SOURCE: An introduction to A Historical Commentary on Arrian's “History of Alexander,” Volume I: Commentary on Books I-III, Clarendon Press, 1980, pp. 1-41.

[In the excerpt below, Bosworth studies several critical issues concerning Arrian's composition of his history of Alexander, focusing on the controversy surrounding the title and dating of the work; the historical method Arrian employed in writing the work; and Arrian's style. Bosworth concludes that while Arrian was a “sophisticated stylist,” his abilities as a historian were somewhat flawed.]

… 2. THE HISTORY OF ALEXANDER

Arrian's major work, the history of Alexander, is universally known by the title Anabasis Alexandri. The title occurs in the codex Vindobonensis, the known archetype of Arrian, as well as in excerpts from Stephanus of Byzantium … and the ‘Suda’. … But the ancient testimonia are not consistent. The majority of later writers, in particular Photius, refer to the work in more general terms, as τα περì Aλεξανδρου.¹ Arrian himself gives little hint of the true title, merely referring to it as his ‘history concerning Alexander’ (vii 3. 1). Now this divergence does not occur in the companion work, the Indica, whose title is twice given by Arrian and is repeated in the manuscript tradition and later testimonia without significant variation.² The divergence in the case of the Alexander history is troublesome. Krüger took it as axiomatic that Anabasis was the original title and assumed that the variants arose from a desire to provide a companion work (and title) for Arrian's History of the Successors. … That seems hardly compelling, and there is at least an equal possibility that the title Anabasis is a fiction of late antiquity. Commentators meeting a work in seven books by a man purporting to be the New Xenophon might well have given it a spurious Xenophontean title. Certainly Arrian's work on Alexander is not notably influenced by Xenophon. His stylistic debts are rather to Herodotus and Thucydides, and his direct references to Xenophon's work are sparse and perfunctory,³ a sharp contrast to the Cynegeticus, a work which is explicitly modelled upon Xenophon and which constantly, both implicitly and explicitly, echoes the subject-matter and phraseology of the master.⁴ The ancient critics were only too aware of the relationship, and on the title-page of the Cynegeticus Arrian's name has been deleted and replaced by ‘Xenophon of Athens, the second’.⁵ They might easily have been dissatisfied with a title as unassuming as τα περì Aλεξανδρου and fabricated a second Anabasis for their second Xenophon.

The date of the work is as controversial as its title. There is both an early date, some time in the reign of Hadrian, and a later date, towards the end of the reign of Antoninus Pius or even in Marcus' reign.⁶ The theories to some extent depend on the view of Schwartz (above p. 4) that Arrian only devoted himself to serious historical inquiry after his retirement from public life, but the arguments which are crucial rest upon the internal evidence of the text and the extant testimonia about the relative order of Arrian's historical works. What is certain is that Arrian has complete confidence in his abilities as a writer. He invites comparison with all extant histories of Alexander, claiming that he is fully competent to do justice to the greatness of his subject, and the success of his work will cement his pre-eminence in Greek letters.⁷ The language presupposes a good deal of earlier publication—and successful publication, if not necessarily in the sphere of historical studies. Such indications as we have suggest that the Alexander history came relatively early in the sequence of historical works. In his digest of the Bithyniaca Photius summarizes Arrian's apology for the late appearance of this history of his native province;⁸ he mentioned other works, monographs on Timoleon and Dion, which preceded the Bithyniaca, and Photius claims that the Bithyniaca was Arrian's fourth historical effort, succeeding the history of Alexander and the two monographs.⁹ The subject-matter seems generally derived from Arrian; and, even if the statement about the relative order of the works comes from Photius himself, he must have had some grounds for the statement. Later works are also quoted and excerpted, the seventeen books of the Parthica and the ten books of the History of the Successors. They were not mentioned in the preface to the Bithyniaca and presumably followed it in the sequence of Arrian's works. There is some corroboration in the literary echoes of the Alexander history which we find in the Parthica. The fragments dealing with Trajan's voyage down the Tigris are strikingly reminiscent of the description of Alexander's voyage down the Hydaspes¹⁰ and the character-study of Arsaces echoes the wording of Arrian's celebrated encomium upon Alexander.¹¹ As far as the evidence goes, it suggests a definite pattern in Arrian's historical development. He began with biographical monographs, culminating in the Alexander history, and then turned to more complex composite works, covering more extended and variegated periods of time.

An absolute date is a more difficult task. Citations in later datable authors give merely a terminus ante quem, and it is rare that we can be sure that Arrian is in fact cited directly. The supposed satire in Lucian's treatise on historiography (c. 165) is not only anonymous; its association with actual passages of Arrian's history of Alexander is very tenuous and depends upon free use of the imagination.¹² It is hardly a basic dating criterion. Similarly Appian's use of the Alexander history in his Syriaca and Book II of the Civil Wars,¹³ even if it is proved, gives no more than the most general indication of date (before 161-3). The relative chronology of Arrian's own work is more helpful. There seem to be echoes of his work on Alexander in the Parthica, and, since that work concluded with Trajan's Parthian Wars of 114-17,¹⁴ there is every reason to assume that it was published before the victories of Lucius Verus in 165/6. The Alexander history appeared some considerable time before, with the Bithyniaca published in the interval, and we must surely go back at least to the middle years of Pius' reign. More significant, however, is the Order of Battle against the Alani, the literary report of Arrian's engagement with the Alani while he was governor of Cappadocia. I have tried to prove that in this opuscule Arrian cast himself in the role of Alexander, and borrowed from his description of Alexander's defeat of the Saca nomads in 329.¹⁵ In any case he uses Macedonian terminology for the Roman army of the 130s a.d. describing legions as phalanxes, his infantry headquarter corps as sωματοpονλακεs.¹⁶ Arrian was certainly familiar with Macedonian military institutions and, in all probability, when he wrote the Order of Battle he had the Alexander history already behind him. Now it is probable that the Order of Battle was composed in the aftermath of the engagement with the Alani, as a literary adjunct to his official report, some time around a.d. 135. If the Alexander history preceded it, it cannot be dated after Hadrian's reign and almost certainly pre-dated the legateship of Cappadocia.

The dating criteria which can be extracted from the body of the Alexander history are largely arguments from silence. It is a feature of Arrian's work that he draws upon his own experience to illustrate or expand material from his sources. The information is provided even where it is not directly relevant, as in the case of the contemporary location of the statue-group of the tyrannicides.¹⁷ Yet there are instances where Arrian ignores experiences which are attested elsewhere in his literary corpus. Ignorance of Cappadocia is particularly striking. His discussion of the location of Prometheus' cave is a case in point. Arrian is aware that the subject is controversial and he explicitly criticizes his sources (Anab. v 3. 1-4). Yet he had been shown Prometheus' purported place of punishment in the Caucasus during his inspection tour of 131/2 (Periplus 11. 5). His experience was materially relevant, but he fails to adduce it. There is also ignorance of the most elementary facts of Armenian geography, which is astounding in a man who had worked and fought in Cappadocia and Lesser Armenia.¹⁸ There is a prima facie case for dating the work before the legateship. But Arrian seems ignorant of Rome also. His interest in the tyrannicide-group at Athens contrasts with his lack of interest in Lysippus' statue-group, which had been removed from Dium in 148 b.c. and transferred to Rome, where it became a noted monument, stationed outside the porticus Octaviae in the southern part of the Campus Martius (i 16. 4 n.). Arrian speaks as though the group were still at Dium—a startling oversight, if he had already held office in Rome. Even the location of the tyrannicide-group at Athens may be imprecise; Arrian seems to have placed it at the wrong point along the Sacred Way (cf. iii 16. 8 n.). If so, it is an error hard to credit in a permanent resident of Athens; it presumably derives from Arrian's faulty recollection of the procession of the initiates, which he will have participated in as a visitor to Athens. Finally the digression on Heracles in the West also contains inaccuracies and a surprising lack of knowledge about the antiquities of Gades, which must have been familiar to him after his proconsulate in Baetica (cf. ii 16. 4 n.). The argument from error and omission is cumulative, and it suggests a date of composition before 125, when Arrian was in his thirties.

There is one strongly autobiographical passage, where Arrian proudly conceals his name and career, claiming that his works are paramount. In the context he claims that he has no need to write his name, nor his homeland, nor the offices he has held εν τη εμαυτον. The reference to offices can only denote offices held in Arrian's homeland, Bithynia (cf. i 12. 5 n.); certainly not Rome, for there is no instance of Greek authors of the second century ad describing Rome as the πατριs of a citizen of eastern extraction. Italy was merely a land of temporary domicile.¹⁹ Now the omission of his senatorial cursus is important in the context. The point Arrian is making is most emphatic; his offices are of secondary importance to his literary themes. It would be most strange if he chose to omit the prestigious senatorial offices and concentrated on the much less distinguished local magistracies. The evidence as a whole suggests a relatively early date of composition. Arrian was a mature writer with a successful record of publication, but not primarily in the area of historical studies. What his earlier works were can only be guessed at, but there is a tradition of specifically philosophical works, portions of which have survived in the meteorological fragments preserved in Stobaeus. Arrian was specifically commemorated as a philosopher in his own lifetime (above p. 5) and it is perfectly possible that his notes on the Diatribae of Epictetus were followed by a series of philosophical dialogues and treatises. Arrian's interest in history, as he claims in his Alexander work and the later Cynegeticus, dates back to his extreme youth,²⁰ but he did not embark on historiography until he was a relatively experienced writer.

Arrian's work on Alexander was the third of a series of biographical monographs. Nothing is known of the earlier works on Dion and Timoleon, but both belonged to a background of philosophical history and exempla which might have rendered them congenial subjects. But there is more. All three individuals had been the subjects of biographies by Plutarch, and the biographies are so interconnected by cross-references that Mewaldt argued that the three pairs were issued simultaneously.²¹ The theory is plausible in itself and has found some acceptance.²² If so, it is hardly coincidental that Arrian produced three successive monographs on figures whose biographies appeared conjointly in Plutarch's publications; and, since Plutarch's biographies (at least the three in question) appeared before 116,²³ there is every possibility that they provided Arrian with his direct inspiration. It was an inspiration to emulate. Arrian harps constantly on the deficiencies of previous histories of Alexander, and, if he knew Plutarch's work, he included it in the general censure. The same will have applied to the lives of Dion and Timoleon. The appearance of Plutarch's work could have given Arrian the stimulus to improve upon it.²⁴ There is no direct influence in the work on Alexander. Arrian worked from different sources, used a different arrangement of material, and is never, it seems, influenced by Plutarch's style or phraseology. Plutarch seems merely to have supplied the theme and the element of literary rivalry. Arrian himself gave a similar stimulus to later writers. A certain Amyntianus dedicated a work on Alexander to Marcus Aurelius and he promised to give an account worthy of Alexander's deeds, language that irresistibly recalls Arrian's own boasts (cf. i 12. 4, vii 30. 3). The classic of one generation became something for the next to surpass.

Arrian's interest in the figure of Alexander certainly went deeper than mere literary emulation of his predecessors. The programmatic statement at i 12. 5 suggests that his involvement with his literary themes goes back at least to his youth. What is more, his generation had witnessed an increasing emphasis on the positive aspects of Alexander's figure. As is well known, the complex and contradictory personality of the Macedonian king made him an obvious source of examples for every type of moral theory; he was a bottle into which could be poured any and every vintage.²⁵ The character of the king was debated in the philosophical schools of the Hellenistic period, but the tradition is too sparse and ambiguous for any clear picture to emerge. There are modern theories of canonical and hostile Stoic or Peripatetic views of Alexander, but those views are largely constructs from scattered passages in later writers, predominantly Cicero, passages which are rarely, if ever, securely derived from a Hellenistic original.²⁶ It is safer to say that the actions of Alexander could be separated into widely different compartments, each compartment serving as an illustration for theories upon monarchy. Certainly there was a tradition of hostile criticism with its keynote Alexander's lack of moderation. Cicero could castigate Alexander for pride, cruelty, and lack of moderation and rate his humanitas far lower than that of his father, a view which Seneca continued more emphatically and consistently, regarding Alexander at best as the victim of his passions and at worst as the type of the cruel tyrant.²⁷ Certain episodes lent themselves to this stock portrait: the murder of Cleitus and execution of Callisthenes, the excess of grief at the death of Hephaestion, the adoption of Persian dress and court protocol, and the claims to be son of Ammon.²⁸ The result was a caricature, but a caricature designed to serve as an exemplum. There were of course other portraits. Alexander the world-conqueror was the symbol which inspired his military imitators—Pompey, Augustus, or Trajan; and in the hands of a Dio Chrysostom Alexander could be portrayed as a defender and emulator of the ideal Homeric kingship.²⁹ It was largely the literary form which determined the treatment for good or ill, and the same incident might be used in very different ways. The adoption of Persian dress could be used either as a sign of pride or immoderation or as an example of conscious stimulation of unity of feeling within the empire.³⁰ Dio Chrysostom varied his treatment of Alexander in his orations on kingship. In the fourth Alexander appears as a youthful interlocutor of Diogenes, basically sound but in need of Cynic deflation. In the second he is the defender and advocate of Homeric virtues, while in the first oration, a general treatise on the ideal king, he appears briefly as a type of the immoderate ruler (Orat. 1. 1). Even in Seneca, whose treatment of Alexander is almost uniformly hostile, there is an inconsistent anecdote: his trust in Philip the Acarnanian is praised as a classic example of moderation (De Ira ii 23. 2 f.).

