Herodotus and the hellenic culture

À CAMBRIDGE UNIVERSITY PRESS . Herodotus and Hellenistie Culture Author(s): Oswyn Murray Source: The Classical Quarterly, Vol. 22, No. 2 (Nov., 1972), pp. 200-213 Dit | | Published by: Cambridge University Press on behalf of The Classical Association Stable URL: http://www.jstor.org/stable/638201 Accessed: 25-07-2015 11:24 UTC Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at birp://www jstor.ore/page/ a info/about/policies/terms jsp a JSTOR is a not-for-profit service that helps scholars. researchers, and students discover, use, and build upon a wide range df content: in a trusted digital archive. We use information technology and tools to inerease productivity and facilitate new forms of scholarship: For more information about JSTOR, please contact supporté jstor.org. ' Cambridge University Press and The Classical Association are collaboraung with JSTOR to digitize. preserve and extend access to The Classical Quarterly. btp ystor.ore

Herodotus and Hellenistic Culture Author(s): Oswyn Murray Source: The Classical Quarterly, Vol. 22, No. 2 (Nov., 1972), pp. 200-213 Published by: Cambridge University Press on behalf of The Classical Association Stable URL: http://www.jstor.org/stable/638201 Accessed: 25-07-2015 11:24 UTC Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Cambridge University Press and The Classical Association are collaborating with JSTOR to digitize, preserve and extend access to The Classical Quarterly. http://www.jstor.org

OUR understanding of the world is not static; it can both expand and contract, and it can also stagnate. In history the expansion of the known universe has come about from various causes, from scientific advance, the slow processes of trade and exploration, from colonization, and especially from conquest. Periods of expansion produce often a re-evaluation of the external world, both that which was already known and that which was previously unknown, or on the fringes of the known. But no one is wholly capable of a direct response to reality: reality as soon as it is experienced is perceived, organized: 'Die Welt ist die Gesamtheit der Tatsachen nicht der Dinge' (the world is the totality of facts, not of things).2 To the understanding of experience, new as well as old, everyone comes with the preconceptionsand prejudicesofhis own environment, and seeks to explain the unknown in relation to it, in termsof marvels,opposites, or contrasts ('The Egyptians in most of their manners and customs exactly reverse the ordinary practices of mankind') :3 in terms of theoretical preconceptions of the noble savage, or the ideal state, or some notion of the structure behind all societies: or in terms of what has been established by previous observersas the proper way of dealing with the unknown, a processwhich often involves the re-evaluation of those previous observers.And to these periods of expansion there correspond periods when the vision of the world stagnates, when it is viewed in terms of a set of stereotypeswhich were once part of a fresh and intelligent response to the new situation, but are now the stale cliches of a generation too traditional to think for itself. These statementsmay be illustrated from many periods of history. The great Arab expansion was followed in the ninth to eleventh centuries by a period in which Arab writers tried to grapple with the new horizonsof knowledge.4They produced a literature based in part on their own observations (the importance of travel was emphasized), on official records, and on written or oral accounts of the travels of others by land and sea. The old world of the Koran was no longer adequate; it even became uncertain whether Mecca was the centre of the world. The theoretical basis of these Arab geographersand ethnographers was not, however, empirical. It was derived from earlier Pahlavi and Greek writings, and especially from the works of Ptolemy and Marinus, now transx This is the full version of a paper read to the Joint Triennial Classical Conference at Cambridge in July 1971. Its purpose, now as then, is to provoke discussion: it is intended as a preliminary attempt to establish lines of approach in an area which I propose to study in greater detail later. My concern here is with an important and neglected aspect of the history of Hellenistic historiography, seen from the wider viewpoint exemplified in the work of Felix Jacoby. I am grateful for the comments of my audience in 1971, in particular Professors Ernst Badian, Moses Finley, Arnaldo Momigliano, Martin Ostwald, and my old tutor J. P. V. D. Balsdon. I would like also to thank Peter Parsons for bibliographical help. 2 L. Wittgenstein, Tractatus Logico-Philosophicusi. I. The context of course is different. 3 Herod. 2. 35; on this pattern of reversal in ancient anthropological descriptions, see S. Pembroke, 'Women in Charge', J. W.C.L xxx (1967), I ff., esp. 16-18; compare also K. Herodotus' emphasis on OavcLa•ca, dergriechischTriidinger, Studienzur Geschichte riimischenEthnographie (Diss. Basle, 1918), 21 ff. 4 See the excellent sketch of S. Maqbul Ahmad, Encyclopediaof Islam, ii (1965), 57587 s. Djughrdfiya.

