THE PROCEEDINGS OF THE AFRICAN CLASSICAL ASSOCIATIONS

Vol. 4- CONTENTS

Page

T. A. DOREY Masinissa, Syphax and Sophoniba University of Birmingham

H. J. ERASMUS Two Notes on the Early History of Sparta 3 University of South Africa.

J. R. HAMILTON The Letters in Plutarch's 9 University of Otago Alexander

L. G. POCOCK Theo9nis, 2 57-66 20 Christchurch

G. VAN N. VILJOEN A Note on Two Details in 22 University of South Africa Pindar's Myth of Pelops P. G. WALSH Liv_y and Au9ustus 26 University of Edinburgh

B. H. WARMINGTON Two Notes on Numidia 37 University of Bristol

Reviews

C. E. Graves, Thucydides, The Capture of Sphacteria (T. F. Carney, Salisbury); K. J. Dover, Greek Word Order (E. L. de Kock, Pretoria); N. E. Collinge, The Structure qfHorace's Odes (0. A. W. Dilke, Grahamstown); E. C. Kennedy, Ca esar: De Bello Gallico I(]. Engelbrecht, Stellenbosch); W. J. Bullick & J. A. Harrison, Greek Vocabulary and Idiom for Hi9h er Forms (M. P. Forder, Salisbury); E. R. Dodds, Plato, Gor9ias. A Revised Text with Introduction and Commentary (J. R. Longrigg, Wellington); A. H. M. Jones, Studies in Roman Government and Law (H.B. Mattingly, Nottingham); J. Wilson and C. Parson, A Basic Latin Vocabulary (D. G. Moore, Birmingham); J. R. Hawthorn and C. MacDonald, Roman Politics 80-44 B.C. (C. P. T. Naude, Pretoria); H. Lloyd-Jones, Menandri D_yscolos (T. B." L. Webster, London).

EDITORIAL BOARD

Dr. E. BADIAN, Durham Colleges, University of Durham (British Editor); Prof. T. B. L. WEBSTER, University College, London; Prof. G. VANN. VILJOEN, University of South Africa, Pretoria; Prof. T. F. CARNEY, University College of Rhodesia and Nyasaland (Managing Editor); Miss M. P. FORDER, University College of Rhodesia and Nyasaland (Editorial Assistant); Mr. L. M. LAMBIRIS, (Technical Editor). MASI NISSA, SYPHAX, AND SOPHONI BA

It is the purpose of this short paper to Here Appian and Dio are supported by examine the accounts given by Appian, Dio Diodorus Siculus (xxvII 7), though Diodorus Cassius, and Diodorus Siculus of the guerrilla suggests that she had actually been married to warfare carried on by Masinissa against Syphax Masinissa before her marriage to Syphax. in the years 205-204 B.C. and the short, fatal Livy, on the other hand, implies that Masi­ marriage of Masinissa and Sophoniba after the nissa had never set eyes on Sophoniba before defeat and capture of Syphax in the following he met_ her in the palace at Cirta after the year; and then to compare these accounts capture of Syphax (xxx 1 2). However, Livy's with the version given in Livy XXIX 2 9-33 account of the whole episode is highly and xxx 1 2 - 1 5. dramatized, and little weight can be attached According to Appian, (Hispanica VII 37; to the details of his story at this point. Punica II Io) Sophoniba had been betrothed The guerrilla warfare between Masinissa to Masi nissa, but the Carthaginians had then and Syphax is described by Appian (Punica II given her to Syphax to secure his loyalty 1 1 -1 2) in far more sober and less dramatic against the Romans. As a result, Masinissa terms than those employed by Livy, and joined Scipio when he was in Spain in 206 B.C. Appian sets out reasons why Masinissa was A very similar account is given by Dio Cassius able to maintain himself against superior (xvu 57, 5 1- 53 ; Zonaras IX 1 1). This version numbers; the whole passage has a Polybian runs counter to that of Livy, who, in his ring about it. However, Appian goes on to account of the meeting between Scipio and say (Punica III 13-14) that, on hearing of Masinissa and the conclusion of their alliance, Scipio's departure from Sicily, the Carthagi­ does not mention any such motive (xxvm 35). nians and Syphax came to terms with Masi­ Moreover, in XXIX 23, Livy describes the­ nissa; but Masinissa had no intention of celebration of the marriage between Syphax keeping his side of the bargain, and imme­ and Sophoniba as taking place while Scipio diately made contact with Scipio; then he was in Sicily, in 204 B.c.; though the words drew Hanno and the Carthaginian cavalry, 'mentio quoque incohata aclfinitatis' (xx1x who had not realised his change of sides, into 2 3, 3) imply that the betrothal may well have an ambush which he had planned with Scipio. taken place some time earlier. However, if This story also occurs in Dio Cassius (xv11 Appian and Dio were right in making the 65-66; Zonaras IX 1 2). betrothal of Sophoniba to Syphax the reason Finally, Appian's account of the marriage for Masinissa's desertion of the Carthaginians, of Sophoniba to Masinissa and her subsequent it would be necessary to reject the repeated death (Punica, v 27-28) is far less dramatic statements in Livy that Syphax was still than that of Livy. According to Appian, after regarded as a friend of Rome until just before the capture of Syphax, Sophoniba sent Scipio sailed from Sicily; to do this would be emissaries to Masinissa to explain that her to regard Livy's account of Scipio's prepara­ marriage to Syphax had been forced upon her; tions in Sicily as fundamentally unreliable. Masinissa accepted her explanation and married However this may be, it does not affect her; when Scipio, acting on the advice of the question of whether or not Sophoniba Syphax, insisted that she be surrendered to had been betrothed to Masinissa once before. the Romans as part of the booty, Masinissa T. A. DOREY sent her poison, telling her she now had the indicates that Appian, Dio, and Diodorus were alternative of killing herself or becoming a all following a common source, and one that captive of Rome. The account in Dio (Zonaras was different from Livy's. IX I 3) is similar to that of Livy, though once It is probable that the original source from again the previous betrothal to Masinissa is which these stories about Masinissa were implied. The remark of Syphax quoted by derived was none other than Polybius. It is Livy, thal it was the one consolation in his certain that Polybius had met Masinissa and defeat that the woman who had ruined him had had the opportunity of talking with him was now likely to ruin his rival too, occurs about the Punic War (Pol. IX 2 5). It is in a very similar form; but the actual ad­ equally certain that Polybius was very inter­ ministering of the po~son is done by Masi­ ested in the personality and career of Masi­ nissa himself, and not, as in Livy and Appian, nissa (Pol. ?(XXVII ro). It is probable, t~ere­ by a messenger. Diodorus (xxvn 7) gives a fore, that the basic material for most of these brief summary of the whole incident, mentions stories was provided to Polybius by Masinissa a previous union between Masinissa and himself. Such material, of course, would be Sophoniba, and says that Masinissa gave her highly subjective. If, then, the account of the poison with his own hands and forced her the reconciliation between Masinissa and to drink it. Carthage is to be assigned to Polybius, he Certain conclusions can be drawn from the must have obtained if from a Carthaginian discussion of the various versions set out source. above. First, the account in Livy of the Finally, apart from the dramatic overtones guerrilla warfare between Masinissa and of Livy's narrative, it differs from that of Syphax and the marriage and death of Appian, Dio, and Diodorus in giving consid­ Sophoniba is confirmed, in outline, by other erable details about the struggle for the accounts that, in part, are contained in an succession in the royal house of the Maesulii extant author, Diodorus, of earlier date than (xx1x 29). These details may have been given Livy. Secondl)'., ignoring minor points of by Polybius, but omitted for some reason by detail, such as whether Sophoniba killed the source followed by Appian, Dio Cassius, herself at Cirta or at Scipio's camp, and and Diodorus; on the other hand it is possible whether Masinissa gave her the poison with that Livy obtained these details from a his own hands or sent it by a messenger, there completely different source, one that had are three main points in which the version in some special interest in N umidian history. Appian, Dio, and Diodorus diverges from Considerations of chronology make it unlikely, that of Livy: first, the story of a previous though not impossible, that the source Livy union, whether betrothal or marriage, between used was one of the historical works of King Masinissa and Sophoniba; second, the story Juba, •but it may well have been the 'libri that the betrothal of Sophoniba to Syphax Punici' of King Hiempsal of Numidia, the while Masinissa was in Spain was what induced contemporary of Sulla, which Sallust had the latter to join Scipio; and third, the story translated (jugurtha XVII 7). of a pretended reconciliation between Masi­ nissa and the Carthaginians just before Scipio T. A. DOREY landed in Africa in 2 04 B.C. and the treacherous deception of Hanno by Masinissa. This Birmingham University

2 TWO NOTES ON THE EARLY HISTORY OF SPARTA

1. A note on the Spartan Rhetra in Plutarch, convincingly refuted by Hammond,(9) that Lye. 6 of Miss Chrimes by Woodward an? den Boer.(10) We are left with the emendation This note is concerned with the powers of Hammond, and that of Treu and Wade­ assigned to the people in the corrupt final Gery. , Several objections may be raised clause of the Spartan rhetra and the limitations against Hammond's emendation: (i) his placed upon these powers by the rider of explanation of how the error arose is

kings Polydorus and Theopompus.(1 ) palaeographically too involved, (ii) The only The corrupt final clause of the rhetra reads reason why the transposition of fl. and T in t yocµwoocvyopLIX.VY)µ"l)V xocl. xp1x-roc;, Any res­ !l.AMOTAN could have taken place is simi­ toration of this passage must take Plutarch's larity in pronunciation, but is this likely in a explanation (Lye . 6, 3) of this part of the familiar word such as oocµ6't""l)c;? (iii) The rhetra into consideration: 't"OU oe 7tA~0ouc; corruption of the completely intelligible cx0poLcr0€V't"Oc; EL7t€LV µe:v ouoevl. yvwµ"l)V 't"WV rlMWV cxyopocv . into an unintelligible yopLIX.V cannot E(jl€L't"O, TI)V o' U7t0 't"WV yep6nwv xocl. 't"WV ~IX.O'L­ be accounted for. Hammond's objections to Mwv 7tp0't"E0efo-CY.V E7tLXpLVIX.L xupLOc; ~v O o~µoc;. the emendation of Treu and Wade-Gery are Numerous emendations have been offered. as follows: "cxnocyoploc does not occur in M. Treu gives ·a list of the earlier emenda­ Greek and its meaning, if it be 'to speak tions, (2) and -below I list th9se offered against the proposals of the Gerousia', is not since Treu: reflected in Plutarch or Tyrtaeus. The form a. M. Treu:(3). oocµep o' CXV't"IX.yop(ocv ~µev xocl. oocµw is inconsistent with Plutarch's text, ·e.g. xpoc't"oc;, "dem Damos aber soil zustehen ~uMocvlou. The emendation also adds two Gegenrede und Entschied (Herrschafts­ letters to the text, introduces the connection gewalt ?) ". oe whereas the other clauses of the rhetra are b". A. von Blumenthal: (4) oocµwowv yopLIX.V ~µev an asyndeton, ~nd affords no explanation for xocl. xpoc't"oc; which he translates "civium the corruption of fl. into r ". However, arbitrium esse et potestatem." palaeographically cxv't"ocyop(oc is probable on e. H. T. Wade-Gery:(s) independently from the assumption that the copyist's eye slipped

Treu also proposed oocµw (or oocµep) CXV't"IX.yo­ from T to r.(11 ) The objection that the p(ocv ~µev xocl. xpoc't"oc; - "the demos shall have word does not occur in Greek is not decisive, the right to criticize (sc. to make counter­ as both Treu and Wade-Gery have pointed proposals ?) and the the final voice". out. (12) The corruption yocµw is more , W. den Boer accepts this emendation but • difficult but, as Wade-Gery has emphasized, with some modification in the interpreta­ some form of oiiµoc; must be restored since tion; he translates, "the people must have Plutarch certainly understood the damos to be the right to contradict and have power". (6) spoken of in this clause.(1 3) Hammond d. K. M. T. Chrimes: (7) oocµwoiiv xuplocv ~µev rightly objects to the gen. oocµw and, with - "confirming to belong to the citizens". Treu, who also considers the gen. an impos­ e. N. G, L. Hammond: (8) oocµo't"

3 H. J . . ERASMUS the ground that the last clause stands in . a tutions are those of Crete and Sparta which strong adversative relationship to the imme­ he had discussed before that of Carthage. diately preceding phrase oi.heu,; dcrq,epeLv -re With den Boer, I believe that this evidence XO(L &q,lcr-rO(cr00(L. The implied subject of this is .conclusive.(15) Wade-Gery defends his phrase is the kings an'd elders, and, after it interpretation of the meaning of &v-rll(yop[O( has been stated that the kings and elders as the 'right to criticize' by the drastic shall introduce proposals, the last clause says: procedure of not only doubting Aristotle's 'but the final decision shall rest with the dependability, but also of regrouping his people.' This special emphasis of the powers text.(1 6) Hammond has advanced a new in­ of the people in the last clause might perhaps terpretation, viz. that the procedure in the be due to the fact that the Spartan assembly, assembly was different when the Gerousia which probably developed out of the gathering was unanimous and when it was not.(17) of fighting men under the early kings, was now Prof. den Boer has shown an acceptable for the first time elevated to a statutory body way out of this difficulty whilst retaining the with defined powers. reading &v-rll(yop[O(. ( 18 ) He points out that Palaeographically, then, the emendation "it should be remembered that at an earlier of Treu and Wade-Gery appears to be the stage of political life the various terms did not most acceptable. On historical grounds, possess the definite technical meaning which however, the word &v-rO(yop[O( seems unten­ they acquired in later periods". Thus

4 TWO NOTES ON THE EARLY HISTORY OF SPARTA terpretation of the verbs oiixo--rpecpi::w and the part of the people, i.e. a safeguard against 1tixpix~ux.~i::cr0ixi, in Plutarch's reason for the abuse of their rights of discussion and introduction of the additional clause, and amendment. I believe that the rider was a the verbs ex-rpbt0V"t"()'. XIXL fJ.€"t"()'.7t0LOUV"t"()'. in his restriction of the powers of the people within explanation of the meaning of the clause. the original limits as laid down by the rhetra: If we assume, with Wade-Grey and others, if the damos should attempt to exceed its that the people had the right of discussion rights and crxo)..icxv epoi-ro the kings and elders and amendment, the meaning would of would withdraw. course be that the people changed the motions The interpretations of neither den Boer put before them contrary to the best inte­ nor Hammond can be sustained. Hammond's rests of the state. But Plutarch and Aristotle interpretation that crxoAL

Lycurgean reform at Sparta took place about second half of the eight century, or with the 600 B.c. ( 2 3) According to this theory there Second Messenian War, the time at which were three 'Dorian' tribes in the time of Tyrtaeus was writing.(30) Tyrtaeus, the Hylleis, Pamphyloi and Dymanes, In 19 !;O Hammond advanced a new inter­ but after 600 the citizens of Sparta were pretation of this fragment.(3 1 ) The main divided into five tribes, the change from the conclusions of his careful study are: in vv. three-tribe to a five-tribe organization being 14-1 6 the emphasis lies with the opening ascribed to the Lycurgean reform. (24) words, and in v. 1 6 the emphasis is fully One of the arguments adduced in support driven home with the words &">·X e:u0uc; of this alleged change in the number of tribes cruµmxne:c;. These words, Hammond contin­ is that before 600 the Spartan military ues, obviously contrast with the only word in organization was based upon the three Dorian the preceding lines which carries such tribes, while after that date we find an army emphasis, viz. :xwp(c; in v. r 2. He suggests consisting of five obal regiments. (25) Thus that the sense may be that "the Pamphyloi Wade-Gery maintains that the Spartan army Hylleis and Dymanes fought separately: 'but which fought at Plataea in 479 B.C. was an we (perhaps the Spartan of Tyrtaeus' day) obal army consisting of five obal regiments. shall obey our steadfast leaders without He believes that the existence of the army flinching, but we shall one and all combine based on the three Dorian tribes before 600 forthwith to beat down (the foe? ... ) as we is proved by Tyrtaeus fr. 1, first published in stand at close quarters to the spearmen'. In 1918 by Wilamowitz-Moellendorff. (26) Wade­ that case a contrast is drawn between the Gery, following Wilamowitz and Gercke,(27) tactics which the Dorian tribes had employed concludes that the papyrus describes a pro­ in the past and those which the Spartans are to spective battle "in which the Spartans are to employ in a future battle". (3 2 ) Hammond go into action :xwptc; II&.µcpu)..o( 't"e: xcd 'Y:1,).e"i:c; concludes that "in this poem, then, reference ~a[e; ~uµiive:c;]. In the Messenian revolt, then, may well be made to the invasion of the Pelo­ the units of the Spartan army were those three ponnese by the three Dorian tribes who are 'Dorian tribes', the same as we find in many mentioned elsewhere in this connection. The Dorian cities, racial or 'kinship' groups which contrast is also a natural one in Tyrtaeus. The purported to be descended from Herakles' change from Homeric tactics took place at three sons, Hyllos, Pamphilos, and Dyman. Sparta not later than c. 700 B.c." W. den The evidence is conclusive, but there is little Boer's only objection to this theory is that trace of it in our tradition: Aristotle in his the contrast may not be between old condi­ Constitution appears to know only the two tions at Sparta and the new tactics, but later stages, the Obal army and the Morai".(2 8) between the condition of the enemy who is Now it must be pointed out, as N . G. L. to be engaged and the conditions at Sparta.(33) Hammond has done, (29) that the evidence is Hammond's interpretation has much to not as 'conclusive' as Wade-Gery maintains. commend it, but, whereas he takes the The author of the fragment is not known and contrast expressed in the fragment as between the suggestion that Tyrtaeus wrote it is proba­ the coming battle and the remote past, I ble but by no means certain. The mention of suggest that the reference is to the immediate the Messenians in v. 66 suggests that the past, to the First Messenian War which fragment is concerned with either the First according to Tyrtaeus took place in the time Messenian War (as Tyrtaeus fr. 4 is) in the of 7tCX't"E:p(uV ~µe:'t"e:pu>v 7tCX't"e:pe:c; (fr. 4, V. 14),

6 TWO NOTES ON THE EARLY HISTORY OF SPARTA and that the main contrast contained in the was certainly accompanied by some change in poem is a contrast in tactics. the army organization. We have no informa­ Hammond(J4) has adequately proved that tion as to the nature of the change, but it the tactics described in vv. 1 6- 24 are those appears quite probable that the change from of hoplite warfare, and, except for one minor tribal to obal army should be associated with detail, he is in agreement with Miss H. L. the change from Homeric to hoplite tactics, Lorimer who has stated that "everything that rather than with an hypothetical constitu­ can be extracted from fragment I is in tional change in 600 B.c. complete accord with the tactics of the The foregoing argument is based upon the phalanx". (35) Miss Lorimer further states, assumptions that the Spartan rhetra in Plutarch "that from c. 700 onwards hoplite equipment Lycurgus 6, in terms of which the obae (what­ was general at Sparta is clear, and, in view of evertheirnature)(37) were apparently created, the interpretation sometimes put on certain must be dated early, and that the obae had been passages in Tyrtaeus or the poetry which goes in existence for some time before they were under his name, the fact is of importance. On incorporated as part of the army organization. the other hand, the evidence does not suggest It must perhaps be emphasized that it is that here any more than elsewhere the new completely unnecessary to associate any change equipment goes back into the eighth cen­ in the army organization with those obscure tury". (36) If the change to hoplite armour words in the rhetra, cpuMc; (j)UAIX~IX'\l't'IX XIXL w~ac; and tactics took place c. 700, the change was w~ix~ocV't'IX. (38) These words need not suggest effected not long before the time of Tyrtaeus. that the Dorian tribes were abolished by the It was such a fundamental change in methods rhetra (whether the participle cpuM~ocv't'oc is of warfare that it is only reasonable to expect derived from cpuM~e:tv or from cpuM't"t'e:tv)(39) that it would be reflected in the literature of and certainly do not suggest that the obae were. the time. Tyrtaeus' poem, composed at a instituted as part of the army organization time of national crisis (if in fact it does refer from the start. Constitutional changes cer­ to the Second Messenian War in the second tainly do influence army organization, but not half of the seventh century), must have had a every change in army organization is ne­ considerable emotional appeal, an appeal cessarily due to a constitutional change - which was strengthened not by a reference to changes in armour and tactics are as liable to a remote, almost forgotten past, but by a affect army organization. reference to the immediate past, to the previous occasion when the security of the state H. J. ERASMUS had been threatened by the Messenians, and University ef South Africa, Pretoria. to the momentous changes in armour and tactics which had since taken place. The ( 1) I cannot accept the view of some scholars that the threat of the Second Messenian War would rhetra and rider are contemporaneous, e.g. H. T. naturally recall memories of the first struggle, Wade-Gery, CQ_, 38, 1944, 11s; K.M.T. Chrimes, and an encouragement to the Spartans that Ancient Sparta, 1949, 477 f.; H. Rudolf, 'Die Lykurgi­ they should face and defeat the foe with their schen Rhetra und die Begriindung des spartanischen new arms and tactics, just as their forefathers Staates', Festschr!fi Bruno Snell, 19s6, 63-64. If the rider is regarded as an integral part of the rhetra, the authority had done with a different military machine, of both Plutarch and Aristotle must be disregarded (the was eminently suitable in the circumstances. mention of Aristotle in Lye. 6, 2 shows that Plutarch is The change to hoplite armour and tactics drawing on Aristotle for his information in this passage,

