Appendix:

Propaganda at the Granicus

THE battle of the River Granicus has at least two special claims on our interest: it was not only the first engagement fought by Alexander on Asiatic soil, but also, apparently, one of the most dramatic. Yet it is, on the whole, poorly documented; and the accounts we possess of it¹ contain inconsistencies and anomalies which have never been satisfactorily explained. Motives remain impenetrable; tactical dispositions range from the wilful to the lunatic. The baffling nature of the evidence was strikingly demonstrated in 1964 by E. W. Davis,² who, after analysing the inadequacies of no less than four previous accounts — those by Tarn, Beloch, Fuller and Schachermeyr — concluded that the problem was, ultimately, insoluble, ‘for with the information at our disposal we cannot read the minds of the Persian leaders’.³ Davis handicapped himself needlessly by his curious assumption⁴ that the Persian army was under the command not of Arsites, but of a committee — perhaps in an effort to excuse the indubitably irrational Persian strategy as reported by our main sources. At the same time his pessimism is all too understandable, and his three basic questions — ‘why the battle was fought, why it was fought where it was fought, and why it was fought as it was fought’⁵ — must be squarely faced by any student of this enigmatic engagement.

The first two points need not detain us overlong: on them there exists a fair (if not unanimous) consensus of agreement. It is the third which has always been the real difficulty. From Alexander's viewpoint, an immediate engagement was essential. He had to secure Hellespontine Phrygia before moving on south; more important, he urgently needed the cash and supplies which only a victorious battle could give him. His debts were crippling. When he crossed into Asia he had a bare seventy talents (perhaps representing two weeks' pay for his troops) and provisions for no more than a month at the outside.⁶ Memnon, well described by Diodorus as ‘famed for his understanding of strategy’, διαβεβοημένος ἐπὶ συνέσει στρατηγικῇ, had accurately assessed Alexander's predicament: hence his shrewd proposal that the Persians should avoid battle, implement a scorched-earth policy, and if possible carry the war across into Greece. Alexander would then be forced to withdraw for lack of supplies.⁷ As his invasion strategy had already made clear, he possessed neither the time nor the equipment to besiege cities en route. If they did not come over to him at once, of their own will, he simply by-passed them.⁸

The Persians, however, rejected Memnon's advice, and chose instead to establish a defensive line on the Granicus River, with the object of holding up Alexander's eastward advance towards Dascylium, and, if possible, of cutting short this Macedonian invasion ‘as it were at the gateway of Asia’, ὥσπερ ἐν πύλαις τῆς Ἀσίας. This may have been, as most modern scholars argue, a mistaken decision; but it was a perfectly understandable one. Pride entered into it: Arsites declared he would not let a single house in his satrapy be burnt. So did distrust of Memnon, the Greek mercenary, who made no secret of his contempt for Persian infantrymen, and was thought, rightly or wrongly, to be ‘deliberately procrastinating over this campaign for the sake of [i.e. to prolong] his commission from the King’,⁹ τριβὰς ἐμποιεῖν ἑκόντα τῷ πολέμῳ τῆς ἐκ βασιλέως τιμῆς ἕνεκα.

Modern scholars have found other additional or alternative explanations, not all equally convincing. Tarn's I will deal with in a moment. Schachermeyr argues that the Persians' aristocratic code forbade them to retreat without a fight, so that Memnon's advice was by definition unacceptable.¹⁰Though the Iranian nobility undoubtedly, like all aristocrats, did observe a strict code of honour,¹¹ this had not prevented them, half a century previously, from using very similar tactics against Agesilaus; and as Davis says,¹² ‘there is no evidence that Persian standards of knightliness had risen noticeably in the interval’. Davis himself suggests, rather more convincingly, that the satraps must answer not merely to their code but also to Darius; that Alexander was, as yet, merely a young Macedonian leader, Philip's son, and not the charismatic world-conqueror of later years; while the threat of revolt by the Greek cities of Ionia would undoubtedly become reality unless a firm stand was taken against the invader.¹³

However, once the decision to fight had been made, the Granicus line, it might well be argued, was the natural one to hold. This river, today the Koçabas, flows in a north-easterly direction from Mount Ida to the Sea of Marmara, through flat rolling country, ringed by low mountains, and ideal for a cavalry engagement such as the Persians were used to fighting. In May, when Alexander made his advance through Asia Minor, the Granicus would be swollen, though still fordable at its main crossing-points.¹⁴ The Persians now advanced from their base-camp at Zeleia (Sari-Keia), and established themselves on the high, steep eastern bank of the river. As Fuller points out,¹⁵ ‘the southern flank of its lower reach was safeguarded against a turning movement from its western side by a lake, now called the Edje Göl.’ Granted the Persians' decision to stand and fight, Arsites and his colleagues had chosen about the best possible terrain for their purpose.

