BMCR 2000.01.28

The Annals of Tacitus, book 3. Cambridge Classical Texts and Commentaries 32

, , The Annals of Tacitus, book 3. Cambridge Classical Texts and Commentaries 32. Cambridge: Cambridge University Press, 1996.

This excellent addition to the Cambridge classical texts and commentaries continues F. R. D. Goodyear’s commentary on books i and ii of the annals. Goodyear once supervised the research of Professor Woodman (henceforth W.) (see the acknowledgement in Velleius Paterculus: the Tiberian narrative [Cambridge, 1977], p. xi), and in the preface to his first volume he described Professor Martin (henceforth M.), an old friend, as ‘of all scholars the most deeply versed in the intricacies of Tacitus’ style’. He would have been delighted that these two scholars have continued his work; and the dedication to his memory will please all his friends. Amongst these the reviewer counted himself and he also apologises for the lateness of this review.

Many readers will think that W. and M. have surpassed the fine achievements of Goodyear on the annals and of Heubner on the Agricola and histories and have published the best scholarly commentary on a book of Tacitus, and perhaps on any Roman historian. An important test of the value of a commentary is whether one returns repeatedly to it with the expectation of learning something: in the years since its publication I have often returned to this book, and only rarely have my expectations been disappointed. Like Goodyear, W. and M. are probably at their best when writing about matters closely related to the interpretation of Tacitus’ Latin, but they also comment on all matters pertinent to the study of Tacitus, whether textual, stylistic, literary, historical, or antiquarian. T.’s style was an especial concern of Goodyear, but in this area W. and M. are at least his equal, and indeed have often been more thorough in collecting the parallels necessary to illuminate Tacitus’ peculiar phraseology. They are more interested than Goodyear in the literary study of Tacitus’ text and offer more subtle analyses of it; they have a stronger interest in the nature of ancient historiography and the place of Tacitus in the historiographical tradition; they have a wider curiosity about the ancient world; and they have been able to benefit from more recent publications on Tacitus.

In one new and very exciting respect W. and M. have an advantage over all previous commentators on Tacitus. While they were preparing their commentary, the discovery was announced of first one, and then a second, copy of the senatus consultum de Pisone patre (henceforth SCPP), by far the most important and exciting Latin text to be found this century, and of far greater importance for the illumination of Tacitus than any other inscription. Although W. and M. were unable to print a copy of the SCPP, its editors very generously allowed them to quote liberally from it. This has given W. and M. both an opportunity and a difficulty — an opportunity in that on many matters or problems raised by SCPP they have become the first scholars to have their views in print, a difficulty in that they have had to commit themselves to these views before a scholarly consensus has been built up concerning the interpretation of the inscription, with the possible danger that that some of their pioneering interpretations may not gain general approval. I shall return to SCPP at the end of this review.

Amongst many wide-ranging notes in the commentary, the following selection will serve to illustrate the diverse interests of W. and M.: 9. 2. Piso’s journey down the Tiber is seen as an inversion of the journey up the river of those hoping for acclaim; 12. 1. It is rightly stressed that speeches in T., even those of Tiberius, are invented; 16. 1. Evidence is produced for the ancient view that seeing was more trustworthy than hearing; 16. 2. It is argued that Tiberius’ asking how Piso spent his last day recalls characters in Greek tragedy asking how someone died; 17. 2. On exsatiare in the context of human death and blood; 25. 2. On altius = ‘from the beginning’; 37. 2. On the cost of building; 45. 2. On the hortatio before battle; 50. 1. On the manner of speech given to Lepidus; 75. 1. On significant names. Many others could be cited.

However, a few general adverse criticisms may be ventured. One concerns the discussion of textual matters, in which W. and M. are thorough and sensible but lack Goodyear’s incisiveness. Another concerns those passages in which the two commentators parade their disagreements, sometimes at length: often one of them seems much less likely to be right than the other (I found myself agreeing with W. at e.g. pp. 97-8, 200-1, 203, 204, 383-4, 402, 472, with M. at e.g. pp. 373, 404, 406, and 432-3), and sometimes his views could have been suppressed or expressed more succinctly. A third concerns the very originality of the commentators, especially W. (for whose approach to Tacitus see now the essays collected in Tacitus reviewed [Oxford, 1998]): they have many new ideas and interpretations to offer, far more than most of us could have managed (see below for some of these); but a defect of this quest for originality is that they are occasionally prone to flights of fantasy which more cautious scholars such as Goodyear would not have countenanced. Some may venture a fourth adverse criticism, that W. and M. have not provided as much historical discussion, especially of campaigning and institutions, as they might have done (contrast the historical commentaries of the kind written so well by Bosworth, Briscoe, or Walbank), and that their book is therefore not an obvious first port of call for someone interested in the politics and administration of Tiberian Rome; but, even though the authors have not set out to write a full historical commentary, an ancient historian would be unwise to ignore their discussion of any passage of T. in which he or she is interested.