201 lated into Arabic. Often there was a conflict between these theoretical structures and the empirical facts gathered by travel and observation; but this conflict was a necessary one. The complexities of the construction of a great geographical literature can be seen for instance in the work of Ibn Khurra(published in two editions of 846 and 885), a director of posts and d.dhbih intelligence under the Abbdsid caliphs, who wrote at the desire of the Caliph, but was also a scholar, a translator of Ptolemy;' or in the Arab Herodotus, al-Mas'fidi, who died in 956, traveller,geographer,historian,who believed that geography was a part of history, and wrote his geographical account as an introduction to and an integral part of his history.z Then from the twelfth century onwards there was a period of consolidation,compilation, and decline, in which the old geographerswere used to propagate a view of the world which the West had rejected by the sixteenth century, but which in the East lasted till the nineteenth. Here we can see too the effect of the transfer of the spirit of exploration from the Arabs to the Portuguese, and so to other western nations. In literary terms, the re-evaluation of previous writers is well illustrated, as Professor Momigliano has shown, by the effect of the reports of European travellers in the sixteenth century, the accounts of the discovery and exploration of America. Fifteenth-century scholars had accepted the view that Herodotus was the father of history perhaps, but also the father of lies. A hundred years later Stephanus in his ApologiaproHerodoto(1566) could point to the comparative study of native customs as showing that Herodotus was basically reliable.3 The influence was reciprocal; Alonso de Zorita in his Relationof the Lordsof New Spain (written before 1570) could discuss the problems of barbarism and civilization in terms of the Greek and Roman response; and Bartolom6 de Las Casas in his ApologiticaHistoria(written in the I550s) could analyse and compare Indian culture, society, and religion with those of ancient peoples, and according to the categories of Aristotle.4 In the ancient world these same phenomena can be observed. During and immediately after the periods of expansion there is a new awareness, a new flexibility of response, to the variety of human cultures. It is no accident that the great writers of cultural history appear in or just after such periods-the firstage of colonization and of contact with the East, the conquestsofAlexander, and those of the middle and late Roman Republic in the West. Ancient writers knew this well: Polybius expressesbest this interrelationshipof conquest and geographical knowledge, the importance of travel, the difficulties of autopsy, and the effect of Alexander and the Romans in a passage too long to quote (3. 57-9).s5 These periods of expansion are followed by periods when the new vision becomes another stereotype to imprison the imagination. But also, in I Text and translation in M. J. de Goeje Arabicorum,vi (ed.), BibliothecaGeographorum (1889). 2 An excellent impression of al-Mas'fidi's main work, the Akhbar ez-zeman (in 30 volumes ,of which only the first survives), can be gained from the abbreviation of it which he wrote, published in nine volumes by C. Barbier de Meynard and Pavet de Courteille, Magoudi,LesPrairiesd'Or, Collection d'ouvrages orientaux, Societe Asiatique (1861-77). 3 A. D. Momigliano, 'The Place of Herodotus in the History of Historiography', Secondo contributo alla storia degli studi classici (I960), 29 ff., esp. 39 ff. (also in Studiesin Historiography[1966], 127 ff.); and 'Erodoto e la storiografia moderna', Secondocontributo,45 ff., esp. 52 ff. 4 See for these authors J. H. Elliott, The Old World and the New z492-165o (1970), chap. 2, esp. 46 ff. s It was a common theme: cf. Strabo I. 2. I, citing and expanding Eratosthenes; and 2. 5. II on the problem of autopsy.

O. MURRAY such periods of change, there is continuity; the writerswho grappled with these new worlds tried to understand them in relation to their own preconceptions, and their predecessors.It is my design in this preliminarystudy to show something of the influence of Herodotus on the most sudden and dramatic extension of the frontiersof the known world which ever faced antiquity, the conquests of Alexander. For a long time the accepted modern view was that Herodotus was not read widely in the Hellenistic world; his style was out of favour, and his reputation a teller of charming stories.He had nothing to was that of a liar, or a muthologos, teach the new scientific ethnography of the Hellenistic period, which mentioned him only in disparagement or for ridicule; and the ordinary reader, finding his delight in the pleasures of the moment, works written E' -rb7rapaa aKOvELV, in romantic stories of distant places, was now better served by Xprn/ the wilder shores of prose romance or utopian travelogues. Nowadays it is usual to dismiss this 'old view'; we know, it is said, that Herodotus continued to be read in the Hellenistic world. But usually this admission is immediately qualified: he was read for his style, but despisedas a historian; he was an author for the schools of rhetoric, not a man read by the true seeker after knowledge; he was of course read, but his influence was slight compared with that of Thucydides-it is Thucydides who dominates the historiographyof the Hellenistic world.' In the face of this attitude, I must begin by offering in general terms some indications of the popularity of Herodotus as an author; for such grudging admissionsscarcely do justice to the central position of Herodotus as one of the most widely read authors throughout antiquity. And it is against this general background, in which it must be assumed that almost every educated man had read Herodotus, and quite as many as had read Thucydides, that we must consider what difference this acquaintance with Herodotus made to the vision of the world created in the early Hellenistic period. First, a general numerical argument. From the provincial capitals of GraecoRoman Egypt have come, according to the last count, no less than twenty-one papyrus fragmentsof Herodotus; he stands fifteenth in the top twenty authors, and sixth among prose writers, after Demosthenes, Plato, Isocrates, Thucydides, and Xenophon. The fragility of such calculations is shown by the fact that in a volume of the Oxyrhynchus Papyri series soon to appear, fragmentsof at least eleven new manuscripts on papyrus will be published, which would raise Herodotus from fifteenth to ninth in order of popularity, roughly equal with Thucydides, and above Xenophon. Moreover, Xenophon was for the Graeco-Roman world a philosopher as much as a historian: in terms of his historical works alone he falls far behind Herodotus. This popularity of Herodotus might be explained by local factors, by the Egyptian Greek'sinterest in RE Suppl.ii (1913), 504-15 ; see also SchmidFor the older view, cf. A. Kirchhoff, Literatur,i. 2 Uber die Entstehungszeit des Herodoteischen Stdihlin, Geschichtedergriechischen (1878), 9; A. Bauer, Herodots (1934), 665-70; and the useful thesis of K. A. Geschichtswerk2 in Biographie (1878), 4. For modern views cf. Riemann, Das herodoteischeGeschichtswerk der Antike (Diss. Munich, 1967). To all of e.g. Momigliano in the articles cited p. 20o these I am deeply indebted for references in n. 3; H. Strasburger, cited below, p. 2 Ii n. 4. The best discussion of the influence of this section. Herodotus in antiquity is still F. Jacoby,