But one point which has worried every student of this battle is the strategy — if we are to believe our sources — which they then proceeded to adopt. They drew up their forces along the river-bank, on a broad front, with high ground behind them. According to Arrian, their infantry was kept at the rear, virtually out of action, and their cavalry posted in front, where it could not charge.¹⁶ fcAs Davis understandably remarks,¹⁷ ‘either error is bad enough, but both together seem almost too much’. The Persians had hitherto acted without comparative good sense, and such a move makes them appear stupid almost past comprehension. It does not need Tarn's assurance¹⁸ to convince us that this was not the proper way to hold a riverbank. Wilcken's comment (‘a glaring error of tactics’)¹⁹is typical of most historians' reaction to this strange aberration, which wasted a perfectly good body of professional Greek mercenaries during the battle, and resulted in its near-annihilation afterwards.

Various attempts have been made to explain, if not to justify, such a move. All, as Davis notes without comment, ‘try to puzzle out some rational explanation as to what could have been the Persians' purpose behind this apparently mad act of folly’²⁰ — i.e. they rest on the initial premiss that our evidence is to be taken at its face value. None is in the slightest degree convincing. Tarn, for instance, argued that the Persian leaders ‘had in fact a very gallant plan; they meant to strangle the war at birth by killing Alexander’.²¹ Elsewhere²² he developed this thesis more fully, claiming that ‘the extraordinary formation they adopted was to induce Alexander himself to charge’. But Alexander, like all commanders of antiquity, led his own troops as a matter of course; nor, granted his position at the Granicus, could he refuse battle even if he so wished. The Persians had no need to adopt a special formation — let along a patently suicidal one — to make him attack, or do their best to kill him when he did.²³

Furthermore, how the king's death would be more surely encompassed by pulling the Persians' only first-class infantry unit out of the fighting-line is left to our imagination. Fuller, with his usual acumen in tactical matters, pointed out²⁴ that ‘if the sole aim of the Persians was to kill Alexander, then the best way to do so was to meet his cavalry charge with a hedge of spears; let him shatter himself against it, and then, should he break through, overwhelm him with javelins.’ Elsewhere²⁵ he spells out just what they should have done by telling us what they did not do: ‘They did not deploy the Greek mercenaries along the eastern bank, with the Persian cavalry on their flanks, and also in their rear to counter-attack any force that might break through the infantry.’ Fuller, likeTarn, takes this failure as fact, and simply casts around for an explanation.

The answer he comes up with is almost identical to that proposed by Schachermeyr, and we may conveniently deal with both together. This is the Military Etiquette or Medieval Tournament theory. According to Schachermeyr, this was to be a formal contest of Junker gegen Junker, where only the cavalry would participate, and both sides would observe rules of knightly warfare: Im Ritterstil hot sich der Gegner zur Schlacht an, im Ritterstil wollte ihm der König begegnen.²⁶ But infantry and light-armed troops did, in fact, take part in the battle, while no knightly code known would require the Persians to adopt the formation they did. Then (we may legitimately ask) why pay several thousand Greek mercenaries for doing nothing? Fuller's answer is that ‘throughout history the cavalry soldier has despised the infantryman, and to have placed the Greek mercenaries in the forefront of the battle would have been to surrender to them the place of honour. Military etiquette forbade it.’²⁷ In support he cites parallels from Taganae (A.D. 552) andCrécy. What he does not emphasize, though it is only too apparent from his own subsequent narrative,²⁸ is the crucial role played by these supposedly despised Greek mercenaries, very much in the forefront of the battle, at Issus and Gaugamela. Nor, obviously, didCyrus have any such social qualms when deploying his forces at Cunaxa.²⁹ Greek mercenaries, in fact, very often enjoyed the place of honour in Persian tactical dispositions, unhampered by any hypothetical requirements of knightly precedence. This theory, then, will not do either.

There are in fact three possibilities, and three only. 1. The Persian commanders were sheerly incompetent. 2. Their known dislike and distrust of Memnon, the mercenaries' commander, were so great that they deliberately threw away a battle rather than let him and his troops win it,³⁰ even while keeping them on what must have been a very expensive payroll. 3. Our surviving accounts of the battle contain, for whatever reason, substantial inaccuracies. 1 and 2, though not by definition impossible, do not readily lend themselves to analytical investigation. Let us see what can be done with 3. The first, and most obvious, fact which emerges from a detailed comparison of our three main versions is that whereas Arrian and Plutarch (with certain exceptions I shall come to in a moment) agree well enough, Diodorus tells a quite different story, and may therefore be assumed to depend, in part at least, on a different source: not necessarily Cleitarchus, as was formerly thought to be the case,³¹ certainly not Tarn's hypothetical ‘mercenaries' source’,³² though perhaps a case of a sort could be made out for Trogus.³³