7 H. J. ERASMUS cf. W. den Boer, Laconian Studies, 19 _s-4, 169) and no originate from the ranks of the people. This interpreta­ scholar has yet advanced a convincing reason for doing so. tion assigns such arbitrary and vaguely defined powers to (2) 'Der Schlussatz der Grossen Rhetra', Hermes 76, the Gerousia that it renders the proper functioning of the 1941, 22-42. assembly almost impossible - any opposition to the (3) Ibid. wishes of the Gerousia could be interpreted as "eine (4) 'Zur Lykurgischen Rhetra', Hermes 77, 1942, Neigung zu misslichen Beschliisse". 212-21s. (21) Op. cit. p. 1So . (s) 'The Spartan Rhetra in Plutarch Lycurgus VI', CQ ( 22) Diehl, Anth. Lyr. Graec. vol. I, 2nd. ed. 6 f. 37, 1943, 62-67; 38, 1944, pp. 1-9, 11s-126 (reprinted (23) The greatest exponent of this theory is no doubt in Essays in Greek History, 19s8, 37-Ss)- H. T . Wade-Gery, CAHm, s62,andCQ_,38, 1944, 11sf. (6) Op. cit. p. 176- 179. In his review ofK. M . T. Chrimes' Ancient Sparta, 1949, (7) Op. cit. p. 478 f. A. M. Woodward states that "in selecting Wade-Gery's (8) 'The Lycurgean Reform at Sparta', ]HS 70, 19so, arguments (in CAH III 19 2 s) as representative of modern 42-64. views and ignoring more recent discussions of the ( 9) L.c. p. 44. problem, whether published by himself or by other histo­ (10) Woodward, Hist. 1, 19so, 633; den Boer, op. cit. rians, especially in German periodicals, the author p. 178-179. appears to some extent to be tilting at abandoned ( 11) Wade-Gery, CQ_, 37, 1943, 64 n. 2. windmills" (Hist 1, 19so, 62s). Woodward is at fault, ( 12) Treu, I.e. p. 33; Wade-Gery, CQ_, 37, 1943, 64 n. 3. for Wade-Gery's articles in CQ_,37, 1943, 62 f. and 38, ( 1 3) CQ_, 37, 1943, 64 n. 2. 1944, 1 f., 1 1sf. certainly support his account in CAH III. ('4) L.c. p. 34 n. 2. In 19 so H. W. Stubbs wrote that the constitutional (15) Op. cit. p. 176-177. settlement at Sparta "is now generally admitted to have ( 16) CQ_, 37, 1943, 7° f. taken place about 600 B.c." (CQ_,44, 1950, 3 2). In 19 s6, ( 1 7) L.c. p. 4s-47. I cannot here consider this theory, in an article entitled 'Die Lykurgische Rhetra und die with which I do not agree, in detail. I can only point Begriindung des spartanischen Staates', Festschrift Bruno out that the passage from Diodorus (XI, so), in which Snell, 1956, 61-76, Hans Rudolph dated the reform in the it is described how Hetoimaridas opposed the general time of Tyrtaeus, "also in das spate 7. Jahrhundert". Spartan wish to go to war with Athens, and which (24) Cf. for example Wade-Gery in CAH III, 560. Hammond cites in support of his argument, does not in (25) CQ_,3 8, 1944, 1 17 f. fact support it. Diodorus merely states mxpcx ,7JV (26) Sitzun9sberichte Kon. Preuss. Akad. d. Wiss., 1918, 1tpoalloxlotv l1te:Lcre: 'r7)V ye:poucrlotv xott ,ov 1Hjµov, and 728-736. there is no suggestion that there was deviation from the (27) Hermes H, 1921, 347-8. procedure which one would consider normal; Hetoi­ (28) CQ_,38, 1944, I 20. maridas first prevailed upon the Gerousia not to go to war (29)}HS 70, 19so, so. with Athens, and, when the rather surprising decision (3°) I accept the second half of the seventh century as the of the elders was brought before the people, he was time when Tyrtaeus lived, cf. RE Zweite Reihe, VII, naturally chosen to persuade the people to support the 194s-46. For the dating of the First Messenian War in decision of the Gerousia. the second half of the eighth century, cf. G. Dickins, ('8) Op. cit. P· 177. }HS 3 2, 191 2, 1 s and K. Beloch, Griechische Geschichte ('9) The emendations f)..oi,o or ot!poi:,o have been 1 2, 262 f. It is here assumed that the Second Messenian proved unnecessary by Wade-Gery, CQ_,37, 1943, 6s, War took place in about the third quarter of the 70. The meaning of &:1tocr-rot-r'ijpotc; I take to be that the seventh century, cf. A. Andrews, CQ_, 32, 1938, p. 96 kings and elders will be "withdrawers", i.e. go away n. 2; W . den Boer, Laconian Studies, 19s4, 133-34 ; G. and dissolve the meeting, cf. Hammond, I.e. p. 4S n. 2 1. Busolt, Griechische Geschichte 1, 590 n . 1. (20) Von Blumenthal, I.e. p. 2 13 was, as far as I know, (3') L.c. 50 f. the first to suggest that lx,pe1tov,ot xott µe:-rot1towuv,ot (32) Ibid. p. s 1. are imperjecta de conatu. The difference between his (33) Op. cit. p. 173 n. 1. interpretation and mine should however be emphasized: (34) L.c. p. so-si. He maintains that in terms of the rhetra the people (3s) 'The Hoplite Phalanx', BSA 42, 1947, 76-138; the possessed arbitrium et potestatem, and that in terms of the evidence of Tyrtaeus is discussed on p. 1 2 1-1 2 8. rider "soil die Gerusia die Versammlung auflosen, sobald (36) ibid. p. 9 3. sie eine Neigung zu misslichen Beschliisse wahrnimmt". (37) Discussions of the nature of the obae include: V. These "missliche Beschliisse" can (I suppose) only Ehrenberg, RE XVII, 1693-1696; A. J. Beattie, CQ_,Ns 1,

8 THE LETTERS IN PLUTARCH'S ALEXANDER

19_p, 46-48; H. T. Wade-Gery, CQ.._38, 1944, 120 f.; no new division of tribes was envisaged by the rhetra; K.M.T. Chrimes, op cit. 424 f.; W. den Boer, op. cit. Hammond, I.e. p. 59, also derives the participle from 172 f.; Hammond, I.e. 59 f.; Rudolph, I.e. 70 f. qiu:i-

THE LETTERS IN PLUTARCH'S ALEXANDER<'>

In the course of the Life Plutarch cites no must have been many reasons for forgery, and fewer than thirty-one letters written by or to Pearson (445) has rightly emphasised that we Alexander. These he evidently regards as are not dealing with a forger in the modern authentic and uses in preference to his other sense of the word. Rather, he suggests, such sources, as e.g. in his account of the operations a writer may wish to provide sensational new against Porus in chapter 60. He treats this material "to refute the judgement of historians correspondence, in fact, exactly as he does the or vindicate the judgement of philosophers". Platonic epistles in his life of Dion. As in To these reasons, as Tarn has pointed out, (3) many cases the letters contain statements we may add the need for ammunition in the about Alexander not to be found in Arrian, propaganda war which followed Alexander's and in others conflict with his account, it is death. Clearly the general position adopted obviously of prime importance to determine by Kaerst and Pearson is strong and I accept whether we are dealing with genuine letters the view of Kaerst that the onus of proof of Alexander and his correspondents, or with lies on the person who would assert the forgeries. Naturally, then, they have been authenticity of any of the letters. But proef, examined many times. In their doctoral in my opinion, does exist in the case of at dissertations at the end of the last century least one of the letters, and there is more to Pridik and Zumetikos came to the conclusion be said in favour of others than has been said. that almost all were genuine, but Kaerst and, Clearly Pridik and Zumetikos claimed too more recently, Pearson have maintained much, and many of their arguments are exactly the opposite. ( 2 ) The truth, I think, rendered valueless by Pearson's acute obser­ lies somewhere between. vation (444) that "the mere fact that a letter Kaerst (605) pointed with justification to includes remarks which are attributed to the widespread tendency in the period Alexander by reputable historians is no proof between Alexander and Plutarch to pass off that the letter is genuine". On the other forged letters as the productions of famous hand it is impossible to be satisfied with the statesmen or philosophers, and cited the position taken up by Pearson (445). "Much correspondence, almost certainly not genuine, trouble", he writes, "has been taken in attributed to Aristotle, Ptolemy and Philip. attempting to sort out the few quotations Few scholars too, I imagine, would argue that which could be genuine from those which all the letters attributed to Plato, Isocrates or probably or certainly are not; but the results Demosthenes were written by them. There of such investigation, even if valid, are of

9 J. R, HAMILTON little importance when the quotations which It is, therefore, unsafe to conclude from this could be genuine are either insignificant or are that a collection of the correspondence was already reported by the historians". But published when Ptolemy and Aristobulus surely it is not of "little importance" to wrote shortly before 280 B.c. Kaerst himself determine whether Callisthenes was implicated once made the suggestion (in his Forschun9en, in the Conspiracy of the Pages, or to learn but not apparently retracted) that the letters what Alexander thought about his divine illustrative of Alexander's character in chapters son-ship. Pearson makes no detailed exami­ 22 and 39-42 of Plutarch's Life may have been nation of the correspondence, but merely taken from Chares' history. This may be advances suggestions about the motives of the . right, although it is not susceptible of proof, forger or forgers. His position is basically but it is important to realise (as Pearson does) this: if the letters contain material found in that Plutarch must have used more than one the historians then they are forgeries based on collection of letters. Even if Chares did have this material; if they contradict the historians access to a collection of letters, they may not then they are designed "to edify or startle have been the sort of letters likely to interest readers with new discoveries", or to present Ptolemy, whose history was clearly very a picture of Alexander as a tyrant. His views different from that of Chares. Ptolemy, it are essentially those of Kaerst, to whom he would seem, avoided controversial questions refers, and it is with the detailed examination and was not concerned to relate intimate of Kaerst that I must take issue. The results details about Alexander.(4) To say that of such an examination are, in my view, because a letter contains information not important, but before I proceed to this I found in Arrian it is probably not genuine is would make a number of general points. then an argument of limited validity. Kaerst ( 6 1 7) has argued that it is extremely The existence of a collection or collections striking that the letters cited by Plutarch were of Alexander's letters was assumed long ago not known to Ptolemy and Aristobulus, or at on the strength of the method of introduction any rate were not known to have been used - 'AAE~IXvapoc; EV -rcr.i:c; E7tLO"'t'OACY.Lc; used by by them, although we know that the letters Plutarch, as also by Athenaeus and Hesychius. written by Alexander, Parmenio etc. were Pearson (448), following Kaerst, carelessly available to the earliest writers on Alexander. refers to "the frequency of his (i.e. Plutarch's) The latter part of the sentence, however, is quotations with his casual reference to merely a deduction from Plutarch Phocion 'Alexander in his letters'", but in fact Plutarch 1 7E. There we are told that "Duris writes uses this form of address only three times. that Alexander, after his victory over Darius, Powell, (S) who also holds that Plutarch made dropped the greeting x_1Xtpe:LV from his letters use of a collection of (forged) letters, likewise except when writing to Phocion and Antipater: calls attention to "the quiet naivete with Chares too has related this". But Duris is which again and again he takes credit for almost certainly using Chares here, and it is making use of the letters for the first time". unnecessary to suppose that Chares had access This, as he says, is the best proof that Plutarch to a published collection of Alexander's used a collection of letters, but of the letters correspondence. In his position as do-1Xy­ he cites two are addressed to Antipater, and ye:)..e:uc;, Court Chamberlain, he may well have the third (in chapter 60) may have been known that Alexander used this form of written to him. Altogether we have 7 letters address only in the two instances mentioned. written by or addressed to Antipater (and

IO THE LETTERS IN PLUTARCH'S ALEXANDER three others without an addressee may be to justified in assuming that the evidence of the him), so that it is probable that Plutarch had "good tradition", i.e. primarily Ptolemy, is before him a separate collection of Antipater' s invariably correct. Even Thucydides has correspondence. It is possible, I suggest, that recently, and in my opinion rightly, come this correspondence was published subsequent under fire for his account of the Athenian to 280 B.C. and was therefore not available to Empire and for his portrait of Cleon, (8) and Ptolemy and Aristobulus. The reign of we have no reason to suppose that Ptolemy Antigonos Gonatas provides a possible date, was a more objective or more truthful and the fact that Gonatas had literary interests historian than Thucydides. W e cannot accept of his own and that he was the son-in-law of his view e.g. of the Conspiracy of the Pages Antipater may be held to provide a reason for or of the operations against Porns without their publication. It is most unlikely that this asking whether he may not have had a motive correspondence is that to which Cicero refers for concealing the truth or may not have made (De Off. II 14,48) and which is almost certainly an honest mistake. forged. More probably Cicero had access to With these preliminary remarks in mind the two books of letters which, according to we may now turn to the letters themselves. Suidas (s.v. 'A,rcfaix't"po~), Antipater left About many of them, especially those to which behind, and which presumably contained only Plutarch refers only briefly, no conclusion letters written by Antipater. Two books is possible. This is true particularly of the after all would not contain many letters. eight letters in chapters 39-42 which Plutarch It is suggested too, or at any rate implied,(6) uses to illustrate various aspects of Alexander's that the existence of collections of forged character, notably his generosity towards his letters during the Hellenistic period makes friends and his solicitude for them. These highly probable the falsity of any collection are not attested elsewhere and we cannot say of letters. Recently the discovery of two whether they are genuine or not. The same papyri has been thought to strengthen this is true of the three letters in chapter 39: that assumption. The former (P.S./. XII, 1285) of Olympias rebuking Alexander's excessive includes two letters from Darius to Alexander generosity to his companions; the letter to and the replies of Alexander, the latter (Pap. Antipater warning him to beware of plots Hamb. 129) contains as well as other cor­ against him; and that from Antipater com­ respondence letters from Darius and Porns plaining of Olympias' interference in Mace­ to Alexander, and from Darius to his satraps. donian affairs. In the case of the first of these These letters, it is clear, formed one of the we do at least have a quotation from the letter sources of the Alexander-Romance. (7) But and Tarn has justification for his statement what is important in the present connection that "it is exactly what any mother of strong is that these letters, together with the vast character, let alone an Olympias, must have majority of these mentioned by Kaerst, are written to a son in Alexander's position". (9) obvious forgeries. Pearson indeed, makes the But this, of course, can hardly be said to prove significant remark that "if Plutarch was its authenticity. Elsewhere when Plutarch familiar with a text of this kind, it must have refers to letters of Alexander he evidently does been a completely different collection". We not have the correspondence before him, but may add, a completely different kind of derives his information from a historical collection. source. This is almost certainly the case in Finally I would stress that we are not chapters 19, 29 and 34. The remaining

l l J. R. HAMILTON

letters demand more extensive examination. letter Alexander has said nothing so outrageous 1. Chapter 7, 7. Plutarch sets out the text as the line he has just quoted from Menander of a letter from Alexander to Aristotle in - he might have added Callisthenes' statement which the King complains that the philosopher (frag. 3 1 Jacoby) that the sea seemed to do has published-rouc; cxxpocx:nxou c; 't'WV Mywv, and prosk_ynesis to Alexander. Menander and that they will henceforth be available to all. Callisthenes go far beyond the oux &veu -rou Gellius (N.A. xx 5, 11 f.) also gives the text 0dou of Arrian, i.e. Ptolemy, and we can of this letter, and adds the text of Aristotle's understand how Alexander might make such reply which Plutarch merely summarises. a remark and still write a letter such as that Kaerst ( 6 14) points to the widespread tenden­ which Plutarch uses. We must also consider, cy to exalt Alexander the student of as Kaerst does not, to whom Alexander sent philosophy above Alexander the king and his letter. There is much force in the supports his argument by reference to Gellius suggestion that whatever view he himself N.A. IX 3,6 and other letters. Almost all took of the matter he would not have been scholars are agreed that these letters are not likely to enlarge upon the marvellous nature genuine and I see no reason to disagree with of the occurrence if he were writing to their verdict. Antipater or other Macedonian leaders. It 2. Chapter 1 7, 8. Plutarch tells us that has also to be remembered that Callisthenes' Alexander's passage along the coast of Pam­ account was probably written later, perhaps phylia afforded many historians material for after the visit to Ammon, and that there is no bombastic description, for they wrote that it evidence that at this time Alexander wished was by some great and heaven-sent good Callisthenes to write as he did. fortune that the sea retired to make way for But there remains one objection to accepting Alexander. He then goes on to refer to a this letter as authentic. In it Alexander made letter of Alexander to an unnamed corre­ a road,(11 ) the so-called Climax, or Ladder, spondent in which the king makes no reference and passed through (? this road) starting from to any such divine intervention. Pearson (447) Phaselis. In Arrian, on the other hand, cites this letter (together with that in ch. 46) Alexander sends part of his army by a road as an example of a forgery designed to through the mountains (i.e. the Climax), "debunk" remarks made by the historians, but while he himself with his entourage goes by does not discuss it further. Kaerst, (1 o) the shore route. Thus if we take aLe:A0e:~v however, gives three reasons for thinking it in the letter to refer to the Climax the letter not genuine : it is much too brief to be must be forged, for Arrian's account based genuine; its contents conflict with Arrian on Ptolemy must be accepted. I am inclined 1 26,2; and it is not in keeping with Alexan­ to think that Plutarch is summarising too der's character to "play down" the marvellous much and that aLe:),0e:~v in the original letter element. But surely everyone except Kaerst meant through the coast passage. We might has realised that Plutarch does not quote expect a forger to make certain that he had letters verbatim (unless he says so) or in the facts straight about Alexander's route. extenso, but selects, perhaps from quite 3. Chapter 1 9, 5. Plutarch recounts that at lengthy letters, those passage which lend the time of Alexander's illness in Cilicia support to the point which he wishes to make. Parmenio sent him a letter bidding him beware Secondly, the conflict with Arrian is hardly of his physician Philip, since he had been proved. Plutarch merely remarks that in his bribed by Darius to kill him. Darius,

I 2 THE LETTERS IN PLUTARCH'S ALEXANDER according to the letter, had offered him a officer in charge of his sea-communications, large sum of money and the hand of his and rebukes him for asking if he should send daughter. Arrian (n 4,8), Justin (x1 8,6) and him two beautiful boys. The second castigates Curtius (m 6,4) all relate that Alexander Hagnon for a similar offer, while the third received such a letter from Parmenio, al­ instructs Parmenio to put to death two though Arrian mentions only a bribe, Justin Macedonian soldiers if-found guilty of rape. only a bribe of money and Curtius says that Neither of the latter two incidents is mentioned Philip was offered 1 ,ooo talents and marriage elsewhere and the first only by Plutarch in with the king's sister. Plutarch has taken the two passages in the Moralia (33 3 A and 1 099C) whole incident, including Parmenio's letter, and by Athenaeus (1 22d), who cites the same from the historical source which · he is letter. Clearly then we cannot say whether following at this point. the letters are genuine or not. Kaerst ( 6 1 6) 4. Chapter 20,9. Plutarch cites a letter holds that they are all forgeries designed to from Alexander to Antipater to dispose of the prove Alexander's chastity and this is certainly assertion by Chares that in the battle at Issus possible, but his assertion that the fact that the king had been wounded by Darius himself. Athenaeus as well as Plutarch mentions the In the letter Alexander does not mention his letter to Philoxenus tells against its authen­ assailant and says that he had suffered no ill­ ticity must be rejected out of hand. If Pridik effects. The wound received by Alexander is and Zumetikos are right in holding that well attested, being mentioned by all the Plutarch found it (? this extract) in Chares, major sources,(12) and Kaerst's reason (611) Athenaeus probably also took it from him, for rejecting the letter is that it contains no since he quotes from Chares no fewer than "individuelles Moment", "nichts Besonderes six times. und Eigenartiges" ! Again his method is at 8. Chapter 2 5, 8. It is impossible to say fault. Even if the letter dealt exclusively with whether the letter sent by Alexander to his the battle, which it need not have done, . tutor Leonidas is genuine or not. Plutarch is using it to contradict Chares only 9. Chapter 2 7, 8. After his visit to the oasis in one particular, the wound sustained by of Siwah Alexander is said to have written to Alexander. He -does not attempt a detailed his mother Olympias that he had received account of the battle, in which he was not certain secret responses which he would tell interested, just as he disposes of the actual to her alone when he returned. Kaerst (61 2) fighting at the river Hydaspes in a few lines. regards this letter as not genuine, as "an Pearson's view (447) is that it was useful to Alexandrian fabrication", but gives no reason be able to quote a letter in which Alexander beyond the prevalence of letters, e.g. Gellius made light of the wound to show the folly N.A. XIII 4, dealing with the relationship of of Chares' exaggeration. Such a possibility Ammon and Olympias.(13) But the contents may be admitted but I cannot think it deserves of this letter agree essentially with Arrian' s to be taken seriously. remark that Alexander heard what he 5-7. Chapter 22, 2-3-4. In this chapter desired.(14) It seems most improbable that a Plutarch assembles a number of anecdotes letter giving no information would have been designed to illustrate Alexander's moral and forged and Tarn(15) very rightly asks what physical continence, and in connection with motive a forger could have had for writing it. the former cites three letters written by the There is no good reason to doubt its au­ king. The first is addressed to Philoxenus, the thenticity. J. R. HAMILTON

1 o. Chapter 2 8, 2. In this chapter Plutarch 16. Chapter 39,7. From Olympias. deals with Alexander's attitude to his divinity 17. Chapter 39,11. To Antipater. and concludes that he had no belief in it. As 1 8. Chapter 3 9, 1 3 . From Antipater. an honest man, however, Plutarch quotes 19. Chapter 41,4. To Peucestas. from a letter in which the king refers to 2 o. Chapter 41 , 5. To Hephaestion. Philip as his so-called father, implying that his 21 . Chapter 41,6. To the physician Alexip­ real father was Ammon. I have dealt with pus. this letter at length elsewhere(16 ) and see no 2 2. Chapter 41, 7. To the physician Pau­ reason to alter my conclusion that it is a sanias. genuine letter written by Alexander towards 2 3. Chapter 4 2, 1 • To an unnamed corres- the end of his life. pondent. 1 1 - 1 2. Chapter 2 9, 7. A letter of Darius 24. Chapter 42, 1. To Peucestas. and reply of Alexander. This is the second 2 5. Chapter 4 2, 1 . To Megabyzus. of the two embassies sent by Darius to As I have written above (p. 2 3), no decision Alexander and is given by Arrian (n 2 5, 1 - 3) about these letters is possible. I list them for in almost identical terms, although he places the sake of completeness. it during the siege of Tyre. Here Plutarch is 26. Chapter 46,3. In this chapter Plutarch probably following Aristobulus. gives a list of those authors who said that the 1 3. Chapter 34, 2. The instructions to the Amazon queen visited Alexander beyond the Greeks and to the Plataeans are attested by no Jaxartes, and then follows this by another list other author and the former nowhere else. of those writers who said that it was a mere The instruction to the Plataeans is mentioned invention. He then remarks that Alexander again by Plutarch at Aristeides 1 1, 8 where we seems to agree with the latter as he wrote are told that Alexander had a proclamation that the Scythian king offered him his made at the Olympic games. The instruction daughter's hand in marriage but did not was presumably issued in the form of a mention the Amazon. Kaerst (608-9) admits diagramma, but it seems unlikely that Plutarch that this letter in itself gives no cause for doubt saw this. More probably he is following a and rejects it merely because he has already historical source. r ejected the letter mentioned in the following 1 4. Chapter 3 7, 3 . Following a lacuna in chapter, of which this is clearly a part. The the text, which must have contained a offer by the Scythian king is attested by Arrian narrative of the capture of Persepolis, Plutarch (1v 1 5 ,3) and Curtius (vm 1 ,9), but there is no relates that Alexander himself wrote that he indication that this was a famous episode such gave instructions for the execution of the as a forger might be expected to know. captives. The words yp&cpe:t y(j'_p ocu't"o~ Plutarch finds (in a collection) this letter to suggest that Plutarch has before him a letter, Antipater dealing in detail with the events of although this is perhaps not certain. The the period but containing no reference to the historians say nothing of the incident and the queen of the Amazons. He therefore concludes authenticity of the letter must be considered that Alexander supports those writers who doubtful, although we must not argue that say the whole thing was invented, but does such an action does not fit Alexander's not venture to assert that the omission proves character. this.(17) Hence his use of fotxe:v. It seems 1 5. Chapter 3 9 ,4. To Phocion - cf. Phocion excessively sceptical to see in this letter the 18,6. work of a forger.