Also suggestive is a comparison with the commentary on annals iv which M. and W. published in the Cambridge Greek and Latin classics in 1989. The new volume contains a great deal more material and discussion, and hence is unquestionably the more important contribution to scholarship and the one without which I should less wish to be; but the briefer compass of the older volume perhaps displays in sharper focus the literary interests of the two commentators and their probing, innovative approach.

I have said that W. and M. are at their best when discussing the interpretation of T.’s Latin. Nevertheless, one may regret that there is no general discussion of the style of T. Although it is understandable that the authors should not wish to repeat what they have already said in the introduction to their commentary on annals iv, this decision makes it difficult to pursue the implications of a comment such as the following on p. 120: ‘T. characteristically uses an appended abl. abs. rather than a main verb to introduce the avenging of Germanicus which is the main theme of this section’.

As in their commentary on annals iv W. and M. take the language of T. very seriously, always searching for its full implications and often arguing for hitherto unsuspected meanings in it. In particular, they are concerned to bring out to the full T.’s richly metaphorical language. Notes of this kind which caught my attention include those on the following passages: 7. 1. exuto iustitio suggests shedding the clothes of mourning; 11. 2. arrecta (or erecta) omni ciuitate is seen as a reference to the pricking up of the ears of the ciuitas; 27. 2. agitandi brings with it latent storm imagery; 55. 2. prolabebantur suggests a falling building. Notes which seemed to me overoptimistic in their quest for metaphor included those on the following passages: 10. 3. haud fallebat Tiberium moles cognitionis : when ‘burden’ (favoured by M.) is such an obvious and apt translation of moles one may doubt whether the meaning ‘storm’ (favoured by W.) makes its presence much felt, and it is even less likely that at 10. 1. cognitionem exciperet there is a preparation for this image; 17. 2. ab imperatore et Augusta defensam Plancinam : I am not attracted by the view that Tiberius conducts ‘a successful military campaign in defence of Plancina, enabling her to resume her own campaign’; 21. 4. spargit bellum : in view of the parallels assembled by W. and M. which suggest that the metaphor is one of sowing, it is hard to follow W. and regard the image as one of a snake spreading poison.

Like other recent Tacitean scholars, W. and M. enthusiastically detect the idiolect of Tiberius (see conveniently their index at p. 513). Since the almost total absence of any writings of Tiberius makes refutation of this tendency very difficult, I limit myself to protesting that N. P. Miller’s article, ‘Tiberius speaks’, AJPh lxxxix (1968), 1-19, is over-rated by Tacitean scholars, and that the not very significant coincidence in phrasing between Cic. Tusc. v. 107 ‘iam uero exilium … quantum tandem a perpetua peregrinatione differt?’ and the ironical comment given to the princeps ( 24. 4), that he was glad that Silanus had returned from his peregrinatio longinqua, does not make it inevitable that ‘the Tacitean Tib. has evidently been reading his Cicero’.

Textual matters are dealt with fully, and there are several attractive new proposals, for example 8. 2. dubitaba[n]tur (Reeve), 38. 3 incusans (with a full-stop after agebat and a comma after sinere) (W.), 46. 2 conuincite for euincite (W.), 53. 4 [lapidum causa] (W.), 70. 3 egregia in publicum ( egregia M.; in publicum W.) for egregium publicum ( egregias in publicum is another possibility, but the following bonas would then be rather flat). However, it is frustrating to find that several conjectures mentioned in the critical apparatus are not evaluated in the commentary, e.g. Fuchs’s terminos ( 12. 2), Held’s exceptat ( 20. 2), Muretus’ priuatisue ( 57. 1), Walter’s decorauere ( 62. 1), Watt’s comprimendi or comprimendi ( 43. 1) and egregium decus publicum ( 70. 3). If a conjecture is worth mentioning in an apparatus, it is worth discussion in an accompanying note. Conversely, at 38. 3 W.’s conjecture is mentioned in the commentary but not in the apparatus. At 11. 2. [is] haud alias intentior populus W. and M. accept the deletion of Acidalius, but also mention in their apparatus that W. S. Watt ( per litteras) suggested ita for is. Although their choice of reading is good, it would have been helpful to have had the reasons for it spelt out (I hazard two: the conjecture of Acidalius postulates an easy dittography, and the parallels which W. and M. quote for haud alias… reveal that expressions of this kind are regularly introduced with asyndeton). Often the place of publication of a conjecture is not given, for instance, for those of Acidalius, Held, Muretus, and Walter just mentioned; but some users of W. and M. will wish to ascertain the reasons why scholars proposed their conjectures.