The literary treatment of Alexander must have continued to be multifaceted and contradictory, but the literature which has survived from the late first and early second centuries is largely encomiastic. The young Plutarch produced two rhetorical speeches On the Fortune or Virtue of Alexander which are undiluted rhapsodies upon the perfection of the king; their object is rhetoric and there is no streak of dark in the entire picture.³¹ Similarly Dio Chrysostom had written an eight-volume treatise on the virtues of Alexander, which was on a much more extended scale than Plutarch's work, but no less encomiastic.³² It was a work which should have influenced his fellow countryman Arrian, and it may have stimulated him as much as did Plutarch's Life. The literary climate was encomiastic and so was the political climate. Trajan's personal interest in Alexander is well attested. He sacrificed to the Macedonian king in the ruins of Babylon and boasted to the senate that in 116 he had advanced further than Alexander.³³ Given such a background Arrian could hope for both official and public approval.

Arrian's emphasis is clear from the outset. It is Alexander's εργα, the res gestae of the reign, which primarily concern him; they have never been adequately commemorated in prose or verse (i 12. 2-4). Arrian is emphatically declaring that his function is to give a record of action and achievement; and there may be an implicit contrast with Plutarch's declaration at the beginning of his Life of Alexander, the statement that the magnitude of Alexander's deeds precludes an exhaustive enumeration of them (Al. 1. 1-2). Arrian, on the contrary, claims to give a full history of his hero and he makes it plain from the outset that its aim will be encomiastic. The stress is upon recording fact, and there is relatively little attempt to use the events of Alexander's life as material for moral discussion. Arrian is of course aware of the role of the historian as moral critic (vii 30. 3), and in Isocratean fashion he inveighs against Alexander at the standard points where his behaviour was used for philosophical exempla. The clearest example is at iv 7. 4, where Arrian is clearly shocked by the barbaric mutilation of Bessus and digresses to give a sermon on the necessity for moderation. Elsewhere he adverts naïvely on the incongruity of Alexander's revenge motive (iii 18. 12, vi 30. 1). But the criticism of Alexander himself is generally muted. When he discusses the Cleitus affair he strongly condemns Cleitus for his contumacy but pities Alexander for falling victim to anger and drunkenness, and he praises the depth of Alexander's repentance (iv 9. 1-2; cf. vii 29. 1). Similarly he condemns Callisthenes for his intransigence and agrees that Alexander had every reason for his hostility (iv 12. 6 f.). The criticism may be muted in other ways. Arrian condemns various aspects of Alexander's grief for Hephaestion (vii 14. 2-7), but he presents the offensive behaviour as reports from untrustworthy authorities, which he himself discounts. Again, Arrian cites with approval the gymnosophists' criticism of Alexander's lust for conquest (vii 1. 5-6), but he adds that Alexander himself approved the criticism, even though he could not resist his ambitions (vii 2. 1-2); and the passage begins with a rapturous description of Alexander's will to conquer, which tends to nullify the criticism. The positive side of Alexander's achievement is presented without qualification. Whereas he uses the unreliability of the sources to discount criticism, he can be quite sophistical where praise is involved. The story of the reception of Sisygambis might be apocryphal, but Alexander deserves praise for having inspired it (ii 12. 8). The cutting reply to Parmenion is similarly presented as a subsidiary report …, but Arrian gives no opinion about its authenticity and praises Alexander for his reasoning and his self-confidence (iii 10. 1-4). The final character study epitomizes the entire tendency of the work (vii 28-30). It begins with an impassioned eulogy, framed in superlatives, and then turns to the various statements of reprobation, which Arrian mitigates firstly by stressing the saving grace of repentance (vii 29. 1-2) and then by suggesting that both his claim to divine filiation and the adoption of Persian dress were sοφιsματα, stratagems to influence the subject peoples (vii 29. 3-4). Finally he says categorically that Alexander's faults are insignificant in comparison with his achievements, and even implies that the critic of Alexander is himself reprehensible (vii 30. 1).

The bias towards encomium explains much of Arrian's work. It is basically a narrative of achievement, with a favourable verdict built into the texture of the narrative (cf i 17. 12, ii 4. 11). Indeed one of the principal reasons why Arrian preferred Ptolemy and Aristobulus as narrative sources may well have been their very favourable attitude to the king (see below p. 30). It also explains the comparative dearth of moral comment. Arrian's verdict was presupposed, and the main function of his comments was to mitigate the criticisms traditionally levelled at the figure of Alexander. He could therefore devote himself to the composition of a laudatory account of Alexander's res gestae and concentrate on the stylistic presentation.

3. ARRIAN'S HISTORICAL METHODS

In the Praefatio Arrian supplies information about his use of sources which is almost unique in ancient works of non-contemporary history. That information concerns both his principles of selecting material from his sources and the reasons for his choice of two basic sources for his historical narrative. The principles of selection come first, and it is best to begin the discussion with them. Arrian states that where Ptolemy and Aristobulus give a consistent account he will follow their consensus as wholly true; where they disagree he will adopt the version he considers the more worthy of credence and the more memorable (praef. 1). One is therefore tempted to assume that Ptolemy and Aristobulus are considered of equal weight and used alternately. But it is clear that Ptolemy is the more important authority. At vi 2. 4 Arrian explicitly singles him out as his principal source …, and it seems that he placed him above Aristobulus in the same way that he places Ptolemy and Aristobulus together above the subsidiary sources (cf. v 7. 1, vii 15. 6). Passages like the account of the Danubian campaign seem extracted in toto from Ptolemy with no additional material added from other sources (cf. i 1. 4, i 4. 6 nn.). The narrative core of Arrian's history is generally agreed to be Ptolemy, but it remains controversial to what degree Aristobulus is used as a control source. There exist speculative analyses by Hermann Strasburger and Ernst Kornemann, but they depend on supposed general characteristics of the lost works, extrapolated from the scanty extant fragments.³⁴ There is rarely any concrete ground for a positive source attribution, and Arrian's text is usually a safer basis for reconstruction than one's impressions of a vanished original source.

Direct citations give some assistance. Like most historians of antiquity Arrian refers to his sources primarily in order to dissociate himself from some particular statement. He may signal a source-conflict by naming his authorities but he also marks out specific statements in the sources which strike him as suspicious by noting the author's name. In the early books it is almost always Ptolemy who is singled out in this way. His figures for the Macedonian casualties at the Lyginus (i 2. 7) and his statement that Perdiccas began the attack on Thebes (i 8. 1) are the only direct citations in Book i and in both cases the detail singled out is an integral part of the context. The surrounding narrative may be attributed to Ptolemy with reasonable confidence. There is no named reference to Aristobulus before Book ii, and, when he is adduced, it is usually to note some aspect of his account which is discrepant from the general tradition.³⁵ Few passages before the end of Book vi can be attributed to Aristobulus by the method of citation,³⁶ whereas whole chapters can be singled out for Ptolemy with relative ease. There is a good example in Book v, where Arrian mentions in passing that the Acesines was the only river whose breadth was noted by Ptolemy (v 20. 8); the narrative continues in an unbroken flow with the crossing of the Hydraotes and the siege of Sangala, culminating with an engagement led by Ptolemy himself (v 23. 7-24. 3). The combination of direct citation and autobiographical material led Jacoby to attribute the entire extended passage to Ptolemy (FGrH 138 F 35), and he was certainly right to do so.³⁷ There is some temptation to infer that Arrian followed the procedure Klotz attributed to Livy,³⁸ namely excerpting Ptolemy's narrative without comment and using Aristobulus as a control source. But Arrian's procedure is certainly more sophisticated. There are passages like the account of the journey to Siwah (iii 3. 3-6) which juxtapose material from Ptolemy and Aristobulus in a way that suggests that material from the two authors was blended in a composite narrative. It clearly made a difference whether an incident was recounted with similar but complementary details or retailed in two different contradictory versions. In the one case Arrian might use the two sources together, selecting material alternately as it appealed to his taste, but in the other he would be tempted to select one version and either suppress the other or note the variants.

Internal inconsistencies also help to isolate the contributions of the two sources. There are variations, and significant variations, in Arrian's use of geographical terms. The people of modern Sϊistan are termed Drangae or Zarangae, and the alternation of names twice helps to isolate the separate traditions in Arrian.³⁹ Similarly the Tapurians of the Elburz are referred to either as Τóπειροι or Ταπουροι (cf. iii 8. 4 n.), and the form Τóπειροι occurs in the Persian army list taken from Aristobulus (iii 11. 4). The different forms suggest different authors. There are also variations in the military terminology, which may also be indicators of the different sources. The battle of the Granicus is a particularly useful test case. The cavalry under the command of Amyntas son of Arrhabaeus is termed both πρóδρομοι and sαριssοφóροι in passages which are widely spaced,⁴⁰ and the account of the Macedonian battle line at i 14. 1-3 has the striking incongruity that the phalanx battalions are termed φαλαγγεs, not ταξειs, the term used in all the other major battle narratives. There is a strong presupposition to identify the source as Aristobulus, for there is no doubt that Arrian's principal source for the battle narratives was Ptolemy. But variations of this nature are infrequent and, when they occur, there is often no means of deciding which variant should be attributed to which author.

Citations in other authors are also a help. Strabo's excerpt of Ptolemy's account of the Danubian campaign confirms that the parallel passage of Arrian is also derived from Ptolemy (cf. i 4. 6 n.), and his excerpts from Aristobulus' version of events at Pasargadae and in Babylonia supplies excellent material for comparison with Arrian's text.⁴¹ But once again the available material is not extensive and there are dangers. It is only too easy to gloss over significant differences and make an attribution on general similarity of narrative. The digression on the Sardanapallus monument has been attributed to Aristobulus on the basis of the parallel descriptions in Athenaeus and Strabo even though the central feature of the description is markedly different (cf. ii 5. 3 n.). The truth is that Arrian's use of his major sources is intricate and varied, and there is no single reliable method to isolate their relative contributions. In the Commentary I have attempted as far as possible to refer specific passages to specific sources, but it must be admitted that the attributions are largely conjectural.

Arrian's use of his major sources is complex but his technique is far from infallible, as one would expect in the first extended work of his historical corpus. There are a number of errors resulting from inefficient conflation or insufficiently close reading of his sources.⁴² On occasion he accepts one tradition but inadvertently combines it with the variant he has rejected. The result can be perplexing. At iii 11. 9 Philippus intrudes as the patronymic of Amyntas son of Andromenes, an error which seems only explicable if Arrian had in mind the name of Philippus, son of Balacrus, who the variant tradition names as the commander of Amyntas' battalion at Gaugamela. Similarly the double placing of Craterus' battalion at the Granicus (i 14. 2-3) is best explained on the assumption that his two sources located it at different points in the battle line and Arrian unwittingly combined the two versions. Occasionally his sources are so divergent that he gives both versions independently without noticing that they refer to the same event. This may occur on a minor scale, as in the case of the double reference to the passage of Drangiana (iii 27. 4-28. 1), but there are also quite serious doublets where the same event is narrated twice at different points in the narrative. The arrival of Phrataphernes and Stasanor is narrated in the winters of both 329/8 and 328/7 (iv 7. 1, 18. 1), and the march of Craterus' column through central Iran is described at two different stages of the journey down the Indus (vi 15. 5, 17. 3).

The occurrence of doublets and conflations inevitably raises a wider question: how careful was Arrian in his use of primary sources? Since neither Ptolemy nor Aristobulus is an extant author, we have little scope for direct comparison. But there is another work, the Cynegeticus, where Arrian was operating with an extant source, the Cynegeticus of Xenophon, and the accuracy of his reproduction can there be checked.⁴³ It is quite clear that Arrian is capable of reproducing his original almost verbatim with only the most insignificant variations of wording (Cyneget. 30. 2; cf. Xen. 7. 3), and in the Alexander history he includes an extremely faithful quotation of Herodotus (iii 30. 8 = Hdt. iv 57). But at other times Arrian merely gives a free paraphrase.⁴⁴ He may distort the sense slightly by variation of expression—and the slightest change in word-order and construction can produce a genuine difference in sense.⁴⁵ What is more, he can superimpose ideas of his own upon his original; from his citation of the exordium of the περì ιππικηs (Cyneget. 1. 5) one would assume that Xenophon disclaimed all considerations of rivalry with Simon but wrote for reasons of general utility. Neither motive in fact occurs in the original. There are obvious repercussions for our assessment of Arrian's use of source material in the Alexander history. We can never be sure how true Arrian is to the wording of his original—and in the case of Ptolemy and Aristobulus he may well have allowed himself freer range than when he adapted the classical masters of prose style. More significantly, he may always alter the original meaning of his source by stylistic reinterpretation or superimpose his own views. The Alexander history is of course far longer than the Cynegeticus and the use of sources is more complex. Arrian is not commenting upon and adding to a single text, but working up a historical narrative from two major sources and a number of subsidiary works. The scope for variation is accordingly greater as is the mass of source material, and there are a number of instances where Arrian can plausibly be argued to have misunderstood his original. These slips range from relatively venial errors such as the placing of the Macedonian Olympia at Aegae instead of Dium (cf. i 11. 1 n.) to serious distortions of historical fact, such as (I believe) the statement that Alexander visisted Ecbatana in 330 (cf. iii 19. 5 n.). The conclusions must impose caution. We cannot usually separate the contributions of Ptolemy and Aristobulus with any degree of certainty. The traditions may be intertwined and intertwined in a misleading fashion. Nor can we be sure that Arrian even in direct citations reproduces the exact wording of his sources or that his reproduction of the original factual material is unerringly accurate. All this may well seem depressingly negative; but the tendency of modern historians is to regard Arrian as a direct reflecting mirror of Ptolemy and his own characteristics as historian and stylist have been relegated to the background—or, more often, ignored altogether. A corrective is both salutary and urgently needed.