14 THE LETTERS IN PLUTARCH'S ALEXANDER

Chapter 4 7, 3. Fearing that the Macedonians apply to the situation on the river Hydaspes. will refuse to continue the expedition, It seems clear from Plutarch's o-rL -rcxu-r' Alexander addresses the picked force he has d1t6v-ro,;

But the matter can be settled without mind. Certainly (chapter 74) we hear of reference to the contents of the letter. (20 ) derogatory remarks about Aristotle's teaching. We need allow no weight to Kaerst' s remark Chares writes that Callisthenes was kept in that Alexander would not have written to bonds for seven months in order that he might Craterus while he was absent on a short be tried ev -rcji cruve:oplep in the presence of expedition. The king may well have wished Aristotle. Ptolemy admittedly says that he to scotch any rumours regarding the extent was tortured and then hanged, but Aristobulus of the conspiracy ; the letter does not refer relates that he was taken along with the army directly to Callisthenes, but merely says that bound in fetters and to this extent supports only the Pages were involved.(21 ) Kaerst lays Chares. If Chares is right in saying that no great stress on the absence of Polyperchon Alexander intended to have him tried in from the list of addressees - "his name could Aristotle's presence, his evidence would easily be left out by Plutarch" - but in fact seem to support the authenticity of the letter. the omission is of vital importance. Thirl­ We may suppose that Alexander intended to wall(22) had already remarked that the punish him by seeing his nephew tried (and genuineness of the letter "seems to be placed presumably condemned), or we may even beyond doubt by its direction, which could consider that Alexander had something worse hardly have occurred to a forger". Certainly in store for him. Zumetikos (p. 52) goes so we cannot credit a forger with the necessary far as to derive the letters from Chares' knowledge to send a letter to only those three history on the grounds that they are closely commanders who were in fact operating interwoven with the narrative. This seems together. For Curtius (vm, 5, 2), whose very questionable; e:u0uc; and ucr-re:pov may well authority was undoubtedly Ptolemy, after be Plutarch's own deduction from the mentioning the dispatch of Craterus against contents of the letters. the rebels, continues "Polypercon quoque 2 9. Chapter 57, 8. By the river Oxus regionem, quae Bubacene appellatur, m Alexander discovered a spring of oil and ditionem redegit". He was therefore oper­ writes to Antipater that he regarded this as ating independently when the letter was sent. one of the greatest of omens, since the seers 2 8. Chapter 5 5, 7. After the previous letter prophesied from it a glorious expedition. Plutarch goes on to say that afterwards Athenaeus (n 42 f.) mentions that Alexander Alexander wrote to Antipater inculpating wrote about the discovery of oil in Asia, and Callisthenes and threatening to punish -rov presumably had before him this letter to crocpicrTT)V and those who sent him out. Kaerst Antipater. This discovery itself is well (608) naturally takes this letter also to be a attested, (23) and it is difficult to see why forgery. It is connected, he writes, with the anyone should want to forge a letter on this rumours regarding the complicity of Aristotle subject, since it can hardly be called a and Antipater in Alexander's murder by "sensational new document". Why the fact poison. We hear, he argues, of no danger to that the letter is mentioned by both Athenaeus Aristotle beyond these threats, while the and Plutarch should argue against its authen­ evidence of Chares, who does not mention ticity, as Kaerst maintains, I do not understand. the danger to Aristotle, is at variance with the 30. Chapter 60, r-r 1. Plutarch bases his letter. These arguments are very weak. account of the operations against Porus Alexander's anger with Aristotle was short­ almost entirely upon a letter of Alexander, lived; he had more important tasks on his and stresses the fact that it is the account of

16 THE LETTERS IN PLUTARCH'S ALEXANDER

Alexander himself - cr.u-ro<; EV 'rCY.L<; lmcr-roACY.L<;. contemplate an easy defeat of the Indians Besides the discussions in Forschungen 1 1 1 f. (xpcr.TIJcre:tv ou x_cr.Amwi;) with cavalry alone. and Philologus 51, 608-611, Kaerst has He never defeated a large army with cavalry devoted a separate article (Phil. 56, 406-41 2) unsupported. An excellent example of the to the examination of this letter and has weakness of cavalry is provided by Alexander's concluded that it is a forgery based on operations against the Malli (Arrian VI 8,6), Ptolemv and Aristobulus. That Arrian's when he pursues them across the river ; account of the operations is much more Hydraotes but on their seeing that he had detailed and much clearer than Plutarch's and only cavalry with him they turned and made that the letter "kein neues Motiv ergibt" a vigorous resistance. Alexander, seeing their (Phil. 51, 609) will surprise no one. Kaerst solid formation, since his own infantry was not himself realises that Plutarch has no taste for present, kept circling round and making this kind of thing. But, in fact, the letter charges without coming to close quarters. contains material not found elsewhere; e.g. Cavalry unsupported by infantry could not Alexander advances 20 stades, and the battle win battles. If Arrian's account did not rest lasts until the eighth hour. Kaerst finds only on the authority of Ptolemy, it would have two important differences between the two been rejected out of hand. It is, of course, accounts. In the first place, after crossing the open to anyone to argue that Arrian is here river Hydaspes Alexander advances with his summarising Ptolemy's account and that it is cavalry and mounted archers, leaving the he who is responsible for the omission of the infantry to follow. Then, according to the reference to an attack by the Indian cavalry. letter, he expected he would easily defeat the In that case one objection to the authenticity enemy cavalry if they should attack, or, in case of the letter vanishes. the enemy phalanx advanced, that his infantry A second reason for rejecting the letter is would come up in time. The version of found in the divergence between Ptolemy Arrian (v 14, 2) is significantly different. In (Arrian v 14,6) and Plutarch concerning the his account, Alexander, being superior in strength of the detachment under Porus' son: cavalry, rode on ahead with cavalry and 2,000 cavalry and , 20 chariots according to mounted archers, intending with these troops to Ptolemy, 1 ,ooo cavalry and 60 chariots overcome the whole Indian force, or at least to according to the letter. Arrian also tells us contain it until the infantry arrived. Zume­ that Aristobulus gave only 60 chariots. Kaerst tikos (p. 60) calls this with justice a "consilium concludes, somewhat rashly, that the letter is temerari um atque praeceps". It is hard to a conflation of Ptolemy and Aristobulus and believe that Alexander did not envisage the talks of "Konkordanz" and "Uebereinstim­ possibility of an attack by a detachment of mung" between Plutarch and Aristobulus. cavalry, even if, as Kaerst says, the main But they agree only in the number of chariots, strength of the Indian force did not lie in this and we have no right to suppose that Aristo­ arm. Moreover did he expect the Indian king bulus may have mentioned cavalry. Ar­ to attack with a force containing elephants? rian' s(24) digression clearly implies that his But there is a weightier objection to accepting account gave only 60 chariots and did not Arrian's account. Although the decisive include other troops. ouoe yixp dx6i;, he attack in the main battle was made by writes, Ilwpov . .. ~uv E~'Y)x6v-rcr.&.pµcr.crtv µ6vot<; cavalry, Alexander needed the support of his hm:µ41cr.t -rov (Y.\J'rOU 7t'CY.L<'ICY. and continues (after infantry and it is incredible that he could the digression) cxMix <'ltcrx_i)..[oui; yixp Mye:t

17 J. R. HAMILTON bmdi:; &.yov-r(I. &qnxfo0a.i -rov Ilwpou 1ta.'i:oa., contrast strikingly with those of the Indian ixpµa.-ra. 0€ exa.-rov xa.t dxom. The emphasis forces facing them - 80,000 cavalry, 200,000 in the last sentence is on the presence of infantry, 8,000 chariots and 6,000 elephants! cavalry, not on the number of chariots. But I suggest that the figures for Porus' troops are the difference between the figures of Ptolemy taken from the letter used in chapter 60. If and those of the letter remains. It is hard to so, it is surely most unlikely that a forger find a convincing reason for it, whether one would have underestimated the totals of the assumes that the letter is genuine or not. Indian forces which Alexander defeated at Certainly it is not altogether satisfactory to the Hydaspes. argue that Ptolemy has exaggerated the 3 1 . Chapter 7 1, 8. After the mutiny at numbers "ad maiorem regis gloriam" in this Opis Alexander sends home those Macedonians minor matter, although he does exaggerate who are unfit for further service and according Persian losses as e.g. at Gaugamela.(25) But to Plutarch writes to Antipater instructing why should the writer of the letter, if it is a him to see that they should have the front forgery, minimise the numbers? Was he seats at games and in the theatre and should trying to depreciate Alexander's success, or wear garlands. Arrian (vu 1 2 ,4) says that merely to be different? This seems even less Craterus was sent to conduct them home and likely than the view that Ptolemy has exag­ that Alexander ordered Antipater to bring out gerated. May we not have here an honest reinforcements. Pearson (446 n. 73) holds difference of opinion? Alexander will then that this contradicts the substance of the have given his estimate of the numbers letter, but this is not necessarily so. Certainly involved and Ptolemy his. ( 26) Arrian does not say, as he asserts, that Craterus Apart from these differences, there is was to look after the resettlement of these nothing to give one pause. The letter men, and it is possible that Antipater might contains nothing at variance with Arrian's do this before he brought out the fresh troops. account of the tactics in the main battle. In But admittedly there is no proef that the letter both it is the presence of elephants which is genuine. Our verdict should be non liquet. determines Alexander's dispositions, and in To sum up, we may say that with regard to both Coenus is sent against the Indian right the letters in chapter 2 2 and chapters 39-42, wing. (27) It is foolish to expect in a veiy most of which deal with Alexander's character, summary account of the battle, made by a our verdict can only be non liquet. This writer who had no interest in its detail, the applies also to the letter to Leonidas (no. 8). accuracy of an Arrian. Nos. 3, 11, 1 2 and probably 13 come not There is, finally, one piece of evidence from collections, but from the historians, and which may be relevant and which, if its we have no reason to doubt their authenticity. relevance is admitted, greatly strengthens But it is the remaining letters which are the case for acceptance of the letter. I refer important and about which the historian must to the passage at the beginning of chapter 6 2 make up his mind. The letter to Aristotle where the Macedonians who had with difficulty (no. 1) is almost certainly not genuine, and defeated Porus' army of 20,000 infantry and nos. 14 and 3 1 must be considered doubtful. 2,000 cavalry(2 8) refuse to cross the Ganges Of the remainder the letter to Craterus, to engage the vast Indian forces on the other Alcetas and Attalus (no. 27) is certainly side. The very moderate totals of 20,000 authentic and the others, mostly to Antipater, infantry and 2 ,coo cavalry under Porus have good claims, I believe, to be considered

18 THE LETTERS IN PLUTARCH'S ALEXANDER

genuine. Each letter must be considered on mountain Climax, and the name has evidently come to be its merits and it must be remembered that transferred from the road to the mountains. ( 12) Arrian 11 1 2, 1 ; Diodorus xv11 34, s; Curtius III alongside the many forgeries which circulated 11, 10; Justin XI 9, 9. in the Hellenistic period there must have ( 13) Kaerst sees in this letter the origin of these later been many genuine letters in existence. letters, but they are more probably due to the writings Collections might be entirely trustworthy, or of Cleitarchus and the well-known addiction of Olympias entirely forged, like the Romance letters, or to mystic cults and to snake-handling - cf. Plut. Alexander 2. anywhere in between. (14) III 4, s - cbcoucrc.t½ ocrc.t c.tu-rci> n:po½ 6uµoG 1)V. J. R. HAMILTON (1S) Op. cit. (n. 3), 348 n. 2. ( 16) CQ..,(Ns) 3, 1953, 151-157. Universiry ef Ota90 (") Pearson (447), however, thinks it was useful to be able to quote a letter in which Alexander did not mention the Amazon but said that the Scythian king (') I am indebted to Mr. G. T. Griffith and to Dr. E. offered him his daughter's hand in marriage. But such a Badian for reading an earlier version of this article and letter would prove nothing. for making valuable criticisms. They are not, of course, (18) See Arrian IV 2 2, 1. responsible for the views expressed here. ( 19) Forschungen 1 1 o. (2) E. Pridik, De Alexandri epistularum commercio, Berlin, (20) On what follows see my article 'Alexander's letter 1 893 ; A. Zumetikos, De Alexandri Olympiadisque to Craterus, Attalus and Alcetas' in CQ_,(Ns) s, I9H, epistularum fontibus et reliquiis, Berlin, 1894; both of the 219-221. foregoing contain references to earlier work. J. Kaerst, (21 ) The -rLvc.t½

TH EOG N IS, 2 5 7 - 6 6

257 "lmtoc; eyw x.ixA~ x.ixl 1xi::0Al'Yj, IXAAa x.&x.t

20 THEOGNIS, 257-66

parents' place. D. takes the point of 261 of the dialogue. The younger man (the teller and 2 6 3 to be "something like this" : 'It isn't of the tale) is treated familiarly - he goes with the wine . .. that is drunk that is any advantage the girl to the well - but without formality to me; on the contrary it is the fact that they or respect, being given nothing but water to want their drink cold that gives me my chance' drink, or so he complains. i.e. to make love to the girl, who has to go 4. The girl's rather indelicate and horsy and fetch water with which to cool it. This metaphor of 257-60 is 'played down' by the seems to me to be throwing a lot of work on young man in his response of 2 6 1 -4; but its the µot's in 261 and 3. I do not think overtones continue, from xixxLcrTov in 2 57, there is any case for D.'s admittedly tentative and xocxov in 260, to xocxlwv in 262; whilst suggestion that lines 257-260 should be XO:A'l) tmtoc; in 2 57 has suggested Te:pdvri in transposed to follow 261-266 as the substance 261 - 'smooth', 'sleek', 'well-groomed' (and of the "soft word" or words ('ri!:pe:v) spoken in the context difficult to translate appro­ by the girl on being kissed. The words in priately); and this again has suggested TEpe:v 2 57-60 seem to me much too vigorous and in 266; the verb xocTlxe:w in 262 re-echoes bitter to be so described. A single word, or a ~vloxov in 260, in the sense of 'bridle', word or two of endearment is rather to be 'control', 'hold the reins' (cf. Aesch. Persae expected from the willing recipient of a kiss. 190 et al.) ; while cpi!:pe:L µe: in 2 64 is similarly I have in short, with respect and due acknow­ reminiscent of &vapoc cpi!:pw in 2 58. ledgement for the stimulus to Professor To render the passage then, with some Davison, a somewhat different explanation to licence of expression: 257-60 '"I'm a good offer for 257-266, as they stand and all of filly and a winner. But I've a rider who's no a piece. good at all, and I hate it. I've often thought 1. The key to it is 257-260. The theme of taking the bit between my teeth, and doing or metaphor of these lines is familiar enough, a bolt, and giving that rotten jockey the e.g. in the ditty "My husband's a jockey, a push." 261-66 "No wine gets drunk for me jockey, a jockey ... ", and it is not to be at that fine (filly's) house: for another fellow's taken any more primly in the present context. in charge there although he is certainly not the 2. The dramatis personae are a young man, man that I am. Her 'dear parents' drink the speaker of 261-264 and the reporter of nothing but cold water for me - so that the tale; and a young woman whose 'dear she has to go draw it and, in spite of her parents' wish another man to enjoy, or rather grumbling, put up with (a jockey like) me to continue to enjoy, her favours. If she were while she's doing it." So saying, I put my a tender maiden she would hardly be so arm around her waist and kissed her neck, outspoken in 257-260, and if she were a and she said a soft word from her lips.' married woman she would not be living at I take yowcroc as referring, more or less home. playfully, to the note of dissatisfaction that the 3. This other man ( older and better off, we girl has voiced, a moment before, in 2 57-60. may assume) they receive with honour and L. G. POCOCK regale with wine - when he comes. It need not be assumed that he is there at the moment Christchurch, N.Z.

2 I A NOTE ON TWO DETAILS IN PINDAR'S MYTH OF PELOPS

In the course of preparing a paper, intended element,(4) which not only purifies the latter for a non-specialist audience, on the myth and so as to render it in itself a suitable symbol the imagery in Pindar's First Olympian Ode(•) or exemplum of areta, on the mythical plane, (S) a few minor details of interpretation occurred but also increases the relevance of the myth to me which it might be worth while bringing by introducing an additional manifestation to the attention of classicists. That minor of inborn areta, namely that of physical details can have great weight in Pindar's beauty.(6) But the reference to Ganymede narratives has been well demonstrated by also serves to bring out the full significance Leonard Illig: Pindar seldom recounts events of Pelops' translation into heaven (already or describes circumstances merely for the joy suggested by 7t0't"L owµIX ~Loe; µE't"IX~~mxt, 42), of telling a story; the objective and picturesque for at the end of the contemporaneous(7) in his myths are seldom there for purely tenth Olympian Ode the poet stresses the aesthetic reasons; in all its details his narrative youthful beauty of the victor &. 7tO't"E I

approach to his subject matter.(2 ) Ganymede was to Pindar a mythical symbol In his version of the Pelops myth Pindar of immortalio/(8) achieved in heaven through rejects, as invidious gossip made up by physical beauty . . jealous neighbours, the current account about Now, in Pindar's reference to Ganymede, Pelops being cut up by his father Tantalus and one is struck by the insistence on two details : served to the divine guests attending his first, Ganymede came to the abode of Zeus banquet at Sipylus. The real cause of Pelops' "at a second occasion"; second, he came there disappearance, Pindar tells us, was that "for the same service to Zeus". According Poseidon, who had fallen in love with him, to Homer (Iliad xx 231/2, v 265/6) and the during the banquet abducted him on his golden Homeric Hymn to Aphrodite (v. 207/8), chariot "to the highest home of Zeus" (37-42). which Krakidis(9) has shown to have influenced Having recounted this, Pindar adds (43-45): Pindar's version in respect of Pelops, Gany­ mede was the son of Tros, i'.e. belonged to the ~V01X oe:u-rlpep xp6vep generation before Pelops, who should be ~A0E XIXL r!Xvuµ~o'Y)c; reckoned as a contemporary of Laomedon z'Ylvt Twih' &1tt xpfoc; (grandson of Tros and grandfather of Hector, "where at a second occasion also Ganymede who was the contemporary of Pelops' came for the same service to Zeus". grandson Agamemnon). In the Little Iliad By this reference to the parallel case of (fr. 6 K), however, a different tradition was Ganymede Pindar, so to speak, quotes the preserved making Ganymede the son of source or authority upon which he has based Laomedon and so belonging to the generation the innovation he introduced into the Pelops efter Pelops. Apparently Pindar's explicit myth. (3) For it is well known that as reference to owTepep xp6vep is a declaration

Poseidon is said by Pindar to have abducted in favour of this latter tradition. (10) But Pelops, so Zeus himself abducted the Trojan why did he so expressly prefer this particular prince Ganymede because of his beauty. On version ?(I 1 ) Naturally in order to ensure for the model of the Ganymede myth Pindar the hero of his present myth priority and introduces into the Pelops myth a new preeminence in being translated to heaven.

22 A NOTE ON TWO DETAILS IN PINDAR'S MYTH OF PELOPE

Pelops thus has the honour of being the first dating from the earliest period portray Zeus to be abducted to Olympus. Had Ganymede's caressing or in the company of a youthful, ascension and immortalisation preceded, sometimes naked, Ganymede playing with a Pelops' subsequent, and temporary, sojourn cock or a hoop, and sometimes with a winged would have shown up rather unfavourably. Eros present; occasionally Zeus is pursuing Pelops, on the other hand, was in any case the youth, showing that the motif belongs "in soon again to lose his elevation so that there den Kreis der Liebesverfolgungen ". (17) It is would then be ample "room" for a succes­ this tradition which Pindar prefers.

sor. (12) So, the parallel case of Ganymede It seems over-ingenious to suppose with was introduced to "authorise" the innovation; Kakridis that Pelops was abducted both for but once the two heroes are juxtaposed, the Poseidon's own erotic purposes and, though present myth requires measures to ensure the this is admitted to be of secondary importance, precedence of its hero. to serve as cupbearer in Zeus' palace.(1 8) In the second detail upon which Pindar This supposition is made from the express insists - "for the same service to Zeus" - he reference in 1to-rt owµix D,,.toc; (42) and its again declares in favour of a younger variant parallelism with the Hymn to Aphrodite tradition. Clearly, according to Pindar, (D,,.toc; Xot't'IX owµot 0e:oLc; emotvoxoe:uOL), and Poseidon abducted Pelops with an erotic also from the assumption that Pelops must motive,· ( 1 J ) an d -rwu-r" ' e:1tt' ' xpe:oc;' can on1 y mean have been serving as cupbearer at his father's that this was Zeus' motive with Ganymede banquet when he was abducted.(19) But too (cf. also cruv Ku1tpoye:ve:i: in OJ. X I 10). though Poseidon may very well have used his So Pindar here expresses his preference for a handsome epwµe:voc; for this and other services, younger tradition above the older epic version that does not enter into Pindar's representa­

according to which Ganymede was abducted tion at all. (20) Could -rwiho in 45, merely 1 merely to serve as a cupbearer in Olympus, ( 4) from the foregoing mention of a banquet at e.g. Iliad xx 234/ 5 ~d otvoxoe:ue:w, I XIX/\/\EOc; Sipylus and the words 7tO't't owµix /;),.toe;, receive e:tve:xot oto; Hymn. Aphr. 2 o 3 / 6 ~pmrne:v 8v the implied additional connotation of service otot~ ' Xot/\/\Oc;'~ ~ "LV ' ()('0 otVot't'Otcn' µe:-re:niI I XIXt' -re: utoc;A ' as cupbearer? Also the express addition of Xot't'IX owµix 0e:oLc; emotvoxoe:UOt. In literature Z"Y)vl in 45 clearly shows that service to Zeus the erotic motive in the Ganymede myth was not intended in the mention of Poseidon was perhaps already found towards the end taking Pelops "to Zeus' home". The reference of the sixth century in Ibycus (fr. 30 Bgk.), to Zeus' home in connection with abduction for this poet is reported by the scholiast on by Poseidon need not upset us: it is merely Apoll. Rhod. III 1 58 to have described the abduction of Ganymede (as well as that of a variation for "O"Auµ1toc; (54), used at the same Tithonus) in his Ode to Gorgias, which was time as an associative link - typical of Pindar's apparently inspired by homosexual love, a manner(21 ) - preparing for the mention of new theme which this poet introduced into Ganymede who was abducted to "Zeus' choral lyric - 1tixtotxot 6µvot. (15) In Theognis house" in the narrower sense, and so anti­ 1 345 f. the erotic motive for Zeus' abduction cipating the theme of immortality which is of Ganymede is clearly stated, though these subsequently suggested more directly by the verses are of disputed authorship and date. ( 16) introduction of Ganymede. An early Ionic-Italic amphora, two late black­ figure vases and a series of red-figure ones * * * 23 G. VANN. VILJOEN