Arrian and Plutarch both make the battle take place in the late afternoon; Diodorus puts it at dawn.³⁴ Arrian and Plutarch describe an engagement where the Persians are holding the high eastern river-bank against a direct assault through the river itself; in Diodorus Alexander gets his whole army across the river unopposed, and draws it up in battle-formation before the Persians can do anything to stop him.³⁵ There are other discrepancies, but these remain by far the most important.³⁶ It is worth noting at this point that though comparatively few scholars have thought the Diodorus version worth serious attention, they include Konrad Lehmann, Julius Beloch, Helmut Berve, and, most recently, R. D. Milns.³⁷ Beloch complained of the difficulty involved in finding an account that was ‘unbeirrt durch den Arrian-Kultus’;³⁸ it is hard not to remember this remark when reading Davis's assertion³⁹ that Beloch ‘contents himself with rewriting the entire battle’ — though in fact Beloch has simply utilized the testimony of Diodorus.

Now Arrian and Plutarch both allude to the possibility of a dawn attack. This was, according to them, the strategy recommended to Alexander by Parmenio when the army first reached the Granicus. It was late in the afternoon; the Persians were entrenched in an extremely strong position; while the Granicus itself, with its steep banks and deep, fast-flowing stream, presented a formidable initial hazard (I am leaving on one side, for the moment, the actual disposition of Arsites' forces). There was, it seems, something of a panic among Philip's old officers, thus called upon to launch an assault under highly unfavourable conditions, while exposed to concentrated enemy fire. Nor would it be the first time their youthful leader had made a dangerous error of judgement: his campaign against Cleitus and Glaucias had come within an ace of ending in total disaster.⁴⁰ Tactfully, they argued that Daisios was a taboo month for Macedonians to fight a battle; Alexander replied by performing an ad hoc intercalation on the calendar, so that the month was now (officially at least) a second Artemisios.⁴¹

This point being settled — again, according to Arrian and Plutarch — battle was joined, and after a hard initial struggle the Macedonians won their great victory. Yet few modern students would disagree with Plutarch's verdict that the strategy which Alexander employed ‘seemed to be crazy and senseless rather than the product of reason’, ἔδοξε μανικῶς καὶ πρὸς ἀπόνοιαν μᾶλλον ἤ γνώμῃ στρατηγεῖν.⁴² In fact the one thing which, so far as we can judge, prevented it ending in total disaster was the even more lunatic strategy adopted by the Persians on the other side. This gives one food for thought, especially since Diodorus offers us not only a quite different picture but an eminently sane one.

Here, beyond any doubt, we have a situation in which Parmenio's advice has been followed. Alexander moves at dawn, and gets his whole army across the Granicus undisturbed — which makes it a virtual certainty (assuming, for the moment, the validity of the report) that during the night he had moved away from the Persian position, and found an easier alternative fording-point. In which direction? Welles (see n. 36) claims that Diodorus, or his source, probably ‘located the battle farther upstream, in the foothills’. He cites no evidence for this view, and the topography of the area is, on balance, against it. There is also the (admittedly ambiguous) evidence of Polyaenus⁴³ to consider in this context. At his crossing of the Granicus, Polyaenus reports, Alexander Πέρσας ἐξ ὑπερδεξίων ἐπιόντας (αὐτοὺς) αὐτὸς ἐπὶ δόρυ τοὺς Μακεδόνας ἀναγαγὼν ὑπερεκέρασεν. The Persians, that is, were advancing, ἐπιόντας, which they could scarcely have been doing in the engagement as described by Ptolemy and Aristobulus; and they were advancing ἐξ ὑπερδεξίων. While this phrase came to mean simply ‘from above’ or ‘from higher ground’ in many cases, its root meaning was ‘from above on the right’, and in various well-attested instances⁴⁴ it could signify ‘from upstream’. Alexander then proceeded to outflank his attackers on the right wing, another significant departure from the canonical version of the battle: ἐπὶ δόρυ … ὑπερεκέρασεν. Whatever meaning we attach to ἐξ ὑπερδεξίων, what Polyaenus would seem to be describing is an engagement fought at right-angles to the river rather than parallel with it, which suggests that he too drew on the Diodoran tradition.

