M. (1926). The subjunctive in Tacitus. Chicago Champlin, E. J. (1991). Final judgments: duty and emotion in Roman wills. Berkeley Chaplin,J. D. (2000). Livy's exemplary history. Oxford Church, A. J. and Brodribb, W. J. (1877). The minor works of Tacitus. 2nd edn. London and New York Clarke, K. (1999). Between geography and history. Oxford (2001). ‘An island nation: re-reading Tacitus' Agricola' , JRS 91:94-112

(= ORT 37-71)

Cokayne, K. (2003). Experiencing old age in Ancient Rome. London and New York Cole, T. (1990). Democnitus and the sources of Greek anthropology. Rev. edn. Atlanta Cornell, T. J. (ed.) (2013). The fragments of the Roman historians. Oxford Cotton, H. (1993). 'The guardianship of Jesus son of Babatha: Roman and local law in the province of Arabia’, /RS 89:94—-108 Courtney, E. (1971). ‘A note on Tacitus Agricola 42', BICS 18:98 (1995). Musa Lapidaria. Atlanta (2012). 'Some passages of Sallust, Nepos and Tacitus’, Prometheus

38:153-4

Cova, P. (1999). 'Il ritratto del buon generale e la fortuna della versione ciceroniana', Paideia 54:199—43 Cramer, F. M. (1945). 'Bookburning and censorship in ancient Rome’, JHI 6:157-—96 Cramsie, P. (2010). The story of graphic design. London Creighton,J. (2006). Britannta: the creation of a Roman province. London

WORKS

CITED

337

Currie, H. MacL. (1957). 'Tacitus, Agricola, 28,2', RAM 100:297—9 Damon, C. (20093). Tacitus: Histories Book [. Cambridge D'Arms,J. H. (1984). 'Control, companionship, and clientela: some social functions of the Roman communal meal’, EMC/CV 28:327—48 Dauge, Y. (1981). Le barbare. Recherches sur la conception romaine de la barbane et de la crvilisation. Brussels Davies,J. L. and Jones, R. H. (2006). Roman camps in Wales and the Marches. Cardift Davies, R. W. (1968). ‘Roman Wales and Roman military practice camps’, Archaeologia Cambrensis 117:103—20 (= Service in the Roman army (Edinburgh 1989) 125-40) Davis, W. (2011). Into the silence: the Great War, Mallory and the conquest of Everest. London Demougin, S. (1988). L'ordre équestre sous les Julio-Claudiens. Rome Devine, A. M. and Stephens, L. D. (2006). Latin word order. Oxford Dihle, À. (1970). Studien zur griechischen Biographie. 2nd edn. Gottingen (1987). Die Entstehung der historischen Biographie. Heidelberg Dilke, O. A. W. (1985). Greek and Roman maps. Ithaca Dise, Jr., R. L. (1997). 'Irajan, the Antonines, and the governor's staff', ZPE 116:273-983 Dixon, S. (1988). The Roman mother. London Drury, P. J. etal. (1984). 'The temple of Claudius at Colchester reconsidered', Britannia 15:7-50 Dueck, D. (2012). Geography in classical antiquity. Cambridge Duft, T. (1999). Plutarch's Lives: exploring virtue and vice. Oxford Du Quesnay, I. M. Le M. (1981). 'Vergil's first Eclogue , PLLS 3:30-182 (1995). 'Horace, Odes 4.5: Pro Reditu Imperatons Caesaris Divi. Filu Augustt', in S. J. Harrison (ed.), Homage to Horace 128—87. Oxford Earl, D. C. (1961). The political thought of Sallust. Cambridgc Eck, W. (2002). ‘An emperor 15 made: senatorial politics and Trajan's adoption by Nerva in 97', in G. Clark and T. Rajak (edd.), Philosophy and power in the Graeco-Roman world: essays in honour of Miniam Griffin 211—26. Oxford Edwards, C. (1993). The politics of immorality in ancient Rome. Cambridge Ehlers, W. W. (1998). La Biographie antique. Fondation Hardt Entretiens sur l'Antiquité classique 44. Vandoeuvres Ekman, V. W. (1904). 'On dead water’, in F. Nansen (ed.), The Norwegian North polar expedition, 1893—1896; scientific results V.1—152. Christiania Evans, E. C. (1995). 'Roman descriptions of personal appearance in history and biography', HSCP 46:43-84 Fehling, D. (1989). Herodotus and his ‘sources’. Leeds Fishwick, D. (1972). ‘Templum divo Claudio constitutum', Britannia 3:164-81

338

WORKS

CITED

Flower, H. I. (2006). The art of forgetting: disgrace and oblivion in Roman political culture. Chapel Hill Forbes, C. A. (1936). ‘Books for the burning', 7APA 67:114-25 Frere, S. (1999). Britannia: a history of Roman Bntain. 4th edn. London Fulford, M. (2004). ‘Economic structures', in M. Todd (ed.), À companion to Roman Britain $09-26. Malden Fussell, P. (1975). The Great War and modern memory. Oxford Gabba, E. (1981). "Irue history and false history in classical antiquity’, JRS 71:50—62 Gaertner, H. A. (1983). 'Massilia et l'Agricola de Tacite', in La Patrie gauloise d'Agrippa au Vieme siécle (Actes du Colloque, Lyon 1981) 89-97. Lyon Gallia, A. B. (2012). Remembering the Roman Republic: culture, politics, and history under the Principate. Cambridge Gambash, G. (2012). “To rule a ferocious province: Roman policy and the aftermath of the Boudican revolt', Britannia 43:1-15 Geiger, J. (1985). Cornelius Nepos and ancient political biography. Historia Finzelschrift 47. Wiesbaden Giancotti, F. (1971). Strutture delle monografie di Sallustio e di Tacito. Florence Gilliver, C. M. (1996). 'Mons Graupius and the role of auxiliaries in battle’,

GR 43:54-07

Gilmartin, K. (1975). 'Arhetorical figure in Latin historical style: the imaginary second person singular', TAPA 105:99-121 Godel, R. (1984). ‘Dorica Castra: sur une figure sonore de la poésie latine', Cahiers Ferdinand de Saussure 38:225-34 Golden, M. (1988). ‘Did the ancients care when their children died?', GR

35:152-63

Gowing, À. M. (2005). Empire and memory. Cambridge Grenade, P. (1950). ‘Le Mythe de Pompée et les Pompéiens Cesars', REA 52:28—63 Grethlein,J. and Krebs, C.B. (edd.)

(2012).

historiography: the *blupast' from Herodotus to

sous les

Time and narrative in ancient

Appran. Cambridge

Grithin, M. (1986). 'Philosophy, Cato, and Roman

suicide' Parts I and 11,

GR 33:64-77, 192—202 (1989). 'Philosophy, politics and politicians at Rome', in M. Griffin and ]. Barnes (edd.), Philosophia togata: essays on philosophy and Roman society (repr. as Philosophia togata 1, 1997) 1-ὁ7. Oxford Gunnewig, B. (1998). Das Bild der Germanen und Bntannir: Untersuchungen zur Sichtweise von fremden Volkern in antiker Literatur und moderner wissenschaftlicher Forschung. Frankfurt am Main Hagg, T. (2012). The art of biography in antiquity. Cambridge Hales, S. (2003). The Roman house and social identity. Cambridge Handford, S. A. (1947). The Latin subjuncttve. London Hanson, W. S. (1987). Agricola and the conquest of the North. London

WORKS

CITED

330

(1991). "Tacitus' Agricola: an archaeological and historical study', ANRW 2.39.9:1741—84 (2007). À Roman frontier fort n Scotland: Flginhaugh. Stroud Hanson, W. S. and Maxwcell, G. S. (1980). ‘An Agricolan praesidium on the Forth — Clyde Isthmus (Mollins, Strathclyde)', Britannia 11:43—9 Hanson, W. S. with Speller, K., Yeoman, P. A., and Terry,J. (2007). Eigin-

haugh: a Flavian fort and its annexe. Britannia Monographs 29. London Hardie, P. (2012). Rumour and renown: representations of Fama in western literature. Cambridge Harrison, S. J. (2007). Generic enrichment in Vergil and Horace. Oxford Hartley, B. and Fitts, L. (1988). The Brigantes. Gloucester Hartog, F. (1988). The mirror of Herodotus. The representation of the other in the wrnling of history. Berkeley Hassall, M. W. C. (1970). 'Batavians and the Roman conquest of Britain', Bntanntia 1:131-6 Haverfield, F. (1913). 'Ancient Rome and Ireland', Eng. Hist. Rev. 28:1-12 Haynes, I. (1999). 'Military service and cultural identity in the auxilia', in A. Goldsworthy and I. Haynes (edd.), The Roman army as a community 165—74. JRA Suppl. 34. Portsmouth, RI Hedrick, C. W. (2000). History and silence: purge and rehabilitation of memory in Late Antiquity. Austin Hellegouarc'h,J. (1991). 'L.e style de Tacite: bilan et perspectives’, ANRW