The passage 67-87 presents Pelops' success­ tradition which assigned a decisive role to the ful contest with Oenomaus for the hand of his treachery of Myrtilus in the contest against daughter, Hippodamia. In his presentation Oenomaus, (24) the connection consisting in of this episode Pindar concentrates attention the curse which M yrtilus was said to have rather on the spirit and mental attitude of called down upon Pelops and his descendants the hero than on the actual contest. After the when the latter killed him, or simply in the detailed report of Pelops' prayer to his blood-guilt caused by the killing. (25) It was divine lover Poseidon (7 r-8 s), the poet Myrtilus' father, the god Hermes, (26) who succinctly states the outcome (86-88): the subsequently sent the lamb with the golden words of Pelops did not remain ineffective; fleece from which resulted the disaffection the god glorified him with a gift of a golden between the two brothers. Now, for obvious chariot and unwearying winged steeds; he reasons, Pindar in his account of events beat Oenomaus and gained his bride. Then completely disregards the role of Myrtilus in one more detail is added (89): favour of the (perhaps older) version based on & -rc:x.e: AIXYE't"IX<; ~~ &pe:-r1Xi:cn µe:µ1X6-r1Xc; urnuc; . the divine aid of Poseidon. (27) And in This point falls strictly outside the limits of the addition to thus tacitly "purifying" the Oenomaus story,(22) and seems to hang narrative of the Myrtilus tradition, Pindar somewhat loosely between, on the one hand, adds, in the end, in verse 89 an implicit the subsequent reference to Pelops' cult at rejection also of the horrors in the subsequent Olympia, serving as a bridge between the history of Pelops' sons. (28 ) Pindar is by his legendary past and present reality (90-9 s, cf. own preceding narrative doubly justified in 23-24b), and, on the other hand, the culminaa doing so, for not only has the absence of tion of the mythical episode about the contest Myrtilus elim.inated a major cause or motive with Oenomaus in 8 8. The reference to the for the Thyestes-Atreus atrocities, but also sons of Pelops could have been introduced as Pindar's correction about the crime of a natural sequel to his marriage, and, as such, Tantalus has removed that version of which an additional honour to the hero(23) - the later crime of the Pelopids was largely a additional, that is, to the honour of receiving parallel. Pindar deals with this point in such the love of Poseidon and a temporary sojourn a vague and indirect way so as to avoid in his in Olympus, and to the honour of his god­ poem an excess of sombre detail. (29) given success against the formidable Oenomaus, and to the honours of the cult which are G. VAN N. VILJOEN mentioned next. However, if this is all that University ef South Africa, Pretoria there is to it, verse 89 still seems to me a rather weak addition. But the verse gains in power and relevance ( 1)' Pindaros se Eerste Olimpiese Ode - Beeld en Simboo I', accepted for publication in Standpunte 3 5, if in the emphasis laid upon the virtues of these 1961 (Cape Town). sons of Pelops ( &pe:-r1Xi:cn µe:µ1X6-r1Xc;) we notice ( 2) L. Illig, Zur Form der pindarischen Erzahlun9, diss. an implicit refutation or suppression of the Kiel 193 1 (Berlin 19 3 2), 6-9 and 68 f. ( ch. 1v) : our well-known tradition about the dispute be­ poet's approach results "aus einer Pindar eigentiimlichen tween the brothers Atreus and Thyestes, and Art, ein mythisches Geschehen nicht nur anschaulich, about Atreus' serving up to his brother the sondern zugleich auch in seiner sinnbildlichen Bedeutung zu sehen und aus der Verpfl ichtung, Uber das bloss flesh of the latter's own children. This Anschauliche hinaus das dem geistigen Auge erfassbare element of the myth was connected with that Sinnbild wirksam zu gestalten ". A NOTE ON TWO DETAILS IN PINDAR'S MYTH OF PELOPE

(3) J. T. Kakridis, ' Die Pelopssage bei Pindar', Philolo- (22) With f).e:v... rcocp6ivov -.e: cruve:uvov (88) the 9us 85, 1930, 463-477, cf. 467 n. 19. compositional ring introduced in 69 by hoIµov (4) cre: 8' cxv·-doc rcpo-.ipwv cp6ta\y~oµocL (36). cxve:cpp6v-.Lcre:v y&:µov is closed and the episode concluded. (5) On this function of the myth see e.g. J. H. Finley, For a recent summary, with the relevant bibliography, Pindar and Aeschylus, '9H, 19 and 40; W. Schadewaldt, of this stylistic feature, cf. B. A. van Groningen, La Der Aujbau des pindarischen Epinikion, Sehr. Konigsberger composition litteraire archaique 9recque, Verh. Kon. Ned. Gel. Ges., ., 9 2 8, 33 8 f. ; L. Illig, loc. cit. Akad., N.R. LXV, 2, 1958, p f. (6) This is how he interprets and links up with the (23) Cf. E. Thummer's Die Reli9iositi:it Pindars, Com­ traditional datum of e}.icpocvn cpoct8Lµov wµov xe:xoc8- mentationes Aenipontianae, 1957, 84, where he stresses µevov ( 2 7). the parallelism between the series of honours received (') Both celebrate victories gained at the 76th Olympiad, by the hero and those of the victor, Hieron. 476 B.c.; on the dating of 01. x in the second half of 474 (24) There is unanimity among scholars about the relative see my thesis Pindaros se Tiende en E!fde Olimpiese Odes, chronology of the two versions. See e.g. C. Robert, Leiden '9H, , 1-14. Griech . Heldensa9e 1, 1920, 209; J. T . Kakridis, Hermes (8) J. Duchemin, Pindare: poete et prophete, 19 H' I 60- 63, 1928, 415 f.; K. Scherling, RE XVI 1152-3 (s.v. 162 ; she attaches to the event also the significance of Myrtilos, 1933), RE Suppl. v11 851-2 (s.v. Pelops, an initiation. 1940). The earliest evidence for the role of Myrtilus (9) Loe. cit. 463-465. is the fifth century Athenian logographer Pherecydes (1°) I cannot share Kakridis' objection (p. 467-9) that (Fr. Gr. Hist. 1, 3, fr. 37). According to Pausanias v 17, considerations of genealogical chronology ("diese 7, the chest of Cypselus portrayed Oenomaus pursuing spitzfindige Entdeckung") would have no influence upon Pelops who had Hippodamia with him on his chariot a poet's mind in deciding such matters. (the contest being in origin apparently a violent (") This question is also implied by Gildersleeve's abduction), which was drawn by winged horses; Pelops remark ad Joe. : "and so the chronology is saved, if it is is also portrayed with winged horses on two black­ worth saving". figure lecythi (Sauer, Arch. Jahrb. 6, 1891, 34 ; P. ( 12) Still, the honour and priority of the present hero, Jacobsthal, Gottin9er Vasen Taf. 6, 2, - quoted by and not the impossibility of having two cup-bearers Krakridis p. 417, n. 4, and Scherling, RE Suppl. vn simultaneously in Olympus (Kakridis, 468-9), is the 861-2). The winged horses seem to indicate the early decisive factor. At any rate there was already Hebe currency (at least sixth century) of a version in which serving in this capacity (Iliad IV 2) ! the treachery of Myrtilus had no place. (Yet the (13) -.ou... ep,focroc-.o 2 5 ,8ocµlv-.oc cppta\voc½ lµta\pep 4,, winged steeds are retained together with Myrtilus by cp!ALOC 8c7>poc Kurcploc½ 75. Pherecydes fr . 37b and Eur. Orest. 988 f. rc-.ocvov µe:v (14) Its difference in this from the later version was 8[wyµoc TCWAWV . . . Mup-.[}.ou cp6vov - a contamination noted by Aristarchus by marking a 8mA'ij against Iliad xx necessitated by ( or originating) the version making 2 34 (see Schol. ad loc. ). Pelops journey across the sea after his victory). But I (15) See W . Schmid, Gesch. Griech. Lit. I, 1929, 494, who cannot accept Scherling's conclusion that Pindar also points out (p. 49 3, n. 3) that to the rare innovations "could not have known the treachery of Myrtilus" (see introduced into myth by Ibycus belong "die neuen also· note 28 below) and that Pherecydes is probably the erotischen Ziige", as e.g. the love of Hypnus for author of the Myrtilus version (RExv11152,66, 1154,40; Endymion (Diehl ad fr. 1), and of Talos for Rhada­ more cautiously at Suppl. VII 8 52 ; cf. W. Schmid, manthys (fr. 32 Bgk). Gesch . Griech. Lit. 1, 711: "Pherecydes will offenbar nur (16) Diehl ad loc. ; Schmid, op. cit. 3 77. das Uberlieferte geben "). There seems to be no valid (1') P. Friedlander, RE vn 742.31 f., 56 f., also 749.1 reason why we may not assume that Pindar knew both (s.v. Ganymedes, 1912). versions, selected that which best fitted the paradeigmatic (18) Loe. cit. 465-6, cf. n . 16 . See previously A. tendency of his narrative, and linked it up with (perhaps Boeckh in his commentary (1821) p. 108, and W. it was a motivating source for) the erotic theme he had Christ in his ( 1896) ad loc. already introduced in the first part of his narrative. If (19) Ibid. 469-470; cf. also K. Scherling, RE Suppl. my interpretation of line 89 is valid, it becomes an vn 8 p,, 5 f. Kakridis also refers to Lucian, De dial. argument in favour of Pindar's having known the dear. 1v 4, v 2, where using Ganymede as cup-bearer is Myrtilus version with its consequences. advanced as a pretext for Zeus' real purpose. (25) Soph. El. 508-p 5 and Eur. Orest. 990 f., 1548-9 (20) Wilamowitz, Pindaros, 1922, 236. consider the killing of Myrtilus as the cause of the woes

( 21) Schadewaldt, op. cit. 301-2, 305-8. of the Pelopids; the curse is mentioned by Apollodorus, P. G. WALSH

Epit. 11 8 and Schol. Eur. Orest. 990, I. 2 5 (Schwartz). we may interpret some not too obviously relevant details ( 26) Eur. Orest. 997 indirectly suggests the role of Her­ as an implicit refutation of a different tradition, con­ mes, and this is made explicit by several scholia ad Joe. necting the founding of the games with Pelops (support (990,26, 995,24-5, 998,11, ed. Schwartz). The schol. for which might in some quarters have been wrongly 9 9 5, 26 f. states that according to the 'cyclographer' inferred from 01 . 1) - e.g. the introductory statement Dionysius ofSamos (Fr. Gr. Hist. 1 15, fr. 7) Euripides (v. 2 5b) that Pel ops' tomb was already &pxcxi:ov when in his account about the lamb follows an early epic, the the games were founded; the statement (v. 52-3) that Alkmaionid, but that in Pherecydes' version the lamb the Cronus hill was nameless and snow-drenched was sent not by Hermes but by Artemis (Fr. Gr . Hist. 1 3, (contrasting with its sun-bathed aspect in the time of fr. 1 33). From this Scherling concludes (RE xvi Heracles, 01. m 25b) "formerly, while Oenomaus 1154,26 f.) that in Pherecydes - and a priori also in ruled", i.e. in the time of Pelops. See my thesis (note Pindar - the killing of Myrtilus had not yet come to be 7), 150- 156, 191-193. connected with the woes of the Pelopids - but this need ( 28) In their commentaries Mezger (1880) and Galiano 2 not follow, for Pherecydes may be contaminating here as ( 19 56) assume that Pindar, like Homer, did not yet know in his account (fr. 37b) combining Myrtilus with winged these horrors. Dissen-Schneidewin ( 1 847) more steeds of Pelops (cf. also W. Schmid, op. cit. 71 2; cautiously say "vides tacere Pindarum. . . nefaria Jacoby, Fr. Gr. Hist . 1 403) or simply disregarding the discidia". different version (Jacoby, op. cit. 424). (29) Gildersleeve (1890) remarks ad Joe . "one cannibalic (27) In the myth of the tenth O lympian Ode, too, incident was enough for one poem, to say nothing of the describing Heracles' founding of the Olympic games as a rule .. a l<

LIVY AND AUGUSTUS

The thesis that Livy's Histories represents such distortion is to be justified in view of the a work of propaganda to stabilise and to purpose to be served. . . That purpose may glorify the regime of Augustus is by no means be described, in the vernacular, as an attempt new.(!) Since 1939 it has received massive to 'sell' the Augustan system. " Cochrane support, partly under the stimulus of, or in further quotes Livy's comment in the Preface, reaction from, German National Socialism "nee vitia nostra nee remedia pati possumus"; and its cultural propaganda-machine. (2 ) In he adds: "To say this, however, is to admit Germany the theory reached an extreme form that the resources of government, in the in the work of G. Stiibler.(3) In 1939-40 ordinary sense, have been exhausted. It is to (the dates are significant) two influential suggest that a situation has arisen which calls books were published in English in which for nothing less than the intervention of a Livy's alleged services to the Augustan second founder. And, finally, it is to point principate were sharply criticised. C. N. to Augustus Caesar as the man." Cochrane(4) states unequivocally: Sir Ronald Syme's memorable The Roman "With Livy we are far removed from the Revolution makes the point more allusively : Thucydidean sense of history as a diligent and "The Emperor and his" (italics mine) "historian meticulous search for truth, conducted with understood each other ... "(s) : "Poetry and due regard for the most exacting standards of history" (Livy is the only historian implicated) evidence. What Livy offers us is rather an "were designed to work upon the upper and unabashed tract for the times; the Augustan middle classes of a regenerated society".(6) version, in fact, of Plato's noble lie. And if Syme further suggests that Livy, like V ergil this involves an element of artistic distortion, and Horace, "had everything to gain from the 26 LIVY AND AUGUSTUS new order", and may have had "private and by making a careful distinction between the material reasons for gratitude to Augustus". traditional Roman Chauvinism, of which The implication is clear, though he concedes Livy is undoubtedly guilty, and the particular that it is inference, resting on no solid propaganda element which might laud the evidence. But in a recent article, which has person and political achievement of Augustus performed a maieutic function for this paper, in order to ease his path to the successful Syme appears at first sight to have modified his manipulation of power. Allegations of pur­ standpoint. He makes it clear that Livy is poseful service to the regime fundamentally "not a flatterer and a timeserver", and even misinterpret Livy's political attitudes. questions the relevance of the label 'Augustan historian' : "All in all, Livy, the pride and I POLITICAL PRESSURES ON LIVY glory of Augustan letters, should perhaps be claimed as the last of the Republican writ­ How far was Livy free to write what he ers". (7) Yet throughout this article Syme wished? The clearest picture of literary makes various allegations of detail to suggest freedom under Augustus can be gleaned from that Livy lent his writings to purposes of Ovid's letters from exile. The literary world propaganda, and finally comes the blunt was circumscribed, and Augustus was suffi­ statement: "Livy's annals of Augustus were ciently acquainted with the greater number written in joyful acceptance of the new order, of the litterateurs to be able to exert strong in praise of the government and its moral pressure on them. Yet there are those achievements". (8) who persue a sturdy independence of action, The student of Livy's Histories must find like Carus and Celsus. (13) Of repressive legal many grounds for criticism of them. Not least action there is little evidence, except in the there is the Chauvinism which excludes two sectors of writings affecting sexual sympathetic assessments of non-Italian na­ morality, and defamation. The case of Ovid tions, (9) and which portrays in idealised illustrates the first, the banishment of Cassius colours the political and military virtues of Severus for pamphleteering against "viros the Old Republic.(10) Yet Livy was more feminasque inlustres"(14) the second. These honest than his annalistic predecessors, and a punishments were exacted in the final years strong case can be made out for his having of Augustus' principate. composed his history in resolute independence Historical studies were little affected. of contemporary political pressures, and in Augustus sat in the audience when the accordance with his own uncompromising outspoken Cremutius Cordus, at a recitatio, political views. It may well be true that the labelled Brutus and Cassius - outlawed by Ab Urbe Condita, "patriotic, moral and horta­ the princeps at a less creditable stage of his tory", could be exploited by Augustus "to career - 'ultimos Romanorum".( 1s) It was honour the memory of ancient valour, revive Cremutius, too, who severely criticised the the pride of the nation and educate coming triumvirs for the proscriptions, again without generations to civic virtue". (11 ) But it is any reaction from the government. (16 ) The important to emphasise that such was the case of the historian T. Labienus, whose traditional function of Roman historiography books were burned by Senatorial decree, must since the time of Sempronius Asellio.(12) however be noted. The elder Seneca's Cochrane's thesis that Livy consciously sought evidence suggests that Labienus, nicknamed to stabilise Augustan autocracy is to be refuted Rabienus for his violent mxpp1Jcrla, was not the 27 P.G . . WALSH v1ct1m of the government but of private earlier, when he became aware of the high enmity: "effectum est enim per inimicos ut literary promise of the Ab Urbe Condita. No omnes eius libri comburerentur".(17) Such evidence however exists to suggest that evidence as there is tempts one to the view Augustus promoted the birth of the work; that there may be an element of exaggeration Livy's case, vis-a-vis Augustus, is utterly in Syme's portrayal of the censorship in different from Vergil's. Augustus' final years: "One symptom was the The imperial friendship left its mark on the suppression of offensive literature. Bonfires poetry of Vergil and Horace. Augustus duly were decreed by vote of the Senate. The appears in the Aeneid - "aurea condet sae­ histories of the Pompeian Labienus were cula" (24) - and the sympathetic understanding amongst the condemned books".(18 ) Dio of the reader is considerably taxed in the states that the bonfires were for scurrilous Horatian poems which celebrate the prin­ libels only. ( 1 9) One need not be an apologist ceps . ( 2 s) Can a similar allegation be directed for the Augustan principate to believe that against Livy's Histories? The evidence from "the Roman government was loath to interfere the extant books must first be considered. with literary freedom". (20) One cannot, of course, dismiss the possibility II THE PREFACE AND THE FIRST PENTAD that Augustus' revival of the long dormant law of defamation afforded the princeps a If the extreme, blatantly untenable view pretext for a wider censorship. But Livy's be discounted that Livy seeks to shed oblique work was virtually completed by the time glory on Augustus by his characterisation of Cassius Severus and Labienus were prosecuted. such heroes as Scipio African us, (26) the Seneca emphasises the unprecendented nature vital areas for consideration are the Preface of Labienus' punishment: "in hoc primum and the first pen tad. Livy states in his excogitata est nova poena. . . res nova et Preface that his history has a didactic function invisitata supplicium de studiis sumi ". Livy in the moral and political spheres,(27) and the was free to frame his own historical inter­ Preface has closest reference to the early pretation. But there is a subtler form of books. (2 8) The political lesson which he censorship, as Syme shows, (21 ) the pressure repeatedly propounds is that Senatorial govern­ of friendship and benign patronage. As with ment, embracing the Ciceronian concordia Vergil and Horace, as with members of Ovid's ordinum and exhibiting enlightened sympathy circle, so with Livy the princeps forged to the plebs, is the ideal. The speeches are amicable relations. When did the friendship the chief media for this lesson. Kingship at commence? Our first indication shows it Rome is anathema: "regium nomen, alibi flourishing in the later years of the reign, magnum, Romae intolerabile esse". (29) Liber­ probably after s B.c., for Livy had reached tas is fundamental: "ea esse vota omnium, ut in his history the career of Pompey.(22 ) qui libertati erit in illa urbe finis, idem urbi Further testimony is provided by the histo­ sit".(3°) The rule of law must transcend any rian's encouragement of the initial studies of individual's political power: "imperiaque Claudius, who was thus congenially detained legum potentiora quam hominum pera­ from the formal appearances of state so gam".(31 ) Libertas is above all dependent on embarrassing to his great-uncle Augustus. the regular transfer of the supreme power : The date was about A. D. 8.(2 3) But "vicissitudinem imperitandi, quod unum exae­ probably Augustus sought out Livy much quandae sit libertatis".(3 2 ) When the Senate

28 LIVY AND AUGUSTUS governs wisely, its popularity exceeds that of an earlier date of composition is likely. any ambitious individual : "itaque haec indul­ Bayet(3B) suggests that an earlier edition of gentia patrum ... adeo concordem civitatem 1-v had been published in 31-29, which tenuit, ut regium nomen non summi magis presupposes a date of composition before quam infimi horrerent, nee quisquam unus Actium. Syme demonstrates the tenuous malis artibus postea tam popularis esset quam tum basis of this thesis, but though he proposes a bene imperando uni versus senatus fuit". (33) later date he considers that much of the (There may be a first-century connotation in pen tad had been written by 2 9. This the employment of 'popularis' here, and assumption also rests on insecure ar­ Julius Caesar is probably in Livy's mind.) Livy guments, (39) and the tone of the later books believes that such aristocratic paternalism (rv especially) suggests that Augustus had at can achieve the concordia which is vital for the the time of composition embarked on his maintenance of libertas and the advancement programme of domestic consolidation. The of Rome's stature: "aeternas opes esse Roma­ date of publication is in short the likeliest nas nisi inter semet ipsi seditionibus sae­ criterion; young historians seeking to establish viant. "(34) Livy addresses this message - the a reputation do not sit on their manuscripts necessity of concordia - directly to his for years . Augustan audience : "mille acies graviores If the first pentad was begun after Augustus' quam Macedonum atque Alexandri avertit return to Rome in 29, and the later part of it (sc. miles Romanus) avertetque, modo sit composed after the ' restoration of the perpetuus huius qua vivimus pacis amor et Republic' in 27, Livy 's anxiety to vvrite civilis cura concordiae ". (35) For the universal didactic history has particular point. He was Augustan longing was for peace, and the politically naive, but he was preaching the horrors of the Civil War dominate Livy 's desirability of a return to Senatorial govern­ description of the clash between Rome and ment, and Augustus' moderation provided Alba: "civili simillimum bello, prope inter deceptive hopes. The Preface succinctly patres natosque".(36) states that a reversion to the old Republican Such passages attest Livy's uncompromising institutions can effect Rome's salvation: senatorial outlook; he is a true son of "inde tibi tuaeque reipublicae quod imitere Patavium. He can have lent support to capias". (4°) But first-century Rome lacks the Augustus only for so long as he believed that necessary moral stamina: "nee vitia nostra

Augustus was striving to reinstate Senatorial nee remedia pati possumus".(4 1 ) What are government. There is no systematic exploita­ the remedia? Syme believes that Livy refers tion of the historical themes of later decades to "the acceptance of centralised government in Augustus' interests; alleged examples of as the only guarantee of Rome's salvation". (4 2 ) such exploitation in the early books merit all But Livy has no such political vision, as his the closer an analysis . Some must be dismissed analysis of first-century history shows. (43) as fanciful. To the rest the prevailing Livy's remedia are to be sought only in the historical circumstances at the time of moral sphere - no less than the idealised bonae composition are closely relevant. artes of the old Republic. He looks backward, The dating of the first Pen tad is thus of more not forward. than academic importance. Book r was Only two explicit references are made to published between 27 and 25,(37) but many Augustus in the first five books. The first believe that the text has been revised, and that is an appendix to his account of Numa's P. G. WALSH

closure of the temple of Janus. The two that since Cossus described himself as consul subsequent closures are cited - the first after before Jove and Romulus he could hardly the completion of the first Punic War, the have been lying. (5°) It comes as a shock to second "quod nostrae aetati di dederunt ut find that Livy is apparently troubled by videremus, post bellum Actiacum ab impera­ Cossus' possible dishonesty, not Augustus'; tore Caesare Augusto pace terra marique but the reader is left with the clear realisation parta ". (44) It needs a scholar's fervid fancy that the literary sources are united against to regard this prosaic comment as a 'palinode' Augustus' 'evidence', and perhaps Livy is to Augustus for the allegedly rough handling satisfied at achieving this in the politest of the gens Julia in Book I. (45) possible way. There may thus be more than The second passage is at 1v 2 o, 7 : "hoc ego absent-mindedness (and more than inadequate cum Augustum Caesarem, templorum omnium revision) in the subsequent statement that conditorem aut restitutorem, ingressum aedem when Cossus won the distinction he was a