203 his own country; but the Egyptian section of Herodotus' work, book 2, is no more heavily represented than any other part'. Of course it is true that such calculations fail to distinguishbetween the Ptolemaic and Roman periods; but I can see no reason why the reading habits of the bourgeoisie of Oxyrhynchus should have radically changed towardsrather than away from Herodotus under Roman rule. And if none of the papyrusfragmentsof Herodotus are Ptolemaic, the same is almost true of Thucydides: against the one third-century B.c. fragment of Thucydides we can set the far more significant evidence, also from a papyrus, that Aristarchus the great Alexandrian scholar of the early second century wrote a commentary on Herodotus-the earliest known commentary on a prose author, and the only known one before Didymus wrote on the orators.2 Nor was Aristarchus the only Alexandrian to interest himself in Herodotus; Aristophanes of Byzantium used him for his Lexeis;and Aristarchus' great rival Hellanicus the grammarian lectured on Herodotus.3 From the papyri then we may perhaps surmise that the two most popular historians in Graeco-Roman Egypt were Herodotus and Thucydides together; or that at least there is no support in the evidence for the view that there was a period when one eclipsed the other. Both were standard authors to be read for pleasure or instruction; and Herodotus at least was the victim of learned Alexandrian discussion and lectures, philological study, and textual criticism. The general popularity of Herodotus is also well shown by the poetry of the period. Herodotus is of course of all historians 'OpPLKwCra7Tos,4and likely to appeal to poets, whereas one can scarcely imagine Thucydides, the political historian of small Greek cities, appealing to the court poets of Alexandria; so the poetic popularity of the two can hardly be compared. But at least the positive conclusion emerges again, that Herodotus was used extensively by Hellenistic poets. His influence has been detected on Callimachus,s and there are clear echoes of him in Apollonius Rhodius.6 Perhaps the most interesting example is Apollonius' account of the Argonauts on Lake Tritonis, and the prophecy of the clod of earth which was to symbolize future colonization in Libya. Here Apollonius does not just use the obvious poetic model; for the story is told in Pindar's FourthPythian.He also uses the variant account in Herodotus book 4, deliberately combining the two in a composite story.7 The list of poets who read and appreciated Herodotus could be extended, with the help of the local epic poets and the Greek anthology; but, for the sake of being controversial, I will add only one more writer. There could scarcely be a greater tribute paid to a historian by a poet than the versification of his narrative. I refer of course to the faithful transmutation into tragedy of the W. H. Willis, 'Census of Literary 288 ff. 3 Pfeiffer, op. cit. 197; Schol. Sophocles, Papyri', G.R.B.S. ix (1968), 212; I am Philoctetes201. grateful to Prof. E. G. Turner for information about the future Oxyrhynchus publication. * Longinus 13. 3; cf. Dionysius, ad Pomp. The Herodotus papyri so far published are 3; Demetrius I2; Jacoby, RE Suppl. ii. collected in A. H. R. E. Paap, De Herodoti 502 ff. reliquiis in papyris et membranis Aegyptiis s E. Howald, Hermes,Iviii (1923), 133-8. 6 E. Delage, La Giographiedans les Argoservatis(Diss. Utrecht, I948). 2 P. Amherst ii, 12 = Paap, no. o; see nautiquesd'Apolloniosde Rhodes(1930), 82 ff., below, p. 204. On the problem of possible 279 in general; in particular pp. 79 f. (Apoll. commentaries on prose authors before I. 591: Herod. 7. 193), P. 255 (Apoll. 4. Didymus, see R. Pfeiffer, History of Classical 1349: Herod. 4. I89). 7 Apoll. 4. I537 ff.: Herod. 4. 179; Scholarship (1968), 224f., 277f.; and on Didymus' methods, S. West, C.Q.xx (1970), Delage op. cit. 261-70.

0. MURRAY Herodotean story of Gyges of Lydia by some unknown writer of the Hellenistic age.' This affection and respect for the first Greek historian also found expression in honorific statues. That might be expected of Halicarnassus,which in florid Asianic style proclaimed (on a statue base which has survived) : 'Nineveh was not allotted the furrow of mankind nor was the self-sown shoot of the Muses reared in India, nor did primeval Babylon bring forth the sweet lips of Herodotus and Panyassis beloved of Hera, but the rocky soil of Halicarnassus; through whose songs she has found great glory among the cities of the Greeks.'Z We possessalso the base of the portrait of Herodotus which stood in the library of the kings of Pergamon, the most famous library of the Hellenistic world after that of Alexandria.3 Such evidence is perhaps enough to show that the appreciation of Herodotus was not confined to natives of Halicarnassuslike Dionysius, who tried to place Herodotus above Thucydides on stylistic grounds; rather Herodotus was read widely by educated people.4 If we ask how he was read, the answer is one which will emerge more clearly later on. But the poets at least, like the philosophers,5read him for information. And Aristarchus'comments (or those that survive in the miserable fragment of excerpts, perhaps also partly in epitome) are surprisinglywide. He does the things we would expect of a grammarian, explaining difficult words, offering parallels (not of obvious relevance), discussingvariant readings.But he is alsoconcernedwith the factual information in Herodotus; he appears to add to it, he offers modern parallels, and in discussing his variant reading he supports it by comparative factual material. This is not a purely linguistic, stylistic, and textual commentary.6 II is however more specific; I wish to show the influence of My present purpose Herodotus on the conception which the Hellenistic age had of the world around it. And especially I wish to argue that it is this influence which lies at the basis of the whole tradition of Hellenistic historical ethnography.The easiest way to demonstrate this is perhaps to investigate those prose authors of the early Hellenistic period who interpreted for the new rulers of the world the alien cultures which now belonged to them; for this small group of writers created a view of foreign civilizations which lasted at least until the differently oriented Roman conquests produced a Poseidonius. Already in the fourth century B.C. the influence of Herodotus on the writing I P. Oxy. xxiii (I956), no. 2382; no. 1707. 2 Pack2, F. Hiller von Gaertringen, W. Peek, Hermes, lxxvi (1941), 220 ff.; presumably the base of a double-herm; Hiller von Gaertringen assigns the poem to Antipater of Sidon, which, in the absence of any more specific statement about the date of the inscription, would suggest that the editors thought it late Hellenistic. It has not yet been observed that this poem seems to contain the earliest reference to the 'Muses' of Herodotus (otherwise first attested in Lucian), and therefore also to the nine-book division (first certain reference Diodorus); the remaining evidence is in Schmid- Staihlin,op. cit. 662 n. 3. Lebas-Waddington, 1618 (second century A.D.) records an early statue of Herodotus which stood in the position of honour in the gymnasium at Halicarnassus (rov 'Hpd3or'ov). v•HrapAavInschr. no. 3 Altertiimer vonPergamon, 199. * Dion. Hal. ad Pomp. 3; H. MqL•aEws ii. III, p. 207 Usener-Radermacher. s Despite the critical attitude of Aristotle, he used Herodotus often, as did other philosophers: some examples, Riemann, op. cit. 36 if.; J. Geffcken, Zwei griechischen Apologeten(1907), 188 n. 3. 6 For Aristarchus' commentary and bibliography, see Paap, no. 10; Packz, no. 483; Schmid-Stithlin, op. cit. 665 n. 8.