2.33.4:2385-453

Hemelrijk, E. A. (1999). Matrona docta: educated women in the Roman elite. London Heubner, H. (1963). 'Manner, Pferde, Watfen’, Gymn. 70:226—30 (1971). 'Vater oder Tochter?’, RhM 114:265-8 Hind,J. G. F. (1985). 'Summers and winters in Tacitus' account of Agricola's campaigns in Britain’, Northern History 21:1-18 Hirt, A. M. (2010). Imperial mines and quarries in the Roman world: organizational aspects 27 BC-AD 235. Oxford Hofter, S. E. (2006). ‘Divine comedy? Accession propaganda in Pliny, Epistles 10.1-2 and the Panegyric , JRS 96:73-87 Hoffman, D. C. (2008). 'Concerning ekos: social expectation and verisimilitude in early Attic rhetoric', Rhetorica 26:1—29 Holmes, N. (1997/98). 'Perinde and proinde', Glotta 74:59—75 Horsfall, N. (1983). 'Some problems in the Laudatio Tunae, BICS 30:85—99 (1985). ‘Illusion and reality in Latin topographical writing’, GR 32:197-208 (1989). Cornelius Nepos: a selection, including the lives of Cato and Atticus. Oxford (1999). 'The legionary as his own historian', Ancient History 29:107—17

340

WORKS

CITED

Hulls, J.-M. (2007). 'Lowering one's standards - on Statius, Silvae 4.2.43, CQ 57:198-206 Immerwahr, H. R. (1966). Form and thought in Herodotus. Chapel Hill Isaac, B. (1988). ‘The meaning of the terms /mes and limitanev, JRS

79:125-47

Jackson, K. (1979). 'Queen Boudicca?', Britannia 10:255 James, S. (2011). Rome and the sword. London

Johnson, W. À. and Parker, H. N. (2009). Anctent literactes: the culture of reading in Greece and Rome. Oxtord Johnston,J. Β. (1934). Place-names of Scotland. London Jones, B. and Keillar, I. (1996). 'Marinus, Ptolemy and the turning of Scotland', Britannia 27:43—9 Jones, B. W. (1992). The Emperor Domitian. London and New York Jones, P. J. (2005). Reading rivers in Roman literature and culture. Lanham and Oxford Kakoschke, À. (2011). De Personennamen im romischen Britanmien. Hildesheim Karlsson, M. (2005). ‘A note on Tacitus Agricola 41:4', Eranos 109:31—4 Kaster, R. A. (1995). De grammaticis et rhetoribus: Suetonius. Oxford (2002). ‘The taxonomy of patience, or When is “Patientia” nota virtue?’,

CP 97:133-44

(2005). Emotion, restraint and community in ancient Rome. Oxford Keegan,J. (1976). The face of battle. London Keitel, E. (1987a).

'Homeric antecedents to the cohortatio in the ancient

historians', CW80:159-72 (1987b). 'Otho's exhortations in Tacitus' Histories, GR 94:73—82 Kraus, C. S. (1992). 'How (not?) to end asentence? The problem of -que', HSCP 94:321-9 (1994). Ltvy: Ab urbe condita: Book VI. Cambridge (2004). 'Historiography and biography’, in S. J. Harrison (ed.), A companion to Latin literature 241—56. Malden (2010). 'History and biography’, in A. Barchiesi (ed.), The Oxford handbook of Roman studies 403—19. Oxford (2011). ‘The language of Latn historiography, in J. Clackson (ed.), A companion to the Latin language 408—25. Oxford (forthcoming). 'Caesar in Livy and Tacitus’, in L. Grillo and C. B. Krebs

(edd.), The Cambridge companion to Caesar. Cambridge

Kraus, C. S., Marincola,J. and Pelling, C. (edd.)

(2010). Anctent historiog-

raphy and its contexts: studies in honour of A. J. Woodman. Oxtord Kraus, C. S. and Woodman, A. J. (1997). Latin historians. Greece & Rome New Surveys in the Classics 27. Oxford Krebs, C. B. (2006). 'Imaginary geography in Caesar's Bellum Gallicum , AJP 127:111-36

WORKS

CITED

341

(2009). ‘A dangerous book: the reception of the Germania', in CCT

280—99

(2011a). A most dangerous book: Tlacitus's Germania from the Roman Empire to the Third Reich. Ncw York and London (2011b). ‘Borealism: Caesar, Seneca, Tacitus, and the Roman discourse about the Germanic North’, in E. S. Gruen (ed.), Cultural identity in

the ancient Mediterranean 202—21. Los Angeles (2012). 'Annum quiete et otio transiit: Tacitus (Ag. 6.3) and Sallust on liberty, tyranny, and human dignity’, in C7'339—44 Lacey, W. K. (1954-5). 'Oblongae scutulae vel bipenni: Tacitus, Agricola 10, 3', PGPS 189:16-20 Laes, C. (2011). Children in the Roman empire. Cambridge Laird, A. (1999). Powers of expression, expressions of power: speech presentation and Latin literature. Oxford La Penna, A. (1975). ‘Rubor e inpudentia da Pompeo a Domiziano', Maia 27:117—10 Lattimore, R. (1942). Themes in Greek and Latin epitaphs (repr. 1962). Urbana Lausberg, H. (1998). Handbook of literary rhetoric. Leiden Lausberg, M. (1980). ‘Caesar und Cato im Agricola des Tacitus', Gymnasium 87:411-30 Lavan, M.

(2011). ‘Slavishness in Britain and Rome

in Tacitus’ Agricola’,

CQ 61:294-305 (2013). Slaves to Rome: baradigms of empire tn Roman culture. Cambridge Lease, E. B. (1903). 'Livy's use of -arunt, -erunt, -ere , AJP 24:408-22 Lee, H. (2009). Biography: a very short introduction. Oxtord Leeman, A. D. (1973). 'Structure and meaning in the prologues of Tacitus’, YCS 23:169-208 Lelis, A. A., Percy, W. A. and Verstracte, B. C. (2003). The age of marriage in ancient Rome. Lewiston and Lampeter Lendon, J. E. (1997). FEmfire of honour: the art of government in the Roman world. Oxford (2001). 'Voting by shouting in Sparta', in E. Tylawsky and C. Weiss (edd.), Essays in honor of Gordon Williams 169—75. New Haven (2009). 'Historians without history: against Roman historiography’, in CCRH 41-62 Leo, F. (1901). Die griechisch-rómische Biographie nach ihrer litterarischen Form. Leipzig Levick, B. (1990). Claudius. London Lewis, M. J. T. (1966). Temples in Roman Britain. Cambridge (2001). Surveying instruments of Greece and Rome. Cambridge Liebeschuetz, W. (1966). ‘The theme of liberty in the Agricola of Tacitus’, CQ 16:126-939 (updated version in ORT 73-94)

342

WORKS

CITED

Lindholm, E. (19931). Stilistische Studien. Lund Lintott, A. W. (1997). 'The theory of the mixed constitution at Rome', in J. Barnes and M. T. Griffin (edd.), Philosophia Togata II 70—85. Oxford Long, O. F. (1911). 'On the usage of quotiens and quotienscumquein different periods of Latin'. Diss. Baltimore Lutz, C. E. (1947). 'Musonius Rufus, the Roman Socrates', YCS 10:3-147 Mackay, L. À. (1952). " Iacitus Agricola 36.9', in M. E. White (ed.), Studies in honour of Gilbert Norwood 224-8. Toronto Maltby, R. (1991). À lexicon of ancient Latin etymologies. Leeds Mann,J. C. (1985). “Iwo “topoi1” in the Agrzcola', Britannia 16:21—4 Marincola, J. (1997). Authority and tradition in ancient historiography. Cambridge (1999). 'Genre, convention and innovation in Greco-Roman historiography’, in C. S. Kraus (ed.), The limits of historiography: genre and narratrve 1n. ancient historical texts 281—924. Leiden Martin, R. (1981).