Feretri lovis quam vetustate dilapsam refecit, military tribune. (5 1 ) Syme is surely right se ipsum in thorace linteo scriptum legisse when he comments that from Livy's view­ audissem, prope sacrilegium ratus sum Cosso point the whole business was "a vexatious spoliorum suorum Caesarem, ipsius templi disturbance in a smooth and satisfactory. auctorem, subtrahere testem". The reference narrative, which had been guaranteed by the to Augustus' rebuilding of the temples consensus of the written sources". (5 2 ) registers Livy's strong approval; pietas towards These direct references apart, a whole field the gods is conspicuous amongst the bonae of speculation, subjective in interpretation, artes on the revival of which Rome's re­ lies open to the researcher. Some have generation depends. But sacrilegium, translated examined Livy's treatment of the Julia gens, carefully, has no reference to the person vainly hoping to discover some flattering of Augustus. (46) manipulation of the tradition.(53) Others Dessau(47) drew attention to the connection analyse the use of the adjective augustus in of this passage with the request of the procon­ association with Rome's early history - an sul Licinius Crassus to be allowed to dedicate unrewarding Wortstudie in view of its in­ the spolia opima. (48) Syme(49) brilliantly creasing currency from the time of Cicero extends the argument to show the unlikelihood onwards. (54) of the survival of the linen corselet from the The most alluring temptation is the tendency fifth century ; and the suspicious coincidence to draw covert references to Augustus from of so convenient a discovery makes it probable the portrayal of the early heroes Romulus, that Augustus found no such evidence. Numa and Camillus. Such investigation is not Livy's reaction, as expressed in the text, is not wholly unrewarding provided that its limita­ without interest. He stresses emphatically tions are recognised. The impact of Roman that the annales maximi - "tam veteres patriotic feeling after the victory at Actium, annales" - and the libri lintei agree that Cossus when talk of a 'second founding' of Rome was was a military tribune at the time. He points widespread, may have communicated itself out that so important a battle could not have to Livy's early books. But such influence is taken place during the year traditionally unformulated and largely unintentioned. To assigned to Cossus' consulship. This reads argue that the description of Romulus, "deum suspiciously like a sceptical rejoinder to deo natum regem parentemque urbis Roma­ Augustus' claim; but finally Livy concedes nae ", (55) has application to Augustus is to

30 LIVY AND AUGUSTUS ignore Livy's vehement objection to the titles lie". (64) Against this view the argument must deus and rex when applied to a living Roman. (5 6) be pressed that support of an avowed monarchy The phrases which describe Camillus - is completely alien to Livy's political con­ "Romulus ac parens patriae, conditorque alter victions. 'Nova imperia' may refer primarily urbis", "diligentissimus religionum cultor"(57) to the administrative developments between - have clearly a closer relevance to Augustus. the fifth and first centuries - the transfer of But the ultimate objection to the 'hero­ the control of justice to the praetors, and the symbolism' theory resides in its selective prorogation of provincial and military com­ approach. Livy sets against his version of the mands. The phrase in aeternum urbe condita dramatic apotheosis of Romulus the "perob­ reflects Livy's essential Chauvinism; he be­ scura fama" that the king was torn limb from lieved that the city was providentially founded limb by attendant senators; (5 8) there are and also supernaturally aided in its growth. (65) likewise less creditable features in the portrayal But if any reference is intended here to the of Camillus. (59) Presumably Augustus himself Augustan scene, the settlement of 2 7, which would have hesitated to draw parallels too seemt,d to be based on traditional lines, may closely! have been in Livy's mind. Far too much has been made of such In sum, the evidence for Livy's having subjective investigation. Livy welcomed the aligned his early books to achieve the purpose pax Au9usta and the 'restoration of the of Augustan propaganda is too thin to merit Republic'. But he explicitly states that his serious consideration. studies are an anodyne, a refuge from the contemporary scene. (60 ) It is only occasionally Ill LIVY AND THE FALL OF THE REPUBLIC that the patriotism expels the pessimism, as in the speech of the tribune Canuleius, (61 ) A second main problem remains. How did probably written shortly after the settlement Livy record the delicate issues of first-century of 2 7. Canuleius seeks political concessions, history? By the time he reached the era of especially the consulship, for the plebs, and the Gracchi, disillusionment about Augustus' parries the conservative arguments. "quid political intentions must certainly have postea? nullane res nova ins ti tui de bet? et dawned.(66) The absence of Livy's own quod nondum est factum - multa enim account, and the tenuous evidence of the nondum sunt facta in novo populo - ea ne si Periochae and later epitomizers, makes a utilia quidem sunt fieri oportet?"(62 ) After precise assessment of his attitudes impossible. citing previous innovations, he patriotically The Periochae are too jejune, and the same can prophesies: "quis dubitat quin in aeternum be said of the most faithful of the compendious urbe condita, in immensum crescente, nova historians, the fourth-century Eutropius. Flo­ imperia, sacerdotia, iura genti um hominumque rus' Epitomae de Tito Livio Bellorum Omnium is instituantur?"(63) On this Syme comments: partially a misnomer; though Livy is often "Livy argues that, since the City is destined to drawn upon, other sources are attested, (67) endure for ever, and will grow all the time, and there are divergences from Livy in content new forms of authority, nova imperia, can be and arrangement. (68) Florus' background expected to emerge. That formulation suits of the Hadrianic principate, for which the avowed monarchy of Caesar's heir - it Augustus was the model, (69) may account for does not have to be assigned to the primacy an apologetic tone on behalf of Augustus of Caesar Augustus in the restored Repub- which is thus by no means necessarily

3r P. G. WALSH

attributable to Livy. Orosius also is over­ contributing to Livy's jaundiced view of the indulgent to the princeps. His Christocentric agrarian agitations of plebeian leaders in the history propounds that Augustus was divinely first decade. (79) But the chief criticisms were chosen to establish tranquillity and universal directed at his position in the state after peace in preparation for the Incarnation of Munda. Eutropius' version, "agere insolentius Our Lord. (7°) Cassius Dio employs Livy coepit et contra consuetudinem Romanae extensively for the period before Augustus' libertatis",(80 ) is echoed by Dio and to a principate, but again the possible use of other lesser degree by Orosius.(8 1) sources, and Dio's own political motivation - For Livy was, as Augustus commented, a reflecting both third-century insight into the Pompeian: "tantis laudibus tulit ut Pompeia­

principate and a maturer political sense than num eum Augustus appellaret".(8 2 ) Were Livy's - have transformed the Livian ver­ such laudes extended to Pompey's early days, sion. (71) On the other hand, Lucan's poem, so that Livy "idealised the early career of

for which Livy is virtually the sole source, (7 2 ) Pompeius, controverting Sallustius" ?(83) It is may express the Pompeian viewpoint with unlikely. True, the unprecedented triumph, more rhetorical force than did Livy. celebrated whilst Pompey was still an eques­ These qualifications cannot, however, ob­ trian, is remarked upon, (84) but so is the scure the central fact that Livy narrated the ineffectiveness of his operations against Ser­ crises of the late Republic with transparent torius. (8s) Whether Livy criticised Pom­ sympathy for the Optimate cause. He is very pey's extraordinary powers is problematic. critical of the Gracchi, both Tiberius and The Manilian law was passed "magna indigna­ Gaius.(73) The blame for the Social War is tione nobilitatis".(86) Dio's account of the not laid on the Senate, but on the methods and debate on Gabinius' proposal centres upon a motives of the tribune M. Livi us Drusus. (74) speech by the moderate senatorial Catulus, Such moral condemnation of individuals, and who criticises the extent of the powers to be the absence of consideration of political conferred, (87) but Eutropius and Orosius programmes and group attitudes, is typically pass no such judgment. Livy may have Livian, revealing one of the historian's marked implicitly criticised Pompey's dissimulation as limitations. The account of the Marius­ an immediate cause of the Civil War. (88) Sulla struggle justifies the Sullan party, with He certainly depicted as undignified and Marius held wholly responsible for the Civil ignominious Pompey's evacuation of Italy War,(75) So far as Marius is concerned, in 49.(89) Livy's attitude to Pompey may "haud facile sit dictu utrum bello melior an have reflected that of Cicero, who supported pace perniciosior fuerit". (7 6) Sulla's victory Pompey as the leader of the Optimates, the is correspondingly acclaimed, but Livy never opponent of revolution, the man who would hesitates to condemn encroachments on civil restore the Republic. (9°) The label Pompeian liberties or acts which militate against may in fact denote not so much a personal concordia, and the proscriptions of Sulla are adherence, but more generally support for

attacked as "unprecedented cruelty".(77) the senatorial government of the Republic.(9 1 ) In his account of 60-44 B.c., Livy was It has often been demonstrated that Pom­ highly critical of Julius. The first Triumvirate pey rather than Julius was Augustus' model ("conspiratio") was a consequence of Caesar's in the manipulation of political power. Syme design to usurp power.(78) The agrarian states that Augustus even "forswore the proposals in the Leges luliae may be a factor memory of Caesar", and after proscribing

32 LIVY AND AUGUSTUS

Republicans he "stole their heroes and Augustus' death, it is possible that Livy had vocabulary". (9 2) Vergil, Horace and Livy completed his history some years before his are cited as presenting a consistent front on death, and that he refrained from publication these matters on Augustus' behalf. The for fear of offending Augustus. (98) argument is overstrained. The sole passage The pentad cxvI-cxx, published in Augustus' from the Aeneid is a denunciation of the Civil lifetime, is thus of particular interest, covering War, in which Pompey - "gener adversis the years 45-43. The portrait of the young instructus Eois"(93) shares the obloquy with Octavian is markedly sympathetic. He arrives Caesar. Nor was the alleged 'party-line' in Rome from Brundisium to the accom­ sufficiently explicit to reach the keen ear of paniment of favourable prodigies. (99) In Ovid, or the notice of Varius.(94) But the face of the intemperance and malice of crowning consideration in all this is the fact Antony, he is active in defence of the that Republican sentiment was inherent in, state. (100 ) The Senate is criticised for and came spontaneously to, the three literary showing insufficient gratitude after the relief giants of Augustan Rome, whose genius had of Mutina, and this is the cause of Octavian's been refined in the death-throes of the making alliance with Antony and Lepidus.(101) Republic. When Livy wrote that it was His appointment as consul is marked by an uncertain whether the state had benefited augury recalling Rome's first foundation.(102) from the birth of J ulius(9s) - he was presuma­ No attempt, however, was made by Livy bly balancing military exploits against the to veil the barbarity of the proscriptions. eversio reipublicae - Augustus may have privately "C. Caesar pacem cum Antonio et Lepido fecit welcomed this interpretation, but it is unjust ita ... ut suos quisque inimicos proscriberent. to hint that he thrust it on Livy. In qua proscriptione plurimi equites Romani, cxxx senatorum nomina fuerunt".( 10 3) Dio does not seek to diminish Caesar's responsibil­ IV THE TRIUMVIRATE AND THE PRINCI­ ity in this, but reasonably suggests that the PATE others, especially Antony, took the ini­ The historian's serious embarrassments tiative. (1o4) Eutropius writes: "senatum pros­ began with his depiction of the second cripsit, cum Antonio ac Lepido rem publicam Triumvirate - in particular with its early armis tenere coepit. Per hos etiam Cicero years, in which the triumvirs't'ix oe: o~ 1tpixyµa't'1X. orator occisus est multique alii nobiles".(10s) 1tp6c; 't'e: 't'O ~OUA'Y)fLIX. XIX.L 1tpoc; 't'O em0uµ'Y)fLIX. 't'O It is doubtful if Florus' special pleading, or E:IX.U't'WV or'ijyov, &cr't'e: xpucrov 't'~V 't'OU Kalcrapoc; the unconscious irony of Orosius, can be µovapxlav (f)IX.V~VIX.L. (96) attributed to Livy. (106) The Periocha of CXXI begins: "qui editus In a long surviving fragment of cxx we can post excessum Augusti dicitur". The ortho­ read Livy's account of the death of Cicero and dox interpretation has suggested that Livy a brief judgment of the man, appended published cxxI-CXLII in the three years A.D. according to his usual method. It has been 14-17. But Syme's important analysis of the alleged that Augustus remained hostile to the biographical evidence suggests that Livy's memory of Cicero, and that Livy dutifully conventional dates as recorded by Jerome echoed the princeps' view. There is certainly may have been postdated by five years. (97) some outspoken criticism: "omnium adver­ If Livy died in A.D. r 2, and the final twenty­ sorum nihil ut viro dignum erat tulit praeter two books were not published until after mortem : quae vere aestimanti minus indigna

33 P. G. WALSH videri potuit quod a victore inimico nil The possibility is worth pondering, there­ crudelius passus erat quam quod eiusdem fore, that Livy's delineation of the events of fortunae compos ipse f ecisset". These two 44-3 -proved privately unpalatable to Augustus, criticisms - Cicero's tendency to luxuriate and that this influenced the historian's decision in his grief, and the immoderateness of his to postpone publication of cxx1-cxu1 until onslaughts on Antony (which certainly suggest after Augustus' death. This procedure enabled that Antony would have followed Catiline had him to recount without embarrassment Cae­ Cicero been sole arbiter) - are not wholly sar's humiliating defeats in the struggle with surprising as a judgment of a historian of Sextus Pompeius. (112) But undoubtedly - independent mind. (101) If, however, Livy had and predictably - in the account of the long said no more, the criticism of Carcopino(108) struggle with Antony Livy is firmly in support and Syme(109) - that Livy was retailing the of Octavian. Due emphasis is laid on the moral officially approved judgment on Cicero - delinquencies of Antony at Cleopatra's court; might seem justified. But Livy adds : "si quis his failure in the Parthian campaign of 36 tamen virtutibus vitia pensarit, vir magnus is attributed to his own folly, (11 3) and the acer memorabilis fuit, et in cuius laudes responsibility for the fresh bout of civil war persequendas Cicerone laudatore opus fuerit". is laid squarely on his shoulders. (114) This is generous praise; analysis of the whole Ten books were devoted to the years passage shows how misleading selective quo­ between Actium and the death of Drusus in tation can be. And again, Livy's pro­ 9 B.c. Why was this stopping-point chosen? Ciceronian sympathies are explicit in his If the suspicion is correct that these final account of the revenge exacted on the corpse books lay completed for years,(11s) Syme may of Cicero. "ita relatum caput ad Antonium well be right in his suggestion that 9 B.C. iussuque eius inter duas manus in Rostris was chosen because Livy felt unequal to positum, ubi ille consul, ubi saepe consularis, recounting the melancholy domestic scene ubi eo ipso anno adversus Antonium, quanta after 6 B.c., especially as it affected Tiberius, nunquam humana vox cum admiratione elo­ Augustus' designated successor. (I 16) quentiae auditus fuerat. Vix attollentes prae The most striking feature of the Periochae lacrimis oculos homines intueri trucidata of cxxxm-cxu1 is the absence of reference to membra eius poterant. " political history at Rome. It is not enough In the same book ( cxx) Livy narrated the to suggest that Livy was thus enabled to omit legislation by which Brutus and Cassius were topics of possible embarrassment to Augustus; condemned on Octavian's proposal in their had Livy been an enthusiastic supporter of absence, and also the operations of Brutus in Augustan autocracy, indications of his appro".al Greece. Undoubtedly Brutus and Cassius would have been manifest. The historian's remained 'under a cloud' in the eyes of deliberate silence reflects a lack of enthusiasm. Augustus, (110) but Livy did not accommodate The sole citation of detail which Syme makes his interpretation to suit the official attitude. in support of his thesis concerns Licinius There is independent testimony in Tacitus of Crassus. He claims that Crassus' Thracian his characterisation of these Republicans: campaigns were discussed by Livy under the ". . . hunc ipsum Cassium, hunc Brutum events of 2 7 not 2 9, in order that Crassus' nusquam latrones et parricidas, quae nunc claim to lay the spolia opima should appear vocabula imponuntur, saepe ut insignis viros after the settlement of January 2 7. This is to nominat".(111 ) visit on Livy the sins of the epitomator.

34 LIVY AND AUGUSTUS

cxxx1v contained the events of 2 9-2 7, and the the evidence of Livy's political views in the epitomator is no stickler for chronological extant books. These reflect a patently order.(! 1 7) Republican bias (which contains large histori­ On the other hand, Livy may have given cal dangers of its own), but there is no ungrudging praise to Augustus in his foreign evidence of Livy's espousal of Augustan campaigns, as he did to Julius. He drew on autocracy, or of propaganda in its interest. the princeps' memoirs for the Spanish campaign Tacitus could be perfectly sincere in praising of 26, and his uncritical attitude towards his not only his eloquence but also his honesty - sources was reflected in the prominence given "eloquentiae ac fidei praeclarus in primis". (120) to Augustus and in the virtual exclusion of P. G. WALSH the concomitant operations.(118) It may well be, however, that the decision of the princeps . University if Edinburgh to embark on the German campaign was criticised. Florus writes: "Germaniam quo­ (') See, e.g., H. Dessau in Festschrift 0. Hirsc/ifeld 1903, que utinam vincere tanti non putasset ! magis 46 1 f. For a recent balanced discussion, see T . F. Carney, turpiter amissa est quam gloriose adqui­ 'Formal Elements in Livy', PACA 2, 1959, 2 f. 2 sita". (119) Livy appended to his account of ( ) Especially the views of A. Rosenberg. (3) G. Stubler, Die Reli9iositiit des livius, 1941. the death of Drusus in 9 B.c. a mention of the (4) Christianity and Classical Culture, 1940, 98 f., 108. disaster of Varus in A.D. 9. He may have (5) p . 3 I 7. been suggesting that the grandiose plan of (6) p. 468. conquest had been ill-advised. But in general, (7) 'Livy and Augustus', HSCP 64, 1959, 71, 53. the achievements of Roman arms abroad will (8) Ibid., 75. (9) G. De Sanctis, Problemi di storia antica, 1932, 230. have been ungrudgingly and patriotically ( 10) P . G. Walsh, Livy, 1961, 151 f. praised. (") R.R. 463 f. Syme's thesis, that "Livy's annals of (12) Ap. Gell. v, 18, 9. Augustus were written in joyful acceptance ( 13) They openly preserved friendship with Ovid when of the new order, in praise of the government others feared Augustus' displeasure: Ovid, Pont . 1v 1 3: Trist. Ill 6, and its achievements" requires severe quali­ ( 14) Tac. Ann. 1 72, fication. In so far as Livy lauded the victories (15) Suet. Tib. 61,3 of Rome and the extension of the empire, he (16) Sen. Ad Marciam 26,1 was motivated by the larger patriotism. But ('7) Sen. Contr. x Praef. 5. 18 the complete exclusion of the political history ( ) HSCP 64, 1959, 72. ('9) Dio LVI 27, 1 of the reign reflects dissatisfaction as well (20) S. G. Owen, Ovid, Tristia ll, 1924, 45. as prudence. ( 21 ) R.R. ch. xxx. Many critics of Livy have proceeded from ( 22) Tac. Ann. 1v 34: Titus Livi us. . . Cn. Pompei um the a priori assumption that Livy's friendship tantis laudibus tulit ut Pompeianum eum Augustus with Augustus spelt collusion and manipulation appellaret; neque id amicitiae eorum offecit. Pompey's death was recorded in CXII: assuming the publication of of the historical tradition. This is a wrong a pentad a year from 27-25, one can hazard a date of basis, reflecting an exaggerated view of the 5-3 B.c. for om. princeps' interference with literature in his (23) Suet. Claud. 41, 1: historiam in adulescentia hortante organization of opinion. It results in a T. Livio, Sulpicio vero Flavio etiam adiuvante, scribere manipulation of subjective factors and a aggressus est. Claudius was born in Io B.C. (24) Aen. VI 791 f. selective treatment of objective evidence. A (25) E. Fraenkel, Horace, 1957, 240. sounder procedure is to base an approach on (26) Stubler, o.c., I.

35 P. G. WALSH

(27) Praej. 9-1 o. (SO) IV 2o, I I, ( 2 8) Dessau, o.c. (S 1) IV 3 2, 4 . (29) XXVII 19, S; cf. II 9, 7; VI 2o, Sf. (S 2) HSCP 64, 19.19, 47. (3°) II IS, 2-4. (53) See Syme's good comments, 48-9. (3 1)111, I. (54) H. Erkell, Au9ustus, Felicitas, Fortuna, 19s2, 19 f. (32) III 39, 8. (SS) 1 , 6, 3. Syme believes that this has reference to the (33) 11 9, 7-8 . fact that in 29 Augustus' name was added to the hymn of (34) 11 44, 8. the Sa/ii (R.G. 10). (35) IX 19, 17. (56) Camillus and Scipio Africanus are both criticised (36) I 23, 1. for arrogating to themselves a stature more than human: (37) 1 19, 3 has reference to Caesar Au9ustus, and hence v 2 3, 4: triumphusque omnem consuetum honorandi was written after January 2 7 : the first but not the second diei illius modum aliquantum excessit. maxime con­ closing of Janus is mentioned, which suggests a date in spectus ipse est, curru equis albis iuncto urbem invectus, or before 2 s. parumque id non civile modo sed humanum etiam visum. (38) Bude, Livy I, XVII f. So xxv1 , 9, 9: multa alia . . . admirationis humanae in (39) Syme assumes (a) that the pessimistic tone of the eo iuvene excesserant modum. On kingship, n. 29 above. Preface indicates early composition; (b) that 1v 20, 7 f. ( 57) V 49, 7; SO, I. (Augustus' famed information that Cornelius Cossus was (58) I 16, 3. consul when he laid the spolia opima) is a later insertion (59) Cf. n. 56 above. because it is inconsistent with the main account; (c) (60) Praej. S · that this information was fabricated at the end of 2 8 to (6 1) IV 3 f. The version of Dionysius (xi .17 ) suggests that foil the ambitions of the proconsul Licinius Crassus. Livy has composed, not merely reproduced, this speech. The first argument is dependent on the assumption that (62) IV 4, J . in Livy's eyes Augustus represents a panacea for moral (63) IV 4, 4 . ills : but see IV 6, , 2 ; x 9, 6 ; XXVI 2 2, 1 s (and I. ( 64) HSCP 64, '9S9, 47. Kajanto, ' Notes on Livy's Conception of History', (6S) I 4, I ; 11 40, I 3 ; Ill 7, I. Arctos, , 9s8, 62 f.) for a refutation of this dangerous (66) Tiberius Gracchus is discussed in Book LVIH, presumption. The second argument raises the question written about 16 B.C. of Livy's method of composition. His pages are crowded (67) 0. Rossbach, RE 6, 2, 2761 f. with inconsistencies through non-reconciliation of (68) C. H. Heyn, De Floro Historico, , 866, 49 f. divergent sources; footnotes are sometimes inserted (69) R. Syme, Tacitus, 19s8, 496. without amendment of the main account (Walsh, Li vy, (7°) See e,g., v1 20, 4: ... ut per omnia venturi 1 4 3 f. ). In short, inconsistencies need not be attributed Christi gratia praeparatum Caesaris imperium compro­ to later insertions. The third argument provides good betur. evidence for dating 1v, but a timelag after the Licinius (") See, e.g. XLVI 34. E. Schwartz, RE 3, 170s, after Crassus incident is obviously possible. A date of comparing correspondences between Dio and the Livian composition well into 27 would also be more in keeping tradition, adds: "durch diese Concordanzen ist allerdings with the description of Augustus 'templorum omnium die Annahme noch nicht ausgeschlossen, dass Dio die conditorem aut restitutorem' (1v 20, 7). livianische Erzahlung aus anderen Gewahrsmannern (40) Praef. , o. erganzt und verandert hat". Or, he continues, the (4') Praej. 9 . discrepancies may be attributable to "eine spontane, (42) HSCP 64, '9S9, 42 . von anderen Gewahrsmannern unabhangige Kritik Dios ". (43) Below, section m . (72) R. Pichon, Les sources de Lucain, , 9, 2. ( 44)1 19, 3. (7 3) According to Orosius (v 8, 3-4), Tiberius' motives (45) Bayet, o.c., XIX. Syme calls the reference ' fulsome' for legislation were revenge on the nobility and a desire (p. 48). for popularity. The laws of Caius were 'perniciosas' (46) Livy means that to remove Caesar as witness would (Per. LX; cf. Flor. 11 3) and Caius was 'magna reipublicae be the equivalent of robbing the shrine of the corselet pernicies' (Oros. v 12, 3). - which would be sacrile9i um, a sacred theft. Syme (74) Per. LXX: 'perniciosa spe largitionum plebem conci­ (p. 43) seems to take the word in a looser connotation. tavit'. Flor. 11 6, 3: 'cupidine dominationis'. Per . (47) 'Livius und Augustus', Hermes 41, 1906, 142 f. LXXI: 'invisus etiam senatui velut sociali s belli auctor' . (48) Dio, u 24, 4 . (75) Per. LXIX; Flor. 11 9, 6: 'initium et causa belli (49) 0 .c., 43 f. inexplebilis honorum Marii fames ... ' LIVY AND AUGUSTUS