205 of history had been strong, especially on Ephorus,whose historyused Herodotus as one of its main sources, and also shows many Herodotean elements.' The historians of the court of Philip of Macedon seem to have been particularly interested in Herodotus-a specific instance perhaps of the growing interest in the Persian empire as an area of expansion. Theopompus wrote an epitome of Herodotus in two books (the firstknown epitome), and his historieswere noted for his Herodotean digressions: he seems to have offered a defence of this addition ofpoikilotisto history, for Strabo recordshim as saying that 'he would write stories in his histories more marvellous than Herodotus and Ctesias and Hellanicus and those who write on India.'2 Callisthenes too is known to have followed Herodotus almost word for word on occasions.3 But it is the early Hellenistic writers who seem most heavily indebted to Herodotus. They never tire of denouncing him, and declaring his information to be unreliable. That was of course necessaryin order to assert (often falsely) their own independence of his narrative, and their allegedly better sources of information: it was a game which had been played with equal dishonesty by Ctesias earlier.4Thus Hecataeus, writing the standard account of Egypt under Ptolemy Soter, can say: 'Now as for the stories invented by Herodotus and certain other writers on Egyptian matters, who deliberately preferred to the truth the telling of marvellous tales and the insertion of myths to please their readers, these we shall omit, and give only what appears in the written records of the priestsof Egypt and has passedour carefulscrutiny.'sReferencesto the unare numerous; as Josephus reliability of Herodotusand his love of stories(6?o80o) other writers attack each 'but other, says, may everyone accuses Herodotus of lying.'6 Indeed I know of only one author to have said anything respectful about Herodotus;Agatharchidescalls him'a tirelessinvestigatorifever therewas one, and with much experience of history'--a significant tribute, for Agatharchides goes on to discuss and refute Herodotus' theory of the rising of the Nile.7 In one sense all the attacks on Herodotus by professional historians and ethnographershelp to show that he was widely read, and all the contradictionsof his detailed information similarly show how carefully he was studied by the very authors who denounce him. But my aim is to go furtherstill, to suggest that for all their denunciations the early Hellenistic writers saw the world through Herodotean eyes, modelled large sections of their works on him, indeed that their achievement in so successfully describing the new world they saw was made possible only with the help of their great predecessor. The earliestwork I wish to discussis that written by Alexander'sadmiraland companion in India, Nearchus.8Here there is no need to sort out in detail the I Jacoby, RE Suppl. ii. 5I0. For the epitome, see F.G.H. 115 T I, F 1-4. 304. The purpose of this epitome is obscure: cf. R. Laqueur, RE 5A (1934), 2188, and below, p. 206 n. I. On the character of the history, Strabo I. 2. 35 = F 381; cf. Triidinger, op. cit. 6o ff. 3 Herod. I. 202: F.G.H. 124 F 38; Herod. I. 175: F 25; cf. F 30; Riemann, op. cit. 50 f.; Jacoby, RE Suppl. ii. 512. 4 The anti-Herodotean polemic is sketched in A. Hauvette, Hdrodote(1894), 63-I80; see also Schmid-Stahlin,loc. cit. (p. 2o02n. I). For Ctesias and Herodotus see F. Jacoby, RE xi 2 (1922), 2041-66; A. Momigliano, Atenee Roma,xii (1931), I5 ff. = Quartocontributo, 181 ff. s Diod. I. 69. 7; cf. the implicit attacks in I. 59. 2 (Herod.2. 1 1); 62. 2 (2. 112); 66. 10 (2. I5I). 6 From Thuc. I. 21-2 onwards; cf. esp. the passing references to Herodotus as 0 pv0o•oMyosin Aristotle, de gen. anim. 3. 5, 756b6; F 248, Rose p. 196;Jos. c. Ap. I. 16. 7 Diod. I. 374. 8 For an up-to-date bibliography on Nearchus (F.G.H. I33), see W. Spoerri, KleinePauly,iv (1970), 33 f. On his relation-