Tacitus. London

(rev. Bristol Classical Press 1994)

(2009). ‘From manuscript to print', in CCT 241—52 Mattern, S. P. (1999). Rome and the enemy: imperial strategy in the principate. Berkeley, Los Angeles and London Mattingly, D. (2006). An tmperial possession: Britain in the Roman empire. London Maxfield, V. A. (1981). The military decorations of the Roman Army. Berkeley Maxwell, G. S. (1990). A battle lost: Romans and Caledonians at Mons Graupius. Edinburgh Mayer, R. (2001). Tacitus: Dialogus de oratoribus. Cambridge (2008). "Iwo tricky transitives’, CO 58:342—4 McDonnell, M. (2006). Roman manliness: virtus and the Roman Republic. Cambridge McGing, B. Ο (1982). 'Synkrisis in Tacitus’ Agricola’, Hermathena 132:15-25 Mendell, C. W. (1911). Sentence connection in Tacitus. New Haven Miles, G. B. (1995). Ltvy: reconstructing early Rome. Ithaca Millar, F. (1977). The emperor in the Roman world. London Millett, M. (1990). The Romanization of Britain. Cambridge Moles,J. (1989). Review of Geiger (1985), CR 39:229-33 Momigliano, A. (1993). The development of Greek biography. Rev. edn. Cambridge, MA Moore, F. G. (1903). 'Studies in Tacitean ellipsis: descriptive passages', TAPA 934:5-306 Morford, M. (1991). 'How Tacitus defined liberty’, ANRW2.94.5:3420-50 Munscher, K. (1920). Xenophon in der griechisch-rómischen Literatur. Philol. Suppl. 19.2

WORKS

CITED

343

Murgia, C. E. (1977). ‘The minor works of Tacitus. A study in text criticism’, CP 72:3293—42 (2012). ‘The textual transmission', in CT 15-22 Nicolet, C. (1960). ‘Consul togatus: remarques sur le vocabulaire politique de Cicéron et de Tite-Live', REL 38:236—63 (1991). Space, geography, and politics in the early Roman emfnre. Ann Arbor Nutting, H. C. (1929). 'The battle at Mons Graupius: Tacitus, Agricola 20-97 , CW 23:65-6 Oakley, S. P. (1997-2005). A commentary on Ltvy Books VI-X. Vols. 1—4. Oxford [references are sometimes given by volume and page, esp. for the Addenda in Vol. 4] (2009a). ' Res oltm dissociabiles: emperors, senators and liberty', in CCT

184-94

(2009b). ‘Style and language’, in CCT 195-211 Ogilvie, R. M. (1971). 'Monastic corruption', GR 18:42—4 O'Gorman, E. (1993). 'No place like Rome: identity and difference in the Germania of Tacitus’, Ramus 22:195—54 (= ORT 95-118) O'Neill, P. (2003). ‘Going round in circles: popular speech in ancient Rome', CA 22:135-66 Osgood, J. (2009). ‘The pen and the sword: wriung and conquest in Caesar's Gaul', CA 28:428-58 (2011). Claudius Caesar: image and power in the early Roman Empire. Cambridge Ostenberg, I. (2009). Staging the world. Spoils, captives, and representations in the Roman triumphal procession. Oxford O'Sullivan, T. M. (2011). Walking in Roman culture. Cambridge Otto, À. (1890). Die Sprichwórter und sprichwórtlichen Redensarten der Rómer. Leipzig Parker, G. (2008). The making of Roman India. Cambridge Parkin, T. G. (2003). Old age in the Roman world. Baltimore

Paterson, J. (1978). ‘Transalpinae gentes: Cicero, De re publica 3.16', CQ 28:452—8 Patterson,J. R. (2000). Political life in the city of Rome. London Pearce, T. E. V. (1970). ‘Notes on Cicero, /n Pisonem', CQ 20:309-21 Pelling, C. (1997). 'Biographical history? Cassius Dio on the early principate', in M. J. Edwards and S. Swain (edd.), Portrails. Biographical representalion in the Greek and Latin literature of the Roman Emfnre 117—44. Oxford (2002). Plutarch and history. London (2011a). Plutarch: Caesar. Oxford

(2011b). 'Velleius and biography: the case of Julius Caesar', in E. Cowan (ed.), Velleius Paterculus: making history 157—76. Swansea Perring, D. (2002).

The Roman house in Britain. London and New York

344

WORKS

CITED

Perrot,J. (1961). Les dérivés latins en —men et —mentum. Paris Peter, H. (1914, 1906). Historicorum Romanorum reliquiae. Vols. 1* and 2. Leipzig Petrocelli, C. (2008). Onasandro: Il generale. Manuale per Uesercizio del comando. Bari Phang, S. E. (2004). 'Intimate conquests: Roman soldiers' slave women and freedwomen', AW95:207-37 (2008). Roman military service: ideologies of discipline in the late Republic and early Principate. Cambridge Piggott, S. (1989). Anctent Bntons and the antiquarian imagination. London Pinkster, H. (1999). 'Ihe present tense in Virgil's Aeneid, Mnem. 52:705-17 Plass, P. (1988). Wit and the wniting of history. Madison, WI and London (1995). The game of death in ancient Rome. Madison, WI and London Powell,J. G. F. (2010). ‘Juvenal and the delatores’, in Kraus, Marincola and Pelling (2010) 224—44 Pratt, M. L. (2008). Imperial eyes: travel writing and transculturation. 2nd edn. London Raepsaet-Charlier, M.-T. (1991). 'Cn. Iulius Agricola: mise au point prosopographique', ANAW 2.39.9:1807—57 Rainbird,J. S. (1969). "TIactics at Mons Graupius', CR 19:11-12 Rawson, B. (2003). Children and childhood 1n Roman Italy. Oxford Rawson, E. (1985). Intellectual life in the late Roman Republic. Balümore Reed, N. (1971). "I he fifth year of Agricola's campaign', Britannia 2:143—8 Rives,J. B. (1999). Tacitus: Germania. Oxtord Rivet, A. L. F. (1988). Gallia Narbonensis. London

Rivet, A. L. F. and Smith, C. (1979). The place-names of Roman Bntain. London Roche, P. A. (2006). 'Selling Trajan’s saeculum: destiny, abundance, assurance', Athenaeum 94:199-229 Rodwell, W.

(ed.)

(1980).

Temples,

churches and religion: recent research in

Roman Britain. BAR Brit. Ser. 77, vols. 1-2. Oxford

Roller, M. B. (2006). Dining posture in ancient Rome: bodies, values, and status.

Princeton and Oxford Romer, F. E. (1998). Pomponius Mela's Description of the world. Ann Arbor Romm,J. S. (1992). The edges of the earth in ancient thought: geography, exploration, and fiction. Princeton Russell, D. A. and Wilson, N. G. (1981). Menander Rhetor. Oxford

Rutherford, R. B. (1989). The Meditations of Marcus Aurelius. Oxford (2010). 'Voices of resistance', in Kraus, Marincola and Pelling (2010)

3]12—90 Rutledge, S. H. (2000). ‘Tacitus in tartan: textual colonization and expansionist discourse in the Agricola', Helios 27:75—95

WORKS

CITED

345

(2001). Imperial inquisitions: proseculors and informants from Tiberius to Domitian. London and New York Saddington, D. B. (1990). 'The origin and nature of the German and British flects', Britannia 21:223-392 (1991). ‘A note on viri militares in the Agricola' , ANRW 2.393.93:1738—40 Sailor, D. (2004). 'Becoming Tacitus: significance and inconsequentiality in the prologue of Agricola’, CÀ 23:199—77 (2008). Writing and empire in Tacitus. Cambridge (2012). ‘The Agricola', in CT 23—44 Saller, R. P. (1984). ‘Familia, domus, and the Roman conception of the family', Phoenix 38:336—55 Salway, P. (1993). A history of Roman Britain. Repr. 2001. Oxford Samuel, R. (1998). Island stories: unravelling Britain. London Sasel Kos, M. (1990). 'Sententiae in the Agricola of Tacitus', Ziva Antika (Antiquité Vivante) 40:83-109 Scanlon, T. F. (1987). Spes frustrata: a reading of Sallust. Heidelberg Scarborough,J. (1969). Roman medicine. London Schadee, H. (2008). 'Caesar's construction of northern Europe: inquiry, contact and corruption in De Bello Gallico , CO 58:158-80 Schama, S. (1995). Landscape and memory. London Schlicher, J. . (1933). 'Non-assertive elements in the language of the Roman historians', CP 28:289-300 Shackleton Bailey, D. R. (1956). Propertiana. Cambridge Shatzman, I. (1974). ‘Tacitean rumours’, Latomus 33:549—-78 Shaw, B. D. (2001). ‘Raising and killing children: two Roman myths’, Mnem. 54:31—77 Shaw-Smith, R. (1979). 'Three notes on Tacitus', CQ 29:224—5 (1995). Notes on lacitus. Keighley (privately printed) Sherwin-White, A. N. (1966). The Letters of Pliny. Oxford Smallwood, E. M. (1966). Documents illustrating the principates of Nerva, Trajan and Hadrian. Cambridge Smith, B. H. (1968). Poetic closure. Chicago and London