(76) Per. LXXX. (IOI) Per. CXJX; Dio, XLVI 40, 1. (") Per . LXXXVIII: crudelitate, quanta in nullo hominum (102) Jul. Obs. 69; Dio, XLVI 46, 2 f. fuit. (103) Per. cxx. (78) Per. cm: eoque consulatus candidato et captante ( 104) XLVII 3 f. rempublicam invadere, conspiratio inter tres civitatis (10S) VII 2, 2. princi pes facta est. (106) Flor. 11 16, 6: haec scelera in Antonii Lepidique (79)1141, 3; .P, 2,etc. tabulis; Caesar percussoribus patris contentus fuit: (80) VI 25 . Oros. VI 1 8, 1o: itaque ne latius atque elfrenatius (81) Dio XLIIl 41, 3; Oros. v1 17, 1. incircumscripta caedes ageretur ( !), cxxx11 senatorum (82) L.c. n. 2 2 above. nomina, etc. (83) R.R., 464. ( 107) Syme believes that Livy was 'prone to benevolent (84) Per. LXXXIX. appraisals' (HSCP 64, 19 59, 70 ), but the evidence (8S) Per. xcn: parum prospere pugnavit; cf. xcm, xcv1. quoted from Quint. 11 5, 19 and x 1, 1o I relates solely (86) Per. c. to stylistic clarity. Perhaps Sen. Suas. 6, 21 was meant. (87) Dio, xxxv1 3 1 f. But Livy is not afraid to make criticisms when necessary, (88) Flor. 11 13, 16: ut daretur consulatus absenti, quern e.g. that of Marcellus at xxv11 27, 11. decem tribuni favente Pompeio nuper decreverant, ( 108) J. Carcopino, Cicero and the Secrets ef his Correspon­ dissimulante eodem negabatur. dence, 1951, 18. (89) Flor. 11 1 3, 20; Dio, xu 1 3, 4. (109) HSCP 64, 1959, 61. (9°) Att. 1x 4, 2; Fam . xvi 11, 3; Att. v119, 3; Att. vm 3,2. ( 110) R.R. 506; HSCP 64, 1959, 60. (91) R.R. 464 n. 2. ( 111 ) Ann. rv 34. (92) Ibid. 3 I 7. ( 112) Per. cxx1x : ex duabus classibus . . . altera quam (93) Aen. VI 831. Caesar duxerat deleta. Cf. Flor. n 1 8, 2. (94) Ovid's eulogy of Julius, Met. xv 745 f.; Varius' epic ( 113) Per. cxxx ' ... quia hiemare in Armenia nole bat, on Caesar's death, Macrobius, v1 1, 39 f. dum ad Cleopatram festinat'. (9s) Sen. N.(L v 1 8, 4. ( 114) Eutrop. v11 7, 1. (96) Dio XLVII I 5, 4. ( 11 S) L.c. n. 98 above. (97) HSCP 64, 19 59, 40 f. ( 116) HSCP 64, 1959, 70. (98) Mr. R. M. Ogilvie first drew my attention to this (1 17) Compare a similar chronological disorder at the possibility. end of Per. cxxxm. (99) Jul. Obs. 68. ( 118) HSCP 64, 1959, 65. (100) Per . cxvn: et sibi et reipublicae ... ; Antony's ( 11 9) Flor. 11 30, 1; cf. Oros. v1 21, 25-7. conduct, Jul. Obs. 68: 'monstrosa malignitate': Flor. ( 120) Ann. IV 34, 3. II 15,2.

TWO NOTES ON NUMIDIA

name. Its western boundary naturally co­ incided with the boundary between the The use of the name Numidia in the later sphere of activity of the proconsul and that imperial period to describe a part of the of the legatus Aug. as established after the proconsular province of Africa to be distin­ division of responsibility by Gaius; it ran guished from the area under the legatus Augusti west of Hippo Regius, Calama, Thubursicu pro praetore leg. III Augustae has been noted Numidarum and Madauros, but to the east on several occasions. ( 1 ) For convenience this of Thibilis. This boundary was probably not part of the proconsular province will be fixed till the second century, because there called Numidia proconsularis in this article, are indications that the proconsul continued though there is no ancient authority for the to act for some time in what was later the

37 B. H. WARMINGTON sphere of the legatus Aug., as in law no doubt which provide the evidence refer to the he was entitled to do. Q. Marcius Barea is spheres of activity of legati proconsulis, about found as proconsul at Cirta in 42/3 and C. whose number and responsibility there is

Paccius Silvanus in the same place in 77/8.(2 ) dispute . (8) In spite of the number of legati The location of these two instances is proconsulis known it is still not proved that the significant; the Confederation of Cirta with its proconsul really had three as Dio says, (9) substantial number of Italian immigrants was though there is nothing either to disprove it. the only part of North Africa west of the In the early principate the duties of the legati Bagradas valley in any sense urbanized and was proconsulis were largely of a formal nature and more suitable for the jurisdiction of the it seems clear that at this stage they had no proconsul than that of the legatus Aug. with fixed areas of responsibility. Later this his almost exclusively military preoccupations changed. Apart from one apparently isolated at this date. It is natural to relate the intestina reference to a legatus prov. Africae dioeceseos dissensio which was one of the reasons for Carthaginiens. proconsulis patris(10) of the time Claudius' choice of Galba to be proconsul of Hadrian, cases of legati proconsulis in charge extra sortem in 44/6 to disputes between legati of specific areas generally called dioecesis occur and proconsules. (3) The position was anomalous from the early third century. The use of the and must have remained so till the elevation word dioecesis to describe such an area in of the area of the legatus Aug. to a province Africa is surprising; one would expect tractus in the full sense by Severus. The Confederation or regio on the analogy of procuratorial of Cirta was certainly included in the sphere divisions in Africa (though there seems to be of the legatus Aug. by the middle of the second no connection between these procuratorial century; a legatus Aug. was patron of the districts and those under the legati proconsu­ Confederation in 140 / 1 and another carried lis).(11 ) In fact the term regio does occur as a out public works in 160/2,(4) and there are synonym of dioecesis. The problem is confused no later records of proconsuls acting there. by variable terminology, but two 'dioceses' On the other hand legati Aug. are found on are recorded in the following forms: ( 1) occasion acting well away from the military leg. Karthaginis, ( 12) leg. provinciae Ajricae zone, being particularly concerned with the dioecesis Carthaginiensium, (13) ( 2) legatus pro­ delimitation of tribal lands; (S) an analogous v inc. Africae dioeceseos Hipponiensis,(14) legatus case is the participation of a legatus Aug. in the provinc. (Afri)cae per Numidiam (Hippon)en­ commission sent by Vespasian to resurvey the siurn, (15) (legatus prov. Aj)ricae regionis Hip­ oldfossa regia.(6) Even after the creation of poniensis. ( 16 ) The diocese of Carthage needs the province of Numidia, the governors were no explanation; the reference to N umidia still generally referred to as legati Aug._pr. pr., Hipponensium shows that Hippo (Regius, not and continued to exercise their authority in Diarrhytos) was the centre of the dioecesis of the desert zone to the south of the proconsular Numidia, which was also sometimes called province and in the hinterland of the Tripo­ the dioecesis Hipponiensis. All these examples litanian cities .(7) are pre-Diocletianic and apart from Carthage Evidence for the existence of an area of the and Hippo they show that Avioccala and proconsular province also called Numidia Thibursicu Bure were in the diocese of dates almost exclusively from the third Carthage. ( 1 7) From the reign of Diocletian century, in other words after the formation come a number of inscriptions referring to an of a province of Numidia. The inscriptions obviously extensive programme of public TWO NOTES ON NUMIDIAS works carried out by the proconsul Ajricae towns Thubursicu Numidarum, Calama, and Aurelius Aristobulus and his legatus Macrinius Madauros, (2 7) but in others further east, though Sossianus between 2 90 and 2 94. The latter is still part of Numidia till Caesar's conquest, referred to as leg. N., clearly an abbreviation namely Bulla Regia, Lares and Mustis. ( 2 8) for leg. Numidiae, at Madauros,(1 8) and as . The appearance of the name N umidia in the legatus (without a geographical designation) third century to describe a portion of the at Calama,(19) Mididi,(20) Thugga Tereben. proconsular province is an example of the thina, (21 ) and an unidentified town at Sidi strength of popular memory and usage in Achmed el Hacheni. (22) The last three places North Africa, like that of the revival of the are in the territory shortly afterwards detached more obscure Byzacium for one of Diocletian's from the proconsular province to make the new provinces. The fact is that in this area new province of Byzacium, but it is by no there were substantial pockets of un•Ro• means impossible that this relatively unur• manized elements surviving till late times, not banized district was earlier in the diocese to mention fragments of the tribe of the of Numidia. Numidae itself. In the literary sources the After the division of Africa Proconsularis un•Romanized elements are mentioned by by Diocletian, the part which continued to Apuleius for the region of Madauros in the bear this name was still divided into dioceses; third century(29) and by Augustine for Hippo there can be no doubt that, whatever Regius in the fifth. (3°) Besides, in geographi• uncertainty exists as to whether there were cal usage, the Numidians came to mean those three or two at an earlier date, there were peoples living north of the desert between the now only two. The terminology became river Ampsaga in the west and Tusca in the more standard in the fourth century, the two east, thus including the Numidian province of officials now being generally referred to as Severus as well as Numidia proconsularis. (3 1) legatus almae Karthaginis and legatus Numidiae. Another curious point is that the boundaries The examples(23) of the former are all in the of the ecclesiastical provinces of N umidia eastern part of the province, in fact east of the and Africa did not coincide with the secular old fossa regia; the only example of a legatus of ones. Quite apart from the fact that as a the Carthaginian diocese active west of this general rule such boundaries did coincide, line is from Thibursicu Bure.(24) It is just we find that in Africa itself elsewhere the possible that there was some reason for this Diocletianic division of provinces was repro· in the early history of the province. Though duced in ecclesiatical provinces. The Chris• west of the fossa, Thibursicu Bure was in an tian churches of Calama, Hippo Regius, area of Marian colonization which included Thagaste, Thubursicu Numidarum, all in the such places as Thugga, Thignica, Uchi Maius secular 'diocese' of Numidia in Africa pro· and Vaga and which always had close connec• consularis, were in the ecclesiastical province tions with Carthage. (25) Such an area may ofNumidia, formed between 256 and 305.(32 ) well have preserved some characteristics One need not assent to all that has been which distinguished it from the area to the written of N umidian particularism to agree west which remained part of the N umidian that this looks like a case in which the churches kingdom till the death of Juba and was not in this area were in tune with popular usage settled till much later. (26) In fact we find and felt more closely tied to those to the west that the legati Numidiae of the fourth century and south west in N umidia proper than to were active not only in the obviously Numidian those nearer Carthage.

39 B. H. WARMINGTON

II the building of a centenarium which he took over from his predecessor in Tri­ The history of the divisions of the African politania(39) seems to be part of a general provinces at the end of the third and beginning plan of strengthening the African limes of the fourth centuries is still obscure. Since which is well attested under Diocle­ the criticism by A. H. M. Jones(33) of the tian. (40) views of W. Seston, (34) which seemed in (c) Whatever may be said about the divisions general to complicate the matter unnecessarily, of Numidia, it is generally agreed that it no fresh evidence has appeared. Attempted was reunited again shortly before 3 20. reconstructions of Diocletian's scheme, espe­ Iallius Antiochus was vir peifectissimus cially where the division of Numidia into praeses of an undivided province in 3 1 5 or

Numidia Cirtensis and Numidia Militiana is later,(4 1 ) and the first consularis, Domitius concerned, remain insecurely based and will Zenophilus, was governor in 320.(4 2 ) not be attempted here. Three points in Zenophilus also had the additional style of connection with the general picture may be seifaxcalis, (43) which often appears sub­ made, however. sequently on inscriptions as an additional (a) The earliest indication of Byzacium as a title of the consularis Numidiae. (44) No province is during the first Tetrarchy. (35) doubt the elevation in status was A faint indication that its formation may connected with the removal of the provin­ be shortly after 294 is to be found in the cial administration of the undivided prov­ extensive activity of the procos. Africae ince from Lambaesis, where it still was Aurelius Aristobulus in the northern part at the time of Diocletian, (45) to Cirta. of the province including a number of This city, which had been sacked by the cities subsequently included in Byza­ troops of Maxentius when they captured cium. (3 6) Aristobulus' four year term of it from the African usurper Domitius office (290-294) is unusual at this date Alexander, was rebuilt by Constantius and and may be regarded as a special charge renamed Constantina. (4 6) A consequence in preparation for the division of the of the removal from Lambaesis will have province. been the separation of civil and military (b) It is difficult to see how the creation of power in Numidia; the latter had still Byzacium could have failed, for purely been retained by the praeses Numidiae as practical reasons, to lead to the simul­ late as 303,(47) as it continued to be, taneous creation of a province of Tripoli­ exceptionally, in Tripolitania and Maure­ tania. The arguments of Jones against tania Caesariensis. (48) In view of the Seston's proposition that Tripolitania was difficulties caused by the brief appearance at one stage associated with Numidia of Numidia Cirtensis and its epigraphical Militiana give weight to the view that the abbreviation N.C., it may be observed province existed under the Tetrarchy. In that during the fourth century the whole further support it may be argued that in province was frequently called Numidia view of the relative importance of the two Constantina, (49) with the same abbrevi­ provinces, Aurelius Quintianus, praeses ation, there being no question of the name Numidiae in November 303,(37) will have applying only to a restricted area like the held the governorship of Tripolitania(38) earlier Numidia Cirtensis. The title was at an earlier stage in his career. Again, still in use as late as 3 8 3/ 392, but on the TWO NOTES ON NUMIDIAS

whole, in view of the number of occasions (19) ILA19. 1, I 79. on which it was omitted from formal (20) Cll vm, 608. 21 inscriptions, was probably a local usage ( ) CIL vm, 11768. ( 22) C/L VIII, 27816. rather than the official designation, and (23) CIL vm, 1277, 1120s, 23849. recalled the favour shown to the city (24) ILA. soB. and the province by Constantine. (25) T . R . S. Broughton, Romanization ef Africa Procon­ sularis, 1929, 32 f., sB f. B. H. WARMINGTON ( 26) Broughton, op. cit. 79 f. Universi9' ef Bristol (27) ILAJ9. 1, 1286, 179, 21 02. ( 28) C/L vm, 2H21, 1782, /LT •H7• 2 ( 1) See especially S. Gsell, inscriptions latines de l' Al9erie, ( 9) Apol. 24. r, 192 2, p. x-xu, A. Albertini, Bulletin de l' Academie (30) Epp. cvm, S, 14; ccrx. 1 d'Hippone, 193O-193s, 27 f., B. E. Thomasson, Die (3 ) See RE, XIV, 2, 2166 f. 32 Statthalter der romischen Provinzen Nord,!frikas von Au9ustus ( ) A. Audollent, in Diet. Hist . Geo9. Eccl . 1, 1912, bis Diocletianus, 1, 1960, 60 f. cols 848 f. (33)JRS 19S4, 20, 21, 27. (2)ILA19. 1, HO, HI. 44, (3) Suetonius, Golba, 7; P. Romanelli, Storia delle (34) Diocletien et la Tetrarchie, I, 1946, 3 26 f. 35 Province romane dell ' Africa, 19s8, 26s-6. ( ) AE 1908, 197. 36 (4) CJL VIII, 703 6, ILAJ9. 11, 631. ( ) Supra, notes 1 8-2 2; see also C/L vm, 624, 464s, 1 (5) ILA/9. I, 2829. 1774, 23413, 2365'7, 236s8, AE 1933, 60; 1946, 119. 6 (37) / LS 644. ( ) en vIII, 23os 4 , 2s967, AE 19 12, 14 s-sr . 38 (7) J. M. Reynolds and J. B. Ward Perkins, Inscriptions ( ) /LS 9JS2. 39 ef Roman Tripolitania, 19p, 8. ( ) /LS 9H2 with AE 1929, 4. 40 (8) See Thomasson, op. cit. 1, p. sS-8 2. ( ) Other centenaria C/L VIII 2O21s, AE 1942/3, 84; (9) Dio Cassius, un, 14, 7. cp. Zosim. n, 34. (41 ) vm, 700s. ( 10) C/L XIV, H99- CIL 42 (11) Contra, Thomasson, op. cit., 78. ( ) Optatus, Append. 1, in CSEL vol. xxv1, 1 8 s f. 2 (43) AE 191s, 30. {1 ) C/L vm, 23831; R. Cagnat, Inscriptions latines 44 d'Afrique, 1923, so8. ( ) E.g. CIL VIII, 17896, 18328, 7034. 45 ( 13) C/L 11, 1 262. ( ) Cll VIII, 2729. (46) Aurel. Viet. Caes. 40, 2 8. ('4) CIL rx, 1s9 2. 47 {15) AE 1933, IH, ( ) AE 1942/3, 84; see also C/L vm, 2S72. ( 16) C/L x, p 78 . (48) Notit. Di9. 0cc. xxrx, xxxx. 49 ( 17) See location of inscriptions in note 1 2. ( ) E.g. ILAl9. n S96, 619; C/L VIII, 8324, 18701, ( 1 8) IL A/9. 1, 2048. 797S, 7979, 201s8; AE 1933, 1s9.

41 REVIEWS

J. Wilson and C. Parsons, A BASIC LATIN J. R. HawthornandC. MacDonald:ROMAN VOCABULARY, Macmillan, 1960. Pp. iv POLITICS 80-44 B.C., Macmillan, 1960. + 59; ls. Pp. X + 259; 9/6d.

It is surely a sign of the times that vocabularies, In this book the classical masters of Bradford usually based on a word-count of the foremost College and Sherbourne School selected a period Latin authors and specifically directed at G.C.E., of Republican history and divided it into a series are appearing in ever increasing numbers. Gone, of crucial topics. Each topic has a brief intro­ apparently, are the days when a Classical educa­ duction followed by extracts from authors - tion was assimilated like the air; but then, gone Cicero, Sallust, Caesar, Asconius, Suetonius, too is the breathing space of those more spacious Gellius, and the Commentariolum Petitionis - as a and leisurely days. basis for study (approx. 92 pages), along with a Following hard on the heels of the same commentary (approx. 100 pages). The book also Company's Latin Word List of K. C. Masterman contains two maps and three plans, a stemma, - a most valuable publication - comes "A Basic indexes of proper names, laws, political and legal Latin Vocabulary". Like all such books, it is a institutions, social and political terms, and somewhat flimsy publication and one wonders if topography. In assessing this publication its such an economy might be a false one. The book primary object must be kept in view, i.e. the claims to contain a collection of a thousand of the reading in secondary schools of good prose common Latin words. Most of the words indeed authors combined with the first-hand study of seem essential but it is surprising to note the history as based on some relevant primary presence of such words as arguo, blandior, daps, material. Additional clarification must be sought manes, rota and ruber while such words as condicio, - or provided by the master - from the basic f,nitimus, inimicus, iniuria, regio and tribunus are literature referred to, e.g. Syme's Roman Revo­ absent. Misprints (such as e.g. 'pugnae, -ae f.' of lution, Scullard's Roman Politics, Badian's Foreign p. 38) are few. English derivations are given where Clientelae, Sherwin-White (JRS 46, 1956 and possible. Greece and Rome 4, 1957), Baisden (JSR 29, 1939 A useful feature of the book is the grouping and 47, 1957), and the relevant chapters by Cary of Latin words which have a common root and and Last in CAH IX. The bibliography is not connected meanings. Helpful too is the table of exhaustive but confined to works "which are constructions at the end of the book. This table likely to be accessible". The central theme is the contains the common constructions with their working of the Roman government in the years uses, examples and notes. The table is followed following the dictatorship of Sulla, studied 'in by the Sequence of Tenses and a short list of action' and for the greater part through the verbs which take a Prolative Infinitive. mouths of eye witnesses. This is an effective All in all, this is a useful publication and could approach and one which will produce insight be of particular value for revision prior to the and a feeling for historical reality such as hand­ G.C.E. examination or to examinations of books alone cannot achieve. similar standing. In the first three chapters, 'The Tribunate', The Equestrians and the Lawcourts', 'Pompey D. G. MOORE and the Tribunate', Pompey's role as a popularis Moseley High School, Birmingham at this stage of his career is clearly brought out.