0. MURRAY work of Nearchus from the accounts of India in Arrian and Strabo. My point is a simple one. Nearchus is a reasonably trustworthyand reliable writer, who reportslargely what he saw, without romantic exaggeration. This is the account of an honest and perceptive soldier, not an intellectual. And yet we know that Nearchus took Herodotus' account of India very seriously, and we can still see traces of its influence, both in detail and in general conception-all this despite the fact that his own experiences must have shown Nearchus how unreliable the account of India in Herodotus was. Nearchus saw India not wholly as the innocent traveller, but through Herodotean eyes; he may very well have travelled with a copy of Herodotus.IAt least when he came to write his account of India and of his travels some time before 312, he wrote with Herodotus in mind; and there is good reason to suspect that he even wrote in Ionic in order to emphasize this connection. He was prepared to go a long way in his trust of Herodotus; thus Nearchus was interested in the gold-digging ants of enormous size which are perhaps Herodotus' most fabulous animals; he said that though he never saw any of these strange creatures, their skins were brought in large numbers to the Macedonian camp, and they looked like panther skins.2Now the Herodotean story is presumably a combination of genuine Indian myth with a trade in skins and gold; but the interesting fact is that Nearchus (and perhaps others on Alexander's expedition) knew of the Herodotean story, and did not disbelieve it, but rather sought to verify it. Again we find Nearchus echoing Herodotus' account of Indian dress,sand in F 17 quoting and generalizing Herodotus' famous description of Egypt as the gift of the Nile, to cover other alluvial plains-in connection of course with the great river plains of North India.4 The parallelism with Egypt is indeed an underlying theme of Nearchus' work: he discussedthe flooding of the Indus valley in relation to the Nile floods, and attributed both to summer rains; he described the flora and fauna of the same area against the background of Egypt.s This comparison between the Indian plains and Egypt is on one level the reaction of an intelligent observer, who saw how similar the natural forces were in both countries. But it is also the observation of a man whose eyes have been adjusted to a river valley environment through the reading of Herodotus; Nearchus' account of India looks not only to Herodotus' short section on the country, but also to the full-scale description of Egypt in book 2. And the Herodotean sections on Arabia and Ethiopia, which occur directly after the account of India, have been used by Nearchus as a model for his description of the desert country and its inhabitants on his voyage to the Persian Gulf.6 ship to Herodotus see esp. L. Pearson, The Lost Historians of Alexanderthe Great (1960), I I8 ff. (though he goes too far in thinking that literary form has distorted the truthfulness of Nearchus' account; and some of his examples are not convincing). I The close relationship between Nearchus and Herodotus raises a wider question: Professor E. Badian pointed out to me on the occasion of this paper that there is some evidence for Herodotus having been a major geographical source for the planning of Alexander's expedition in general. Was Theopompus' epitome an epitome of the early books of Herodotus rather than of the late ones, and designed as a field-book for the campaign which Philip planned? 2 Herod. 3. o02, 105: Nearchus F 8. On this story see the excellent note of How and Wells ad loc.; W. W. Tarn, The Greeksin Bactria and Indiaz (195I), io6 ff. is useful for references, but somewhat too rationalistic in 3 F 1i. his approach. 4 Herod. 2. 5; 2. Io: F 17; cf. Arrian, Anab. 5. 6. 3-8. On Nearchus' use of Herodotus here, see Pearson, op. cit. I 18-20. 5 F 18-20; Pearson, op. cit. I20-3; Triidinger, op. cit. 66 f. 6 F. i. I omit any discussion of Onesicritus: his inventiveness and strong philo-

207 And yet Nearchus is perhaps closer to Herodotus in spirit than other Hellenistic writers are; for he still retains something of the innocent eye of the traveller without preconceptions.That innocent eye was lost in the work of the writer who influenced most decisively Hellenistic history and ethnography, Hecataeus of Abdera, the first author to write for and under the patronage of one of the Diadochoi. His work on Egypt, largely preserved in book I of Diodorus, was written very early in Ptolemy Soter's control over Egypt, probably between 320 and 315, before the essential character of the Ptolemaic state apparatus had been established, and when native Egyptian attitudes were more respected; its purpose was primarily to glorify the land of Egypt, to present it as the source of all civilization, and the ideal philosophical state.' Hecataeus' work transformedthe writing of ethnography. It was arranged on logical principles, carried out for the first time with total consistency. Firstly the archaeologia, mythical period (theology and prehistory or theologoumena-the in Greek fashion mythology being typical equated, and explained as a reflection of early history). Then perhaps a geographical section, though this could be disputed. After the mythical period, the historical; finally a systematic description of the customs of Egypt. Hecataeus' account is based on his own investigationsand informationgathered from Egyptian priests, to be sure: he is insistent on his access to the written archives of Egypt, and on the new standards of accuracy he has imported. But it is also clear that he used previous Greek writers; and of these writers he used most extensively Herodotus. Before he embarks on his description of the customs of Egypt, he delivers the explicit attack on Herodotus as an unreliable source, quoted above. But the whole section on the history of Egypt, just before that, is in fact for the most part taken with only the smallest alterations from Herodotus. Again Hecataeus' rationalistic attitude to Egyptian religion and its relation to Greek is modelled on that of Herodotus; and even in the section on customs, where he is so insistent on his own superiority, it is clear that he has used Herodotus extensively. Hecataeus may have known of the lost fourth-centuryworks on Egypt; but it is obvious that he regardedhis greatestpredecessoras Herodotus. He thought that he was superseding Herodotus because Herodotus was too credulous for the modern taste, and because he was insufficiently'scientific'in his approach; instead of organizing his material, he wrote as caught his fancy. Hecataeus imported structureinto his account, a structurebased on philosophical theories -theories of the ideal state which for Hecataeus was exemplified in old Egypt, and theories of the nature of ethnographic description. This marriage of the Herodotean innocent eye with genuine local tradition on the one hand and on the other with Greek philosophy, set a standard and a pattern for the writers of the next two centuries: it was the renewal of the Herodotean approach adapted to a more sophisticated age. Hecataeus' work provoked immediate competition from the other successor sophical interests make him a special case, and put him outside the main stream of serious ethnographic historians with which I am concerned here. Nevertheless the influence of Herodotus can of course be shown. I On Hecataeus, see the works cited in my article, 'Hecataeus and Pharaonic Kingship', J.E.A. Ivi (1970), 141-71. ments for and against the date there proposed, see the discussion between myself and M. Stern in J.E.A. lix (1973) (forthcoming). On Hecataeus and Herodotus see also the comments of J. Vogt, Tiibinger Beitrage z. Altertumswissenschaft,v (1929), 132 f. For argu-