Smith, C. and Powell, A. (edd.) (2009). The lost memoirs of Augustus and the development of Roman autobiography. Swansea Smout, T. C. (2000). Nature contested: environmental history in Scotland and northern England since 1600. Edinburgh Solodow, J. B. (1978). The Latin particle 'Quidem ; University Park, PA Sorabji, R. (2006). Self: ancient and modern insights about individuality, life, and death. Chicago Sorbom, G. (1935). Vanatio sermonis Taciter aliaeque apud eundem quaestiones selectae. Uppsala Stadter, P. (2007). 'Biography and history’, in CGRH 2:528-40 Starr, R. J. (1981). 'Cross-references in Roman prose', AJP 102:431—7

346

WORKS

CITED

Stevens, C. E. (1951). 'Claudius and the Orcades', CR 1:7-9 Stewart, P. C. N. (1995). 'Inventing Britain: the Roman creation and adaptation of an image', Britannia 26:1-10 Strasburger, H. (1965). 'Poscidonius on problems of the Roman Empire’,

JRS 55:40-53

Strocchio, R. (1992). I significati del silenzio nell'opera di Tacito. Turin Stuart, D. R. (1928). Epochs of Greek and Roman biography. Berkeley Syme, R. (1958). Tacitus. Oxford (1986). The Augustan anstocracy. Oxtord Talbert, R. J. A. (1984). The senate of imperial Rome. Princeton (2008). ‘Greek and Roman mapping: twenty-first century perspectives', in R. J. A. Talbert and R. W. Unger (edd.), Cartography in antiquity and the Middle Ages: fresh perspectives, new methods g—27. Leiden and Boston larrant, R. (2012). Virgil: Aeneid Book XII. Cambridge Thomas, R. (1992). Literacy and orality in ancient Greece. Cambridge (2000). Herodotus in context: ethnography, science, and the art of persuasion. Cambridge Thomas, R. F. (1982). Lands and peoples in Roman poetry. The ethnographical tradition. Cambridge Philological Society Suppl. 7. Cambridge (1986). 'Virgil's Georgics and the art of reference', HSCP go:171—8 (2009). "Ihe Germanta as literary text', in CCT 59—72 lierney, J. J. (1959). ‘Ptolemy’s map of Scotland’, JHS 59:1932—48 Till, R. (1943). Handschriflliche Untersuchungen zu Tacitus Agricola und Germania. Berlin lomlin, R. S. O. (1992). 'The Twentieth Legion at Wroxeter and Carlisle in the first century: the epigraphic evidence', Britannia 23:141—58 (1998). ‘Roman manuscripts from Carlisle: the ink-written tablets’, Britannia 29:31—84 (2004). 'Inscriptions', Britannia 35:335—49 losi, R. (2003). Dizionario delle sentenze greche e latine. Milan

Traub, H. W. (1953). “Tacitus’ use of ferocia' , TAPA 84:250—01 lreggiari, S. (1991). Roman marriage. Oxford Tuck, S. L. (2005). ‘The origins of Roman imperial hunting imagery: Domitian and the redefinition of wirtus under the principate', GR 52:221—45 Ulery, Jr., R. W. (1986). ‘Tacitus’, in F. E. Cranz, V. Brown and P. O. Kristeller (edd.), Catalogus translationum et commentariorum. Mediaeval and Renaissance Latin translations and commentaries: annotated lists and guides Vol. 6, 87-174. Washington Vardi, A. D. (1996). 'Ditudicatio locorum: Gellius and the history of a mode in ancient comparative criticism', CO 46:239-60 Vielberg, M. (1987). Pflichten, Werte, Ideale: eme Untersuchung zu den Wertvorstellungen des Tacitus. Stuttgart

WORKS

CITED

347

von

Fritz, K. (1954). The theory of the mixed constitution in antiquity. New York Vout, C. (1996). ‘The myth of the toga: understanding Roman dress', GR 49:204-20 Wallace-Hadrill, A. (1982). ‘Cinilis Princeps: between citizen and king , /RS 72:92—48 (1989). Suetontus: the scholar and his Caesars. London Walser, G. (1951). Rom, das Reich und die fremden Vólker in der Geschichtsschreibung der frühen Kaiserzeit. Baden-Baden Walsh, P. G. (1901). Ltvy: his historical atms and methods. Cambridge Watt, W. S. (1983). “Two notes on the text of Tacitus, Agricola [195; $9.4]', Britannia 14:274 (1987). 'Five notes on Tacitus, Agricola [2.1, 15.2, 19.4, 44.2, 44.4] , AJP 108:465-6 Weinstock, S. (1960). 'Pax and the "Ara Pacis"', JRS 50:44—58 (1971). Divus Julius. Oxford Wellesley, K. (1969). Review of O-R, /RS 59:266-9 West, D. A. (1965). 'Haurire, haustus (Lucr. 5.1069)', CQ 15:271-80 Whitaker, I. (1981-82). ‘The problem of Pytheas' Thule', C/ 77:148—64 Whitmarsh, T. (2006). '"Ihis in-between book": language, politics and genre Ίη the Agricola , in B. McGing andJ. Mossman (edd.), The limits of ancient brography 305—39. Swansea Whitton, C. L. (2010). 'Pliny, Epistles 8.14: senate, slavery and the Agricola’, JRS 100:118-39 (2011). 'Dubitatio comparativa: a misunderstood idiom in Pliny (Natural History 7.150), Tacitus (Histories 4.6) and others', CQ61:267—77 (2012). '"Let us tread our path together”: Tacitus and the younger Pliny', in CT 345-68 (2019). Pliny the Younger: Epistles Book Il. Cambridge Wiedemann, T. E. J. (1986). 'Between men and beasts: barbarians in Ammianus

Marcellinus’, in I. S. Moxon,J. D. Smart and A. J. Wood-

man (edd.), Past perspectives 18g—201. Cambridge (1989). Adults and children in the Roman empire. London Wilkinson, L. P. (1963). Golden Latin artistry. Cambridge Wille, G. (1983). Der Aufbau der Werhe des lacitus. Amsterdam Williams, G. (1957). Review of Shackleton Bailey (1956), JRS 47:240—7 (1968). Tradition and originality in Roman poetry. Oxford (1978). Change and decline: Roman literature in the early Emfnre. Berkeley, Los Angeles and London (1980). Figures of thought in Roman poetry. New Haven and London Williams,J. H. C. (2001) Beyond the Rubicon: Romans and Gauls in republican Italy. Oxtord (2007). 'New light on Latin in pre-conquest Britain’, Britannia 38:1—12

348

WORKS

CITED

Wills, . (1996). Repetition in Latin poetry. Oxtord Wirszubski, C. (1950). Libertas as a political idea at Rome. Cambridge Wiseman, T. P. (1979). Clio's cosmetics. Leicester (ed.) (1985). Roman political life οο BC-AD 69. Excter (1992). Talking to Virgil. Exeter Wolfson, S. (2008). Tacitus, Thule and Caledonia. The achievements of Agncola’s navy in their true perspective. BAR Briush Series 459. Oxford Woodman, A. J. (1977). Velletus Paterculus: the Tiberian narrative (2.94-131). Cambridge (20092). 'Introduction', in CCT 1-14 (2009b). ‘Tacitus and the contemporary scene’, in CCT 31—43 (2010). 'Community health: metaphors in Latin historiography', PLLS 14:43-61 (2 ΡΗ 162-80) Woolf, G. (1998). Becoming Roman: the ongins of provincial ctvilization in Gaul. Cambridge (2011). Tales of the barbarians: ethnography and empire in the Roman West. Malden, MA and Oxford Woolliscroft, D. J. and Hoffmann, B. (2006). Rome's first frontier: the Flavian occupation of northern Scotland. Stroud Yardley, J. C. (1973). ‘Sick-visiting in Roman elegy', Phoenix 277:283-8 Yegul, F. (2010). Bathing tn the Roman world. Cambridge Yellin, K. (2008).