Some more specific comments on textual notes. 11. 2. W. and M. well bring out how hard it is to decide whether arecta should be emended to arrecta or erecta. I do not censure the final choice of arrecta; but, since an indirect question follows, it would be reassuring to know that there are parallels for this after arrigere. For this construction with erigere, cf. e.g. Liv. xxvi. 22. 5 erectis omnibus exspectatione quidnam postulaturus esset. 19. 2. is finis fuit ulciscenda Germanici morte : W. and M. show that ulciscenda … morte is untranslatable, and I wish that they had either obelized or adopted one of the conjectures which they mention. 25. 1. de moderanda Papia Poppaea : ‘commentators seem not to remark on the ellipse of lege, for which we have found no parallel’. In which case one ought to give serious consideration to reading either Papia Poppaea or Papia Poppaea. 26. 2. postquam exui aequalitas et pro modestia ac pudore ambitio et uis incidebat : W. and M. may be right to defend incidebat by suggesting that it evokes the image of seed being scattered on the ground, but the passage perhaps reads more easily with Lipsius’s incedebat; compare several of the passages cited at TLL vii. 1. 856. 68-78 in which abstract nouns such as libido and timor are subject of incedere. 59. 3. bellum scilicet aut diuerso terrarum distineri, litora et lacus Campaniae cum maxime peragrantem. Lipsius proposed bello for bellum, a conjecture which I find more attractive than W. and M. do. 63. 1. auditae aliarum quoque ciuitatium legationes; quorum copia fessi patres . . . consulibus permisere : some commentators take quorum as neuter, referring to the content of the legationes, but W. and M. prefer to take it as masculine, referring to legatorum, which they extract from legationes. It might be easier still to emend to quarum.