0. MURRAY kingdoms. Megasthenes was on several occasions the ambassadorof Seleucus to the court of Chandragupta in India. The work he wrote in the first decade of the third century on the basis of his experiences was a direct reply to Hecataeus, and modelled on the method, form, and content of his book-a systematic account of Indian culture with geography, flora, fauna, and the people (book I), the system of government and nomoi(book 2), society and philosophy (book 3), archaeology, mythology, and history (book 4).' This work is the source of Diodorus' section on India, as of much of Strabo and Arrian. Megasthenes was a well-read man; and, like Hecataeus, he aimed to write a definitive account. He relied on his own travels and the information he had carefully gathered from Indians, but also on the accounts of previous Greek writers; it was perhaps unfortunate that their stories were so unreliable. But he knew at least on general grounds that he should prefer Herodotus to Ctesias.2Apart from Megasthenes' debt to other writers on India, the work contains a large number of obvious reflections of Hecataeus, for it is an attempt to show that India is an even better land than Hecataeus' Egypt, a Platonic ideal state with a rigid caste system and philosophers on top, and that all civilization springs from India not Egypt. The book also looks back through Hecataeus and Nearchus to Herodotus' great set piece of ethnographicdescription, the land of Egypt. The other great early Seleucid writer is in a rather different situation. Berossus was a bilingual priest of Baal, whose Babyloniakawas addressed to Antiochus I.3 As the work of a non-Greek it is not surprising that his book shows no direct acquaintance with Herodotus. But it still belongs within the same tradition, for it too is arranged according to the principles established by Hecataeus, and can be seen as yet another reply, compiled perhaps under direct Seleucid patronage, to the court historian of the Ptolemies. Its three books described the land of Babylon, the origins of civilization, and Babylonian mythology in book I, the ten mythical kings of Babylon in book 2, and the history in book 3. Berossusis of course a good deal more accurate than either Hecataeus or Megasthenes, because he knew the Babylonian records, and kept I The available accounts of Megasthenes (F.G.H. 715) are not particularly satisfactory: bibliography in J. D. M. Derrett, Kleine Pauly, iii (1969), 1150 ff.; add the important discussion of F. Altheim, WeltgeschichteAsiens im griechischenZeitalteri (1947), 257-64. The long and discursive article of O. Stein, RE xv (1931), 230-326 contains also the most balanced appreciation, especially on his relationship to earlier Greek thought, 237-67. But the central importance of the relationship between Megasthenes and Hecataeus has not so far been realized. It is clear in general (much of the evidence collected by Stein 272 ff. and Triidinger, op. cit. 75 if. is relevant here). In detail see for instance Megasthenes F 32. 54 on the perfect nature of Indian society, F 14 and F 4- 36. 4 on climate, population, colonization, conquest, (an implicit contrast or comparison with much in Hecataeus, e.g. Diod. I. 28), or the 'Stoic' description of the nature of the (F 33. 59) in identical words to KO'ato! Hecataeus 264 F I. 2 The gap between the evidence of the fragments and the harsh judgement of Strabo on Megasthenes' reliability (T 4, a evusoAo'yo!; contrast Arrian, T 6) is best explained by seeing Megasthenes as an accurate reporter, but uncritical of his predecessors and informants: cf. the somewhat unsympathetic account of T. S. Brown, A.J.P. lxxvi (1955), 18 ff. On Megasthenes' relationship to Herodotus, see esp. F 23: Herod. 3. o02(the gold-digging ants again) ; further examples in Stein, op. cit. 237 f- On the relative importance of Herodotus and Ctesias for Megasthenes, see Stein, op. cit. 243. 3 Bibliography on Berossus (F.G.H. 68o) in W. Spoerri, Kleine Pauly, i (1964), 1548; see esp. P. Schnabel, Berossosu. die babylonischhellenistischeLiteratur(1923) chaps. 1-2. On the dedication, see T 2.

2og9 closely to their form. But the arrangement of his material and his explanations of it fit the pattern established by Hecataeus. Berossuswrites as a Babylonian trying to provide his Greek masterswith an explanation of his culture in accordance with their preconceptions. One vital section is however missing, that on the customs of Babylonia: a native perhaps could not distance himself enough to be able to describe his culture from the outside, as a foreigner would see it. Berossus was in turn followed by Manetho, the Egyptian high priest of Heliopolis;' he is so confidently placed after Berossusthat it may be that he specificallyreferredto Berossusin his works.His attitude at least is very similar. His work was probably produced under royal patronage, and may well have been addressedto Ptolemy.2Even more than Berossus,he was a product of the fusion of Greek and native cultures, as is shown best by his collaboration with Timotheus, of the Athenian priestly family of the Eumolpidae, in creating a theology and cult of the new Graeco-Egyptian ersatz god, Sarapis; Josephus calls him 'a man well acquainted with Greek culture'.3His history shows more strongly the same limitations as that of Berossus: it is based directly on the sacred books of Egypt ;4writing thus, Manetho sought to correct the mistakes of Greek writers who had access to these records only indirectly. His work is therefore primarily concerned with what appeared in those records-the history of the kings of Egypt; and Manetho's account is so accurate that it is still the foundation of Egyptian chronology. Manetho does not seem to have written a complete Hecataean ethnography, but merely to have covered the same ground as his historical section, and (both in his history, and more especially in other works) to have discussed the theology of Egyptian religion. Within these limits it is still clear that Manetho followed the Hecataean structure, the threefold division of the early kings of Egypt into gods (identified with physical elements), divine kings, and human kings; and he accepted the rationalisticphysical explanation of the Egyptian gods provided by Hecataeus.s But despite these signs of Hecataeus' influence, that author is not named in the extant fragments. Rather we hear that Manetho's history was ostensibly directed to the correction of Herodotus: according to Josephus, 'he convicted Herodotus of making many mistakes about Egyptian affairs because of ignorance.'6 Manetho may even have written an independent work Against I On Manetho (F.G.H. 609) see R. Laqueur, RE xiv (1928), Io6o ff. 2 The relationship between Manetho and Berossus is unclear, because it is not certain when the remarks of Syncellus refer to Manetho and when they refer to a pseudonymous work which he also knew; this is merely a special instance of the general problem of interpolations and pseudoManethos. T I Ib states that Manetho is later than Berossus, T I Ic and I Id show him contradicting Berossus. Two pseudonymous works portray him addressing Ptolemy II: T IIa+F 25; T I2. None of this evidence seems to relate directly to the real Manetho; nevertheless this consistent attitude of depicting a court writer under Ptolemy II may well imply that the genuine works possessed similar characteristics. The fact that Plutarch on the establishment of the Sarapis cult (T 3) puts Manetho in the later years of Ptolemy I does not of course imply a date for his literary works earlier than Berossus, or even under Ptolemy I; see on all this Laqueur, op. cit. IO63 ff. 3 C. Ap. 1. 73 = T 7a. 4 F I; cf. F 9. 105; F Io. 229. The last two passages divide the sources of Manetho into the sacred records and a selection EK7Tv adEacrroTTcaJw lyvoAoyov~'vw v 'from popular legends'. The first category certainly represents the real Manetho; but the second may be the distinguishing mark of an antiJewish interpolator. s6 See J.E.A. Ivi (1970), 167 f. Loc. cit. (n. 3). Other explicit references to Herodotus, F 3 pp. 16 f.; F 2/3 PP. 22 f. ; cf. Jacoby's apparatus, p. 42 note on I1-13, p. 103 note on 8-12.