Now in Diodorus' account, the Persian order of battle, far from being a mere unaccountable whim, makes very good sense indeed. Here it is only after Alexander has crossed the river, and deployed his forces,⁴⁵ that Arsites and his fellow-commanders decide to counter the Macedonian attack with an all-out cavalry front, and to hold their infantry in reserve. This plan bears some resemblance to Darius' battle order at Gaugamela (see above, pp. 289–90), and was adopted for very similar reasons. In the first place, Persian infantry (or indeed any infantry if sufficiently outnumbered) was ‘unsuitable for a pitched battle in the plains either against hoplites or charging horsemen’.⁴⁶ Secondly, and more important, in cavalry the Persians were overwhelmingly stronger than their opponents, a fact which went some way to balance out their shortage of first-class foot-soldiers.

To calculate the actual number of troops which the Persians had available at the Granicus is a highly conjectural task, but in ways a most revealing one. Arrian (1.4.4) states that they had 20,000 cavalry and 20,000 infantry, the latter consisting exclusively of mercenaries. Diodorus (17.19.5) gives the figure as over 10,000 cavalry, plus 100,000 infantry. This latter figure, improbably enough in itself, is contradicted by Arrian's statement elsewhere (1.13.3) that the Persian infantry was ‘outnumbered’, and thus even at an outside estimate lower than the overall Macedonian total of 43,000⁴⁷ — some at least of whom were probably on line-of-communication duties. Plutarch gives no figures at all, while Justin (11.6.11) offers an all-in total of 600,000 (sic).

Let us now compare these figures with the casualty lists. Diodorus (17.21.6) claims that the Persians lost over 2,000 cavalry and 10,000 infantry. Plutarch (Alex. 16.7) places the infantry losses at 20,000, those of the cavalry at 2,500. Arrian (1.16.2) makes no assessment of infantry losses at all, except to say that the Greek mercenary phalanx, all but some 2,000 men, was totally wiped out. Diodorus further records the number of prisoners taken — and in the context it is clear that means infantry prisoners — as 20,000. In contrast, Macedonian losses, according to our sources, are unbelievably small. The highest cavalry losses recorded (Justin 11.6.12) are 120; Arrian (1.16.4) puts the figure at 60, including 25 Companions, while Plutarch (16.7), on the authority of Aristobulus, cites the 25 Companions alone. Infantry losses, on the testimony available, were even smaller: thirty, according to Arrian, no more than nine by Plutarch's and Justin's reckoning. The historian, remembering the circumstances in which the battle was putatively fought, may perhaps permit himself a brief smile of incredulity.

There is, however, one even more striking and paradoxical fact which instantly stands out about these figures. In an engagement where the Persians are often said to have relied exclusively on their cavalry, their heaviest losses — or so we are asked to believe — took place among the infantry. Yet according to the same sources, these troops, except for the Greek mercenaries, put up little resistance: they fled in a rout, and there was no pursuit (Plut. Alex. 16.6; Arrian 1.16.1–2). This would at once seem to dispose of those 10,000 corpses and 20,000 prisoners: the first law of propaganda is to make your story consistent. Yet in sharp contrast to this, the cavalry losses recorded are, as we shall see in a moment, perfectly plausible. What, one well may ask, lies behind so striking and blatant a discrepancy?

First, let us see if we can find any evidence from which the true size of the Persian forces can be deduced. Diodorus (17.19.4) gives Arsites' order of battle in some detail, certainly as regards the cavalry: whatever source he is here utilizing at least had access to Persian as well as to Macedonian records, if only in the form of captured intelligence-files (always presuming that such things existed in the fourth century, for which there is little evidence).⁴⁸ On the left wing was Memnon, with his Greek mercenaries: an exclusively mounted contingent, it is assumed. Next to him came Arsamenes with his Cilicians; then Arsites, commanding the Paphlagonians; then Spithridates, with the eastern cavalry from Hyrcania. At this point Diodorus has a moment of infuriating vagueness: the centre, he says, is also occupied by ‘other national cavalry contingents, numerous and picked for their valour’, τὸν δὲ μέσον τόπον ἐπεῖχον οἱ τῶν ἄλλων ἐθνῶν ἱππεῖς, πολλοὶ μὲν τὸν ἀριθμὸν ὄντες, ἐπίλεκτοι δὲ ταῖς ἀρεταῖς. Beyond them the right wing was held by 1,000 Medes, 2,000 Bactrians,⁴⁹ and 2,000 unidentified horsemen under Rheomithres.