Arrian also refers to his subsidiary sources in the Praefatio. There is material from historians other than Ptolemy and Aristobulus which strikes Arrian as memorable and not unconvincing and he has included it under the heading of tales told about Alexander. … The motif recurs repeatedly in the body of the narrative,⁴⁶ and it is tempting to infer that whenever an incident is introduced by a formula like λεγεται it is a detail extracted from Arrian's subsidiary sources. But, once again, Arrian's usage is more sophisticated than it first appears. The formula has a wider purpose, to mark out a detail which Arrian does not find wholly convincing and does not wish to relate on his own authority. Such detail may well come from his principal sources.⁴⁷ At vii 20. 1 he introduces the information about the gods of the Arabians with the sceptical formula λογοs δι κατεχει, but it is clear from Strabo (xvi 1. 11 (741)) that the material comes directly from Aristobulus. Again, it seems certain that the narrative of the Danubian campaign is taken directly from Ptolemy, yet it proceeds in oratio obliqua for some eight lines before reverting to direct narrative (i 1. 4-5). In this case the introductory λεγεται δο does not express scepticism; it seems to be a conventional introductory formula (cf. i 1. 1 n.). At other times the λεγουsι-formula is used in a general sense to indicate passages of general agreement and to prepare the way for variants. At ii 12. 3 he begins the story of the Persian captives with a general reference to the statements of some Alexander historians; at 12. 5 he identifies those sources as Ptolemy and Aristobulus and adds a logos from his subsidiary sources dealing with the confusion between Alexander and Hephaestion. Similarly at vi 24. 1-2 Arrian first refers to the general tradition … and then singles out Nearchus' unique story of rivalry with Cyrus and Semiramis. True logoi are relatively rare and they illustrate Arrian's methods of selection. The stories of Hephaestion's actions at the tomb of Patroclus pave the way to the logos of Alexander envying Achilles his Homer, which in turn leads naturally to Arrian's famous piece of self-advertisement (i 12. 1 with notes); the logoi are chosen to lead naturally to the justification for the history.⁴⁸ Again, the logos at ii 12. 6-8 is explicitly introduced to illustrate the virtues of Alexander and the story of Parmenion's advice to launch a night attack leads to an appreciation of Alexander's generalship (iii 10. 2-4). Other stories are included for pure sensationalism, like the Bacchic procession through Carmania, which Arrian is at pains to deny—in true Herodotean manner.⁴⁹

Arrian's procedure is, as always, flexible. In the Praefatio he says merely that he will present material from his subsidiary sources as legomena. That does not exclude his using oratio obliqua to present material from his principal sources. Indeed there is every reason to believe that Arrian followed Herodotus in referring to his source material, as he followed him generally in matters of style. Now Herodotus has a rich variety of constructions in oratio obliqua, designed to indicate reservations towards his subject matter.⁵⁰ His method centres on constructions with the infinitive; and infinitives may intrude in subordinate clauses, follow sequentially in a clause introduced by οτι/ωs, or even do duty for indicatives in narrative passages. Arrian uses the narrative intrusive infinitive, notably in the account of the happenings at Nysa, and the usage appears Herodotean.⁵¹ His deployment of intrusive infinitives is nowhere near as sophisticated as Herodotus', and often accusative and infinitive constructions alternate with οτι-clauses, with no motive apparent other than desire for variation.⁵² But the alternation is occasionally more complex. At iii 26. 1 he begins his account of the Philotas affair with the joint account of Ptolemy and Aristobulus, presented as a οτι-construction. But when he presents Ptolemy's story alone the construction switches to accusative and infinitive—with a single vivid clause in the indicative underlining the ground of Philotas' conviction (iii 26. 3). At iii 27. 1 the joint account resumes, beginning with accusative and infinitive, but changing to oratio recta after the first clause. The most elaborate instance is the extended passage beginning with the Cleitus affair (iv 8 ff.). Arrian first states that he will narrate the episode out of chronological sequence and plunges immediately into an extended construction in oratio obliqua, with a single outburst into direct speech introduced for dramatic effect (iv 8. 8). Now it is clear that the story is not taken in its entirety from Arrian's subsidiary sources, for he cites an apologetic variant from Aristobulus (iv 8. 9). The oratio obliqua occurs because Arrian wishes to distance himself from his material; it was a stock negative exemplum hardly congenial to his encomiastic purposes (above p. 14), and Arrian is reluctant to retail it as fact on his own authority. Accordingly he switches to an infinitival construction and distances himself from the narrative. We cannot therefore infer from the construction that the episode was not recounted by Ptolemy; it is perfectly possible that he gave the story much as Arrian tells it.⁵³

Arrian's citations of his source material are complex and varied, but there is no reason to doubt what the Praefatio implies, that the majority of his narrative was built upon the works of Ptolemy and Aristobulus. It remains to see how justifiable that choice was. Such an investigation properly requires detailed examination of all the extant fragments together with detailed study of the general characteristics of Arrian's narrative. That clearly lies outside the scope of an Introduction, and in what follows I can only distil the arguments expounded at greater length in the body of the Commentary.

Ptolemy's history is almost entirely known from citations in Arrian.⁵⁴ Arrian provides all but four of the fragments which Jacoby accepts as genuine; and of those four one is a piece of comic romance preserved by Synesius,⁵⁵ another is at best a second-hand citation by Stephanus of Byzantium, the third a passing reference to his figure for Alexander's invasion army, cited by Plutarch in a list of variants, and finally Strabo's brief reference to the story of the Celtic embassy.⁵⁶ It is a reasonable assumption that Ptolemy's work was ignored and largely unknown in antiquity. It was Arrian's service to exhume his work, and modern scholars have been unstinting in their gratitude. But, even so, the outlines of Ptolemy's work remain vague. We have no idea of the number of books or how the material was distributed. More seriously we have no information when Ptolemy wrote, except for Arrian's statement that the history appeared after Alexander's death. It is commonly believed that the work fell after 305/4, when Ptolemy assumed the title of king, but we cannot assume that Arrian's reference to him as a king is taken from Ptolemy's preface or is even strictly accurate for the time of writing. Arrian may have regarded him in timeless terms, the kingship a permanent characteristic. Nor can it seriously be alleged that Ptolemy had no time for serious writing until he abdicated in favour of his son in 285;⁵⁷ such an argument would make it a priori impossible for, say, Caesar to have written his Commentarii or Augustus his Autobiography. It would help if it could be proved that Ptolemy knew Aristobulus' work, for that author wrote after 301 in his old age (see below), but the hypothesis is at best unprovable. The key passage (v 14. 5) is not evidence for direct criticism of Aristobulus. Arrian merely places Aristobulus and Ptolemy side by side and (for once) explains his preference for Ptolemy; the formula αλλα … γαρ marks the return to Ptolemy's statement about the forces with Porus' son, after an anticipatory argument justifying the figures (and rejecting those of Aristobulus) on grounds of general probability.⁵⁸ Nothing seriously suggests a late date of composition, but there are certain positive and negative aspects of the work which suggest a period close to Alexander's death. The strongly autobiographical character of many of the attested fragments indicates a period when it was in Ptolemy's interest to emphasize his closeness to Alexander, and the well-known polemical bias against Perdiccas and his associates suggests a period not too far removed from Perdiccas' death during the invasion of Egypt in 321.⁵⁹

The characteristics of Ptolemy's narrative are equally hard to grasp. His traditional reputation rests largely on the assumption that he preserved in his narrative archival material derived from the Ephemerides or Court Journal of Alexander himself. This is a timehallowed theory⁶⁰ but impossible to sustain. The Ephemerides as such are cited in the sources six times in all—once for Alexander's hunting habits, three times for his excessive drinking, and most extensively in the reports of Alexander's last illness, which Arrian and Plutarch take from the Ephemerides with varied and sometimes inconsistent details.⁶¹ It seems most probable that Arrian's extract is in fact taken indirectly from Ptolemy and/or Aristobulus. He concludes his digest of the Ephemerides with the observation that Ptolemy and Aristobulus took their narratives no further (… vii 26. 3). The alternative interpretation, that their account was not substantially different from that of the Ephemerides, is most unlikely.⁶² Arrian is delaying mention of his principal sources (as at ii 12. 3-6); he begins with a reference to the account of the Ephemerides (vii 25. 1), repeats it at vii 26. 1, and then ends his report of the official version with a remark that Ptolemy and Aristobulus take the story no further. He can then add the vulgate story of the bequest ‘to the strongest’ (vii 26. 3). The account provided by the Ephemerides was recounted by Ptolemy (or Aristobulus) as the prime source for Alexander's death; but the document is not cited outside the context of the last illness and it is a reasonable assumption that the information it provided did not extend beyond the last months of Alexander's life.⁶³ But even if the Ephemerides were, as some have supposed, a complete day-to-day record of the official business of the reign, the fact that Plutarch refers to them proves that they were not unique to Ptolemy and the variants in his account suggest that they existed in several recensions—and were not necessarily an unimpeachable source of fact.

Nor is the so-called archival material unique to Ptolemy's narrative, as reproduced by Arrian. The ‘vulgate’ tradition of Diodorus and Curtius Rufus provides details of appointments, promotions, arrival of reinforcements, and reception of embassies which often corroborate the material in Arrian and not infrequently supplement it.⁶⁴ There seems to be a central core of information about the public business of Alexander's court which was available to all sources. It may stem originally from some document like a court journal, but there is no positive evidence for the supposition and equally nothing to controvert the suggestion that the ‘archival’ material was transmitted first by contemporary historians like Callisthenes, recording events from their own experience, and then used by subsequent writers as basic fact. Ptolemy's only advantage over the other sources is that he was closer to Alexander and presumably had more direct experience of events at court, but it remains conjectural whether his memories and files were any more accurate or detailed than Callisthenes' contemporary account of events as they happened. Perhaps the most seductive (and certainly the most fallacious) argument is based on the ‘diary-like’ style of Arrian, which is thought to reflect—at two removes—the dry, factual record of the original court journal. But, if anything is certain about Arrian as a writer, it is that his style is his own, and, if his narrative has lucid clarity and an apparent diary-style, that is a carefully contrived result and contrived by Arrian himself. There are passages in Thucydides which are as dry and detailed as any in Arrian, but nobody would claim (I hope) that the style is based on some official chronicle of the war which he copied out.⁶⁵

There is another hypothesis that Ptolemy wrote ‘to set the record straight’, that is, to correct the extravagant fantasies of earlier writers by presenting a sober record of fact. Once more there are some serious difficulties. In the first place there is no evidence that Ptolemy indulged in polemics against other writers. The two passages of Arrian which might be adduced under this heading are the criticism of Aristobulus' figure for the forces with Porus' son (v 14. 3-6) and the criticism of the vulgate tradition that Ptolemy saved Alexander's life at the Malli town (vi 11. 7-8), but in both cases the criticism is Arrian's own; he places Ptolemy's statement against the other tradition and uses it in an argument from probability. If, then, Ptolemy's critique of Alexander historians was not explicit, it was implicit, leaving his readers to draw their own conclusions from his unvarnished record of fact.⁶⁶ Such a hypothesis is difficult either to support or subvert. If Ptolemy's narrative differs from the rest of the tradition it is indeed possible to argue that he was attempting to correct it—but it is also possible that the majority tradition is correct and that Ptolemy had motives for altering the record. The question can only be answered by continuous detailed analysis of the tradition such as attempted in the Commentary, and even then the results are ambiguous. There are cases like the treatment of Parmenion (cf. i 13. 2, iii 15. 1 nn.) where Ptolemy does seem to be implicitly rebutting some aspect of the received tradition, but there is not sufficient evidence to single out such implicit criticism as a primary reason for his writing. More often when his version is discrepant there appear to be personal or propagandist motives at work. Nor can it be seriously argued that Ptolemy reacted against the romantic, novel-like aspects of earlier writers. It seems difficult to maintain that he avoided all excursuses on natural history, for the Elder Pliny names him alongside such authors as Onesicritus and Ephippus as an authority on exotic types of trees.⁶⁷ His account of Issus included the romantic picture of a ravine choked with enemy dead (cf. ii 11. 8 n.), and it is difficult to accept his story of guidance to Siwah by snakes as anything more than romantic fiction (iii 3. 5 n.).