In the manner of reviewers of commentaries I offer now some miscellaneous observations: introd. p. 11. It is noted that Aeneid vii resembles annals iii in beginning and ending with a woman, but the point and significance of the comparison is not made clear. 1. 3. On ex alto W. and M. begin their note thus: ‘”out at sea”, as Liv. 44.7.10 ex alto conspecta classis (30.10.10) and Pan. Lat. 11.10.2 Poeno ex summis Alpibus uiso‘. The first passage is certainly parallel, but the third merely provides analogy for this use of ex, and on inspection the second turns out to contain the phrase ab alto. 1. 3. For the coupling moenia ac tecta cf. e.g. Liv. ix. 4. 12, xxii. 52. 7, and xxvi. 30. 9. 2. 2 and 16. 2. One may doubt whether the mss of the Latin historians are good enough to allow us to determine whether or not they preferred the form plebes to plebs or ferme to fere. 2. 3. When Drusus and Claudius go right out to Tarracina to meet Agrippina and the ashes of Germanicus on their return to Rome, they perform what was a regular compliment to a great personage arriving. For a private individual being greeted in this way, cf. Acts 28. 15 οἱ ἀδελφοὶ ἀκούσαντες τὰ περὶ ἡμῶν ἦλθαν εἰς ἀπάντησιν ἡμῖν ἄχρι Ἀππίου Φόρου καὶ Ταβερνῶν; for members of the imperial family cf. August. res gest. 12. 1 ex senatus auctoritate pars praetorum et tribunorum plebi cum consule Q. Lucretio et principibus uiris obuiam mihi missa est in Campaniam, qui honos ad hoc tempus nemini praeter me est decretus and Suet. Gai. 3. 4 (the populace go twenty miles from Rome to meet Germanicus). 2. 3. consules … et senatus ac magna pars populi : as W. and M. note, this is a grand way of saying ‘everyone’. They offer no real parallel, but this manner of writing was rather affected by Valerius Maximus: cf. e.g. ii. 7. 7 imperiosissimi duodecim fasces, penes quos senatus et equestris ordinis et uniuersae plebis summum decus erat and iii. 7. 1 et senatum totum et uniuersum equestrem ordinem et cunctam plebem … comitem habuit (on Scipio Africanus). 2. 3. consules M. Valerius et M. Aurelius (iam enim magistratum occeperant) : M. and W. (p. 78) note the unusual fact that book iii may not begin at the beginning of a consular year, but they might have pointed out more clearly that the parenthesis allows T. to signal the beginning of a new year without disrupting the flow of his narrative; for a good parallel cf. Liv. ix. 42. 10 noui consules P. Cornelius Aruina Q. Marcius Tremulus (hi iam enim creati erant)4. 1. sine insignibus magistratus : W. and M. quote Cons. Liv. 186 adspicitur toto purpura nulla foro and explain that ‘the point of the custom was that on the day of burial no official business could be transacted’. Whilst this observation is true, it should also be noted that as a sign of grief a magistrate would dispense with his lictors and insignia : see e.g. Cic. Planc. 98 (Plancius greeting the exiled Cicero) statim ad me lictoribus dimissis, insignibus abiectis, ueste mutata profectus est. 4. 1. concidisse : W. and M. well write that ‘the verb is appropriate to persons as well as constitutions ( OLD 1, 4a) and thus suggests that Germanicus had personified the state in the same way as did an emperor’. They might also have observed that a parallel metaphor with stare is very common ( OLD 16a), and that the state might be said to stand on one individual (see e.g. Liv. vi. 1. 4 ( urbs) innixa M. Furio principe stetit, with my note or that of Kraus). 6. 3. When T. makes Tiberius refer to quotiens populus Romanus clades exercituum, interitum ducum, funditus amissas nobiles familias constanter tulerit, he alludes to an important aspect of the Roman national myth: see e.g. Liv. ix. 19. 9 Romanum, quem Caudium, quem Cannae non fregerunt, quae fregisset acies? For constantia as a Roman virtue see esp. Liv. xxx. 7. 6 Romanae in rebus aduersis constantiae; also e.g. ann. xv. 20. 4, Liv. xxvi. 12. 13, and xxxiv. 58. 11. 7. 2. In the discussion of T.’s use of the genitive of the gerundive I miss a reference to P. Petzke, Dicendi genus Tacitinum quatenus differat a Liviano (Diss. Königsberg), 42-7. 9. 1. eaque res agitata rumoribus, ut in agmine atque itinere crebro se militibus ostentauisset : W. and M. acutely suggest that ‘Piso’s rumoured behaviour is typical of a general attempting to win favour’, for which they cite one excellent parallel ( hist. i. 23. 1) and one passage ( hist. v. 1. 1) which is barely even analogous. 9. 2. quia pauidis consilia in incerto sunt : for another Tacitean generalisation about fear, cf. hist. iii. 84. 4. 9. 3. domus foro imminens : an excellent note on houses which were tyrannical, or potentially tyrannical, is supported by numerous parallels, to which one may add Sen. Thy. 455-6 non uertice alti montis impositam domum | et imminentem ciuitas humilis tremit and should subtract Sall. Cat. 40. 5 (W. and M. do not make clear that the domus of D. Brutus was of note not because of its commanding position but because it was the house of Sempronia and near the forum). 9. 3. An obvious reason for the conuiuium and epulae in the house of Piso was that such celebrations were common when someone returned from a long sojourn away from home: see the introductory note of Nisbet and Hubbard to Hor. carm. i. 36. 12. 2. For the commonplace that private enmities should be subordinated to the common good, see also P. A. Brunt, The fall of the Republic and related essays (Oxford, 1988) 368-9, with further examples. 13. 2. acie uictum : the stress here is on acie : Piso went so far that he even took part in a pitched battle against Roman forces. 15. 2. durat mentem : a famous analogy perhaps deserves to be cited alongside the parallels for mentem durare : Catull. 8. 11 sed obstinata mente perfer et obdura. 15. 3. It is noted that multam post noctem is an unparalleled expression, but Caes. Gall. iv. 36. 3 post mediam noctem should have been separated from the other analogies cited, since it does not illustrate the collocation multa nox. 16. 1. audire me memini ex senioribus uisum saepius inter manus Pisonis libellum, quem ipse non uulgauerit; sed amicos eius dictitauisse litteras Tiberii et mandata in Germanicum contineri : W. and M. write: ‘the passive contineri implies that Piso had transcribed any original litterae and mandata into his libellus; but transcriptions would have been of little value as evidence in the circumstances’; but T.’s insinuation would lose much of its point if the document possessed by Piso was of only limited evidential value. Since libellus can be used almost as an equivalent of epistula (Plaut. Pseud. 706 [with which contrast epistula used of the same document at 690 and 691], Brut. ap. Cic. fam. xi. 11. 1 Antonius … ne de Planco quidem spem adhuc abiecit, ut ex libellis eius animaduerti qui in me inciderunt, and some of the other passages cited at TLL vii. 2. 1262. 71-1263. 6), and since it is regularly used as a generic term for correspondence with a princeps ( TLL 1264. 26-82), I tentatively suggest (admittedly without a clear parallel) that here it perhaps means a bundle of documents, which included litterae and mandata. 18. 4. quanto plura recentium seu ueterum reuoluo : there may be a stronger image than W. and M., who gloss reuoluo as either ‘go back over’ or ‘think back on’ and cite OLD 2c-d, allow. We should perhaps imagine the historian turning back scrolls of history ( OLD 2b) or conceivably the fasti (Hor. serm. i. 3. 111-12 iura inuenta metu iniusti fateare necesse est, | tempora si fastosque uelis euoluere mundi; perhaps also Claud. 20. 60 prisca recensitis euoluite saecula fastis). 22. 