If this catalogue is at all trustworthy, we can make a very fair guess at the size of the Persian cavalry arm. Seven regiments are named and described; the other ‘national contingents’ provided at least two more, probably three. We read of two that are 2,000 strong, and one of half that number. If we strike a (conservative) average of 1,500, we obtain a round total of about 15,000 — a median figure, as it happens, between the estimates given respectively by Diodorus and Arrian. Losses of 2,000+ or 2,500 (i.e. of 14–16 per cent) would be just about what one might expect.⁵⁰ When we turn to the infantry, however, it is a very different matter. To begin with, there can be no doubt that Arrian (or Ptolemy) has vastly exaggerated the numbers of mercenaries involved.⁵¹ When Memnonwas first commissioned by Darius, he got no more than 5,000 mercenaries; Polyaenus puts the figure as low as 4,000.⁵² It is unlikely that the troops at his disposal were substantially increased until he obtained the supreme command in western Asia Minor; and Darius lost no time in recalling what mercenaries he did have immediately after his death — which shows that, as a commodity, they were still in short supply.⁵³ Indeed, it was only in 333, when Alexander had already conquered most of Anatolia, that the Great King began recruiting in earnest. By the time of Issus he had arguably raised the number of mercenaries to 30,000, and the force on his payroll later reached an attested total of 50,000.⁵⁴

But in May 334, when Alexander reached the Granicus, it is doubtful whether Darius had more than 15,000 Greek mercenaries all told, in Egypt, Asia Minor, or anywhere else, including the eastern provinces. 5,000, in fact, would be just about what he could spare Memnon to deal with Parmenio's advance force, and it is doubtful whether, at this stage, he thought Philip's untried son dangerous enough to justify any further reinforcements. There are two additional points to bear in mind here. That Alexander massacred 18,000 out of 20,000 mercenaries at the Granicus is not an absolute impossibility per se; but it is, to say the least, unlikely.⁵⁵ The sack of Thebes, a far more general and unrestrained piece of mass-slaughter, produced a death-roll only one third the size;⁵⁶ even the butchery of the Athenians at the Assinarus was on a lesser scale.⁵⁷ Secondly, it is quite incredible, on any reckoning, that the Persians, with so wide a variety of units to draw upon, should have had no infantry whatsoever apart from the mercenaries; and indeed neither Aristobulus nor Diodorus' source assumed this to be the case.⁵⁸

On the other hand, if we are in search of hard figures, the case is almost hopeless. Arrian's 20,000 is the only remotely plausible estimate: we should not reject it out of hand because of Ptolemy's assertion that it consisted of mercenaries alone. But even this figure may well be too large. Justin's overall estimate of 600,000 is so ludicrously inflated that it suggests textual corruption rather than propaganda. At some point a scribe might well have misread  (30,000) in his Greek source as  (600,000); but though this would give us a very plausible round figure, it is not a theory on which one can build with any confidence. If we allow for an infantry force of, say, 15–16,000, of which up to one third were Greek mercenaries, that is about as close as we are likely to get.

Let us now turn back to the battle itself, as reported by Ptolemy and Aristobulus. Against Parmenio's considered advice, and amid general reluctance on the part of his Macedonians, Alexander disdainfully insisted on pressing home the attack (Arrian 1.13;Plut. Alex. 16.1–3). He then, according to Aristobulus (16.3), plunged precipitately into the river with no more than thirteen squadrons accompanying him. Ptolemy, on the other hand, makes him order his whole battle-line in a way that agrees with Diodorus' account of the dawn engagement,⁵⁹ and emphasizes at the same time the disposition of the Persians: lined up along the bank, cavalry to the fore, infantry in rear — again, duplicating Diodorus. One or the other of them, it is fair to assume, has mistaken his occasion. At this point, according to Ptolemy, there was a short pause, while both sides eyed each other and did nothing. Then Alexander sent the Scouts, the Paeonians, one Companion squadron and one file of infantry ahead, and followed in person at the head of the whole right wing, advancing obliquely with the current towards the Persian centre. This seems a far more deliberate and well-organized manoeuvre; it also sounds far more appropriate for a normal land-battle.

Both sources are in general agreement as to what happened next. The Macedonian spearhead found itself up against the Persian cavalry, who were, very gallantly but for no good apparent reason, doing a job that could have been done far better by Memnon'shoplites and light-armed javelin-men (cf. Fuller, above). Curiously, it is javelins (ἀκοντία, βέλη) which now rained down on them from the banks; the Persians are described as ἐσχκοντίζοντες, while the Macedonians resist with spears — δόρατα or ξυστά. When Alexander is struck⁶⁰ he is ἀκοντισθείς. We may note, however, that when there is a specific reference to the Persian cavalry, these are not, apparently, their weapons. They, like their Macedonian counterparts, use spears, δόρατα, and the sword, ξίφος, when their spears are broken. Some of them are also armed with the scimitar or sabre, κοπίς, a traditional cavalry weapon. Diodorus also mentions the σαυνίον.⁶¹ Only Ptolemy refers to παλτά in this context,⁶² and though it remains uncertain just what kind of spears or javelins these were, they are specifically associated with cavalry usage.