SC

Battle exhortation: the rhetoric of combat leadership. Columbia,

Zoido,J. ( I. (2007). 'The battle exhortation in ancient rhetoric’, Rhetor-

ica 25:141—58

INDEXES GENERAL

auxiliary troops 157, 227, 269; Cohors VIIII Batauorum 272; Cohors I Tungrorum 272

ablative, of duration of time 165 ablative absolute 30, 32, 78, 125, 131, 148, 207, 210, 307; compressed 19O, 271 abstract for concrete 175, 282, 290, 306, 312, 320 adjectives, in -bundus 280; in ' ab urbe condita construction' 294 adverbs, used attributively 93, 217 adynaton 142, 265 Agricola, historical reliability of 25—-30 alae 157, 183 alliteration g1, 107, 119, 145, 151,

barbarians see ethno-geography beneficiarii 193

biography 1—5, 6-10, 35, 65. 67, 77, 98, 127, 159, 236, 309, 322, 325

blushing 320 book-burning 78, 79 British Empire 23

Caesar, Julius 16, 25, 29, 94, 122, 123, 127,139, 137, 138, 140, 142, 156,

161, 177, 178, 190, 194, 210, 211,

160, 166, 180, 182, 187, 193, 198,

225, 235, 237, 239, 241, 242, 244, 246, 250, 256, 273 allusion (s) 34, 68, 68 n.2; see also

200, 215, 218, 220, 234, 237, 255,

278, 267, 270, 273, 311; Caesarian language and expressions 131, 136, 147, 148,

Caesar, Cato the elder, Catullus, Cicero, Livy, Lucan, Martial, Pliny

the younger, Sallust, Seneca the younger, Silius Italicus, Statius, Valerius Flaccus, Virgil anagram 6g, 161, 180, 329 anaphora go, 32, 104, 105, 167, 181, 188, 193, 242, 256, 315, 318 apostrophe 320 appearance - reality 206 'appendix' sentences 32—3, 78, 125,

151, 201, 209, 224, 226, 262, 269

cannibalism 230 casualty figures 280 Cato the elder 67, 249, 292 Catullus 319 cavalry 274, 275 n.21 centurions 168, 193 character descriptions 123, 211, 310 chariots 149, 270, 275 n.21 chiasmus, chiastic arrangement 31, 32,

148, 190, 207, 210, 307

archaising 101, 105, 134, 164, 174, 185, 188, 246,

249, 276,

65, 72, 78, 80, 86, 88, 106, 113, 118,119, 122,131, 142, 161, 162, 171, 174, 178, 179, 180, 183, 191,

281

assonance, paronomasia, word play g1,

194, 196, 197, 210, 211, 229, 232, 234, 235, 237, 238, 242, 244, 253, 261, 262, 271, 281, 287, 290, 300,

69, 107,119, 151, 161, 162, 190, 194, 221, 225, 232, 235, 237, 238,

241, 243, 244, 245, 252, 253, 254,

304, 309, 313, 320, 328

256, 260, 266, 295, 301, 304,

Cicero 1, 4, 7,12, 16, 17, 34, 92, 232, 233, 237; Ciceronian language and expressions 70, 76, 98, 101, 102,109,113, 115,118,119, 121, 158, 167, 175, 177, 178, 187, 233, 238, 239, 248, 251, 292, 295, 297, 303, 307, 311, 312, 315, 316, 320,

329 asyndeton 87, 88, 106, 119, 121, 144, 151, 170, 179, 184, 187, 191, 201, 210, 211, 241, 244, 246, 249, 277,

287, 327, 329

attraction 73, 250

auantia 155, 157, 170 autobiography 1, 73, 74

321, 324, 325, 326, 329

circumlocutions 95, 188 circumnavigation of Britain 3, 13, 136, 137, 139, 141, 171, 227, 283

‘automimesis’ g, 10, 22, 68; see also

‘transference’ autopsy 25, 99, 129, 136, 141, 152

349

INDEXES

civilisation, history of (Kulturgeschichte)

derivations see etymology

1997207

despatches, military 286; see also Latin

168, 171, 197, 238, 239, 241, 243,

digression (s) 2, 18—19, 125, 126, 146, 158-60, 180, 181, 215

clausulae 82, 92, 138, 148, 158, 167,

259, 256, 260, 262, 263, 303, 304;

see also rhythm client kings 164 Codex Aesinas 35—6 commonplaces see topoi

comparison, authorial device

128; in

ethno-geographical writing 135, 151, 215; see also synkrisis comparative expressions: adjs./advs. 101, 310; compendious 152, 215, 287;in sententiae etc. 107, 117, 211 concrete for abstract 98 conditional sentences 324; imperf. subjunc. in protasis 82, 212; ‘mixed’

100, 161, 248, 276;

pluperf. subjunc. in apodosis 109; protasis suppressed 110 coniunclio 71, 73, 155 conjunctions 98, 141, 154, 188, 219 conquestio 5, 321; 566 also consolatio,

death

consolatio 5, 232, 308—30

topoi of: the deceased achieved much in a short life 311; his beauty 310; escape from pains of life 315; his life to ΡῈ measured by his virtue 312; opportunitas mortis 312, 320; his political honours 312; verbal repetition 328; see also conquestio, death

constructio ad sensum 219, 313

counterfactual statements 173, 224, 276, 278, 321 cross-references 187

cum, inverted 183, 218, 221, 222, 223,

235> 257 dative, of agent 77, 104, 128, 131; of

point of view/of person judging 135, 146, 190; predicative 73 ‘dead water' 139 death, topoi of: the abode of the dead 323; to die with one's family surviving 312; final embraces etc. 322; final gaze 323; funerary

formulae 324; nouissima uerba 320;

last wishes 322; presence of relatives at death-bed 323; see also conquestio, consolatio

Words Index s.v. laureatae

discipline 119, 197 ‘disjunctiveness’ 69 ‘divided self’ 90; see also ‘dualities’ dress and clothing 293 ‘dualities’, dual perspectives 66, 70, 159, 311, 322

ellipse 75, 81, 88, g6, 105, 193, 249, 262,

286, 287,

290, 302

enallage g8, 275 encomium 1, 3—-4, 26, 67, 77, 211 topoi of g4, 98, 99, 129; non silebo 295; laudandus makes proper use of his advantages 311; operates in difficult terrain 126; is as great in peace as in war 200; restores good things 82; is unique 185, 192, 222, 271; performs even better

than expected 189; conquers the unknown 137, 214, 241; see also ideal general Ennius 270 environmental determinism 99, 145 epic 256, 257 Epicureans 100 ethno-geography 12-15, 32, 125-7, 129, 142, 151, 215 topoi of: climate 145, 215; conjecture 146; land and produce 154, 215; measurements 136; naming 138, 207; origin of people 143; shape of country 132; situs 126, 127, 215; authorial

superiority 143; 'then — now' 149; transferable

(Wandermotive) 12,

28; see also comparison

barbarians: like beasts 261;

cannibalism 230; initially bold, ultimately cowardly 147; defiant 117, 147, 247; disunited 150, 171, 216, 226, 237, 281; ignorant 143;

ingenium 173; large and red-haired 144; quarrelsome leaders 216; addicted to plunder 249; raucous 255; wield huge but ineffective swords 271; defective in swordsmanship 273; warlike 148; woaded 147 ; fight in the

INDEXES

presence of women 226; women as leaders and warriors 173 etymology 31, 128, 138, 143, 153, 155, 179, 183, 198, 227, 234, 246, 272, 301 exemplarity 191, 325 exitus illustrium uirorum 8 fama

g—10, 94, 123, 124, 190, 301,

305, 309 ‘firsts’ 26, 129, 137, 160—1,

207, 260

163, 201,

fleet 218 forests 220

freedmen 192, 297 freedom,

libertas 15—25, 80, 84, 148,

217, 236, 256, 303 'fronting' 32, 146, 183, 277, 305, 321

funeral

oration

308

generic combinations (Kreuzung der Gattungen etc.) 4

genitive, defining 112; descriptive 96, 97; of reference 162, 307; hyperbaton of 130; 4 abstract noun (= adj. - noun) 205, 257, 320; - neuter substantival adj. 112, 153, 220 gerund