1. The note on P. Sulpicius Quirinius curiously omits a reference to the passage which has made him in fame second only to Pontius Pilate amongst Roman provincial officials: ‘And it came to pass in those days, that there went out a decree from Caesar Augustus, that all the world should be taxed. (And this taxing was first made when Cyrenius was governor of Syria.)’ (Luke 2: 1-2). 26. 3. quidam … postquam regum pertaesum, leges maluerunt : for the contrast between rule by kings and rule by law, cf. Liv. ii. 3. 2-4 and Lact. inst. vii. 15. 14 = Sen. mai. fr. 1 ( HRR) (on Rome) reiecto superbae dominationis iugo maluisse legibus obtemperare quam regibus. I wish that W. and M. had been even firmer in their rebuttal of the argument of Syme and Townend that T., in a digression on the history of legislation which he places in AD 20, echoes a speech made by Claudius in AD 48 and now preserved on an inscription found at Lyons. This speech is famous to-day largely because it survives; but T., even though he later chose to adapt it ( ann. xi. 24. 1-7), could doubtless read many such speeches. 27. 1. multa populus parauit tuendae libertatis et firmandae concordiae: populus puzzles W. and M., who well observe that ‘while populus had long since become a synonym for plebs… its combination in this sense with patres is, if not unparalleled, at least highly unusual’. Is it conceivable that T. has used populus rather than plebs because he thought that the Roman populace was basically united in this period (note concordiae) and that only a minority of patricians engaged in factio ? It might have been observed that factiones were associated especially with patricians or nobiles : cf. e.g. Cic. rep. iii. 23, Liv. vii. 32. 12 (with my note), and ix. 26. 22. 27. 1. The note on aequum ius does not remark on its emotional significance for plebeian ideology; Brunt, loc. cit., pp. 334-8 might well have been cited. 27. 2. nec minus largitor nomine senatus : since accusations of largitio were generally made against popular politicians (see R. J. Seager, CQ n.s. xxii [1972] 335, xxvii [1977] 378), T. may be making the point that supporters of the senate also resorted to this tactic. 29. 1. For the phenomenon of a simple verb continuing from a compound, rather more bibliography could be cited: see e.g. E. J. Kenney on Lucr. iii. 261 and J. Diggle, Studies in the text of Euripides (Oxford, 1981) 18. 29. 1 non sine is not only a mannerism of T. and Suetonius but also of Horace, who uses it sixteen times. 30. 3. incolumi Maecenti is interpreted ‘while M. was still alive’, but it is perhaps not quite so easy to exclude ‘while M. was still influential’, especially as Maecenas’ fall from grace could have opened the way for Sallustius. 32. 2. socordem is an archaism: see H. Tränkle, WS ii (1968), 111-12. 33. 4. In the note illustrating the impotentia of women a famous passage has gone missing: Hor. carm. i. 37. 10-11 ( regina) quidlibet impotens sperare; note also Sen. Phaedra 699 (Phaedra speaking) sed mei non sum potens. 39. 1. alarios equites ac leuis cohortium : Tacitean uariatio is noted so very regularly (see conveniently the long entry in the index on pp. 504-5) that here the lack of comment on the change from the accusative to the genitive surprises. 45. 1. certantibus inter se signiferis, fremente etiam gregario milite, ne suetam requiem, ne spatia noctem opperiretur : for sentiments similar to these ascribed by an historian to Roman troops, cf. Liv. iii. 27. 6-8. 46. 1. oppidani neque oculis neque auribus satis competebant perhaps needs further explanation; presumably the Gauls were frightened by the gleam of the armour of the legionaries and by their war-cry. 47. 1. W. and M. argue that T. ‘presents the emperor himself as a historian of the war’, an interesting interpretation which cannot be refuted but which I should not myself wish to adopt. 47. 3. After the senate has decreed vows and supplications for the return of Tiberius solus Dolabella Cornelius, dum anteire ceteros parat, absurdam in adulationem progressus, censuit ut ouans e Campania urbem introiret. W. and M. do not comment on whether the senate adopted Dolabella’s sententia. At first glance, solus suggests that they did not, but the following rebuke of Tiberius shows that they must have done. In general, it may be suggested that, unless T. records that a sententia was modified (as at 17. 4-18. 1, 18. 3) or rejected (as at 18. 2, 33. 1-34. 1, iv. 20. 1, 30. 1), we should assume that it was passed (thus at iv. 20. 1 and iv. 20. 4 too we should assume that the mild sententia of Lepidus and the the stricter sententia of Cotta Messalinus were both adopted). 55. 1-4. There is a full discussion of the chronology implicit in this famous digresson on the decline in luxury in the first century A.D. However, since famous men began to be killed from c. A.D. 30 onwards, and since noui homines began to arrive in the senate throughout the Julio-Claudian period, the view that sect. 2-3 refers to the period before Vespasian is perhaps rejected too briskly. In which case sect. 4 sed … Vespasianus is a chronological marker; and it may not be fanciful to take sect. 2 conuertere as an historic infinitive (they began to turn to more prudent activity after killing began in the later Julio-Claudian period). 55. 3. It ought to have been observed that industria is a stock characteristic of the nouus homo : see conveniently T. P. Wiseman, New men in the Roman senate 139 B.C.-14 A.D. (Oxford, 1971) 109-11. Particularly worth citing would have been Cic. IIVerr. iii. 7 odistis hominum nouorum industriam, despicitis eorum frugalitatem, as it also illustrates the topos of the parsimony of the nouus homo, about which W. and M. might have said a little more. 65. 1. exsequi sententias haud institui nisi insignes per honestum aut notabili dedecore, quod praecipuum munus annalium reor, ne uirtutes sileantur utque prauis dictis factisque ex posteritate et infamia metus sit. W. and M. devote most space to an exposition of W.’s view, published in MH lii (1995), 111-26 = Tacitus reviewed, 86-103 that the quod -clause is a parenthesis and that we should translate ‘It has not been my practice to go through in detail senatorial sententiae except those conspicuous for honour or of notable shame (which I reckon to be a very great responsibility of history), lest virtues be silenced and so that … ‘. This interpretation is possible, and they rightly reject the common view that quod is a relative pronoun which finds its antecedent in the main clause as a whole and is explained by the following ut and ne clauses. However, the view, espoused by M. himself in Tacitus (London, 1981) 126, that quod means ‘because’ or ‘for’, could be right. At the very least, W. and M. have not shown that it is wrong; they say that it requires praecipuum to have full superlative force (‘the highest function of history’), but this is not true, and ‘a particular duty of history’ is a perfectly possible translation. 66. 4. Bruttedium artibus honestis copiosum et, si rectum iter pergeret, ad clarissima quaeque iturum : the context of a man making moral choices as to how to conduct his political career suggests an allusion to Sall. Cat. 11. 2 nam gloriam honorem imperium bonus et ignauus aeque sibi exoptant, sed ille uera uia nititur, huic quia bonae artes desunt, dolis atque fallaciis contendit. 67. 1. ille quaestor Silani, hic legatus suggests the disloyalty of Gellius and Paconius. 70. 3. bonas domi artes. W. and M. write ‘Elsewhere bonae artes can mean ‘moral qualities’ … but such a meaning seems precluded by what is said about Capito at 75. 2′; but T. may refer to Capito’s private life. 74. 4. For the theme of a general given (or wanting) a reward despite his failure to finish a war, cf. also Liv. viii. 12. 9.