The initial attack suffered badly, as we might expect (how this setback is reconciled with the minuscule Macedonian casualty-list remains a mystery) and part of the credit for the repulse is specifically attributed to Memnon.⁶³ There follows another interesting discrepancy between Ptolemy's version and that of Aristobulus. While the cavalry was engaged upon this heroic hand-to-hand struggle, the latter tells us, ‘the Macedonian phalanx crossed the river and the infantry forces on both sides engaged’ (Plut.Alex. 16.6). But according to Ptolemy, the Persian infantry (whether mercenaries or not) remained in rear of the cavalry throughout. Which of them is telling the truth? And who (if Aristobulus is correct) are these ghostly foot-soldiers, with their javelins and darts, that we glimpse here for a moment (under Memnon's orders, it can scarcely be doubted), first resisting Alexander's cavalry charge, and then grappling with the phalanx: καὶ συνῆγον αἱ πεζαὶ δυνάμεις?⁶⁴ In the next sentence we read that they ‘did not resist vigorously, nor for a long time, but fled in a rout, all except the Greek mercenaries’ — a clear enough statement that Memnon's troops were not the only infantry fighting on the Persian side.

Ptolemy is at least consistent: according to his version, Alexander only dealt with the enemy infantry after the main cavalry engagement had been won — a view, be it noted, which is also that of Diodorus.⁶⁵ But Diodorus, as he makes very clear, is dealing with a battle which supposedly took place at dawn the following morning, and in very different circumstances: not across the river, but in the open plain on the far side — in campis Adrasteis, as Justin says (11.6.10); a small pointer, but not without its significance. Nevertheless, once Alexander and his men are up the further bank of the Granicus, and firmly established — it is just at this crucial point, suggestively enough, that the narratives of Ptolemy and Aristobulus become momentarily blurred in detail — the three accounts all go forward in close agreement. We have the famous duel between Alexander, Mithridates, and Rhosaces; Alexander's split-second rescue by Black Cleitus; the final rout and victory. Alexander himself is handled a little more roughly, a little less like the invincible hero, in Diodorus' version: at one point he seems actually to be down on the ground, with Spithridates and his royal kinsmen assailing him from all sides.⁶⁶ But that all three sources are from now on dealing with the same battle seems beyond dispute.

It will be convenient, before proceeding further, to recapitulate the facts that have emerged in the course of this investigation. Firstly, we have two separate (and on the face of it irreconcilable) accounts of the battle which Alexander fought at the Granicus. In the one he is advised to wait until dawn rather than launch an impossible frontal assault against heavy odds; he refuses the advice, attacks, and ultimately triumphs. In the other, he does wait till dawn. In the one he attacks across the river and up a steep bank on the farther side; in the other he gets his troops across unseen by the Persians (at least till the very last minute) and then fights a classic Macedonian-style engagement. In the one, both sides' tactics are ill-advised, and those of the Persians flatly incredible; in the other they are appropriate and excite no comment. Up to the crossing of the river, Ptolemy and Aristobulus disagree not only with Diodorus, but also, on occasion, with one another, in a way which suggests that they may well be suppressing vital evidence (e.g. the possible role played by Memnon's infantry during the initial assault). After the crossing, their account of the battle merges smoothly into that given by Diodorus, though the latter is, on the face of it, describing a quite different occasion. Lastly, we have the remarkable exaggeration of Persian infantry numbers and losses, together with a suggestion on Ptolemy's part that they were all Greek mercenaries; and, balancing this, an estimate of Macedonian losses so small that it can hardly be explained away as propaganda.Propaganda, after all, is meant to be believed.

What are we to make of all this? We may argue, and with some confidence, that Diodorus' version of events has a good deal more to be said for it than is generally allowed. This at once raises the question of why most scholars dismiss it out of hand. The most illuminating answer to this question is contained in Davis's criticisms of Beloch:⁶⁷

The Arrian — Plutarch version of the battle he dismisses as merely a romantic picture designed to exhibit Alexander in the light of a Homeric hero. What he is doing here is not merely preferring the poorer to the better authority; he is also setting the Granicus against the evidence of Alexander's whole career. He is making Parmenio out of Alexander the Great. Why should this be the one occasion when Alexander chose the more cautious over the bolder course? And it is impossible to explain either the rest of Alexander's career or the history of the years after his death if Alexander is reduced to a mere colorless competence. Alexander was a Homeric hero.

Now whatever our feelings about a mechanical reliance on ‘better’ as against ‘worse’ sources, we may willingly concede Davis's central point. The Diodorus account does indeed run counter to Alexander's known life-style in every possible way. But does that justify us in rejecting it out of hand, without further consideration? I think not. Circumstances may arise in which even an Alexander is forced to act against his own wishes, or, worse, to admit a serious error of judgement. On such an occasion his immediate instinct will be to falsify the record in his own interests. Our problem, I would submit, is a more complex one than merely deciding between two alternative traditions. What we are faced with here is deliberate, unmistakable, and systematic manipulation of the evidence.