O. MURRAY though this may rather be a characterizationof his main work.' But it is perhaps worth noting that Herodotus is accused not of falsehood, but of Why Manetho sought to correct Herodotus ignorance (tir' dayvo&la E~evapLevov). rather than Hecataeus I do not know; perhaps he did not wish to attack the respected protege of the Ptolemies: perhaps he recognized that Hecataeus' historical section was so derivative on Herodotus that it was better to concentrate on the mistakes of the original. For some reason at least, he felt that Herodotus was still worth attacking, despite the existence of a more modern account equally open to criticism. It was not just ethnographersand cultural historians who used Herodotus. One of the greatest of the early Hellenistic political historians, Hieronymus of Cardia, wrote the standard account of the Diadochoi-pure political and military history. And yet, for instance, his account of the expedition of Antigonus against the Nabataean Arabs in 312 is organized and arranged like a Herodotean logos,with a detailed description of the habits and customs of the Arabs before the account of the actual fighting.2 With Berossusand Manetho we begin to see the development, the branching out of different tendencies which can be traced back through Hecataeus to Herodotus. The analysis of this Herodotean tradition could be extended to the later Hellenistic writers, to Timaeus (the Herodotus of the west), to Agatharchides (who admired Herodotus), to Eratosthenes (who was always quoting him), and to Poseidonius. But this will be enough to show how the Herodotean legacy affected the world view of the Hellenistic period, and how important he is for an understanding of Hellenistic historiography. It would hardly be too much to say that the early Hellenistic period saw the new world of Alexander through Herodotean eyes, and sought to give the Herodotean tradition a more systematic basis. But without the example of Herodotus the achievement of the writers under the Successor kingdoms in recording and understanding the would have been very different, and more difficult. oikoumene Only in the late Hellenistic period does Herodotus seem to have suffered a slight eclipse as a seriouswriter to be studied. Dionysius of Halicarnassusmight praise his literary merits; but Diodorus did not use him; and the great majority of the references to Herodotus in Strabo come not from his own reading but through earlier writers, like Eratosthenes.3 How does this sketch which I have given of one of the most important areas of Hellenistic historiographyfit into the standard modern accounts? It shows, I believe, that modern attempts to write the history of historiographyin this period are seriouslydistorted and inadequate. So far interest has concentrated on Polybius, on his theoretical statements about the writing of history, and on his polemic against his predecessors. But, as F. W. Walbank has shown, this polemic is remarkably selective, and often selective on grounds which have F 13. 2 Diod. 19. 94-o00 (not printed in F.G.H. 154) ; cf. F. Jacoby, RE viii (1913), 1559. 3 The exceptions are Strabo 7. 3. 8 and perhaps 17. 2. 5. The fact was demonstrated by W. Althaus, Die Herodotzitatein Strabons Geographie(Diss. Freiburg i. B., 1941), which I know only through Riemann's discussion, op. cit. 47-55. Whether this establishes Althaus's conclusion, that Strabo himself did not use Herodotus as a geographical source, is more dubious. Riemann rightly points out that the explicit references to Herodotus are mostly in polemical passages, but that there are a number of passages where Strabo reproduces information in Herodotus without mentioning his name: it is unlikely that all these passages come through an intermediary. The problem needs further investigation.

21r nothing to do with the writing of history, but rather with the discrediting of rivals to Polybius' own reputation, like Timaeus and Pytheas of Massalia, or with historians who belonged to the wrong states or held the wrong political views for Polybius.' One type of history, however, Polybius does attack with some justification, a type which has been called by modern scholars 'tragic history'. The discussion of this genre of tragic history, its existence or nonexistence, its importance, its origins, has dominated the study of Hellenistic historiographyfor seventy years, since Eduard Schwartz first raised the problem.2 The central idea in this type of history is that of mimesis;the reader must feel the emotions which the actors themselvesfelt-it is a sort of psychological version of the theory of Croce and Collingwood, that all true history is contemporary history, the re-enactmentof past experience, relevant to the modern age. Now many political historians of the Hellenistic period did aim for such effects; and if there is also a theory of history behind this practical aim, it may well be an application of the Aristotelian theory of tragedy to history. And Polybius was at least right to object to such tendencies when they were allowed to obscure the search after truth. But my point here is that tragic history is essentially a form of political history; it is concerned with the sufferingsand actions of men and peoples at war, emotions in a night battle or at the sacking of a city. To set up the theory of mimetic history as central to the Hellenistic period is to accept the limitations of Polybius' polemic in an even more important way than Walbank has pointed out. For Polybius was a political historian, in the tradition of Thucydides: he is not particularly interested in other types of history, though he could of course achieve remarkablestandards of cultural history with his description of the Roman constitution or his geographical sections, especially we may suppose in book 34. But since Polybius is primarily a political historian, his polemic is directed against political historians for the most part (and also against professional geographers):3 it simply ignores a whole area of Hellenistic historiography.Among the great and famous authors whom Polybius loves to attack, one name is wholly absent. In all the work of Polybius, there is no reference at all to Herodotus. H. Strasburger, in an important article called Die Wesensbestimmung der has pointed to a basic dichotomy Geschichtedurchdie antike Geschichtsschreibung,4 in ancient historiography, indeed in all historiography: between 'kinetic history', the history of battles, events, political decisions on the one hand, and on the other 'static history', the history of culture, of civilization, and of areas at peace. Of the former Thucydides stands as the archetype, of the latter Herodotus-at least in the early books, for Herodotus really fused the two. This distinction is I think useful. But Strasburgergoes on to accept the traditional F. W. Walbank, 'Polemic in Polybius', J.R.S. lii (1962), i ff. 2 The best introductions to the bibliography and problems raised in this discussion are F. W. Walbank, B.LC.S. ii (0955), 4 ff.; Hist. ix (1960), 216 ff. ; cf. also C. O. Brink, P.C.P.S. vi (i960), 14 ff., and for a more favourable view Strasburger, op. cit. (n. 4) 78 ff. Interesting for Schwartz's approach is his unfavourable judgement, contrasting Ionian and Hellenistic ethnography, in his article on Demetrius of Callatis, RE iv (1901), 2807. 3 For an assessment of the growth of Polybius' geographical interests, see P. Pedech, La Me'thode historiquedePolybe(1964), chap. 12. 4 SB. d. Wiss. Gesellschaftan der Johann WolfgangGoetheUniversitdit Frankfurt/Main,v (1966), no. 3; see my review in C.R. xviii (1968), 2 i8 ff., which contains indeed the germ of the present article. There is therefore no need for me to stress how much I owe to Strasburger for clarifying the problems I here discuss.