On some passages where an expression or idea in annals iii may be paralleled in Livy vi-x I have already adduced material in my own commentary on those books (book vi, 1997; books vii-viii, 1998; books ix and x, forthcoming) which is not found in W. and M. Putting the needs of brevity before those of modesty, I do not repeat myself here but simply refer to the note in question, observing only that commentators on Livy and Tacitus must cover much similar ground and that from W. and M. I have learnt much of which I should otherwise have been unaware. 6. 3. nil opus uetustioribus exemplis : for new exempla being preferable to old, see Livy ix. 34. 14 quid ego antiqua repetam? and n. 19. 2. adeo maxima quaeque ambigua sunt : for concluding epiphonemata of this kind see Livy vi. 41. 11 n. 20. 1. It is noted that uagae populationes are typical of non-Romans, but not that the contrast (here implicit) between plundering and warfare constitutes a regular topos (see Livy vi. 31. 6 n.). 26. 4. religionibus … deuinxit : for further plays on the etymology of religio, see Livy vi. 1. 9 n. 28. 1. For further parallels for the topos of remedies being worse than illness see Livy vi. 16. 7 n. 30. 2. Good parallels are cited by W. and M. for the expression aditus ad honores, but it might have been worth noting that metaphors of this kind are common in Roman political contexts: see Livy vi. 35. 3 n. 32. 2. maioribus suis dedecorum : for the Roman expectation that a man should try to be worthy of his ancestors, see Livy vii. 10. 3 n. 39. 2. neque aciem aut proelium dici decuerit : for historians passing over unmemorable items, see Livy vii. 10. 5 n. 39. 2. sine nostro sanguine : contrast expressions such as hist. ii. 15. 2 ne Othonianis quidem incruenta fuit uictoria (for which see Livy vii. 8. 7 n.). 41. 3. pugnam … ciens : see Livy vii. 33. 12 n. 43. 2. inferendis ictibus inhabiles: habilis is regular in the context of weapons and armour: see Livy vii. 10. 5 n. 45. 1. uiderent modo aduersos et aspicerentur : for the importance of the eyes in combat see Livy vii. 33. 17 n. 60. 3. magnaque eius diei species fuit : cf. Livy x. 30. 4 magna eius diei … fama est. 66. 4. spretis quae tarda cum securitate, praematura uel cum exitio properant : for the topos of haste preferred to safety see Livy ix. 32. 3 celeriora quam tutiora consilia magis placuere and n. 72. 2. manente tamen nomine Pompei : not to place one’s name on the monuments of others was considered good imperial practice, and discussion of the practice of individual emperors is a regular topos of imperial historiography: see Livy viii. 30. 9 n.