What, then, can be deduced about the characteristics of Ptolemy's history? In the first place it was strongly autobiographical. Ptolemy laid great emphasis on his services under Alexander and wrote them up in vivid and elaborate detail, particularly in the section dealing with the Indian expedition.⁶⁸ What is more, he certainly exaggerated his role, particularly in his earlier appearances; his command at the Persian Gates is apparently distorted in respect of its importance and function (iii 18. 9 n.) and his role in the capture of Bessus is built up into a full-scale repetition of Alexander's pursuit of Darius (iii 30. 5 n.). Such an emphasis is only natural in a survivor of the campaigns (one need only think of Nearchus);⁶⁹ and it had an immediate purpose in the period of the Successors, when prestige was enhanced and positions were cemented by reference to services rendered under Alexander.⁷⁰ But natural though the emphasis is, it should put us on our guard against the idea that Ptolemy is an unsullied repository of historical fact. There are also occasions where a hostile bias can be demonstrated. Ptolemy's treatment of Perdiccas is particularly vulnerable in this respect. Perdiccas' achievements are omitted (iii 15. 2 n.), particularly his promotion in 324 to the vacant chiliarchy of Hephaestion, and there is a degree of calumny, notably in the allegation that the Macedonian reverse at Thebes was caused by an unauthorized attack on his part (i 8. 1 n.). The same bias can be detected in Ptolemy's treatment of the lieutenants of Perdiccas, particularly Aristonous who is almost totally ignored in Arrian⁷¹ and the younger sons of Andromenes, who are portrayed in a most invidious light (cf. iii 11. 9, 14. 5, 27. 2 nn.). Antigonus' achievements in Phrygia are also omitted, but in this case there is no detectable malice (cf. i 29. 3 n.). Once again the bias has an immediate explanation in the period after Alexander's death; Ptolemy was suppressing services rendered by his political enemies and suggesting that their role was harmful. But the bias is again an obvious warning against accepting every detail from Ptolemy as the literal truth. The assessment of the rest of the narrative is a more intricate problem, since what we have is an amalgam of Ptolemy and Aristobulus; but that amalgam has certain general characteristics which can be analysed and appreciated. The appreciation must, however, follow consideration of Arrian's use of Aristobulus.⁷²

Aristobulus' work appeared relatively late, after the battle of Ipsus in 301 (vii 18. 5); and according to the pseudo-Lucianic Macrobioi he stated in his preface that he began writing in his 84th year.⁷³ He is also described as a citizen of Cassandreia,⁷⁴ founded in 316 b.c. on the site of Potidaea, and it seems very probable that he wrote in Macedonia after the death of Cassander (298). Details of his life are very sparse, and we know only that he served with Alexander. In what function is uncertain. The supposition that he was an architect and/or engineer⁷⁵ rests on his commission to restore the pillaged tomb of Cyrus (vi 29. 4-11; cf. Strabo xv 3. 7 (730)), but that was largely a cosmetic operation; the only feature requiring anything like engineering skills was the sealing of the door (cf. vi 29. 10). His other commission in India is totally obscure (Strabo xv 1. 19 (693)). All that can be said is that Aristobulus served as an under-officer, close enough to Alexander to receive direct commissions but outside the immediate court entourage.

In antiquity Aristobulus was much better known than Ptolemy. Outside Arrian there are citations in Plutarch, Athenaeus, and Strabo. Strabo in particular seems to have used him as a primary source for his description of India. Even so the content of his work is most imperfectly known. No book-number survives and there is no indication how the narrative was distributed. Various characteristics appear in the surviving fragments, but it is conjectural how prominent they were in the non-extant parts of the work; Aristobulus is cited by name for divergent or sensational details, and it is unsafe to extrapolate from them characteristics of the whole work. There seems an interest in botanical curiosities. Aristobulus described the oaks of Hyrcania (F 19), the asafoetida of the Hindu Kush (iii 28. 5-7 = F 23) and the plants of the Gedrosian desert (vi 22. 4-8 = F 49). Rivers are also prominent in the extant fragments. Aristobulus noted the course of the Oxus (F 20), the phenomenon of the disappearing rivers of Central Asia (F 28), the monsoon floods in the Punjab (F 35), the fauna of the Nile (F 39), and the drainage system of Mesopotamia (vii 21; cf. F 56). Such preoccupations were not unique to Aristobulus; Onesicritus gave a detailed (and inaccurate) description of the banyan trees of India⁷⁶ and the monsoons are a common feature of Alexander historians. But the citations of Aristobulus' data about natural history come from a variety of sources, both from Strabo and Arrian; and botanical and geographical excursuses are likely to have been a recurrent feature of his work. We can go further. Some of Aristobulus' information is given by other sources in different contexts. Aristobulus mentioned asafoetida as a natural curiosity of the Hindu Kush, but another source, excerpted by Strabo (xv 2. 10 (725)), includes a description of the plant in an account of the hardships suffered during the transit of the passes (cf. iii 28. 6 n.). Similarly the description of the monsoon rains is presented as a catalogue of natural phenomena with no attempt to trace their effect (which was devastating) upon the Macedonian troops.⁷⁷ Again Aristobulus' description of the flora of the Gedrosian desert is a bland survey of curiosities—even the giant thorns being portrayed as more comic than dangerous (vi 22. 7-8), whereas in Strabo's parallel account (probably from Nearchus) the desert plants are described as poisonous, contributing to the fearful hardships suffered during the desert journey.⁷⁸ There seems a tendency to shy away from descriptions of hardships and to concentrate on the curiosities of the march. That tendency was not unique to Aristobulus. One of Hermann Strasburger's great services has been to call attention to a series of reports of hardships (‘Strapazenberichte’) which are totally ignored by Arrian,⁷⁹ and the most obvious and plausible explanation is that they were omitted by both his primary sources. Stories of suffering and casualties on the march did not apparently suit the taste either of Ptolemy or of Aristobulus.

An associated tendency of Aristobulus' work is his inclination towards apologetic, to exculpate or mitigate aspects of Alexander's behaviour which were susceptible to criticism. This tendency has often been noted, perhaps too emphatically;⁸⁰ we know too little of what happened at Gordium or at the Cydnus to be sure that his variant stories are apologetic in nature (cf. ii 3. 7, 4. 7 nn.). But some things are indisputable. Aristobulus unequivocally made Cleitus alone responsible for his murder (iv 8. 9) and denied that Callisthenes was ever executed (iv 14. 3). He also reacted against the contemporary allegations that Alexander drank himself to death and maintained that the king did not drink excessively but kept late hours for conversation's sake.⁸¹ This apologetic tendency was noted in antiquity. Some of the references are of late and weak authority and may be discarded,⁸² but it is hard to see how Lucian's story of Alexander's rebuke to Aristobulus (de hist. conscr. 12=T 4) could have taken shape unless his history was commonly thought to be a highly flattering one.⁸³ Clearly Arrian was aware of the reputation of his source but he thought the bias justified. Because Aristobulus wrote after Alexander's death he wrote what he did without fear or favour; if, then, his work has a bias towards eulogy, it is a bias justified by the facts. The argument may be naïve and simplistic, but it does not disprove the view that Aristobulus' work was adulatory and apologetic. Again Ptolemy shared the tendency, to some degree at least. The tendency to exculpate is sharply illustrated by their common account of the trial of Philotas, which both treat as a clear-cut case of treason and even, it seems, adduce admissions extracted by torture as historical fact (iii 26. 1 n.); and they state unequivocally that Callisthenes was involved in the Pages' conspiracy (iv 14. 1).

There was clearly much material common to both Ptolemy and Aristobulus. Their joint account is sometimes referred to, as are joint omissions (cf. ii 12. 6, iii 26. 1, iv 14. 1, v 7. 1, vi 11. 5, vi 28. 2, vii 13. 3). Occasionally variants are cited which are so trivial that they suggest that the two accounts were otherwise unanimous (cf. iv 3. 5, v 20. 2). It has even been argued that Aristobulus drew upon Ptolemy,⁸⁴ but it is hard to see why he should have followed the same general lines and chosen the most trivial points for disagreement. It seems more probable that both authors followed a common tradition which diverged in detail. Arrian may even have chosen to report only those variants which he was unable to decide between and in other cases made his own selection without noting the variants. The disagreements may well have been wider and more radical than the text of Arrian implies. But even so, the narrative of Arrian has certain general characteristics which were certainly common to both sources. Arrian's narrative can be placed alongside the so-called vulgate tradition, the tradition common to Diodorus xvii, Curtius Rufus, and Justin, which in all probability derives from the first-generation historian Cleitarchus;⁸⁵ and from critical comparison of the two traditions various traits emerge.

Firstly, it is undeniable that there is a strong encomiastic trait in Arrian, a trait that depicts both the king and his army as invincible and virtually superhuman. The campaign narrative reads as an unbroken catalogue of success, and contrasts sharply with the more chequered record of the vulgate. There is little or no record of setback and losses—and what defeats occur tend to be attributed to others. The reverse outside Thebes is laid at the door of Perdiccas (i 8. 1) and the disaster in Sogdiana attributed to the incompetence of other officers (iv 5. 7-6. 2). Alexander is uniformly successful, except in the first assault at Tyre (ii 22. 6-7 with notes) and the first advance on the Persian Gates (iii 18. 3 n.), and these comparative failures are given the briefest coverage, with no suggestion that casualties were widely sustained. In the large-scale siege descriptions, particularly those of Halicarnassus, Tyre, and Gaza, the selection of episodes is overwhelmingly biased towards reports of Macedonian successes, particularly those led and inspired by Alexander himself,⁸⁶ and there are significant instances where reverses sustained are transformed into victories for the attackers.⁸⁷ They are not merely victories but effortless victories. There is an additional tendency. Alexander is made to accept every challenge offered him. At the Granicus he accepts at once the challenge of the Persian defence along the river bank (i 13. 2 n.) and at Issus the whole tenor of the narrative is that Alexander moved at high speed to attack the Persians on their chosen position at Sochi; there is no hint of the rival tradition that Alexander deliberately forced the Persians into his chosen ground in the coastal defiles (ii 6. 1 n.). Natural obstacles also presented challenges to be overcome, and the motif is epitomized in the comment about the ‘impregnable’ position of Gaza (… ii 26. 3; cf. iv 21. 3, 28. 3-4). There are also exaggerations of the natural difficulties, notably the absurd figures for the height of the walls of Tyre (ii 21. 4 n.) and the description of the citadel at Celaenae (i 29. 1 n.). Alexander conquers against all obstacles and conquers effortlessly with the favour of the gods. That tendency is perhaps best illustrated by the brief narrative of the passage of Mt. Climax (i 26. 2 n.), but the presentation of Arrian as the favourite of heaven is recurrent and consistent (cf. i 17. 6, ii 3. 8, 6. 6, 7. 3, 14. 7, iii 3. 4). The king fought with the gods on his side and was naturally invincible, as the casualty figures strikingly demonstrate. Nowhere else in the extant sources is so much carnage wrought upon the enemy at such trivial cost as in Arrian's narrative (cf. i 16. 4, ii 11. 8, 24. 4, iii 15. 6 with notes).

The encomiastic aspect of Arrian's narrative is patent, as are the distortions it produced, and it seems that much of the tradition was already provided by Alexander's first historian, Callisthenes of Olynthus.⁸⁸ Callisthenes' work was notoriously eulogistic and flattering to the king, and some of the known features of his work recur in Arrian. His description of the phenomena encountered in the journey to Siwah was repeated more or less completely by Ptolemy and Aristobulus (iii 3. 4 n.). More significantly Arrian's account of Issus not only follows the main lines of Callisthenes' narrative (in so far as it can be restored from Polybius' critique) but reproduces and even improves upon the encomiastic distortions which it contained (ii 10. 1, 11. 8nn.). The motif of divine favour is also likely to derive from Callisthenes, who depicted the Pamphylian sea offering proskynesis to Alexander (F 31; cf. i 26. 2n.), portrayed him as son of Zeus (F 14 a), and reported his prayer at Gaugamela—which was duly answered (F 36=Plut. Al. 33. 1-2). The eulogistic bias is entirely explicable, since Callisthenes was broadcasting the king's achievements from court, and the emphasis upon invincibility had an obvious propaganda value at a time when there was serious unrest in the Greek world. Nor is it surprising that Ptolemy used and improved upon the distortions in Callisthenes. He was the custodian of the mummified body of Alexander, and his regime dependent upon Alexander's right of conquest. Not only that. Survivors of Alexander's campaigns were well represented in his armies, and he had some interest in stressing that they had been invincible.⁸⁹ Aristobulus' bias is less easy to explain. It is usually argued that he was influenced by genuine friendship and admiration for Alexander, and that may be true. If so, his experiences under Cassander may have had some effect. Cassander was bitterly hostile to the memory of Alexander, annihilated his family, and ceremoniously restored Thebes.⁹⁰ The negative aspects of Alexander's reign will have been stressed in the propaganda of his reign, and Aristobulus may well have been left with an obsession to correct the emphasis at all costs. If he were looking back at Cassander, his tendency to apologetic is perfectly explicable.