(abl.) 201

gerundive (abl.) 245; (dat.) 246, 247 glory, gloria 9, 15, 16, 93, 101, 120,

285, 287, 298, 309, 311 'greatness' 10, 189, 310 grief 232—3, 324

hendiadys 98, 101, 150, 194, 195, 209, 221, 228, 241, 245, 269, 274, 304,

314, 319 Historia Augusta 7 n.2q historiography 1—5, 29, 35, 126, 132, 159, 204, 227,

228, 236, 256,

267 topoi of: autopsy 99; future work(s) g1, 93; greatness of topic 228; impartiality 72; importance of topic 71, 228; literary inadequacy 92, 123; majority of sources 128, 161; novelty 129 Homer 236, 322 Horace 161, 288, 329 hortatio 161, 170, 232, 236—55, 256—66 topoi of 236, 237; abduction of children

168; advantage

(ut?litas)

167; ancestors 170; same enemy 264; hope 247; patria, parentes etc. 169, 252; posterity 255; unequal numbers 169, 253 humanitas 205 hunting 265, 278 hyperbaton 130, 230, 312, 318 hypercharacterisation 255 hysteron proteron 155, 327 ideal general 26, 267 topoi of: affability 211; awareness and knowledge 186, 193, 222; is cautious 276; takes appropriate counter-measures 222; deploys troops wisely 268; directs operations in all quarters 271; maintains discipline 197; importance of first impressions 186; known to his troops 104; learns from the experienced 104; moderation 102; keeps men in

order 126; counteracts ofium 177;

chooses advantageous positions 197, 208; preparation 187; ensures supply of provisions 209; keeps troops in reserve 276; self-restraint 104; severity 194; combines severity with clemency 194; performs the role of soldier 271; deals with difficult terrain 180; knowledge of terrain

104;

welcomes toil and danger 190; see also encomium, topoi of imperialism 15—25, 198, 254 inertia 111 infinitive, historic 31, 104, 166, 183, 193, 197, 267, 272, 277, 281 inscriptions: (CIL 5.7817) 16, (CIL

6.41069) g6, (CIL 6.41106) 5, 78, (CIL 14.3612 = ILS 1025) 223, (ILS 216 — IRB 22) 162, 163, (ILS 8704a — IRB 25) 95, (IRB 28) 201, (IRB 286)

284, (RIBg1



IRB 137) 164, (RIB 662—3 — IRB 126) 203, (Tab. Vind. 2.154 — IRB 273) 272, (Tab. Vind. 2.164 — IRB 274) 147, 275 n.21 interpretatio Romana 149 Isocrates 7 itineraries, zfneraria 14 n.54 ius liberorum 108, 110, 111

Juvenal 161, 203, 216, 317

352

INDEXES

wounds 113, 232, 279, 322; yoke

Latin 202 legions 102, 252, 268, 269, 331;

169, 249

glittering appearance 223, 253

II Adiutrix

115; II Augusta

102, 116,

162; III Augusta 298; VIII Augusta

162; IX Hispana 27, 29, 102, 106, 162, 179, 222, 224, 248; XIV Gemina 102, 115, 162, 165; XX Valeria 102, 116, 162, 165, 180,

2839 letters of appointment 290 libertas see freedom Liburn(ic)ae 228

life expectancy 8g, 110

Livy 3, 12, 25, 27-9, 34, 92, 105, 132,

134, 236, 267; Livian language or expressions 78, 102, 114, 125, 141, 181, 225, 252, 274, 206

167,168, 182, 196, 226, 2209, 255, 258, 276, 277,

174, 218, 238, 266, 279,

177, 221, 243, 268, 281,

179, 222, 247, 269, 287,

180, 224, 251, 273, 292,

‘loaded alternative' 306 Lucan g7, 267, 282, 287 Lucretius go luxuria 170 maniple 228 manuscripts 35—7; see also Codex Aesinas maps 14, 134 n.11, 145, 225 marriage, girls' age at 125 Martial 7o, 87, 202, 207, 226, 303 marvels, miracula 229 memory 82, 128 Menander Rhetor 3-4, 5, 67, 95, 98,

99, 159, 312, 313, 326

metaphors 35; metaphorical equivalence between story and discourse 126 bonds 251; circumdare (wrapping) 196; dawn 83, 313; death go; exuo 122; felling or pruning 191, 192, 195; famulatus 246; fire 99, 118; gale 219; gladiatorial arena 256, 266, 277; goads

107; hands of

Death 321; hunting 256, 267, 277; medicine and disease 83, 86,

87,115,

233, 240, 254, 288;

metonymy 69, 191, 318 militares uiri 295 mimetic syntax 274 mining 155

modestia, moderatio etc. 10, 23, 70, 93,

102, 232, 293

Naevius 325 names, inversion of 78; significance of 34, 69, 85, 133, 183, 191, 225, 235, 316 (see also etymology); tria nomina 95 narrative pattern

180, 226, 232

novel 227, 231 Nuremberg Chronicle 12

obsequium 10, 23 omission see ellipse ‘only’ frequently omitted in Latin 139, 152, 169, 316 oral history 26 otium g, 10, 94, 110, 148, 177, 200,

292, 209 Ovid 269, 329

participles: ‘ab urbe condita construction' 114, 137, 145, 184, 220, 321; ‘aoristic’ 78, 103, 162, 178, 232; future as substantive 194; Grecising dat. + esse 183; pregnant use of 75; present 98, 120; substantival

144, 231, 240,

202 peace 243 pearls 156 peripeteia, dramatic reversal 159, 221, 229,224, 225, 267, 277 personification 133, 142, 153, 166, 173, 244, 273 philosophy 100, 102, 118 physical appearance 309, 310 pirates 231

Pliny the younger 6, 8, 11, 69, 84, 88,

89, 112, 302, 304, 308, 316, 317, 317 n.22, 318, 325, 329; possible allusions to T. 6q, 92, 93, 105, 120, 202, 270, 326

resurrection go; roads and travel

plural, authorial 77; ‘collaborative’ 8,

spatium 118, 311; sun 83; taming

254 poisoning 305 polyptoton 190, 195, 296

70, 71, 72, 261, 302; sowing 282; 158; war 219, 220; wild animals

35, 214, 253, 260, 261, 265;

80; ‘patriotic’

150; rhetorical

106,

INDEXES

polysyndeton 315 praefectus 211 praeteritio 123, 141 praetors 112 preface, prefatorial statements 65_7’

71,

75,

92, 93,

125,

sentence structure 32—3; periodic 30, 184, 270; see also 'appendix' sentences, word order

127,

128 procurators 95, 168, 194 proverbial expressions 86, 150, 237, 239, 252, 253, 263; see also topoi provincial governors 121, 123, 168, 189, 192, 195 quaestors 109 quies g, 10, 70, 94, 110, 184, 200, 299 reading 69; retrospective see e sequentibus praecedentia; readerly

collaboration 88, 140, 146, 152,

211, 310 relative clauses, antecedent of 96; gender of pronoun 114 repetition 91, 109, 127, 130, 144, 171, 259, 289, 328 repraesentatio 167 Res Gestae, of Augustus 16 rhythm 86, 124, 136, 257, 323, 324, 325, 329; see also clausulae ring composition 2, 63, 67, 125, 158, 163, 181, 231-2, 267, 295, 299,

308,313, 324, 326, 329

rivers 207 Rudge Cup 14 rumour 305

'saecular' language 83 saeuitia (saeuire), savagery 75, 78, 90, 91, 167, 288, 298; see also tyrant Sallust 2, g, 12, 13, 27—, 31, 32, 33,

34> 19, 91, 92, 120, 123, 127, 267; Sallustian expressions 72, 107, 119, 123,

353

94, 104, 107, 117, 158, 237, 242, 243, language or 80, 89, 105, 106, 129, 143, 144, 147,