42 REVIEWS

The restoration of the juries and of the tribunicia obstruction (p. 195), and constant reference to potestas go together (p. 15 f., Cicero, in Verrem). the maps of Rome and the forum. The value of The next chapter, 'Optimates and Populares', Asconius' commentary is revealed e.g. on the fulfils the essential function of presenting the issue underlying the trial of Milo and on Cicero's major factions in their true perspective - the excellence as an advocate on the technical side inner ring of consulares viri and the opposition (p. 93). This is the outstanding chapter in the which sought to break down their exclusiveness. book. The prominent feature of the chapter on It sheds valuable light on Cicero's thought, as a the civil war (ch. IX) is the ambivalence of Pom­ supporter of the 'true' populares in the sixties pey and Cicero's disillusionment: nihil actum est (p. 35 f., de Leg .Agr.; contrast populares isti, p. 65, a Pompeio nostro sapienter, nihil fortiter (p. 121 in 59 B.C.), and a follower of the ideal optimates f.; cf. p. 125. Contrast Pompei us, nostri amores, in the fifties. In view of the central theme of the p. 65 - letter to Atticus in 59 B.C.). The crisis, book the selection from the Commentariolum and attempts to reach a compromise, are clearly Petitionis is happily included in this chapter for set forth (p. 109 f.; correlate 114 f.). It is good to what it reveals of amicitiae and the raw materials study Cicero's letter of Dec. 50 B.C. (p. 115) to of political power and influence. In the next appreciate the righteous indignation at Caesar's chapter (ch.V) the legal aspects of the execution demands and the irony of his Gallic command - a of the Catilinarian conspirators are clearly dealt truer reflection of how men felt than the speech with, and the extracts selected illustrate Cicero's de Prov. Cons. (p. 103 f.). point of view and justification for the extreme It is oot clearly stated which were the powers measures taken (p. 58 f., in Cat. IV). The con­ of the tribunes Pompey is likely to have restored, spiracy is appropriately presented in its context, or the effect of this, since the intercessio was i.e. in the aftermath of the Sultan regime (cf. the never removed (Caesar makes much of this, p. tertium genus, p. 52 f., in Cat. II). A long chapter, 120) and the disqualification from holding further 'Force and Fraud in Politics', is devoted to the office had already been abolished by the lex years 58-52 B.C. The theme is the struggle for Aurelia. Admittedly we can only surmise (cf. the preservation of the Caesarian legislation of CAH IX, p. 292 f.). Similarly the brief note on 59 B.C. (cf. p. 106 and 211), which gives back­ patrician sanction for laws (p. 139) does not ground and unity to this period; it accounts for satisfy. The inclusion of chapters 136-7 of the the sustained opposition to Caesar throughout pro Sestio (ch. IV) - the plea for admission to the the decade and the intransigence which produced highest order ab universo populo on the basis of the civil war in the end (cf. p. 115, letter to industry and merit - or a reference to them in Atticus, Dec. 50 B.C.), and which Caesar felt so the commentary would have underlined the bitterly (p. 118). Our attention is duly drawn unity of Cicero's outlook as a novus homo and to the 'new generation of optimates' and 'ex­ basically a popularis, linking up with his ideal of tremists' (p. 71 and 188). Caesar fares badly; cum dignitate otium as based on aconcordia ordinum he is virtually responsible for the new era in (p. 156, in conjunction with p. 167). Attention bribery and physical violence (p. 70 and 179). A should perhaps be drawn more specifically to fine note on Messius' proposal explains the Cicero's antipathy to Sulla, and the reasons for constitutional implications of Pompey's commands it - to the illorum temporum dolor (p. 53) - but (p. 188 f.). This period comes alive with the aid particularly to the constitutional aspects of the of information supplied on public assemblies new kind of dictatorship as Cicero would have (p. 197), their meeting places (p. 200), procedure seen it (cf. p. 127). The generalisation on spee­ and places of voting (p. 182 and 201), methods of ches (p. 166) is apt to mislead. It reflects unfavour-

43 REVIEWS ably on Thucydides, about whom the last word H. Lloyd Jones, MENANDRI DYSCOLUS, in this respect has not yet been spoken; and the Oxford: Clarendon Press, 1960. Pp. 84 historicity of Tacitus' version of Claudius' speech including INDEX NOMINUM and INDEX in the light of the Tabula Lugdunensis must be VERBORUM; 1Ss. borne in mind: "II correspond ... a sa tenue essentiellement historique" (P. Fabia, La table The new Regius Professor deserves our c/audienne de Lyon, 1929, 203). In the chapter on gratitude for producing an inexpensive text of the First Triumvirate (ch. VI) not enough is the Dyskolos quickly. The more general aspects said of the 'constitutional' background to of the Dysko/os have been discussed recently in Pompey's dispute with the senate over his acta this journal (PACA 2, 1959, 27) so that this (p. 171); in the light of its·prerogatives the senate review is confined to the text. After the manner had a strong case for its attitude. Similarly the of the Oxford Classical Texts introduction and request of the tax-farmers was outrageous - and critical notes are in Latin. The author adds Cicero thought so too (Att. II, 1, 8). We need dramatis personae, scenic notes, references to more on the progressive deterioration of public book fragments, etc., index of names, index of order because of the fact that Rome did not have words. On the whole the work is excellently a police force (cf. p. 24, Ascon., in Corn.); the done and this is the best text yet available. The reader may gain the impression that it began with only general criticism to be made is that in a text Caesar (p. 70). There seems to be an omission which will be set for school and university in the text in regard to the circumstances which examinations obelised words and gaps might gave Caesar his Gallic command (the note on p. have been further avoided where a reasonable 205 is not clear as it stands); and perhaps the emendation or supplement could have been constitutional procedure in the allocation of suggested, e.g. 43, 44, 48, 89-96, 114, 142-3, 201, consular provinces can be given more fully 247, 406, 496, 550, 596-9, 647, 817, 836-41. (p. 208, 30). Lastly, an introductory note in the I add some notes on detail in the hope that commentary on Asconius would not have been some points may be reconsidered in a second out of place. edition. The Latin Dramatis Personae defeats me : Criticisms and suggestions are made primarily how can CHAEREA, SICO, PARTHENIS, turba in view of the teaching function of the book, and iuvenum ebriorum belong to the fami/ia Callip­ in the light of the many other technical matters pidis? 1. 5: ad sinistram Cnemonis (ut vidit which are dealt with in detail and with superb Quincey) domum videmus. I should take this to clarity. Among the commendable features of the mean 'on the audience's left' (which was Martin's work I would include notes such as those on correct interpretation of Pan's EitL ae~l()() but lawyers (p. 143), cum dignitate otium (p. 156), Quincey says 'the audience's right' because he the s.c. ultimum (p. 164), topography (p. 172), believes that Getas calls 'Right wheel!' in 909 salva republica (p. 175 and 233), edicta (p. 176), (the solution to 909 may be that Getas is looking CCCC (p. 182), the imperium maius (p. 189 f.), towards the house). Does Lloyd Jones mean 'we and the lex annalis (p. 212). see Knemon's house on Pan's left'? 35: I hope This book may with profit be prescribed in the better second thought in the footnote will university undergraduate courses as well as for prevail over the emendation. 126: Quincey's secondary schools. It should not however take t'.moauvwµ evo ~ might have been quoted. 134: the place of the study of individual authors. Exit Chaerea. This must be right, and the para­ C. P. T. NAU DE graphos shows that the speaker changes at the University of South Africa, Pretoria. end of the line. ·144: An scaenam linquit Byrrhia?

44 REVIEWS

I think he does 'not remain on the stage as far as Knemon may have been brought out on a possible from Knemon' but retires into the wheeled couch, but dcrxuxA e:L-r' is not the only Nymphaion (cf. Rylands Bulletin, 42 1960, 507). possibility in 758 as Quincey has seen. 727: 149: 'He shouts Ares' is out of key for Sostratos. Szemerenyi's 81te: p av (X/\/\Oc; x&aLXCX LCuO"' is the The traces of letters are compatible with Martin's best solution, if a verb is felt desirable. 736-7 : cxAM ycx p (resuming after a digression). 173: I prefer to punctuate: 'You have wits by the gift ).e:w cp 6pov (Fraenkel, Post, Dale independently) of heaven. You care for your sister as is natural.' is at least Greek. Here an obelus in the text 746: I am not convinced that oux fits the traces would be justified. 201: Dale's &vumxc; y' better than

45 REVIEWS alteration of substance might fittingly have might have been taken of a reissue to imp ove marked this interval, the turn of the half-century the Notes occur at pages 51 er.L

Greek is much freer and (ii) less predictable than should be altered to read pNNpqNCppNpC). that of the majority of European languages and In short an excellent book and the best guide (iii) th<'t all patterns of order which are describa­ on Greek word order to put into the hands of ble in syntactical terms are secondary phenomena. students and teachers ( ?). In a very thorough and basically sound chapter I found the second chapter especially con­ on the nature of the problem, D. declares that vincing, where D., as no one before him , as far this problem should never be approached from as I know, has succeeded in showing the tendency the angle of only one determinant, but that of certain words to behave according to definite various determinants should be considered. D. patterns. (Think of the postpositives (symbol q, then discusses the most important determinants i.e. i.a. ixpct, ycx p, &.v, µe:, µo u, µm) and the governing word order in Greek and finds three: prepositives (symbol p, i.a. &tJ.ix, ~. xd , µ~ (lest), i) the tendency of certain specified words to relatives and o (the)) in their relation to M ( = take a constant position (cf. esp. c. II (lexical and mobile words, apt to be found anywhere in a semantic determinants)), Greek clause). A few patterns discovered by D., ii} certain types of logical relation between the but which are already implicit in the work of sentence and its context (= c. IV, cf. also p. 4 Wackernagel, are of more than passing impor­ for a definition of logical determinant), and tance: historically seen, the tendency of Mqq .•. iii) the tendency to ad here to familiar patterns. M > MqMq ... , with the second last sentence of In a separate chapter the syntactical deter­ p. 14 ("in particular, q which 'belong' in parti­ minants (= c. Ill) are discussed and via i.a. cipial and infinitive clauses are commonly associ­ statistical data D. concludes: "It is clear that ated with the words to which those clauses are these statistics are very far from establishing for subordinate"), is of the utmost importance. 'Classical Greek' simpliciter anything worth Equally important is the tendency of p and M to calling a syntactical rule of word order" (p. 31); coalesce (with the q apt to follow, thus giving and later: "all patterns of order which are pMq). This complex (pM) tends to be disrupted describable in syntactical terms are secondary by an intruding q (pMq > pqM) - the q of the phenomena"; "In Greek the primary logical connecting particles being especially fond of principles 'weighted the scales' in favour of an disrupting this complex (cf. esp. p 16. on the increasing dominance by syntactical patterns of various likes and deslikes of different authors). order"; and "We find in fact that in the language This tendency (pMq > pqM) is however often of the New Testament rules of order are much halted because "q are not necessarily placed more easily defined in syntactical terms than they after the leading p or M of what would traditio­ are in Classical Greek" (p. 68). nally be defined as a 'clause', but may occupy a The greatest merit of D.'s work, as has already similar position within one of the word-groups been stated, is that he considers various deter­ which constitute the clause" (cf. p. 17 with good minants, and he has produced a work which examples). Equally convincing and important though brief - perhaps too brief - is always is - I choose quite at random - D's remark that clearly written, always well arranged, always " chiasmus is not necessarily a literary embellish­ fundamentally - whether you agree with his ment" (p. 54) ; his very thorough and sound principles or not - well-thought-out. D. knows judgement on the reliability and dangers of mere his bibliographical material and is in full control statistical data (p. 60); and the contentious - but of everything important done by his predecessors. to my mind correct - first 12 lines on p. 67. Moreover, the book is almostfreeofmisprintsand The task of the critic is not only to flatter but inaccuracies (the last two symbols on p. 46 also to evaluate and to criticise. Without

47 REVIEWS

detracting in any way from my very high esteem Ships, metre and poetical style, apart from the for D.'s work, I would, though hesitatingly, formulaic character, certainly were also impor­ suggest the following: tant determinant factors. D. in his evaluation has to my mind, not taken this sufficiently into i. consideration. And is this not also true of I believe that D. in his search for patterns has Aristophanes' Ra. 1434 (p. 19)? overemphasized the role played by the other two determinants (esp. the lexical and the logical) to the detriment of the role played by syntacti­ iii cal factors right from the era of the 1.-E. age. In his chapter on logical determinants D. by D. has succeeded in proving the existence of creating new terms (Nucleus and Concomitants patterns - also patterns which viewed histori­ (= N and C, cf. his definition p. 40)) gained cally could be termed as the product of a certain much; amongst other things he hereby eliminated 'chance' (p. 59) - but he has failed to explain to a large extent the subjective factor. The latter convincingly why all patterns of order describa­ of course is always present when applying ble in syntactical terms are necessarily secondary 'emphasis' as a criterion. He indeed gained phenomena. The relationship Subject followed much. I am also convinced that a certain paral­ by Verb (and not V-S) approximates to the lelism does exist between the behaviour of q character of a 'rule' - dangerous as this dictum and C - D's analyses on pp. 41-6 being particu­ may be - to a greater extent than D. is inclined larly convincing. I am , however, inclined to to admit - and this is confirmed by D's own and think that D. in his enthusiasm has gone much also by other statistics. Although i.a. on p. 65 too far - sections ii and iii as a whole being not reasons are given for the preponderance of the so convincing. To mention but one example: is S-V pattern (cf. also p. 26 f. on the behaviour of Herod. 1,1,1 (symbols N (=Mb)C(=Mb)C(=Mq) preferential words (symbol Ma) vs ordinary C(=Mq)pC(=Mb)) really "clearly" (see p. 47) words (= Mb), one feels that more attention modelled upon M : MqqM? And is the break should have been paid to the possibility that, after the leading M (symbol M : or N : ), a break after all , pure syntactical reasons were in this necessary for D's pattern, really free of all case of greater import than any other possible subjectivity (cf. e.g. in this connection the last determinant. two breaks in Herod. 11,26,2 (see p. 44))? The preferential treatment of concomitants ii (see p. 49 f. )is according to D. mainly determi­ D's motivation for his choice of material (esp. ned by antithesis (explicit or implicit). But can from Herodotus and documentary inscriptions) antithesis really always be discovered on mere is hardly satisfactory. What he says esp. of formal or logical lines without involving emphasis Herodotus, on the authority of Pohlenz (see as a major criterion, and thus bringing in sub­ p. 11), is a questionable truth. Do we indeed get jectivity through the back door? (In this I see a more ' natural' Greek from the older Greek a real danger - and I say this in spite of the fact prose? Or is this also a fact that can never be that I personally on the whole agree with his proved without the risk of circularity? D. analyses - those on p. 52, as well as the argument prefers prose to poetry. But the important there, being less convincing). And the sentence conclusions on p. 64-65 re the conflict pattern­ on p. 51 ("the preferential treatment of words principle is to a large extent based on Homer. which have some emotional force") confirms my And even though it be from the Catalogue of fear of at least a certain amount of subjectivity. REVIEWS

iv. In conclusion: in many respects this is a unique work. The defects are mainly on the more formal Perhaps the fact that the work originated from level, and do not basically affect the lasting value lectures might be responsible for traces such as the following which underline defects on a more of the book. It is to be warmly recommended. formal level: on p. 13 among reasons given why E. L. DE KOCK cre:, crou , croL could not unconditionally be classed University of Pretoria. as q, D. mentions, quite correctly, i.a. "the danger of circularity inherent in the concept 'emphasis"'. But this argument can only be fully grasped by the reader when he has also read D's exposition N. E. Collinge, THE STRUCTURE OF of 'emphasis', which comes on p. 32 f. Here HORACE'S ODES, London 1961, Oxford should perhaps also be classed i.a. 'n words' University Press (University of Durham

(p. 3), M1q + M2 > M3q (p. 15), the symbols : , I, 11 Publications). Pp. ix + 1S8; 2Ss. - which are not explained explicitly (as is done This book aims at tracing the design of Horace's in the case of pq 1Mq 2 (cf. p. 16)) though their meaning can be gathered from the context. Odes, and at finding sense and beauty in that Would this work, which is already rich in symbols, design. Since Horace bel ieved in the equal not have gained by using separate symbols for importance of ors and ingenium, and since he was 'complex N' and 'complex C', by which the unremitting in his zeal for lyric perfection, there global pattern would undoubtedly have gained is some point in submitting his words, his ideas in clarity (cf. esp. p. 42 and 47)? Personally I and his whole poems to detailed analysis. In shou Id also have preferred to have the com parable recent years Fraenkel, Wilkinson, Campbell and statistics on p. 28, 29 SY vs VS etc. placed directly others have thrown light on many of the Odes, beside each other thus saving the reader the and Mr Collinge both acknowledges his debt to trouble of waiting for the conclusions at the them and criticises some of their interpretations. bottom of the page. One would only wish that, in outlining his aims in his Preface, he did not overstate his case (p. viii): 'The Odes ... are very largely exercises v. in form, to the exclusion of content, or at least I know no better survey to put in the hand of to the exclusion of fresh invention.' In reality, students, but from the student's point of view apart from rare instances like iv 7, a palpable I miss two things: first, a section or short chapter re-casting of i 4, each poem contains its own in which the long history of the problem is thoughts, its own inventio as orators would have briefly stated and the different points of view are said, appropriate to the addressee or the topic. mentioned, thus giving the student more scope The mere fact that many of the ideas of a par­ to appreciate and evaluate D's own merits. t icular ode may be found elsewhere does not Secondly, a concluding chapter offering and mean that the ode is lacking in content. analysing a few Greek passages, thus discussing There are four chapters, an analytical appendix the various problems and the arbitrariness of the to Chapter Ill, and indexes of names and Horatian word order with reference to the role of all passages. Chapter I is entitled 'Words and three (or more) of the determinants (instead of images : the mechanistic approach.' On iv only one of them at a time). This would have 5, 17-24, a 'long series of strictly parallel asyndetic enabled the student to see what a real expert, sentences, each of them filling a line' (Fraenkel), as D. is, can make of it. Collinge remarks that Horace has alternated

49 REVIEWS plain statement in the odd verses (tutus bos, etc.) the melic poets. Sometimes these blocks of with a goddess and divine personifications in thought are 'responsive', e.g., the second may the even, so that not only Faustitas but Mos, Lex, answer the first, and the third tie first and second Nefas, Culpam and Poena should be given capitals. together. In Greek choral verse such responsion With Wilkinson (C.Q. (N.S.) 9, 1959, 181-192) the is tied up with strophe, antistrophe and epode; author takes the iunctura of the Ars Poetica as in Horace it is not necessarily tied up with 'collocation', J. R. Firth's term for the company metrical divisions. Moreover the same ode may a word keeps. By comparing the vocabulary of have responsive and non-responsive deployment iii 1, domos fastidit, aequora, a/tum, molibus, of thought; for exam pie, in i 9 (vi des ut a/ta) caementa, fastidiosus, moliar, with other passages the first three stanzas show a closely linked in the Odes, one can see how Horace builds up triad of ideas, whereas the last three have only such iuncturae and gives them a personal signifi­ an informal, progressive train of thought, the cance. two halves being 'welded by the associative, Chapter II is headed 'Contrast-technique I: indeed ambiguous phrase donec virenti canities the order of the odes.' It is held that, although abest,' Four main types of the responsive mode Horace mostly avoids setting side by side odes in of writing are distinguished: strophic, e.g., the same metre, we have no positive metrical i 19 (mater saeva Cupidinum); patterned, e.g., guide other than the variety of the 'Parade' odes iii 1 (odi profanum vulgus), whose pattern is (i 1-9) and the alternation of sapphics and alcaics roughly x/ /a/ bb/ cd/ dd/ /cc/ cd; symmetrical, e.g., in ii 1-11. One may only wonder whether it is iv 1 (intermissa, Venus, diu), with 8 + 12 + 12 a coincidence that a sapphic ode appears second + 8 lines; interwoven, e.g., i 14 (o navis), which in each book except Ill, where it appears second intertwines 'the perils to which the allegorical after the Roman Odes; that (among others) i 10 ship is exposed' with 'the resources the ship has and 12, 20 and 22, 30 and 32; iii 8, 11, 14 and 18, for withstanding them or inspiring confidence' 20, 22 are sapphic. More interesting is the (Tracy). Odes consisting of two non-responsive question of linkage of motif between adjacent sections are subdivided by Collinge into four odes. The first three Roman Odes are closely categories, according to whether the thought of linked; Fontaine and Fraenkel have recognised a each part is static or progressive. For example, structural nucleus in iv 7, 8 and 9, while ii 4 and the first half of ii 4, 'that wicked little piece of 5, 6 and 7, 13 and 14, iii 22 and 23, iv 2 and 3, banter ne sit ancil/ae tibi amor pudori', is static, 10 and 11 have obvious affinities. In considering whereas in the second Xanthias is subjected to the relationship between these and certain other a neat process of deflation (Postgate). Eight odes adjacent odes, the writer stresses the necessity are said to have responsive and non-responsive to apply the criterion of atmosphere, depending sections; six of them have these in that order, two on the reader's mental associations and frame of in the reverse. Where two responsive sections mind, and concludes that there is a little more are combined, one would expect a treatment to complexity in the arrangement of the odes than be possible analogous to that where two non­ is generally allowed. responsive sections are combined. But since The third chapter, 'Contrast-technique II: only three such examples emerge, each with its thought-structure within the odes', is the longest own complications, it is perhaps true that no and most important. Many of the poems have further subdivision is profitable; the three are blocks of thought within them, a feature which the astrological ode ii 17, iii 3 (iustum et tenacem) Horace inherited from Greek choral lyric, where­ and iii 6 (de/icta maiorum). as his method and mental atmosphere come from Chapter IV is an analysis, somewhat in Fraenkel's REVIEWS

style but at greater length, ofii 7 (o saepe mecum), put it, 'dum ... scandet soll das emphatisch voran­ in which Horace welcomes back his old comrade gestellte usque nicht einschranken, sondern Pompeius who fought with him at Philippi. The veranschaulichen: wie Rom ewig ist, so des preceding poem, to Septimius, also has mecum Dichters Ruhm.' (5) The analysis of i 12 on pp. in the first line and a form of amicus at the very 104-5 starts by approving Christ's rather arti­ end; it too is addressed to a comrade in stress ficial division into 5 x 3 stanzas, but this is not and danger; both have a theme of war-weariness. the writer's last thought on the matter, for he But there is a difference; Septimius is young and later ('to be honest') changes to 3, 3, 2; 3, 1, 3. active, Pompeius is Horace's age and under a The printing is very good; one might wish, strain. Again, whereas i 36' (et ture et fidibus) however, that where a sentence begins with a has many similarities with this, Numida there reference to a new ode ( e.g., 'i. 36 shows .. .' on takes the initiative; on the other hand Pompeius p. 84) there were a space or a dash after the here is shown up as a diffident man. Next, the preceding full-stop, or else some system of black ode is contrasted with ii 3, in which the turncoat type, to warn the reader that a new ode is being Q. Dellius is put in his proper place. In contrast, analysed. Pompeius had clearly been under a cloud for In conclusion, this book offers much stimulating many years since Philippi and was hesitant about reading, and goes a fair way towards justifying returning. Structurally the ode is compared with the claim that categorizing, however provisional, i 4 (so/vitur acris hiems): it is combined of a has its rewards. The fact that it is provisional responsive and a non-responsive mode, the latter will be made abundantly clear by contrasting being strongly progressive. In geometrical terms Collinge's rather bald treatment of ii 1 on p. 119 we have in each a circle followed byastraightline; with Campbell's fuller treatment and somewhat but is it not an exaggeration to call them 'perhaps different analysis in Horace, pp. 230 f. An exami­ the most ambitious erections of Horace among all nation of Book IV shows that Horace did not, [his] architectural experiments'? with time, materially change his structural A few smal I criticisms: (1) the interpretation methods, which certainly represent an important of dare c/assibus Austros as 'mutua casuum permu­ contribution to the lyric ode. tatio' (pp. 29-30) shows too naive an attitude to 0. A. W. DILKE syntax; see C. J. Fordyce in CR 54, 1940, 96; Rhodes University, Grahamstown. while as to the suspected 'double hypallage' of iii 30, 13 f. (p. 32 n.), carmen was indeed a Latin song in origin, but no one will suppose that it E. C. Kennedy, CAESAR: DE BELLO GAL­ refers to any other than a Greek song in several LICO I, C.U.P., 1960. Pp. 200 with notes and passages. (2) On p. 56 we are told that within one vocabulary; 6/-. ode there is no variety in the elevation of the language; what about iv 12, whether it is One must agree with Mr. Kennedy that addressed to Virgil, as Collinge believes, or not? "changing conditions in schools and especially (3) On p. 61, the statement 'the sapphic stanza ... in Latin ... seem to indicate that there may be a is basically not a unit of sense' should be qualified: place for another edition for boys and girls .. .'' the thought-breaks given by the writer for The summaries at the head of each ch;,pter sapphic odes always fall at metrical divisions, e.g., should prove most helpful; the notes are ade­ 3 + 3 + 1 + 3 + 3 stanzas for i 2. (4) To quate and, with a few exceptions, clear and to interpret iii 30, 1-14 as a 'run-down' is to the point. The introduction, however, follows degrade it to the level of ii 4; as Kiessling-Heinze the pattern of Peskett and Shuckburgh (1879 and REVIEWS