It would be wrong to claim that the tradition of Ptolemy and Aristobulus is worthless. As one would expect in the writings of first-generation historians with access to contemporary materials, there is a great deal of uncontroversial factual material—details of itineraries, embassies, appointments, and the like. But the information is not unique; the vulgate tradition often confirms and sometimes corrects. Above all there is a constant bias to encomium which needs the constant corrective of the vulgate tradition. It therefore needs to be asked why Arrian chose precisely these two sources and did not adduce other contemporary historians, notably Cleitarchus who also participated in the campaigns (it seems) and wrote after the king's death. His work is usually stigmatized as rhetorical and romantic, and Arrian has received many compliments for his preference for the supposedly sober and factual narrative of Ptolemy. But it is now clear that Ptolemy's narrative was as romantic as any other—and Arrian was far from averse to romance.⁹¹ His account of Parthian pre-history is both sensational and romantic and contrasts most unfavourably with the alternative tradition preserved in Strabo and Justin/Trogus.⁹² It is more likely to have been the over-all view of Alexander which attracted him. His task was to commemorate Alexander's res gestae, and the material provided by Ptolemy and Aristobulus, a virtually unbroken catalogue of success, was particularly adapted to his purposes.

Arrian did occasionally use sources other than Ptolemy and Aristobulus in his narrative portions, particularly in the sections on India which overlap his work in the Indica. His geographical material seems taken from Eratosthenes (cf. v 3. 1-4, v 5. 1, 6. 2), but significantly Eratosthenes seems not to have been used before Book v, for Arrian uses the new material to correct an earlier inaccuracy (cf. iii 28. 5n.). Nearchus is also used from the beginning of Book vi both as a narrative source and a control source.⁹³ An appreciation of his material is out of place here and must be reserved for the second volume, but once again it must be stressed that there is no trace of Nearchus before the Indus journey. Arrian only used the portions of his work which were relevant for the Indica.

The sources for the logoi are wholly uncertain. Arrian only once mentions specific authors; the supposed Roman embassy is reported according to the accounts of Aristus and Asclepiades (vii 15. 5). Both authors are obscure, Asclepiades totally so. Aristus was known to Strabo as a historian much later than Aristobulus and Onesicritus, resident in Salamis in Cyprus.⁹⁴ A handful of citations survive, none of any significance—apart from the report of the Roman embassy, and nothing can be inferred of the content of his work—or even of its date, beyond the fact that it appeared before Strabo and long after Alexander's reign. The possibilities remain open. Arrian could have selected the accounts of Aristus and Asclepiades because they gave the fullest version of the Romans' reception. Alternatively he could have used a later work which singled out Aristus and Asclepiades as sources. The latter is probably the case. There is little trace of intensive investigation of sources elsewhere in Arrian, and he seems to imply, wrongly, that only Aristus and Asclepiades referred to a Roman embassy; had he researched the matter closely he must have found the reports in Cleitarchus which were known to Pliny.⁹⁵ Similarly the subsidiary report of the introduction of proskynesis runs exactly parallel in substance and language to Plutarch's narrative excerpted from Chares of Mitylene.⁹⁶ Plutarch is not used directly, for Arrian has a little additional detail;⁹⁷ but the similarity of wording is so close that we must accept that the two authors either used Chares direct or drew on precisely the same intermediary source. There is no other valid evidence. The rest of the logoi have little in common. Some are frequently attested throughout the tradition (e.g. ii 12. 6f., iii 10. 1), while others reflect a variant or even unique version (e.g. i 12. 1, iii 2. 1). The variety of sources at Arrian's disposal was far greater than the number known to us, whether extant or attested, and he could logically have used any of them in any combination. All that we can say with any confidence is that logoi are not likely to have been taken from a large number of direct sources.

Arrian's qualities as a historian can be called into question both in his selection of sources and in his use of them. It may also be doubted how extensive his historical knowledge was. He had none of the encyclopaedic erudition of Polybius, and when he leaves the narrow confines of Alexander and his chosen sources the knowledge displayed is at best superficial. The excursus on the fall of Thebes is based almost entirely on the two great classics, Thucydides and Xenophon; and Arrian's presentation is neither wholly accurate nor wholly lucid (cf. i 9. 3nn.). Whatever originally stood in his sources, the double attribution of the Peace of Antalcidas to the reign of ‘Darius’ shows scant appreciation of the facts of fourth century history (cf. ii 1. 4n.). His final character-study of Darius shows total ignorance of his reign, where its history did not impinge on the actions of Alexander (iii 22. 2n.), and refers casually to the battle ‘at Arbela’, despite the fact that he devotes an excursus to correcting that very error (iii 8. 7, 22. 4nn.). The combination of carelessness and lack of erudition is startling in a writer who has been regularly praised for his reliability and who himself claimed to be composing a definitive factual history (cf. vi 11. 2, 8).

4. ARRIAN'S STYLE AND POPULARITY

Arrian prides himself on his style above all, and he claims that the success of his history of Alexander will win him primacy in the sphere of Greek letters (i 12. 5nn.; cf. praef. 3). Half a millennium later his claims were fully justified by Photius, who ended his précis of the History of the Successors with a glowing appreciation of his style (cod. 92: 72^b40ff. = Roos, Arriani Scripta Minora lxvif.): ‘this man is inferior to none of the best writers of history’. He is praised for his conciseness and avoidance of newly coined vocabulary. Photius admires his skilful use of figures to vary his narrative and states explicitly that it is in the composition of the narrative that his innovations lie. The result of his writing, he thinks, is a happy mixture of stylistic virtuosity and conventional vocabulary, which produces an impression of lucid clarity. That is fair comment. Anyone who examines an extensive passage of Arrian's narrative, such as the review of the armies before Gaugamela (iii 11. 3-12. 5), will be immediately struck by the complex variation of word-order, construction, and vocabulary. Even in the most lengthy catalogue the construction is so subtly arranged that it is rare for two consecutive clauses to match each other in language. What in particular impresses Photius is Arrian's use of ellipse; he does not omit periods, rather individual words—and it is stylistically impossible to provide a supplement. Modern historians, more concerned with fact than style, might be less charitable in their reactions to Arrian's ellipses, which can pose baffling problems of interpretation,⁹⁸ but Photius' judgement remains a valuable corrective. Arrian was no slavish copier of sources, as modern scholars have tended to suppose, but a very expert and sophisticated stylist in his own right.

Arrian's prose is a very personal and artificial creation. It has been described as Attic, but it is far removed from the academic Atticism of the second century a.d. There is a strong trend towards archaism; Arrian uses compounds in ξυν-, the reflexive pronouns ον and sφειs, and the archaic Attic plural-ηs.⁹⁹ Some usages contravene the strict canons of Attic prose; αιχμαλωτιsθεsαν was a notorious inelegancy (cf. iii 22. 4n.), καιτοι is used with the participle instead of καπερ, and πρίν with the infinitive occurs after a negative main clause.¹⁰⁰ Significantly the last two usages, like many other departures from strict Attic, have parallels in Herodotus, and it cannot be too strongly emphasized that Arrian's chief stylistic debt is to Herodotus.¹⁰¹ Most of the particular characteristics of the Ionian historian are reproduced in Arrian's style. There is the same superfluity of expression,¹⁰² the same use of epanalepsis,¹⁰³ the same use of resumptive clauses at the end of key stages of the narrative.¹⁰⁴ There is the same looseness of construction, which Demetrius characterized as λεξιs ειρομενη; clauses are strung together by connecting particles with little or no binding from subordinate clauses¹⁰⁵ and anacoloutha occur frequently, the construction often changing in mid sentence.¹⁰⁶ Equally striking is a wide range of vocabulary shared between the two authors and often not attested elsewhere in Greek prose usage.¹⁰⁷ One can speak of Arrian's style as literary reconstruction of Herodotus. The borrowing is of course explicit in the Indica where Arrian resurrects the Ionic dialect but it is implicit throughout most of his other work; the Ionicisms are eliminated but the style is distinctively Herodotean.

Arrian has debts to other authors, not least Thucydides. There are repeated echoes of Thucydides, particularly in speeches or historical excursuses. The comment on the fall of Thebes is strongly Thucydidean in phrasing; not surprisingly, since most of the examples chosen are based on his work (i 9. 1-3 with notes). But Thucydides can appear less intrusively; the debate at Miletus (i 18. 6-9) contains reflections of the speeches before the battle in the Great Harbour,¹⁰⁸ and Alexander's speech before Gaugamela is virtually a piece of free composition in Thucydidean style (cf. iii 9. 6n.). Key phrases from Thucydides also occur at appropriate points in Arrian's work (cf. i 12. 2 and vi 11. 8 with Thuc. i 97. 2). But Thucydides is a more occasional influence than Herodotus, who is an all-pervasive influence on style and vocabulary, and, perhaps surprisingly, Xenophon is even less obtrusive. Admittedly the opening of the narrative proper seems modelled upon the Cyropaedia (cf. i 1. 1 n.), but the stylistic debt to Xenophon is not pronounced. His characteristic use of και with a demonstrative pronoun referring to the antecedent clause occurs frequently enough in Arrian¹⁰⁹ as do his temporal conjunctions … and the repetition of a subordinate verb as the main verb of the following clause.¹¹⁰ There is also some imitation of Xenophon's vocabulary,¹¹¹ but at a much lower level than his imitation of Herodotus.

These remarks might give the impression that Arrian's style is a pastiche. That would be unfortunate. Arrian had an artificial style, it is true, and the elements which constitute it can be isolated, but his work displays none of the flagrant parody such as is found in the passages of the ‘histories’ of Verus' Parthian Wars which are stigmatized by Lucian.¹¹² On the contrary, his style is unitary, and, even though it is dominated by Herodotus, the effect is one of re-creation not crude borrowing. It is, if anything, a tribute to him that modern historians have so often been misled by the seeming simplicity of the style, and treated his carefully constructed narrative as an excerpt at two removes from Alexander's archives. Arrian might not enjoy the pre-eminence in Greek prose writing that he and Photius claim for his work, but he is without a doubt a most accomplished and sophisticated stylist and he should be treated as such.

Arrian regarded his work as definitive (vi 11. 2), and he seems to have been regarded as a model historian in antiquity. Lucian refers to him as a precedent for his attempt to write up the career of Alexander of Abonuteichos:¹¹³ even Arrian, a leading Roman and a life-long devotee of culture, wrote a history of the brigand Tillorobus. Nothing is known of this work, but, since the brigand in question ranged over Mysia and Mt. Ida, there is a fair probability that it formed part of the Bithyniaca or was issued as a supplementary opuscule like the Indica.¹¹⁴ Lucian's formal justification lies in the rank and literary prestige of Arrian, but there are wider implications. He begins his work with a contrast between Alexander of Abonuteichos and Alexander son of Philip; the one was as pre-eminent in vice as the other in virtue. The reference to Alexander no doubt evoked the name of Arrian, and, Lucian implies, if the leading historian of that paragon of virtue could stoop to the biography of a brigand, there was every justification for his description of the career of the prophet of Abonuteichos. Lucian's remark in its context is ample evidence of Arrian's prestige immediately after his lifetime, as is the fact that Appian drew directly upon him for material on Alexander and the mysterious Amyntianus parodied his claims to literary supremacy.¹¹⁵ So far the evidence deals with the reign of Marcus Aurelius and its immediate sequel. But Arrian's reputation remained high in the third century. His fellow countryman, Claudius Cassius Dio Cocceianus,¹¹⁶ wrote a life of ‘Arrian the philosopher’, commemorated his achievements in Cappadocia, and used his Parthica as the primary source for Trajan's Parthian Wars.¹¹⁷ That emphasis was to be expected in a native of Bithynia, but Arrian's prestige stood high elsewhere, notably in his adopted city, Athens. About the middle of the third century the Athenian historian P. Herennius Dexippus wrote a history of events after Alexander's death. Photius gives a summary of his version of the Babylon settlement, which follows very closely his summary of Arrian and adds the note that the rest of the work likewise agreed largely with Arrian.¹¹⁸ Dexippus clearly used his History of the Successors as a primary source for his own work; it must have been the definitive treatment of the subject.

In the fourth century the influence of Arrian is less explicit but it is none the less detectable. Themistius writing in 384 was well acquainted with the facts of Arrian's career and adduces him as a celebrated example of a philosopher active in public life.¹¹⁹ More importantly there exists from the time of Constantius II (340) an anonymous tract named the Itinerarium Alexandri.¹²⁰ It was composed at the beginning of Constantius' Persian expedition and purports to give the itinerarium of two earlier conquerors, Alexander and Trajan. The portion relating to Trajan has been lost, and what remains concerns Alexander alone. It is a brief and scrappy history, carelessly written and, despite the author's protestations, very rhetorically presented. But it is clear that the bulk of it (chs. 16-109) is extracted almost wholly from the first four books of Arrian's history of Alexander (thereafter the Itinerarium reverts to the tradition of the Alexander Romance). The selection of material is capricious and sporadic but there is no doubt that Arrian was used, however carelessly, as a direct source (cf. Itin. 18 with i 11. 3-5; Itin. 42 with ii 20. 1-2; Itin. 48 with iii 1. 1-5). Indeed there are peculiar additions which suggest that the text used was different from that of our archetype (see the notes to ii 14. 3, iii 2. 1). But, whatever minor divergences occur, the main fact is clear that the Itinerarium is a derivative of Arrian, and the author makes it clear that he has followed the sources whose reputation traditionally stood highest.¹²¹ It seems unavoidable that the reputation of Arrian and his Alexander history was undiminished in the fourth century.