154, 155, 161, 162, 166, 172, 174,

176, 185, 187, 189, 196, 197, 198, 221, 222, 225, 239, 249, 261, 262, 263, 269, 277, 278, 279, 280, 300, 301, 311, 324 Seneca the elder 220, 305 Seneca the younger 237; Senecan language or expressions 8g, go, 142, 242, 268, 300, 301, 311, 321, 325, 328

sententiae 32, 107, 119, 124, 151, 156, 158,161, 170, 172, 200, 211, 234, 301, 329 e sequentibus praecedentia 777

servitude, seruitus etc. 15-25, 81, 206,

236, 246, 256 shields 273 sick-visiting 306 Silius Italicus g4, 87, 242, 267, 280 simile 265, 277; see also metaphors slogans 84, 85 sources 132 speech 4, 20, 22, 32, 35, 159, 171, 236; examples in 169, 248; captatio beneuolentiae 2 38; 566 also hortatio spoils 168 Statius g4, 87, 102, 107, 108, 178, 196,

325

Stoics, Stoicism 77—8,

312

100, 303, 311,

structure 2—3, 65, 109, 119, 120, 126, 127, 129, 146, 150, 176; see also

chiasmus, ring composition, tricolon subjunctive: in generalising clauses 122, 158, 230; imperfect to express futurity 298; jussive 302; ‘polite’ use of perfect g1, 141, 263, 305, 326; present retained in repraesentatio 167; second-person singular potential 87, 146, 152, 211, 242, 252, 301, 310, 326, 327 suicide 303 superlative expressions 105 syllepsis 107, 219, 244, 260, 278, 284,

315 synchesis 131, 162, 237, 251 synecdoche 240 synkrisis4, 123, 159, 296; 566 also comparison taxation 194 technical terms used non-technically 121 teichoskopia 252, 282 'theme and variation'

107, 153, 161,

181, 251, 253 Thucydides 25, 33, 105, 158, 206 toga 204

354

INDEXES

topoi 28, 267; aduentus of a great man 291; annihilation of enemy

‘only one battle' 174; blunt

186;

soldiers 121; bravest are safest

239; bravest all die 265; cavalry fighting on foot 279; 'this day' 238; literary decline 87; decoration impressive but useless 253; no sooner discovered than conquered 138, 260; lamentation of women

and children 281;

mothers against philosophy 101; longa pax 147, 233; battle ended by night 280; night the time for planning 282; provincial frugality 09; lemeritas ducum 295; bloodless victory 269; see also consolatio, death, encomium,

ethno-

geography, historiography,

hortatio, ideal general, proverbial

expressions traductio 322 tranquillitas 292 ‘transference’ g; see also ‘automimesis’ tribuni militum 102

tricolon 79, 105, 144, 145, 155, 157,

178, 184, 188, 200, 226, 266, 303,

317, 326, 327

triumphalia ornamenta 289 tyrant 8o, 89, 241, 287, 288, 291, 293, 2094, 304, 317, 319; see also saeuitia ua?iatio, variation g1, 33—4; advs. 230;

conjunctions 98; forms (positive ~ comparative) 113; names 33, 131; prepositions 105, 214; tenses 309; vocabulary g1, 99, 145, 151, 177, 224; Word order 105, 200, 262, 313 abl. abs. — clause 124 — finite verb 33, 145 abstract — concrete 33, 197, 216, 219,

228, 233,

263, 281,

adj. ~ abl. phrase 176

322

~ genitive 210 ~ prepositional phrase 72 - relative clause 307 adv. — adj. 113, 140, 164 — prepositional phrase 233, 255 asyndeton - co-ordination 170, 254, 281 gerundive ~ prepositional phrase 244 noun - conditional clause 268 ~ infin. 118 — participle 230 prepositional phrase — noun 74, 109, 113, 123, 171, 193, 211, 273, 295, 300, 327 ~ clause 33, 129 — participle 105, 300 ut ~ adv. 233 uirtus 68, 70, 72,94, 111, 120, 129, 148, 285, 292 Valerius Flaccus 87 verbs, present tense 19 n.68; perfect tense in generalising clauses 147; simple rather than compound 174; compound followed by simple 194 vigintivirate 108 Virgil 16, 17, 34, 113, 151, 156, 196,

198, 235, 257, 267, 278, 301 viticulture 154 vivid narrative (enargeia, euidentia etc.)

149, 173, 267

warning figures 28, 222 wills 307—8 ‘window reference’ 68 word order, parallel 31, 194, 197, 219, 238, 262, 293; see also chiasmus,

sentence structure, synchesis word play see assonance

zeugma 85, 157, 250, 278, 318; double 133

LATIN

WORDS

ac 100 ad with numerals 280

bipennis 133

124 adulescentia gg alius ending a list 155 amplector 217

citra 13 claritas/claritudo 320 codicilli 290 cogito/reputo 255

aduersus ‘in relation to’ 217; /aduersum

celerum 142, 191, 217

cowinnanus 270 cnnis 145 cunctus 100 cupido/cupiditas 107 curo 176

35

ob 105

obsidio/obsidium 210 ortus 94

patientia 81 perhibeo 139 plerique 73 potius quam 221 praeuehor/praeteruehor 229 proinde 255, 263; /perinde 139

donec 272; + subjunctive 196 dux 288 enim/nam 169

-ere/-erunt 134 exterreo/terreo 33, 276

quamquam without finite verb 70; t subjunc. 83; / quamuis

felicitas 84 ferocia/ferocitas 147 fidus/fidelis 33, 164 forem/essem 174

83

-que at sentence-end 138, 155; -que et 188 quid si... 88 quidem 179, 281, 306, 311 quis/quibus 281 quo — ul 185 quominus — quin 198, 226 quotiens 70

grandis 89 hau1io 99 iactantia/iactatio 220

igitur 114 ignarus/nescius 186 in + accus. ('for the sake/purposes

referre ad 104 reor/puto 226

of’) 105

incertum + alternative indir. qu. 117 incunosus + genit. 70 infestus 217; / infensus 294 inuicem 109

scutulum 133 segnitia 104 seruitus/seruitium 81, 234 si quidem + subjunc. 214 socordia 249 sunt qui + indic. 231

inuidia 71, 120, 300

iste 290 iuuenta 99.

supersum 310

Suus 114, 221, 244

laureatae (noun) 190 luxus/luxuria 113

tolero 158

tollo (of children) 110 tum/tunc 106 lurmae 275 n.21

memoro 187; /commemoro 128

-mentum, noun formation 198 miles (collective singular) 177 mortales 143 mos 67 ; moris est 255

uastus 282 ueniam petere 5 uinclum/uinculum 251 ultra quam 100 usque 164

namque 97 ni 101, 276 non quia 4- subjunc. 326

ut (causal) 221

NAMES

Dates are AD unless stated otherwise

Adamklissi 273 Aemilius Scaurus, M. (cos. 115 BC) 74 Africa 13, 285, 298, 301, 332

Agricola, Cn. Julius (cos. 76), career summarised g—10, 332 Aldborough (Isurium) 178

Q

INDEXES

356

INDEXES

Alesia, battle of 29 Alexander the Great 233, 238, 241,

244, 263

Anglesey (Mona)

14, 166, 180, 186

Anteia, wife of the younger Helvidius 318 Antonine Wall 213 Antonius, M. (cos. 44 BC) 293 Aphrodisias 244 Aquitania 120 Aristotle 17 Arminius 169 Arria, wife of Thrasea Paetus 316 Arulenus Rusticus, Q. Junius (cos. 92) 6, 66, 70, 75, 76, 77, 110, 298,

316, 318 Asia 109, 285, 298, 301 Atilius Rufus, T. 290

Atticus, A. (?) Julius 281

Augustine, St 17 Augustus

16, 161, 258, 287, 288, 291

Baebius Massa 317

Batavians 188, 268, 272 Bodotria (Firth of Forth) 212,217

13, 181, 208,

Boece, Hector 234 Boudica

105, 159, 172, 222, 237, 242,

246, 248, 255, 257

Brigantes 13, 173, 177, 178, 248 Britain, Britons 11—15,

18—23,

129,

130, 133, 152, 212, 227, 238, 242, 244; Britanni 147; Britannia 103 Burrium see Usk Caerleon (Isca) 331 Caesar, C. Julius (cos. 59 BC) see General Index Caledonia

135, 207, 212, 213, 220,

225 Calgacus 20—5, 159, 227, 235-57 Calleva Atrebatum see Silchester Callimachus 207 Camboglanna (Castlesteads) 14 Fig. 1 Camulodunum see Colchester and Slack Cantium

(Kent)

149

Caratacus 164, 179, 237 Carlisle (Luguvalium) 14, 179, 198, 2839 Cartimandua 173, 177, 178, 180, 248

Castlesteads see Camboglanna Cato the elder see General Index

Cato, M. Porcius (praetor 54 Bc), the younger g4, 117, 120, 122, 123, 189, 190 Chatti 223, 286 Chester (Deva) 331

Chesterfield, Lord 258 Chrysippus 324 Cicero, M. Tullius (cos. 63 Bc) 566 General Index Civica Cerealis, C. Vettulenus

c. 76) 298 Civilis, C. Julius 272

(cos.