Finally, I come to the use which W. and M. make of SCPP. They have studied the document very carefully indeed and (so far as I could see) have cited it at virtually every point for which its evidence is relevant. Their introductory notes at pp. 67-77, 77-9, and 110-18 on the episodes for which it provides new evidence are models of lucidity. As regards its relationship to T., they have seized on the two main problems posed by the inscription.

The first concerns the date of the trial. T. places the end of the trial of Piso before the ouatio of Drusus, which we know from the Fasti Ostieneses to have taken place on 28th May; SCPP is dated to 10th December. This problem is much discussed in literature on SCPP (for which see below) and a variety of solutions to the difficulty have been propounded. None is more plausible than that of W. and M., who suggest (p. 73) that in T. the relative positioning of the trial and of the ouatio is due to T.’s subordinating strict chronology to the artistic purposes of his narrative.

The second concerns the difference in tone between Tacitus and the inscription. As W. and M. observe, in a remark already quoted with approval in several articles on the inscription, (p. 117) ‘T. characteristically converts the monument’s monotonous confidence into discrepancy and doubt’. Whereas the inscription gives a relatively simple picture of harmony in the Julio-Claudian house, not least between Tiberius and a beloved nephew and adopted son; of a Piso who had quarrelled with Germanicus, corrupted his troops, raised the sceptre of civil war by setting Roman against Roman, and committed suicide before he could receive condign punishment; and of a princeps and senate united in their view of the punishments due to Piso and his family, T. allows many insinuations into his narrative: that Tiberius removed Silanus, a friend of Germanicus, from Syria (ii. 43. 2); that Piso thought that he was sent to Syria to restrain Germanicus (T. carefully does not state that these were the instructions of Tiberius) (ii. 43. 4); that some believed that Tiberius had sent secret orders to Piso (ii. 43. 4); that the disciplined behaviour of Drusus towards Piso had been brought about by a careful warning from Tiberius (iii. 8. 2); that Tiberius knew the potential for hostile comment on his own conduct (iii. 10. 3); that some people believed (again T. does not vouch for the statement) that Piso did not commit suicide but was killed mysteriously the day before he was about to reveal to the court the secret instructions which he had received from Tiberius (iii. 16. 1); and that Tiberius’ toning down of several over-severe proposals for the punishment of Piso and his family was due to his sympathy with them (17. 4-18. 4). All this allows T.’s readers to construct a version of events very different from that implied by SCPP, in which Tiberius was deeply suspicious of Germanicus, was fearful of the support which Germanicus received at Rome and amongst the legions, sent Piso to watch over him and perhaps even to poison him, was shocked by the mourning for Germanicus at Rome, and was determined that no hint of any quarrel between himself and Germanicus should ever form part of the official version.

I had always viewed T.’s depiction of the lamentation at Rome after the death of Germanicus as overdrawn; but in the late summer of 1997 his account found an extraordinary parallel in the emotion unleashed in Britain at the death of Diana, Princess of Wales, emotion which latter-day British equivalents to Tiberius found undignified and absurd. And the parallels go further: T.’s insinuation that Piso poisoned Germanicus at the behest of Tiberius is mirrored by the view, widely held even in educated circles, that the Princess was killed by the British Secret Service. We shall never know whether Tiberius was as hostile to Germanicus as our literary sources suggest (although it is worth noting that Germanicus was the son of a brother to whose memory Tiberius was devoted) but we do have the evidence which T. and others provide for the popular reaction to his death; and I for one am now readier to accept T.’s insight into the effect which charismatic royal personages can have on popular emotion. Furthermore, the extraordinarily sycophantic tone of SCPP seems intended to counter rumours of the kind which T. reports.