Thus we cannot, like Gulliver, opt for one end or other of the egg, since propaganda (contrary to popular belief) avoids direct lies whenever possible. It normally prefers to save the appearances, aided by those two time-honoured devices suppressio veri andsuggestio falsi. The carefully slanted half-truth is far more effective than any mere fabrication, if only because it becomes much harder to expose for what it is.^a If we provisionally accept the hypothesis that our main account of the Granicus has been doctored to conceal some kind of initial failure, then a completely new light is shed not only on Alexander's behaviour, but also on the supposedly divergent testimonia, which it may prove possible to reconcile in an unlooked-for fashion. What we seem to have here is, on the one hand, the ‘official’ version of the Granicus battle; and on the other an independent account which, while accepting some of the ‘official’ record's more dubious claims (e.g. those concerning Persian infantry losses), nevertheless disagrees with it at several crucial points.

If we ask ourselves who was ultimately responsible for doctoring the record utilized by Ptolemy and Aristobulus (both of whom, incidentally, must have been well aware of the truth), the only possible answer is Alexander himself, aided in all likelihood byEumenes, his chief secretary, and the expedition's official historian, Callisthenes. So much seems clear enough. But our most important task is to find out not only how the truth was distorted, but also why. After all, the battle of the Granicus was won: that fact remains solid and undeniable. But it also poses an obvious dilemma. If Alexander won in the way suggested by Diodorus, why should he bother to make up a completely false version of events which does no credit to his strategic sense?⁶⁸ And if Ptolemy and Aristobulus are telling the truth, how did the eminently sane and unromantic account utilized by Diodorus ever get into circulation at all? Diodorus, significantly, makes the king out as Homeric a figure as anyone could wish during the actual battle (whenever and wherever that may have taken place); it is only beforehand that caution comes to the fore.

Here we may pertinently recall Davis's question: ‘Why should this be the one occasion when Alexander chose the more cautious over the bolder course?’ Might it not be that in the first instance he did nothing of the sort, but acted, characteristically, like the Homeric hero on whom he modelled himself, and with disastrous consequences? A hypothesis of two battles at the Granicus,⁶⁹ one, abortive, in the afternoon, the second, overwhelmingly successful, the following morning, would not only enable us to reconcile our conflicting evidence; it would also provide the strongest possible motive for Alexander to falsify the record afterwards. An initial defeat, at the very outset of his Asiatic campaign — even though recouped immediately afterwards — would make the worst possible impression, not least on the still undecided Greek cities of Asia Minor. Delphi had pronounced Alexander ἀνίκητος,⁷⁰ unconquerable, and ἀνίκητος he had to be, on every occasion. Herein lay the ultimate secret of his extraordinary personal charisma: the quasi-magical belief that he could not fail, that his leadership in itself guaranteed victory.

Now throughout his life, as we have seen,⁷¹ Alexander reacted very badly indeed to any direct thwarting of his will and ambition. His instinct was to destroy those who stood in his path; he would, if need be, wait years for an appropriate and satisfying revenge. A setback, even a temporary one, at the Granicus would bode ill for all persons responsible once victory had been secured. The most competent and experienced troops fighting on the Persian side were, of course, Memnon's Greek mercenaries. Can we regard Alexander's special, and singularly vicious, animus against this particular unit as mere coincidence? He slaughtered them wholesale, and sent the survivors, chained like felons, to forced labour in Macedonia, at a time when common sense would have suggested acquiring their valuable services for himself at preferential rates. Moreover, this was an isolated action: from then on he enrolled Greek mercenaries whenever he could get hold of them.⁷²

His ostensible reason (published by Ptolemy and accepted by most modern scholars) was that ‘they had violated Greek public opinion by fighting with orientals, as Greeks, against Greeks’ — ὅτι παρὰ τὰ κοινῇ δόξαντα τοῖς Ἕλλησιν Ἕλληνες ὂντες ἐναντία τῇ Ἑλλάδι ὑπὲρ τῶν βαρβάρων ἐμάχοντο.⁷³ In other words, he was making a gesture as captain-general of the league. But Greek public opinion was something of which Alexander took notice only when it suited him; and the league served him as a blanket excuse for various questionable or underhand actions, the destruction of Thebes (see above, pp. 147 ff.)⁷⁴ being merely the most notorious. A little good publicity in Greece never came amiss; but it is improbable, to say the least, that this was his primary motive.Aristobulus tells us that Alexander was ‘influenced more by anger than by reason’, θυμῷ μᾶλλον ἤ λογισμῷ,⁷⁵ and this sounds far more like the truth. His behaviour, indeed, bears all the signs of that terrible rage which could, at times, sweep away the last vestiges of his self-control, and was invariably caused by some personal insult, some thwarting of his destiny, some affront to his will, dignity, or honour.