Arrian remained a classic in the Byzantine period.¹²² Stephanus drew on his historical works, particularly the Bithyniaca and the Parthica, but also to a lesser degree the Alexander history. In the mid ninth century Photius read the Alexander history, the Bithyniaca, the Parthica, and the History of the Successors, and, as we have seen, placed Arrian in the first rank of historians. Arrian's works were a mine for the compilers of florilegia and commentaries. The ‘Suda’ refers repeatedly to all the historical works (with the exception of the Bithyniaca), and the Parthica is largely reconstructed from the fragments so provided. The same liberal use of Arrian persists as late as Eustathius in the latter part of the twelfth century ad; Eustathius drew upon the Bithyniaca and to a lesser extent the Alexander history in his commentaries on the Iliad and Odyssey and Dionysius Periegetes. However harshly we may judge Arrian's talents as a historian, his works were undoubtedly used and admired as classics for a millennium after his death. …

Notes

1. Phot. Bibl. cod. 58: 17^a24; cod. 91: 67^b23; cod. 93: 73^b12. See also Eustath. ad Dion. Per. 907, 976; Anecd. Bekkeri i p. 129. 27; Schol. ad Procl. Comm. in Timaeum (i 469 Diehl). …

2. Arr. v 6. 8; vi 16. 5. Cf. Phot. cod. 91: 68^b1; Steph. Byz. s.v. Μαssακα.

3. i 12. 3, ii 7. 8, ii 8. 11 (see my notes ad locc.), vii 13. 4. Herodotus is far more to the forefront of his work.

4. See the excellent discussion by P. A. Stadter, GRBS xvii (1976) 157-67. In the Alexander history the echoes of phraseology are confined to Herodotus (cf. iii 30. 8 with Hdt. iv 57).

5. Cf. Roos, Arriani Scripta Minora xii; Stadter, GRBS viii (1967) 157.

6. The locus classicus for the late dating is E. Schwartz, RE ii 1230-6. See also G. Wirth, Historia xiii (1964) 209-45; Studii Clasice xvi (1974) 188-209, esp. 199f.; E. Bowie, Past and Present xlvi (1970) 24-7. For the early date see F. Reuss, Rh. Mus. liv (1899) 446-65; A. B. Bosworth, CQ xxii (1972) 163-85, and ‘Arrian and Rome’, section iv.

7. i. 12. 2-5 (see the notes ad loc.); cf. praef. 3, vi 11. 2; vii 30. 3.

8. Phot. Bibl. cod. 93: 73^a32ff. = Bithyniaca F 1. For the interpretation see Bosworth, CQ xxii (1972) 178-80 (contra Wirth, Studii Clasice xvi (1974) 199 n. 128).

9. …

10. Parthica F 60-3 (Roos) = vi 3. 3. The alternative suggestion (Wirth, Philologus cvii (1963) 298), that Arrian excerpted both passages from Nearchus, gives too little credit for stylistic originality.

11. Parthica F 19 = vii 28. 1-3.

12. For the supposed echoes see C. E. Gleye, Philologus liii (1894) 442-8; Wirth, Historia xiii (1964) 232-45.

13. App. Syr. 56. 285-91 = vii 22. 5; App. BC ii 152. 639f. = vii 18. 1-4; BC ii 153. 642-44 = vii 16. 5-17. 2; 21. 1-22. 1. Cf. Bosworth, CQ xxii (1972) 176-8.

14. Parthica F 1. 2-3 (Roos). The seventeenth (and last) book apparently dealt with the siege of Hatra in 117 (F 17).

15. Bosworth, HSCP lxxxi (1977) 247-55. The most convenient and accurate text of the Order of Battle is in Roos, Arriani Scripta Minora, pp. 177-85. For the parallels with Alexander's engagement see Bosworth, op. cit. 252.

16. Ectaxis 5-6; 15 (phalanxes); 22. … Cf. Bosworth 248-51.

17. Cf. iii 16. 8 with note.

18. At vii 16. 3 he refers to majority opinion for the simple fact that the river Araxes flows into the Caspian (contrast Strabo xi 16. 3 (527)), and there is no hint of autopsy in his reference to the domicile of the Amazons along the Thermodon (vii 13. 4-6; cf. Periplus 15. 3).

19. Dio F 1. 3; cf. F. Millar, Cassius Dio 10.

20. … i 12. 5; Cyneget. 1. 4. The Cynegeticus speaks only of interest in sοφια, but the term was general enough to include history, and some regarded history as its most important component (cf. Dio xxxviii 28. 1).

21. J. Mewaldt, Hermes xlii (1907) 564-78. Dion 58. 10 refers to the Timoleon, whereas Timoleon 13. 10 and 33. 4 refer to the Dion. Brutus 9. 9 refers to the Caesar, whereas Caesar 62. 8 and 68. 7 refers to the Brutus.

22. Mewaldt's thesis was attacked extensively by C. Stolz, Zur relativen Chronologie der Parallelenbiographien Plutarchs (Lund 1929), but he failed to shake the arguments for the interrelation of the Dion and Timoleon and the Caesar and Brutus. Cf. P. A. Stadter, Plutarch's Historical Methods 32 n. 1; C. P. Jones, JRS lvi (1966) 66 f.; J. R. Hamilton, Plut. Al. xxxv ff.

23. Cf. C. P. Jones, JRS lvi (1966) 69; Hamilton, Plut. Al. xxxvii.

24. It need not have been an immediate stimulus. Literary rivalry with Plutarch persisted throughout the second century. Appian, BC ii 149-54. 619 ff, supplies a comparison between Caesar and Alexander, which Plutarch had omitted; and Amyntianus also wrote parallel lives of figures omitted by Plutarch (Phot. Bibl. cod. 131 = FGrH 150).

25. The metaphor of A. Heuss, Antike und Abendland iv (1954) 102, developed by E. Badian in Alexandre le Grand: image et réalité 280 ff. (with respect to modern interpretations).

26. The theory of Peripatetic hostility to Alexander is based on interpretation of Theophrastus' lost work, Callisthenes (cf. Schwartz RE iv 1899; Tarn ii 319), and its supposed derivatives in later literature. But nothing is known of the content or economy of the Callisthenes, and the other passages cited cannot be referred to specific Peripatetic sources. Cf. Badian, CQ viii (1958) 153 ff.; E. Mensching, Historia xii (1963) 274 ff.; Bosworth, Historia xix (1970) 407 f. The Stoic portrait was a construct of J. Stroux, Philologus lxxxviii (1933) 222-40, and its foundations have been recently undermined by J. R. Fears, Philologus cxviii (1974) 113-30. See further P. A. Brunt, Athenaeum lv (1977) 30-48.

27. Cic. de Off. i 26. 90; ad Att. xiii 28. 3; Sen. de Ben. i 13. 2-3, ii 16. 1-2, v 6. 1, vii 2. 5-6 (cf. Lucan x 20 ff.); on Seneca see Heuss (above n. 25) 88 f.

28. For a typical exposition see Livy ix 18. 3-5.

29. Dio Chrys. Orat. 2; for discussion see Heuss 92 f.

30. Cf. Livy ix 18. 3-4; Curt. vi 6. 4; Justin xii 3. 8 f. (negative); Plut. Al. 45. 1; de Al. f. i 8, 330 a (positive). Arrian seems to espouse both views: iv 7. 4 f., vii 29. 4.

31. See the analysis by Hamilton, Plut. Al. xxiii-xxxiii.

32. ‘Suda’ s.v. Διων ο Pαsικρατουs = FGrH 153 F 6.

33. Dio lxviii 29. 1, 30. 1. Cf. W. Weber, Untersuchungen zur Geschichte des Kaisers Hadrianus (Leipzig 1907), 8 ff.; F. A. Lepper, Trajan's Parthian War (Oxford 1948), 193 ff.; G. Wirth, in Alexandre le Grand: image et réalité 197 ff.

34. H. Strasburger, Ptolemaios und Alexander (Leipzig 1934); E. Kornemann, Die Alexandergeschichte des Königs Ptolemaios I. von Ägypten (Leipzig 1935). Of the two Strasburger's scheme is the more flexible, and he begins with an analysis of Arrian as a historian (8 ff.), but the explicit ‘Ausgangspunkt’ is appreciation of the extant fragments of Ptolemy (16 ff.), and much depends on subjective presuppositions of the contents of the lost original (e.g. ‘Spuren seines Stils finden sich nicht’: 40). Kornemann's work is far more schematic and operates on the assumption that Arrian had no style of his own and largely copies Ptolemy. For appreciations see Strasburger's review, Gnomon xiii (1937) 483-92; Pearson, LHA 195 f.; Wirth, RE xxiii 2467 f.

35. e.g. ii 3. 7, 4. 7, iii 30. 5, iv 6. 1-2, 8. 9.

36. From the end of Book vi Aristobulus becomes dominant: cf. vi 22. 4-8, 28. 3-4, 29. 4-11, vii 17. 5-22. 5.

37. Cf. Pearson, LHA 202-4; Seibert, Untersuchungen zur Geschichte Ptolemaios' I. (Munich 1969), 23-5.

38. A. Klotz, Livius und seine Vorgänger (Leipzig/Berlin 1940-1); cf. P. G. Walsh, Livy (Cambridge 1963), 138 ff.

39. iii 25. 8, 28. 1, vi 15. 5, 17. 3. See further CQ xxvi (1976) 128 f.

40. i 12. 7, 14. 1, 14. 6; see the notes ad locc.

41. Strabo xv 3. 7 (730) = vi 29. 4-11; xvi 1. 11 (741) = vii 20.

42. For a fuller exposition see CQ xxvi (1976) 117-39.

43. See the excellent discussion of P. A. Stadter, GRBS xvii (1976) 159-67.

44. For examples see Stadter 161.

45. Cf. Arr. 2. 2 with Xen. 5. 29; Arr. 31. 2 with Xen. 7. 5.

46. Cf. ii 12. 8, iii 2. 1, vii 15. 6.

47. The classic discussion is by Schwartz, RE ii 1240-3; see also Kornemann 21-30 (much oversimplified).

48. Compare the discussion of the Amazons at vii 13. 2-6, where the logos is a peg for an antiquarian excursus.

49. See also the discussion of Alexander's reaction to the death of Hephaestion (vii 14. 2-8).

50. See the detailed analysis by Guy L. Cooper, TAPA civ (1974) 23-80.

51. v 1. 4 … ; 2.2 … ; 2.5. … In all these cases the intrusive infinitive follows directly after a genuine accusative and infinitive construction, and the closest parallel seems to be Hdt. i 59. 3 (cf. Cooper 72).

52. Cf. iii 2. 3-6, iii 10. 1, etc.

53. Similarly we cannot assume that he omitted the stories of the Gordian Knot and Philip the Acarnanian (ii 3. 8, 4. 7 notes).

54. The fragments are printed by Jacoby, FGrH 138. For a survey of literature see Seibert, Alexander der Grosse 19-21; note particularly the discussions of Pearson, LHA 188-211 (with Badian, Studies 256-8) and G. Wirth, RE xxiii 2467-84.

55. F 11. Cf. E. Rohde, Rh. M. xxxviii (1883) 301-5; Jacoby, FGrH ii D. 504; Pearson, LHA 189.

56. F 5, 4, 2. Curt. ix 5. 21 refers to Ptolemy's statement that he was not at the Malli town (more fully in Arrian vi 11. 7-8 = F 26a) and Plutarch, Al. 46. 2 includes Ptolemy in a list of sources not mentioning the Amazon queen (more fully in Arrian vii 13.2-3 = F 28^b).

57. Cf. Kornemann 7 f.; Tarn ii 43; Pearson, LHA 193.

58. So Hamilton, PACA iv (1961) 19 n. 24; Badian, Studies 257, against Pearson, LHA 172 f., following the traditional view (cf. Schwartz, RE ii 916; Kornemann 13-15).

59. So R. M. Errington, CQ xix (1969) 241.

60. Adumbrated by Droysen i² 2. 383-6 and established as a canon by Wilcken, Philologus liii (1894) 84-126. I have examined the theory briefly in Alexandre le Grand: image et réalité 3-6. See also Brunt, Arrian xxiv-xxvi.

61. The extant fragments are printed by Jacoby, FGrH 117. For analysis of the reports of the illness, with the variants, see Pearson, Historia iii (1955) 432-4; cf. also Bosworth, CQ xxi (1971) 120 f.

62. For a review of opinions see Pearson, art. cit. 438 n. 37 (add Jacoby, FGrH ii D. 507; Strasburger 48 f.). Like Herodotus Arrian uses πορρω in a spatial sense, even when he is metaphorical; so iv 11.5 … = ‘further than is sufficient’ (cf. v 20. 10: … = ‘not far from the truth’; Strabo xv 1. 35 …). ‘Far from this’ in the sense of ‘different from’ is unlikely and one would expect some explanatory qualification. Elsewhere Arrian tends to be explicit; cf. iv 14. 1. …

63. For hypothetical explanations of these peculiarities see A. E. Samuel, Historia xiv (1965) 1-12; Bosworth, CQ xxi (1971) 117-23. Pearson 432-9 argues that the so-called document is in fact a Hellenistic forgery.

64. Note the records of arrangements in Babylon and Susa (iii 16. 4 and 9-10 with notes). For a baffling case of disagreement over what must have been officially recorded and widely known see the variant reports of Greek embassies at the Persian court in 333 and 330 (ii 15. 2 n.).