O. MURRAY account of Hellenistic historiography: there Thucydides, kinetic history, was supreme, and was exemplified in two schools, the mimetic and its opposite, the factual Polybian approach. Static history was forgotten until the tradition of Herodotus revived with Agatharchides (a curious choice), and especially Poseidonius, who united again for the first time the two strands. For all its attempts at balance, this analysisseems to me to accept the traditional myths of the importance of mimetic history and the neglect of Herodotus in the Hellenistic world. Certainly Polybius is the greatest surviving political historian of the period, and our accounts will always to some extent be biased towards his preoccupations. But anyone who has read Diodorus, and reflected on the remarkablyhomogeneous school of Hellenistic philosophical historieswhich he used as his sources,will see that it is possible to avoid excessive concentration on Polybius. The only full-length work to discussthis period of history-writingis indeed S. Mazzarino's stimulating and perverse II pensierostoricoclassico.IIn volume ii of that work, he chooses as his models for part of that area of Hellenistic history which I have been discussing,not Herodotus, but two fourth-centuryhistorians of Persia, Deinon (father of the historian Cleitarchus), about whom we know little more than that his work was romantic but reasonablyreliable and covered both Persian history and Persian customs,2 and Deinon's predecessor, the notoriously unreliable Ctesias. Certainly Ctesias was widely read and had considerable influence on Hellenistic prose, but on historical romances rather than on serioushistory.3It would not be possible to gather for Ctesias the great mass of references and allusions in serious historianswhich can be offered for Herodotus. Moreover Ctesias and Deinon are themselves romantic and unworthy representativesof the Herodotean tradition; and they covered only one It seems unnecessaryto set up these two as area of the new Hellenistic oikoumene. the source of a type of Hellenistic history, when their predecessor Herodotus is such a much more obvious candidate. In 19o9 Felix Jacoby wrote a famous article giving the theoretical justification of the structure of his proposed 'Fragments of the Greek Historians'.4 This was to include virtually all forms of non-fiction prose writing, not just what we should narrowly call history, but also mythography, ethnography, chronography, biography, literary history, and geography. His reason for this 2 On the I S. mysterious Deinon (F.G.H. 690), Mazzarino, II pensierostoricoclassico, to the brief article of E. Schwartz, RE v ii. I (1966), 14-26; compare his emphasis on the importance for Eratosthenes of Heca(1905), 654 add F. Jacoby, RE xi (1922), taeus, rather than the far more obvious 20o69. 3 On the influence of Ctesias (F.G.H. 688) Herodotus (42 ff.). See in general the see F. Jacoby, RE xi. 2045, 2066 ff.; E. review of A. Momigliano, R.S.L lxxix (1967), 206 ff. = Quartocontributo,59 ff. Schwartz, Fiinf Vortrdgeiiber den griechischen The standard work on Greek ethnography Roman2(I943), 84 ff. The novelistic characteristics of Ctesias have been amply conis the thesis already cited, K. Triidinger, Studien zur Geschichteder griechisch-rdmischen firmed by the new papyrus fragment, P. Oxy. Ethnographie(Diss. Basle, 1918); despite its xxii (I954), 2330. On the general problem of great value, it has serious limitations. It is the relationship between historiography and in the origins of the novel, see the survey of obsessed with problems of Quellenforschung Herodotus and Poseidonius; and it is B. P. Reardon, Courantslittiraires grecs des specifically confined to ethnographic digres- Ile et Iiie siecles apris J.-C. (I971), 315 fz. sions in historians rather than the main 4 Klio, ix (i909), 8o ff. = Abhandlungen writers on ethnography. It is therefore some- griechischenGeschichtsschreibung, 16 ff. what peripheral to the real problems.

213 wide sweep was that Greek history-writing in its origins did not distinguish between these different types; and the various developed forms sprang from a complex interaction between the conflicting interests of these undifferentiated attitudes of mind. When Jacoby died exactly fifty years later, his work, the greatest philological work of this century and the greatest work on Greek history for all time, was only three-quartersfinished, and there seems no sign of anyone taking up the task. To many, his work has seemed marred by a principle of organization which was unnecessarilycomplicated and with little point. Yet I believe thatJacoby was right in his central insight. It is impossible to consider Greek historiographyand its development unless ethnography,geography, and so on are included. And this is not just because the early Ionian writersfailed to distinguish these genres. It is because one of the central strands of Greek historiography remained this Ionian tradition, a tradition which dominated cultural history, as Thucydides dominated political history, and did so as a direct consequence of the continuing influence of Herodotus. No satisfactory history of Hellenistic historiography will be written until this dual legacy is recognized, and each side given its proper prominence. Nor can the Hellenistic world view be understood without appreciating the importance of Herodotus. Balliol College,Oxford OSWYNMURRAY