Claudius 16, 137, 162, 220, 222, 244 Clota (River Clyde) 14, 181, 212, 214

Clyde, River see Clota

Colchester (Camulodunum)

106, 164,

201, 248, 254 Columella 330 Corbridge (Coria) 209 Corbulo, Cn. Domitius (cos. 39) 107, 115, 295 Coria see Corbridge Cornelius Fuscus 295 Cornovii 283 Crag Lough 14 Crantor 232 Culloden 235

Dacia, Dacians 295 Dalginross 224 Damnonii 224 Danube 231, 295, 296 Decantae 283 Demetrius, grammarian on Agricola's staff 141, 203 Deva see Chester Didius Gallus, À. (cos. 39) 165 Dio Cassius 226, 227, 230, 237, 242,

246, 255 Domitia Decidiana, Agricola's wife 108, 232, 326

Domitian 5, 8-11, 18, 21, 33, 66, 74,

75, 76, 115, 127, 211, 223, 244, 259, 265, 285-321

Eboracum see York Elginhaugh 213 Enoch of Ascoli 36 Epidii 214 Exeter (Isca) 102, 331 Fabius Rusticus 132, 134 Fannia, daughter of Thrasea Paetus 78, 316, 318

INDEXES

357

Firth of Forth see Bodotria Flavius Clemens, T. (cos. 95) 316 Forum Iulii (Fréjus) 95, 115 Frilford 201 Frisii 231

Kent

Gaius Caligula 97, 162, 286, 309, 315

Licinius Mucianus, C. (cos. 65 (?), 70,

Galba, Ser. Sulpicius (cos. 33) 113 Gask Ridge 208, 225 Gaul, the Gauls 19-20, 202

Germany 169, 22 Glevum see Gloucester Gloucester (Glevum) 116, 165, 331 Gratilla, wife of Arulenus Rusticus 316 Guarnieri, Stefano 36 Gulf of Lyon 215 Hadrian's Wall 14, 14 n.54, 178 Helvidius Priscus (cos. before 87), the younger 318 Helvidius Priscus, C. (praetor 70), the elder 6, 21, 66, 70, 75, 76, 78,

303, 316

Herennius Senecio 6, 66, 70, 75, 76, 78,298, 316,317, 318 Herodotus

126, 222, 229

Hersfeld, abbey of 36 Himmler, Heinrich 36 n.122 Horace see General Index Iceni 172, 248 Inchtuthil 3351 India 24 Intimilii 114

178

Johnson, Dr Samuel 258 Josephus 238 n.19 Julia Procilla, Agricola's mother 97 Julius Classicianus 175 Julius Frontinus, Sex. (cos. g8, 100)

83, 120, 179

Julius Graecinus, L., Agricola’s father 96, 330

La Turbie 16 Leicester (Ratae)

165

Licinius Crassus, L. (cos. 95 BC) 303

72) 115

Lincoln (Lindum) Lindum

see Lincoln

102, 179, 248, 331

Livy see General Index Lockleys 176 Londinium

5ee London

London (Londinium) 254

Lucan see General Index Lucretius see General Index

Lugi 283 Luguvalium see Carlisle

Marcomanni 295 Martial see General Index Massilia (Marseille) 98, 137 Mela, Pomponius 141, 144, 15 215, 270 Menander Rhetor see General Index Mettius Carus 316 Moesia 295 Mollins 213 Mona see Anglesey Mons Graupius 2, 10, 14, 16, 20, 21,

29, 35, 159, 213, 221, 226, 232-83 Mull of Kintyre 214 Musonius Rufus, C. 109 n.1o

Inveresk 209 Ireland (Hibernia) 131, 214, 215, 216, 240 Isca see Caerleon and Exeter Isocrates see General Index

Isurium (Aldborough)

see Cantium

Kipling, Rudyard 24 n.8g

Junius Mauricus 318 Junius Silanus, M. (cos. 15) 97 Junius Silanus, M. (cos. 19) 97 Juno Sospita 134 Juvenal see General Index

Nepos, Cornelius 1, 4, 7, 65, 71, 94, 309

Nero g, 107,111,113,

288, 311, 315, 319

115, 162, 192,

Nerva 8, 21, 66, 83, 87, 88, 318

Newstead (Trimontium) 209, 272 Ocean

131, 137, 139, 140, 161, 169,

220 Onasander 102 n.g, Oppius Sabinus, C. Orcades see Orkneys Ordovices 13, 182 Orkneys (Orcades) Orosius 280 Ostorius Scapula, P. 163, 183 Palatium 291 Pannonia 295

165, 209, 271 (cos. 84) 295 14, 137 (cos. before 47)

INDEXES

Peduceus Priscinus, Q. (cos. 93) 309 Petilius Cerialis Caesius Rufus, Q. (cos.

Taexali 283 Tassiesholm 209

Petronius Turpilianus, P. (cos. 61) 175 Plautius, A. (cos. 29) 163 Pliny the elder 141, 206 Pliny the younger see General Index Plutarch 1, 3 n.12, 7 Pompeius Collega, Sex. (cos. 93) 309 Pompey the Great 189, 202, 287, 297, 302, 320 Posidonius 141

217, 225 Tay, River see Taus Tencteri 227 Thrasea Paetus, P. Clodius (cos. 56)

70 (), 74,83 (?)) 118, 178

Ptolemy

(Claudius Ptolemaeus),

geographer 14 n.55. Puteolanus 37 n.127, 234 Pytheas 98, 130, 137, 138, 141

Quadi 295 Quintilian 87 Quintilius Varus, P. (cos. 13 BC) 169 Ratae see Leicester

Roscius Caelius, M. (cos. 81) 116 Rutilius Rufus, P. (cos. 105 BC) 74

Rutupiae see Richborough

201, 254

Sallust see General Index Sallustius Lucullus 291 Salvius Otho Titianus, L. (cos. 52, 69) 109 Sarmatians 295 Selgovae 207 Seneca the elder and younger see General Index Silchester (Calleva Atrebatum)

Silius Italicus see General Index Silures 13, 144, 179, 183 Slack (Camulodunum)

14, 180, 207, 208,

6, 21, 66, 70, 75, 76, 77, 303,

316 Thucydides see General Index Thule (Shetland?) 14, 98, 138, 153, 209

Tiberius 161, 192, 233, 287, 288, 294,

19 Togiäumnus 149, 164 Trajan 8, 66, 83, 85, 87, 88, 313; Trajan’s Column 269 Trebellius Maximus, M. (cos. 55) 116, 175,

201, 266

Trimontium see Newstead Trucculensis 14 Tungrians 268, 272

Rhine 231, 296 Richborough (Rutupiae) 284

St Albans (Verulamium)

Taus (River Tay)

201

248

Smertae 283 Springhead (Vagniacis) 201 Statius see General Index Strabo 200 Strageath 224 Suebi 231 Suetonius Paulinus, C. (cos. c. 45) 103, 159, 165, 266 Suetonius Tranquillus 7, 94, 309 Sutcliff, Rosemary 223 Syria 285, 290 Tacitus, career of 5-6, 8—9, 11, 332; wife of 5, 316, 322, 324, 326

Usipi 137, 226—32, 253 Usk (Burrium) 102, 165, 331 Vacomagi 283 Vagniacis see Springhead

Valerius Catullus Messalinus, L.

(cos. 73, 853) 316, 318 Valerius Flaccus see General Index Varro 141, 309 Velleius Paterculus 5, 25, 125, 180, 182 Venicones 207 Veranius, Q. (cos. 49) 102 n.9, 165 Verulamium see St Albans Vespasian

114, 115, 120, 163, 178,

295, 303

Vettius Bolanus, M. (cos. 66) 116,117, 177, 266

102, 108,

Vindolanda 15 n.57, 203 n.14, 272 Virgil see General Index Viroconium see Wroxeter Votadini 207 Wallsend 14 Whin Sill 14 Wroxeter

(Viroconium)

165, 331

102, 116,

Xenophon 7, 25, 68, 101 Yeavering Bell 207 York (Eboracum) 179, 203, 223, 331