65. There are obvious similarities with Wilamowitz's famous theory that the entire framework of Athenian local history was provided by an official chronicle of the exegetai. Cf. Jacoby, Atthis (Oxford 1949) passim, esp. 5: ‘an ingenious thesis … which moreover begins with the monstrosity in method that the first Atthidographer must be eliminated from Atthidography’.

66. ‘So ist sein Werk Korrektur und Polemik durch Darstellung und Stillschweigen’ (Strasburger 55); ‘silence was Ptolemy's method of dealing with untrue stories; he never explicitly rejects or argues’ (Tarn ii 268).

67. Pliny NH i 12-13 (T 2); cf. Badian, Studies 256; CW lxv (1971) 38.

68. See, in general, the survey by C. B. Welles in Miscellanea Rostagni (Turin 1963), 101-16; his arguments are challenged and modified by Seibert, Untersuchungen zur Geschichte Ptolemaios' I. 4-26, but the conclusions reached are similar.

69. For the bias of Nearchus see Pearson, LHA 131-9, and most recently Badian, YCS xxiv (1975) 147-70.

70. See the independent observations by Bosworth and Errington in Alexandre le Grand: image et réalité 14-16, 159-62. Cf. also Seibert, Untersuchungen … 152-6.

71. See the cogent analysis by Errington, CQ xix (1969) 235 f.

72. Fragments of Aristobulus may be found in Jacoby, FGrH 139, and a survey of modern opinion in Seibert, Alexander der Grosse 21-3. See most recently, P. A. Brunt, CQ xxiv (1974) 65-9.

73. [Luc.] Macrob. 22 = T 3. Brunt, art. cit. 65, suggests that the author may have been deceived by a manuscript corruption; that seems improbable.

74. T 2, F 6, F 47 (Plutarch and Athenaeus).

75. The assumption is well-nigh universal; cf. Jacoby, FGrH ii D. 508; Berve no. 121; Pearson, LHA 151; Hamilton, Plut. Al. liv.

76. FGrH 134 F 22; cf. T. S. Brown, Omesicritus (Berkeley 1949), 81 ff.; Pearson, LHA 100 f.

77. Cf. Diod. 94. 1-3; Strabo xv 1. 27 (697). Nearchus also described the Acesines in flood, and Strabo significantly connects his account with the effects on the Macedonians: they were forced to move camp (xv 1. 18=FGrH 133 F 18).

78. Strabo xv 2. 7 (723) with Theophr. HP iv 4. 13. The fundamental treatment of the passage is by H. Strasburger, Hermes lxxx (1952) 461-5. His attribution of the variant account to Nearchus has not gone unchallenged (cf. Pearson, LHA 178 n. 151; Badian CW lxv (1971) 50), but it should be clear that the passage on the poison plants does not come from Aristobulus.

79. Strasburger, art. cit. 470-3. The most striking of these instances is the death march to the Oxus (cf. iii 29. 2 n.).

80. Cf. Schwartz, RE ii 917 f.; Strasburger 13 f.; Pearson, LHA 150, 157 ff. Tarn (ii 37-43, 131 f.) protested in favour of Aristobulus, but his own taste for apologetic was remarkably similar.

81. For the contemporary propaganda see Bosworth, CQ xxi (1971) 114-16 and for a possible additional example of Aristobulus' apologetic see below i 9. 1-8 n.

82. Cf. Brunt, CQ xxiv (1974) 65 ff., disposing successfully of the evidence from the Byzantine epitome on rhetoric (T 5), which most likely does not refer to the historian Aristobulus.

83. Brunt 68 f. attacks the content of Lucian's story as romantic fabrication. That is no doubt true, but it is significant that the romance was attached explicitly to Aristobulus. It is a counsel of desperation to suggest that it was wrongly fastened to him in later tradition.

84. Cf. Strasburger 15 f., cautiously approved by Badian, Studies 257.

85. The existence of a common source has long been known and it is an incontrovertible fact, acknowledged as such since at least the first edition of de Sainte-Croix, published in 1770 (cf. Seibert, Alexander der Groβe 26-8; and for a revealing list of parallels see Schwartz, RE iv 1873 f.). The identification of that source as Cleitarchus is still contested (cf. E. N. Borza, PACA ii (1968) 25-45; P. Goukowsky, RÉA lxxi (1969) 320), but there seems no other viable contender and the identification has been strongly pressed in recent years (cf. Schachermeyr² 658-62; J. R. Hamilton in Greece & the E. Mediterranean 126-46). The dating of Cleitarchus' work is controversial and depends in part on intricate arguments of borrowings which at times border on the metaphysical. I would accept with great caution the arguments lately adduced which would place his work in the vicinity of 310 b.c.; cf. Schachermeyr, Alexander in Babylon 211-24; Badian, PACA viii (1965) 1-8; Hamilton, Historia x (1961) 448-58.

86. Cf. i 21. 5, ii 4. 4, ii 10. 3, ii 22. 4, 23. 5, iii 14. 3, etc.

87. Cf. i 21. 3, 22. 2, ii 22. 6 with notes ad locc.

88. For the fragments of Callisthenes see Jacoby, FGrH 124. The foundation article is still that by Jacoby in RE x 1674-1707. See also T. S. Brown, AJP lxx (1949) 225-48; Pearson, LHA 22-49; Hamilton, Plut. Al. liii f.; Seibert, Alexander der Groβe 11 f.

89. This argument is expanded in Alexandre le Grand: image et réalité 25-9.

90. Diod. xix 52. 4-5, 53. 2 ff.; cf. Errington, in Alexander le Grand: image et réalité 151 f.

91. Cf. v 5. 1: a major heading of the Indica is to be ατοπα ξωα.

92. Parthica F 1; cf. Strabo xi 9. 2 (515); Justin xli 4. 4-10 with the excellent analysis of J. Wolski, Berytus xii (1956/7) 35-52.

93. vi 13. 4, 24. 2, vii 3. 6, 20. 9. Cf. Schwartz, RE ii 1239.

94. Strabo xiv 6. 3 (682), xv 3. 8 (730) = FGrH 143 T 1, F 1. For a brief description of the fragments see Pearson, LHA 254 f.

95. Pliny, NH iii 57 = FGrH 137 F 31. Cf. Schachermeyr, Alexander in Babylon 218 ff.

96. iv 12. 3-5 = Plut. Al. 54. 4-6 (Jacoby prints both versions as FGrH 125 F 14).

97. He gives the father's name of Demetrius (Pythonax); Plutarch 54. 6 gives his pseudonym Pheidon.

98. Note particularly vii 6. 4-5, where Arrian's extremely elliptical phrasing totally obscures the crucial question how far barbarian cavalry had been drafted into the ranks of the hetairoi: cf. P. A. Brunt, JHS lxxxiii (1963) 43 f.; E. Badian, JHS lxxxv (1965) 161.

99. The archaic form only occurs sporadically in the manuscripts of Arrian, but it appears unambiguously in the second-century papyrus fragment of the History of the Successors (PSI xii 2. 1284, col. 82. 5; cf. K. Latte, Kleine Schriften (Munich 1968), 597).

100. For a convenient list of non-Attic usages see the edition of K. Abicht, 16-18. For the divergent uses of καιτοι and πριν compare Hdt. viii 53. 1, i 165. 3 (cf. Powell, Lexicon to Herodotus 317).

101. See particularly the excellent dissertation by H. R. Grundmann, ‘Quid in elocutione Arriani Herodoto debeatur’, Berliner Studien ii (1886) 181-268. This work supplements and largely supersedes earlier dissertations by E. Meyer, De Arriano Thucydideo (Rostock 1887) and C. Renz, Arrianus quatenus Xenophontis imitator sit (Rostock 1879).

102. Dozens of examples collected by Grundmann 200-2.

103. Cf. ii 10. 4 … ; iii 18. 12 … See further Grundmann 206 ff.

104. Cf. i 19. 11, 20. 7, 25. 10, iii 12. 1, etc. Further examples Grundmann 211.

105. Good examples at i 1. 1 and ii 11. 5. Full discussion Grundmann 214 ff.

106. Cf. iv 4. 4 with the further examples in Grundmann 230 ff. For a particularly striking use of anacolouthon see the imitation of the Herodotean nominative absolute at i 9. 5 (cf. my note ad loc.). The Herodotean intrusive infinitives (above p. 21) should also be included in this category.

107. Note particularly πόθοs λαμβάνει αοτόν (i 3. 5 n.) and ονδεν αχαρι παθεν (i 17. 9 n.), both phrases which have been claimed for Ptolemy. See also the notes on the Herodotean verb … (iii 2. 1-2). For a full and impressive list of such borrowings see Grundmann 248-56.

108. Cf. Thuc. vii 62. 2, 67. 2.

109. i 12. 7, 16. 4-5, 21. 2, etc. Full list of instances in Grundmann 184-6.

110. i 10. 6, 17. 2, etc. See further Grundmann 186.

111. Perhaps most strikingly at ii 10. 3. … See further the far from impressive list in Grundmann 191.

112. Cf. Luc. de hist. conscr. 15 (Creperius Calpurnianus and Thucydides); 18 (copying from Herodotus in bogus Ionic). See the commentary by H. Homeyer, Lucian: Wie man Geschichte schreiben soll (Munich 1965).

113. Luc. Alexander 2 = T 24 (Roos). The work was written after the death of Marcus in 180 (cf. Al. 48).

114. Tillorobus is only known from the reference in Lucian. There is no way of dating him and his depredations may well have occurred before the Romans annexed Bithynia—the terminal point of Arrian's Bithyniaca.

115. See the references above pp. 12.

116. For the full nomenclature see AÉ 1971. 430.

117. For the biography see ‘Suda’ s.v. … (T 1) with G. Wirth, Klio xli (1963) 221-33; Bosworth, CQ xxii (1972) 166; for Arrian in Cappadocia see Dio lxix 15. 1 with Bosworth, HSCP lxxxi (1977) 218 ff.; for the use of the Parthica see A. G. Roos, Studia Arrianea (Leipzig 1912); K. Hartmann, Philologus lxxiv (1917) 73-91; Lepper, Trajan's Parthian War 1 ff.; Wirth, Studii Clasice xvi (1974) 194 ff.

118. Phot. cod. 82: 64^a21 ff. = FGrH 100 F 8. For the life and background of Dexippus see F. Millar, JRS lix (1969) 12-29.

119. Themistius 34. 8 = T 13. Cf. Bosworth, HSCP lxxxi (1977) 229 ff.

120. For general discussion see Kubitschek, RE ix 2363-6. The most accessible (though far from satisfactory) text is that following Müller's edition of Ps.-Callisthenes in the appendix to Dübner's Arrian (Didot 1846: pp. 154-67). The edition of D. Volkmann (Programm Schulpforta 1871), which is allegedly a great improvement, has unfortunately not been accessible to me. …

121. Itin. 2: nec … vilibus usus auctoribus, sed quos fidei amicissimos vetus censura pronunciat.

122. For documentation see the painstaking indexes of Roos (Anabasis xxxvi-xli; Scripta Minora 315-20).

Abbreviations and Bibliography of Short Titles

AAA: … Athens Annals of Archaeology

Abh. Berl. Akad.: Abhandlungen der preuβischen Akademie der Wissenschaften, phil.-hist. Abteilung

ABSA: The Annual of the British School of Archaeology at Athens

AE: L'Année épigraphique

AHR: American Historical Review

AJA: American Journal of Archaeology

AJP: American Journal of Philology

Alexandre le Grand: image et réalité: Alexandre le Grand: image et réalité, Entretiens sur l'Antiquité classique xxii (Fondation Hardt, Geneva 1976)

Andreotti: Il problema politico R. Andreotti, Il problema politico di Alessandro Magno (Parma 1933)

'Aρχ. 'Εφ.: 'Aρχαιολογικη 'Εφημεριs

Ath. Mitt.: Mitteilungen des deutschen archäologischen Instituts: athenische Abteilung

AS: Antike Schlachtfelder in Griechenland, ed. J. Kromayer and G. Veith (4 vols.: Berlin 1903-31)

ATL: The Athenian Tribute Lists, by B. D. Merritt, H. T. Wade-Gery, and M. F. McGregor (4 vols.: Cambridge and Princeton 1939-53)

Badian, Anc. Soc. & Inst.: E. Badian, ‘Alexander & the Greeks of Asia’ in Ancient Societies and Institutions: Studies presented to V. Ehrenberg (Blackwell, Oxford 1966)

Badian, Studies: E. Badian, Studies in Greek and Roman History (Oxford 1964)

BCH: Bulletin de Correspondance hellénique

BE: J. and L. Robert, Bulletin Épigraphique (published initially in REG and reprinted as a series with indexes, Paris 1972- )

Bellinger: A. R. Bellinger, Essays on the Coinage of Alexander the Great (Numismatic Studies xi: New York 1963)

Beloch: K. J. Beloch, Griechische Geschichte (2nd edn.: 4 vols.: Strassburg-Berlin and Leipzig 1912-27)

Bengtson, Strategie: H. Bengtson, Die Strategie in der hellenistischen Zeit (Münchener Beiträge, vols. xxvi [1937]; xxxii [1944]; xxxvi [1952])

Berve: H. Berve, Das Alexanderreich auf prosopographischer Grundlage (2 vols.: Munich 1926)

BMC Arabia: G. F. Hill, Catalogue of the Greek Coins of Arabia, Mesopotamia and Persia. British Museum Catalogue (London 1922)