1901 respectively - or should one say respect­ Historic Present: " ... can be followed U) by fully?): Early Life of Caesar, The Gauls, The Roman either primary or historic subjunctives" cf. Army, Pay and Equipment and The Roman Camp. Draeger, Hist. Synt. der lat. Sprache, 1878-81, 352: The introduction is good as far as it goes when the subordinate clause precedes the main (some phrases such as " ... at the age of 40, he clause Cicero and Caesar regularly use an found that he had a genius for war" are vaguely Historic Subjunctive in the subordinate clause. irritating), but an extra chapter on the content When the main clause precedes the subordinate of this particular book would have been very clause there is no fixed rule; cf. vii 45, 1 and 3. welcome. The incidental information scattered The note on i 8, 2 (p. 96) is, for the same throughout the Notes is excellent, but should reason, vague to the point of being misleading. rather have been collected under one heading There must be some other reason for the mixture with the purpose of leading pupils to find out of primary and historic subjunctives. whether "on the whole ... he (Caesar) wrote an I wish here to refer to i 31 and i 44. In the unbiased and accurate account of his campaigns" 0.0. the tenses originally used by a speaker are (Intro. p. 16). More use should be made of the often retained. This is known as repraesentatio. Commentaries as a primary source for the study Instead of using Imperfect and Pluperfect Sub­ of the historical background and of Caesar junctives only after historic main verbs, the himself as a politician, general and man of letters. Present and Perfect Subjunctives are often used Discussions based on quotations such as i 10, if a Present or Perfect Indicative has been used in 5; 12, 5-7; 31, 12 and 35,4 will throw some light the 0.R. for the sake of "vividness". In i 31 on whether Caesar was trying to justify an act there are eleven regular Imperfect and Pluperfect of aggression and whether he was guilty of Subjunctives after the historic main verb. Then treason according to Sulla's Lex de Maiestate. If follows a Perfect Subjunctive followed in turn by the text is handled properly Caesar can be yet another eleven Imperfect/Pluperfect Sub­ presented as a person who did have ambitions, junctives. From "Ariovistus autem ... " we find who was an opportunist and who did have to four Perfect and ten Present Subjunctives. justify his actions in an age of propaganda. An In i 8, 2 conarentur is Prospective and at the edition for boys and girls should stress these same time indicates an action prior to that of points. possit (i.e. it is not subjunctive because it is in The period is a fascinating one. Kennedy quite the 0.0.). And it is quite possible that conentur rightly discusses aspects of the Roman army, but (i 8, 3) could have been subjunctive in the 0.R. the emphasis should be on the importance of the as well. army during the time of Caesar. It is significant i 13, 3; 44, 13 et al.: statements such as: "The that soldiers took the oath of personal loyalty Pluperfect Subjunctives represent the Future to the general himself, with as result the politi­ Perfect of direct speech" could be very confusing cisation of the army's high command and army to pupils if no further explanation is given, and involvement in metropolitan politics over the especially when they come across: question of demobilisation, (cf. R. E. Smith, (a) i 27, 3 (p. 121): " ... so that perfugerunt Service in the Post-Marian Roman Army, 1958, 38-43). becomes perfugissent". It should be sufficient to Pupils should find the Notes extremely useful. point out that this is the verb of a dependent There is a (welcome) tendency to describe rather relative clause and that the Pluperfect Subjunc­ than to prescribe, but occasionally the rule-of­ tive here indicates a prior action. thumb-method of explaining the useofsubjunctive (b) i 31, 15: "The Perfect Subjunctive represents tenses is unsatisfactory, e.g. i 3,2 (p. 89): the the Future Perfect of direct speech."

.P REVIEWS

The names given to the various tense-forms a gain in clearness by restricting the inflected often do not conform to their functions. These tense to its function as a mere praeterite and by names are very often meaningless when seen in using a combination of habeo with the P.P.P. to connection with their functions, especially in express the meaning of the Greek Perfect. si-clauses. i 34, 2 is a case in point: si quid opus Coegi came to mean / brought together, and esset ... venturum fuisse. The O.R. is: 'si quid coactum habeo, I have brought together. There are opus esset ... venissem' (the Pluperfect Sub­ twenty-three examples of this construction junctive of an unfulfilled past conditional in direct in the two Commentaries. One wonders speech). Venissem means much more than 'I whether such sentences are not lapses into the should/would have come': there is no indication use of the spoken language of that time. from the context that Ariovistus had been The note on i 42, 5 (p. 147), 'si quid opus facto unable to come. It implies that he would have esset', is interesting. A reliable grammar des­ come (but no longer intends to). Nor does the cribes opus est with the Ablative of the P.P.P. as Pluperfect here refer to an impossible condition an alternative for opus est plus Present Infinitive in the past, but refers to some future action as (cf. Cic. Fin. 5,16: nihil opus est hoc facere seen from a point of time in the past. The rules for longius), which brings us to the realm of Obli­ sequence of tenses, although useful as a general gation. guide, cannot be considered as binding in every J. ENGELBRECHT, construction. Paul Roos-Gimnasium, Stellenbosch. The Subjunctive was very often used instead of an Indicative (i 23, 3, p. 116) when something different was to be expressed from what would W. J. Bullick and J. A. Harrison, have been conveyed by an Indicative. Quod ... GREEK VOCABULARY AND IDIOM FOR existimarent need not be "illogical" ('the Helvetii HIGHER FORMS, Macmillan, 1960. Pp. probably thought .. .'). This could also be an viii 105; Ss. example of attraction into the subjunctive, + something which occurs in the earliest Latin The book is divided into three parts, dealing (cf. Handford, The Latin Subjunctive, 1946, 148-51, respectively with the vocabulary of Greek drama, on the weakened force of the Subjunctive). vocabulary for prose composition, and idio­ A construction which is of some interest is the matic usages. It is intended as a means to the use of habeo plus the P.P.P.: exercitum quern rapid acquirement by the student of an adequate coactu m habebat (i 15, 1 ). The reference to the basic vocabulary for reading Greek drama Pluperfect (p. 104) is not clear and can be rather (especially Euripides and Sophocles) and a more confusing. Other editors who comment on this than basic vocabulary for composition. The construction merely state that continuous action compilers suggest that the student make himself is indicated (and that it is much more than a familiar with the two vocabularies by working periphrastic form of the Perfect Indicative). This through them constantly, and the student who usage is, in fact, the result of the ambiguity of does so will certainly find himself with an im­ the Perfect tense. The Latin Perfect is in origin pressive vocabulary at his disposal. a confusion of two prehistoric tenses, corre­ The compilers have deliberately arranged the sponding respectively to the Greek Perfect and vocabularies in such a way that they will not Aorist. The popular (Vulgar) Latin (which had serve as a substitute for a dictionary: the first already been established by this time and from section (drama - Greek-English), although it is which the Romance languages descend) achieved alphabetical in arrangement, groups words of

53 REVIEWS similar meaning under the initial letter of the ship both wide and exact, a great clarity in first word of the group, without cross-reference exposition and the shrewdest of common-sense. to the other words of the group under their The work is designed to meet the needs both of initial letters; e.g. s.v. &daw are included the professional scholars and of undergraduates. It verbs µiA1tw and uµviw and the nouns ?l~ and aims to fulfill a long-felt need by replacing the µiAoi;, but there is no entry of these words in Victorian editions of Thompson and Lodge, both theµ-, u-, or w- sections of the vocabulary. In the now out of print and out of date. It admirably second section (prose - English-Greek), though fulfills its purpose! there is some cross-referencing between words of Dodds discovered that the standard Greek text similar meaning, it is by no means complete; of the Gorgias provided no adequate basis for a e.g. "maiden" is in the list, but not "girl"; modern edition and furthermore that the Oxford "inasmuch as" is there, but not "because" (indeed text of Burnet was in many respects quite false there appears to be no mention of wi; as contras­ due to some misunderstanding between its ted with &-n:). author and Kral, the collator of F. Accordingly, The same principle operates in the third he undertook a new examination of the evidence, section: the idiomatic usages are classified collating fully for the first time two important according to the form the Greek is to take, not MSS, F, the sole representative of the second the form in which the English phrase presents major family, and W of the first. He also takes itself, so that the English phrases have a rather into account the indirect tradition, which is miscellaneous appearance. occasionally our only authority for a good reading, In the hands of a diligent and intelligent student and is the first editor to employ all four Gorgias the book should prove extremely useful. It is papyri. As a result of these researches the compact, clearly set out in large type, and in­ Platonic scholar has now at his disposal a text of cludes a number of blank pages for the student's the Gorgias which is far superior to any previous additions. If a criticism must be made, it is that one. It must be noted, however, that very few the decision to specify the genders of only some of the textual problems discussed by Dodds nouns (mostly neuters ending in -oi; and -µ

55

.,, REVIEWS sophist are unconvincing. Contra Dodds, Cal­ details Plato supplies in the case of Callicles would licles does not say, in Gorgias' presence, that have little point if he were a fictitious character. sophists are "worthless people" (520a1), but only Dodds suggests that the reason why the latter that those who reach ocpETIJ are - and these left no mark on the history of his time may well Gorgias himself ridiculed. Nor does Meno 95c have been because his vigorous ambition and his support Dodds' claim. The context here surely frank avowal of it brought him to an untimely implies that Meno did regard Gorgias as a.sophist, end in the desperate years at the close of the albeit with the peculiar characteristic that he Peloponnesian War. When Socrates is made to did not claim to teach ocpETIJ (-rwv rx.).J.wv at 95c3 say to Callicles at 519a7 crou oe: fowc; em),~~ov-rixi, is inclusive and means "the rest of the sophists"). eixv µ~ Eut.ix~?i, Plato, Dodds points out, may well If Gorgias is not a sophist, he is so like them as to be putting into his mouth a prophesy post be indistinguishable from them! eventum. Dodds sees in the sympathetic portrait In his discussion of the relationship between of Callicles something which Plato had it in him­ Gorgias and Empedocles Dodds might also have self to become (and perhaps would actually have examined Diogenes Laertius' statement (viii 57) become had it not been for Socrates). The that Aristotle in his lost work, the Sophist, Calliclean way of life Dodds well describes as described Empedocles as the inventor of rhetoric. the poisonous fruit which grew from the seed This role is traditionally ascribed to the Sicilians, of Gorgias' teaching. It is for this reason that the Tisias and Corax, but the value of the information dialogue is more aptly called Gorgias than as evidence should not be disregarded on this Callie/es. account. It seems quite possible that Aristotle's The Commentary is extremely full (it occupies claim may well have been a good deal less specific nearly 200 pages) and is a positive mine of infor­ than Diogenes' comment implies (cf. Sext. Emp. mation. In his note on 508a3 Dodds, discussing Adv. math. vii 6 and Quint. Inst. or. iii 1, 8 the usage of the term x6crµoc;, is surely right in his DK31A19). (In the same passage Aristotle is argument against the views of Kranz and Kirk. held to have described Zeno as the "inventor of The former holds that this term was used in the dialectic". This comment, too, contains a certain sense of "world" by the Milesians in the sixth amount of truth and may also have been less century, ignoring the doxographical tradition specific than Diogenes implies. For Zeno's (Philo/. 93, 1938/9, 430 f. and in his monograph method has one important element in common Kosmos, 1955); whereas Kirk believes that it with dialectic as conceived by Plato and Aristotle, was first used for "world-order" late in the namely, the practice of refuting an opposing fifth-century and in the sense of "world" later thesis by deducing intolerable consequences still (Heraclitus, the Cosmic Fragments, 1954, 311 from it.) f.). Dodds himself believes that the truth lies The section on Callicles is the most exciting of somewhere between these extremes. Kirk, we all. The only information we have about him is may note, in a later consideration admits that what Plato tells us in the Gorgias itself. Dodds the term has the sense of "world" in Empedocles' argues persuasively both against the view that Fragment 134 1,5 (Kirk & Raven, The Presocratic he is a "mask" for some historical figure and that Philosophers, 1957, 159). Parmenides' Fragment 4 he is a fictitious character. He points out that 1,3 surely confirms Dodds' standpoint. x6crµoc; there are no clear instances of fictitious characters must have the meaning of "world" here. The with personal names introduced as speakers in text could hardly mean "scattered in order". conversation with Socrates or of real persons (More recently Kahn, strangely ignoring the introduced under fictitious names. In any case the evidence of Aetius (ii I, 1 DK14A21) altogether, REVIEWS has argued, like Kranz, for a Milesian origin of the or new, is marked by its author's habitual clarity term (Anaximander and the Origins of Greek and practical commonsense. He wants to know Cosmology, 1960, Appendix I p. 219).) how the machinery actually worked. Ancient and The work has been most carefully brought to modern blueprints are ruthlessly scrutinised and press. The slips we detect are but few. On p. 13, rejected if they appear unworkable. The mastery n. 3 read Jaeger, The Theology of the Early Greek of evidence is enviable. Philosophers, 186; p. 101, 1.1 cx.u-r4'> & 1.4 ecr-r~v; Inevitably Augustus is central to many of the p. 243, last line 'personal'; p. 303, 1.33 '485e3' studies. Jones argues for a full consulare imperium and in the note on 456b6 'Phil' instead of 'Ant'. for Rome and Italy and three tenures of censoria potestas, despite the apparent denial in Res J. R. LONG RIGG Gestae, 6. Dio may be right and he certainly Victoria University, Wellington. deserves so able a champion. Many, however, will remain unconvinced. Augustus was surely supremely opportunistic and limited himself to A. H. M. Jones: STUDIES IN ROMAN the practical minimum. From 19 B.C. he enjoyed GOVERNMENT AND LAW, Macmillan, some of the consuls' prerogatives, but how much of their power beyond sharing responsibility for 1960. Pp. 243; 30s. public order, as his twelve fasces proclaimed? Professor Jones' new book consists of seven In 8 B.C. and A.D. 14 he needed special grants of published articles, a radical revision of )RS 26, imperium consulare like Claudius later (Dio, Ix, 1936, 223-35 (on the dediticii) and two fresh 23, 4) - not simply for the census, but in order to contributions - 'The Censorial Powers of raise levies in Rome (note Tac. Ann., i 31). E.T. ' Augustus' and 'Procurators and Prefects in the Salmon ably expounded this view in Hist. 6, 1956, Early Empire' (pp. 19-26 and 115-25). We must 471 f. Jones might well have answered him in a surely be grateful for having this impressive body special note and similarly replied to other critics of work gathered conveniently between two of his articles. He did recognise the need to covers. Here are Jones' mature, considered views allow for new evidence and make minor changes, on the most vital issues of early imperial history. but this is hardly enough. Since he still clearly A common theme runs through the work - the stands by what he wrote, we want to know how organic development of institutions from Republic he disposes of later objections. He connects Dio's to Empire - and thus it may not safely be ignored censoria potestas very closely with the three by those whose interests are bounded by the senatorial /ectiones (which he dates 29, 19 and 11 age of Cicero. B.C.) and indeed finds no other evidence for its In the second of his new studies Jones concludes use. But can we be sure that Augustus' senatum that praefectus was the regular title for all ter legi applies to the strange affair of 19 B.C. equestrian governors until Claudius suddenly (ipsorum arbitratu: Suet. Aug. 35)? Dio moreover, and tactlessly introduced procurator, a term of links his /ectio of 11 B.C. with a census. Perhaps private law under the Republic. Only Egypt's he has misplaced both from 8 B.C., where he viceroy retained the old style, but then his whole unaccountably omits the census known from position, with its proconsular affinities, was quite Res Gestae, 8 (consu/are cum imperio!). exceptional. A study of apparent minutiae leads Jones shows conclusively that cons~lar and typically straight to one of the arcana imperii, praetorian elections were genuinely fought though Jones is content here with lifting only a under Augustus until in A.D. 5 he created a corn~r of the veil. The work indeed, whether old senatorial/equestrian electoral body, which voted

57 REVIEWS as a reinforced praerogativa. The new system virtually unquestioned coercitio against lesser gave the equestrian elite effective control of folk. But were _such cives legally degraded to these elections through their superior numbers quasi-peregrine status, as Jones argues (p. 61 f.)? in the college and they could be trusted to It is possible that appeal survived in some form, favour the kind of novi homines that the Emperor but was rendered inoperative normally by fraud needed; but it became a mere formality in A.D. or such devices as Verres discovered long before 14 when Tiberius transferred the real decision (see Jones, p. 54). to the Senate. This is a bold attempt to reconcile Dominium in provincial land, Mommsen held, the tabula Hebana with Tac. Ann. i 15. Perhaps, was vested in the Roman people. Jones surely however, we should recognise Tacitus' trouble­ dealt the coup de groce to this theory in his 1941 some comitia as the elections for the lower article ()RS, 31), here reprinted. Provincials were offices, which henceforth (save briefly under not then in strict law all dediticii. Having cleared Gaius) were the sole concern of the Senate. We the ground Jones proceeds to a substantially new could then treat the tabula Hebana as seriously attack on the Constitutio Antoniniana. The crucial as Jones treats the /ex Valeria. The Emperor clause did not exclude existing dediticii from the continued to get his way with a show of real citizenship itself, though the status was main­ /ibertas. For how long did Augustus' creation tained after A.D. 212 for certain slaves and survive? It is not an academic question on this barbarian settlers. If the Aegyptii were still view, as it is for Jones (p. 46). dediticii, I think that we must accept this anoma­ Augustus' civil and criminal jurisdiction is lous view, since Caracalla gave them civitas. derived by Jones from his tribunicia potestas May not Severus' municipalisation of Egypt, and the lex Julia de vi publica, which extended however, have changed their status to that of exercitio iudicii publici at least to proconsuls. attributi? Jones would deny this (p. 136 f.), but Since this law is surely due to Caesar (Cic. Phil. how decisive is his evidence? Whether we are to i 9, 21-3) not Augustus, I would prefer to construe xwpl~ -rwv OEOE:L'rLX.LWV with olowµL or recognise the lex lulia de iudiciis publicis as µevov-ro~ (as Jones prefers), Caracalla probably the regulating statute (11 B.C. ?) and assume excluded all dediticii from his grant and this sense that the consuls were definitely included, not can surely somehow be obtained by restoration. brought in by loose interpretation (see Jones According to Jones there was no one central p. 97 f.). Otherwise Jones' view is cogent and imperial treasury called the fiscus in the early accounts well both for the emergence of the Empire. The word designated either any one of Senate as a High Court and for the fact that the many provincial chests or, more specifically, proconsuls sometimes executed capital sentences the fiscus patrimonii (as in Sen. De Ben., vii 6,3). on cives without allowing appeal. Imperial In the Flavian period the word acquired its governors may have subsequently acquired the modern meaning, but even later such money as same power (pace Jones; p. 56 f. and 91 f.); Tac. the Emperor held in Rome was stored in a Ann. xii 60 points this way, while the offence of number of fisci (ludaicus, Alexandrinus, Asiaticus Florus and Galba (Jos. Bell. lud. ii 308: Suet. etc.); the Fiscus was the central organisation. Golba 9) was surely the imposition of forms of Severus organised the patrimonium and the new, punishment from which cives were exempt, sinister res privata in a single central office and rather than refusal of appeal. This right suffered blurred the distinction between the Emperor's a melancholy decline in the third century. It personal funds and the State monies which he remained a reality perhaps only for honestiores handled. All this seems as cogent as Jones' after A.D. 212, since governors exercised insistence that the Julio-Claudians drew regularly REVIEWS on the Aerarium for funds, which explains their times entertaining account of the lower Roman interest in its management. Less certainly he bureaucracy from the late Republic to Justinian. claims that Vespasian diverted the main revenues It calls for less comment, but repays careful of the Aerarium into the fiscus Alexandrinus and attention because of Jones' strong sense of how the fiscus Asiaticus. These may have held only things really get done. The Republican governors Vespasian's new taxes, exactly as the fiscus were not so short of staff as the text-books ludaicus, not the whole provincial revenue. suggest, while one cause of the Byzantine Empire's Hence Jones' picture of a depleted, insignificant resilience lay in the character of its bureaucracy, Aerarium is open to serious criticism. He is also superior to that of the West. Throughout the perhaps too positive in calling Augustus' rationes book we have this blend of close argument and imperii "a general balance-sheet of the Empire" firm generalisation. We are forced to think and (p. 105). The breviarium totius imperii (Suet. Aug. rethink and if we disagree it is with an uneasy 101) was perhaps confined to the public monies feeling that the last word may yet lie with its handled by the Princeps: "quantum pecuniae in author. Here we can judge the full worth of his aerario (sc. the aerarium mi/itare) et fiscis et contribution to early imperial history and salute vectigalium residuis". This ratio is perhaps an achievement that will not be outdated for a contrasted with the patrimonium in De Ben. very long time. vii 6, 3. HAROLD B. MATTINGLY Jones' last study is his penetrating and some- University of Notting ham

59 Printed in The Netherlands by Royal VanGorcum Ltd., Assen THE PROCEEDINGS OF THE AFRICAN CLASSICAL ASSOCIATIONS, a journal for original contributions in any aspect of Greek or Roman studies, will accept such contributions primarily, but not exclusively, from scholars within Africa.

Material submitted for publication must be type-written, double spaced, and have ample margins. Contributors of articles are entitled to receive 2 5 copies of their respective contributions free; reviewers receive Io copies of their reviews.

Proceedings of the African Classical Associations is published annually in August or thereabouts. Articles for publication must reach the Managing Editor not later than April 30th. Advertising copy should be submitted to him by June 1st.

The yearly subscription rate and/ or the price for individual volumes is 1 6 shillings (numbers of volumes available are limited). Individual off-prints will be sold at cost prices. Cheques etc. should be made payable to: The Treasurer, Classical Association of Rhodesia and Nyasaland, Private Bag 167 H, Salisbury, S. Rhodesia.

Articles intended for publication, books for review, subscriptions, remittances and other editorial communications should be addressed to the Managing Editor: Professor T. F. Carney, University College of Rhodesia and Nyasaland, P. Bag 167 H, Salisbury, S. Rhodesia. CAMBRIDGE BOOKS OF INTEREST TO CLASSICISTS

TEXTS: Plato's Meno Edited by R. S. BL U CK Late 1961. About 6os. net Propertius: Elegies, Book I Edited by W. A. CAMPS I2S. 6d. net The Per sae of Aeschylus Edited by H. D. BROADHEAD 45s. net Towards a Text of Cicero Ad Atticum By D.R. SHACKLETON BAILEY 27s. 6d. net GENERAL: Livy: His Historical Aims and Methods Edited by P. G. WALSH 40s. net The Political Thought of Sallust By D. C. EARL Late 1961. About 35s. net Greek Word Order By K. J. DOVER 15s. net Plato's Later Epistemology By W. G. RUNCIMAN Late 1961. About 30s. net Documents of the Principates of the Flavian Emperors, A.D. 68,96 By M. McCRUM and A.G. WOODHEAD 27s. 6d. net The Woodwork of Greek Roofs By A. TREVOR HODGE 501. net Obtainable from all bookseller, U.K. prices shown

CAMBRIDGE UNIVERSITY PRESS BENTLEY HOU SB, 200 EUSTON ROAD, LONDON, N.W. I