W. and M. do not much discuss the wider implications of the discovery of SCPP for our study of T., but on one matter, central to our understanding of Tacitean historiography, its evidence is vital. Even though T. is notably well informed about what happened in the senate during the reign of Tiberius, W. and M. do not accept the view that he consulted the acta senatus (see esp. p. 26 of their commentary on book iv, where they point to the general failure of ancient historians to exploit archival material). Moreover, on several occasions (but see especially chapters 1 and 5 of Tacitus reviewed), W. has argued that large portions of Tacitus’ narrative derive not from anything which modern scholars would accept as evidence but from his own invention. Yet the inscription confirms so many details of T.’s narrative, and especially of his reports of what happened in the senate (the deference of both Tiberius and the senate to the power of Livia, to take an obvious example), that no one should doubt the basic authenticity of this material; and this conclusion is supported by the more limited evidence offered by the Tabula Siarensis. If T. did not himself exploit the acta senatus, then we must conclude that he used an otherwise unknown writer who did consult these records; and that merely pushes back the problem one stage further, since there is no reason to think that the views of T.’s sources on archival research differed much from his own.

What then of Tacitean invention? To deny its existence would constitute a foolish and unwelcome return to the days when historians used Tacitus without proper concern for the literary nature of his text; but perhaps large-scale invention is to be found mostly in areas for which T. had only limited evidence, such as the description of battles. For politics at Rome the abundance of ‘official’ documentation (whether he himself consulted the acta senatus or merely derived information from them through another source) provided him with a more secure framework. But nevertheless it was only a framework. The list of doubts and insinuations assembled above from the story of Piso, and above all the faulty date for the ouatio of Drusus, show how, even when his material was at its most abundant, T. is still able to put an individual slant on it. Although the evidence of Suetonius and Dio elsewhere makes it certain that many insinuations in T. go back to the common sources of all three writers, and the same is likely to be true for the story of Piso (to which Suetonius and Dio devote only a little space), it is doubtful whether these common sources account for all this material. T. was always able to mould official material and earlier historical narrative to suit his own purposes, often discovering motives for which (in the view of many later scholars) he had no evidence. The discovery of SCPP now allows us to discern his way of operating as never before, and the further reflections of W. and M. on these problems would be most welcome.

In 1996, the year of publication, this commentary seemed notably up-to-date, and for the last two-thirds of the volume this impression remains. But scholarly comment on SCPP is now pouring from the presses. In addition to the two editiones principes of the inscription by W. Eck, A. Caballos, and F. Fernández ( Das senatus consultum de Cn. Pisone patre [Vestigia xlviii] [Munich, 1996], and its Spanish equivalent), see M. T. Griffin, ‘The senate’s story’, JRS lxxxvii (1997), 249-63, J. S. Richardson, ‘The senate, the courts, and the SC de Cn. Pisone patre‘, CQ n.s. xlvii (1997), 510-18, E. A. Meyer, ‘The senatusconsultum about Cn. Piso senior (the father)’, CJ xciii (1997-8), 318-24, D. S. Potter, ‘ Senatus consultum de Cn. Pisone‘, JRA xi (1998), 437-57, A. Cooley, ‘The moralizing message of the senatus consultum de Cn. Pisone patre‘, GR xlv (1998), 199-212, T. D. Barnes, ‘Tacitus and the senatus consultum de Cn. Pisone patre‘, Phoenix lii (1998), 125-48, C. Damon, ‘ Relatio vs. oratio : Tacitus, ann. 3. 12 and the senatus consultum de Cn. Pisone patre‘, CQ xlix (1999), 336-8, and the works in the special issue (cxx. 1) of AJPh devoted to the inscription: J. Bodel, ‘Punishing Piso’ (pp. 43-64), D. S. Potter, ‘Political theory in the senatus consultum Pisonianum‘ (pp. 65-88), R. J. A. Talbert, ‘Tacitus and the senatus consultum de Cn. Pisone patre‘ (pp. 89-98), H. I. Flower, ‘Piso in Chicago: a commentary on the APA / AIA joint seminar on the senatus consultum de Cn. Pisone patre‘ (pp. 99-116), E. Champlin, ‘The first (1996) edition of the senatus consultum de Cn. Pisone patre : a review’ (pp. 117-23); J. González, ‘Tacitus, Germanicus, Piso, and the tabula Siarensis‘ (pp. 123-42), and C. Damon, ‘The trial of Cn. Piso in Tacitus’ annals and the senatus consultum de Cn. Pisone patre : new light on narrative techique’ (pp. 143-62). It is to the credit of W. and M. that most of what they have said emerges unscathed; but let us hope that the Cambridge University Press will request a second edition, which will allow the commentators to take account of this and other work and to guarantee an even longer usefulness for their researches.