BMC Caria: B. V. Head, Catalogue of the Greek Coins of Caria, Cos, Rhodes, etc., British Museum Catalogue (London 1896)

BMC Phoenicia: G. F. Hill, Greek Coins of Phoenicia, British Museum Catalogue (London 1910)

Briant, Antigone le Borgne: P. Briant, Antigone le Borgne: les débuts de sa carrière et les problèmes de l'assemblée macédonienne (Paris 1973)

Brunt, Arrian” P. A. Brunt, Arrian i (Loeb Classical Library: Cambridge and London 1976)

Busolt-Swoboda GS: G. Busolt and H. Swoboda, Griechische Staatskunde (3rd edn.: Munich 1920-6)

Casson, Ships and Seamanship: L. Casson, Ships and Seamanship in the Ancient World (Princeton 1971)

CIL: Corpus Inscriptionum Latinarum

CIS: Corpus Inscriptionum Semiticarum

Cook, Troad: J. M. Cook, The Troad: an archaeological and topographical study (Oxford 1973)

Corinth: Corinth; results of excavations conducted by the American School of Classical Studies at Athens (Cambridge and Princeton 1929-)

C. Phil.: Classical Philology

CQ: Classical Quarterly

CRAI: Comptes rendus de l'Académie des inscriptions et belles-lettres

CW: The Classical World

Delbrück: H. Delbrück, Geschichte der Kriegskunst im Rahmen der politischen Geschichte (3rd edn.: Berlin 1920)

Denkschr. Akad. Wien: Denkschriften der Akademie der Wissenschaften in Wien

Droysen: J. G. Droysen, Geschichte des Hellenismus (2nd edn.: Gotha 1877-8)

Ehrenberg: Al. & the Greeks V. Ehrenberg, Alexander and the Greeks (Blackwell, Oxford 1938)

Ellis: Philip II J. R. Ellis, Philip II and Macedonian Imperialism (London 1976)

FGrH: F. Jacoby, Die Fragmente der griechischen Historiker (Berlin and Leiden 1923-)

Fuller: J. F. C. Fuller, The Generalship of Alexander the Great (London 1958)

Gomme: HCT A. W. Gomme, A Historical Commentary on Thucydides (vols. i-: Oxford 1945-)

Grazer Beitr.: Grazer Beiträge: Zeitschrift für die klassische Altertumswissenschaft

GRBS: Greek, Roman and Byzantine Studies

Greece & the E. Mediterranean: Greece and the Eastern Mediterranean in History and Prehistory, Studies Presented to Fritz Schachermeyr, ed. K. H. Kinzl (Berlin 1977)

Grundmann: H. R. Grundmann, Quid in elocutione Arriani Herodoto debeatur (Berliner Studien II [1886])

Habicht, Gottmenschentum²: Chr. Habicht, Gottmenschentum und griechische Städte (Zetemata XIV: 2nd edn.: Munich 1970)

Hamilton, Al.: J. R. Hamilton, Alexander the Great (London 1973)

Hamilton, Plut. Al.: J. R. Hamilton, Plutarch, Alexander: a Commentary (Oxford 1969)

Hammond, Epirus: N. G. L. Hammond, Epirus: The Geography, the Ancient Remains, the History and the Topography of Epirus and Adjacent Areas (Oxford 1970)

Hammond, Macedonia: N. G. L. Hammond, A History of Macedonia i (Oxford 1972)

Head, HN²: B. V. Head, Historia Numorum (2nd edn., Oxford 1911)

HSCP: Harvard Studies in Classical Philology

HZ: Historische Zeitschrift

IG: Inscriptiones graecae (1st edn. Berlin 1873—2nd edn. Berlin 1913-)

IG: Bulgaria Inscriptiones Graecae in Bulgaria Repertae, ed. G. Mihailov (4 vols.: Sofia 1958-70)

IGR: Inscriptiones graecae ad res romanas pertinentes

ILS: Inscriptiones latinae selectae, ed. H. Dessau (Berlin 1892-1916)

Inschr. Magn.: O. Kern, Die Inschriften von Magnesia am Maender (Berlin 1900)

Inschr. Priene: F. Hiller von Gaertringen, Die Inschriften von Priene (Berlin 1906)

Instinsky: H. U. Instinsky, Alexander der Groβe am Hellespont (Godesberg 1949)

1st. Mitt.: Mitteilungen des deutschen archäologischen Instituts: Abteilung Istanbul

Janke: A. Janke, Auf Alexanders des Groβen Pfaden (Berlin 1906)

JDAI: Jahrbuch des deutschen archäologischen Instituts

JHS: The Journal of Hellenic Studies

JKF: Jahrbuch für kleinasiatische Forschung

JÖAI: Jahreshefte des österreichischen archäologischen Instituts

Jones, CERP²: A. H. M. Jones, The Cities of the Eastern Provinces (2nd edn.: Oxford 1971)

JRAS: Journal of the Royal Asiatic Society

JRS: The Journal of Roman Studies

Julien: P. Julien, Zur Verwaltung der Satrapien unter Alexander dem Groβen (Diss. Leipzig 1914)

Kaerst: J. Kaerst, Geschichte des Hellenismus (2 vols.: Leipzig 1927 [vol. i, edn. 3] and 1926 [vol. ii edn. 2])

Kirchner: PA J. Kirchner, Prosopographia Attica, 2 vols. (Berlin 1901-3)

Kornemann: E. Kornemann, Die Alexandergeschichte des Königs Ptolemaios I von Aegypten (Leiptzig 1935)

Kromayer-Veith: J. Kromayer and G. Veith, Heerwesen und Kriegführung der Griechen und Römer (Munich 1928)

Lanckoroński: K. Graf Lanckoroński, Städte Pamphyliens und Pisidiens, unter Mitwerkung von G. Niemann und E. Petersen (2 vols.: Vienna 1890-9)

Lane Fox: R. Lane Fox, Alexander the Great (London 1973)

Larsen: GFS J. A. O. Larsen, Greek Federal States (Oxford 1968)

Launey, Recherches: M. Launey, Recherches sur les armées hellénistiques (2 vols.: Paris 1949-50)

Leuze, Satrapieneinteilung: O. Leuze, Die Satrapieneinteilung in Syrien und in Zweistromland von 520-320 (Schriften der Königsberger Gelehrten Gesellschaft: Geisteswiss. Klasse xi [1935])

LSJ²: H. G. Liddell, R. Scott, and H. S. Jones, A Greek-English Lexicon (rev. ed. Oxford 1968)

Magie, RRAM: D. Magie, Roman Rule in Asia Minor to the end of the Third Century after Christ (Princeton 1950)

Marsden: E. W. Marsden, The Campaign of Gaugamela (Liverpool 1964)

Mederer: E. Mederer, Die Alexanderlegenden bei den ältesten Historikern (Stuttgart 1936)

Meiggs/Lewis: R. Meiggs and D. Lewis, A Selection of Greek Historical Inscriptions to the end of the Fifth Century B.C. (Oxford 1969)

Meyer, Forschungen: E. Meyer, Forschungen zur alten Geschichte (2 vols.: Halle 1892-9)

M. Helv.: Museum Helveticum

Milet: Milet: Ergebnisse der Ausgrabungen und Untersuchungen seit dem Jahre 1899 (Berlin 1906-)

Miscellanea Rostagni: Miscellanea di studi alessandrini in memoria di Augusto Rostagni (Turin 1963)

Moretti: L. Moretti, Iscrizioni storiche ellenistiche (Florence 1967-)

Müller, GGM: C. Müller, Geographi graeci minores (2 vols.: Paris 1855-61)

NC : Numismatic Chronicle

OGIS: Orientis graeci inscriptiones selectae, ed. W. Dittenberger (2 vols., Leipzig 1903-5)

PACA: Proceedings of the African Philological Association

PCPS: Proceedings of the Cambridge Philological Society

PdP: La parola del passato

Pearson, LHA: L. Pearson, The Lost Histories of Alexander the Great (Philological Monographs XX: Am. Phil. Ass. 1960)

PIR²: Prosopographia imperii romani, saec. I, II, III, 2nd edn., ed. E. Groag, A. Stein, L. Petersen (Berlin and Leipzig 1933-)

P. Oxy.: Oxyrhynchus Papyri, ed. B. P. Grenfell and A. S. Hunt (London 1898-)

PSI: Papiri greci e latini (Pubblicazioni della Società Italiana per la ricerca dei papiri greci e latini in Egitto)

P. Tebt.: Tebtunis Papyri, ed. B. P. Grenfell, A. S. Hunt, J. G. Smyly, E. J. Goodspeed (3 vols.: London-New York-California 1902-38)

RE: Realencyclopädie der classischen Altertumswissenschaft, ed. Pauly, Wissowa, Kroll (Stuttgart 1893-)

REA: Revue des études anciennes

REG: Revue des études grecques

RFIC: Rivista di filologia e di istruzione classica

Rh. Mus.: Rheinisches Museum für Philologie

Robert, Études anatoliennes: L. Robert, Études anatoliennes: recherches sur les inscriptions grecques de l'Asie Mineure (Paris 1937)

Robert, Hellenica: L. Robert, Hellenica, recueil d'épigraphie, de numismatique et d'antiquités grecques (vols. i-: Paris 1940-)

Rostovtzeff, SEHHW: M. Rostovtzeff, The Social and Economic History of the Hellenistic World (3 vols.: Oxford 1941)

RSI: Rivista storica italiana

SB Berlin: Sitzungsberichte der preuβischen Akademie der Wissenschaften, phil.-hist. Klasse

SB Heidelberg: Sitzungsberichte der Heidelberger Akademie der Wissenschaften, phil.-hist. Klasse

SB Munich: Sitzungsberichte der bayerischen Akademie der Wissenschaften, phil.-hist. Abteilung

Schachermeyr²: F. Schachermeyr, Alexander der Groβe: das Problem seiner Persönlichkeit und seines Wirkens (Sitzungsberichte der österreichischen Akademie der Wissenschaften, phil.-hist. Klasse, cclxxxv [1973])

Schachermeyr, Alexander in Babylon F. Schachermeyr, Alexander in Babylon und die Reichsordnung nach seinem Tode (Sitzungsberichte der österreichischen Akademie der Wissenschaften, phil.-hist. Klasse, cclxviii. [1970])

Schaefer: A. Schaefer, Demosthenes und seine Zeit (2nd edn.: Leipzig 1886)

Schwyzer: W. Schwyzer, Dialectorum graecarum exempla epigraphica potiora (Leipzig 1923)

SEG: Supplementum Epigraphicum Graecum

Seibert, Alexander der Groβe: J. Seibert, Alexander der Groβe (Erträge der Forschung X: Darmstadt 1972)

SGDI: Sammlung griechischer Dialekt-Inschriften, ed. H. Collitz and F. Bechtel (4 vols.: Göttingen 1884-1905)

SIG³: Sylloge inscriptionum graecarum, ed. W. Dittenberger (3rd edn.: Leipzig 1915-24)

Staatsverträge: Die Staatsverträge des Altertums, ii: Die Verträge der griechischrömischen Welt von 700 bis 338 v. Chr., ed. H. Bengtson (Munich 1962), iii: Die Verträge der griechisch-römischen Welt von 338 bis 200 v. Chr. ed. H. H. Schmitt (Munich 1969)

Strasburger: H. Strasburger, Ptolemaios und Alexander (Leipzig 1934)

TAM: Tituli Asiae Minoris

TAPA: Transactions of the American Philological Association

Tarn: W. W. Tarn, Alexander the Great (2 vols.: Cambridge 1948)

Tod: M. N. Tod, A Selection of Greek Historical Inscriptions, vol. ii (Oxford 1948)

von Schwartz: F. von Schwartz, Alexanders des Groβen Feldzüge in Turkestan (Munich 1893: 2nd ed. 1906)

Walbank, Philip V: F. W. Walbank, Philip V of Macedon (Cambridge 1940)

Walbank, Polybius: F. W. Walbank, A Historical Commentary on Polybius (vols. i-iii: Oxford 1957-79)

Welles, Diodorus: Diodorus of Sicily, vol. viii, ed. C. B. Welles (Loeb Classical Library: Cambridge and London 1963)

Welles, RC: C. B. Welles, Royal Correspondence of the Hellenistic Age (Yale 1934)

Wilcken: U. Wilcken, Alexander the Great, with Introduction and Notes by E. N. Borza (New York 1967)

Wilcken, Berliner Akademieschriften: U. Wilcken, Berliner Akademieschriften zur alten Geschichte und Papyruskunde (1883-1942) (2 vols.: Leipzig 1970)

Wilcken, Grundzüge: L. Mitteis and U. Wilcken, Grundzüge und Chrestomathie der Papyruskunde (4 vols.: Leipzig-Berlin 1912)

Wilcken, UPZ: U. Wilcken, Urkunden der Ptolemäerzeit (2 vols.: Berlin 1922-37)

Wüst, Philipp II: F. R. Wüst, Philipp II von Makedonien und Griechenland in den Jahren 346 bis 338 (Munich 1938)

YCS: Yale Classical Studies

ZPE: Zeitschrift für Papyrologie und Epigraphik

N.B. References to the ancient sources follow the standard conventions. Where Diodorus Siculus is cited by chapter and section alone the reference is to Book xvii.