The falsification of the record in this respect is highly suggestive. The infantry were made out to be more numerous than they were; in Ptolemy's account (see above) they are no mere Persian conscripts either, but highly trained mercenaries to the last man. We have already seen how improbable a claim this was. As propaganda, however, its meaning is clear. The threat which the Greek mercenaries represented was to be highly exaggerated, and the glory of overcoming them correspondingly increased. Yet at the same time any part they may have played in the actual crossing of the Granicus was to be deleted from the official account, even if it meant crediting the Persians with a wholly unbelievable battle-plan. This double reaction, coupled with Alexander's savage treatment of them afterwards, suggests that they somehow thwarted his plans in a way which showed him up in a very bad light, and which he was determined should be forgotten. In any case the odds against him were to be dramatically increased: if he had failed, he was determined to show that no mortal man could have succeeded.

Now if Alexander had in fact simply followed Parmenio's advice, crossed the river at dawn, and won his victory, there would have been no pressing need for him to invent the long dramatic rigmarole recounted by Ptolemy, with its wealth of circumstantial detail: the Macedonian panic, the intercalation of a calendar month, the argument with Parmenio, the details of that first suicidal assault across the river. These things really happened; and they happened in the late afternoon, just as Ptolemy says they did. If, at this point, we are prepared to argue that Diodorus' account islikewise substantially true, then the nature of Alexander's propaganda at the Granicus at once reveals itself, and all the apparently unmotivated discrepancies fall into place. Here, then, is a reconstruction of what I believe may have been the true course of events.

When Alexander reached the Granicus, he found that Arsites had made his dispositions not perversely but all too well. He did, indeed, have his cavalry along the river-bank, since this was by far his strongest native arm; but it was not alone. At the crossing-point itself he had placed Memnon's redoubtable mercenaries, just as any competent commander might be expected to do. The Persians knew the strength of their defences; they simply sat tight and waited to see whether Alexander (whose dashing reputation, clearly, had preceded him) would be rash enough to try a frontal assault. They had gauged their man well. Alexander was determined to cross the river at once; any further delaying tactics on the part of his officers would leave the man who used them facing a charge of cowardice, if not of treason.⁷⁶ For the second, and last, time in his life, the king's youthful impetuosity, coupled with the dire need to force an engagement at all costs, got the better of that cool strategic head. Parmenio suggested, hopefully, that the enemy might decamp during the night.⁷⁷ This, of course, was the one thing Alexander had to prevent, and it was probably a major factor in deciding him to reject his second-in-command's advice.

Besides, his Homeric destiny was summoning him to achieve heroic deeds, like his exemplar Achilles; and where better, here and now, than across the Granicus River, in the face of fearful odds? He charged headlong into the stream, and thirteen squadrons went with him. Perhaps the phalanx followed; just possibly it did not. There had been panic in the ranks; Parmenio's advice had been flouted; and almost every key command — including those of the Hypaspists and the Companion Cavalry — was held by one ofParmenio's sons, relatives, or personal nominees.⁷⁸ If there was a power-struggle between Alexander and Parmenio from the first, Burn asks,⁷⁹ why did the army not simply ‘make a Uriah’ of Alexander at the Granicus? Nothing, he adds, could have been easier. In fact, I would submit, they may well have attempted to do so; but Alexander, as his subsequent exploits make abundantly clear, had an even more remarkable talent for survival than his father Philip.⁸⁰

For a while, with furious resolve, Alexander and his squadrons battered at Memnon's mercenaries, while a deadly blizzard of javelins rained down on them.⁸¹ If other Macedonian units, whether of foot or horse, supported this attack, they still made very little headway. At last, forced to admit defeat, they turned back across the river. This is the central fact which Ptolemy and Aristobulus are at such pains to conceal. Alexander's first brush with the Persians had ended in humiliating failure. Worse still, Parmenio had been proved right; and with all the weight of his sixty-five years behind him, he would not be slow to emphasize the fact. Yet Alexander, though he never forgot or forgave an injury, was also a realist, who never lost sight of his ultimate goal. He swallowed his pride; it must have taken some doing. During that night the army marched downstream and forded the Granicus. Perhaps Alexander simply intimated to his staff that if the troops distinguished themselves in battle next morning the matter would be regarded as closed. After all, he had as much reason for wanting the first assault forgotten as anyone.

So, indeed, it turned out: the Macedonians, perhaps a little ashamed of themselves, won an overwhelming victory. But that, from Alexander's point of view, was by no means the end of the matter. There were scores to settle, and an episode to be hushed up. Not for several years yet would the king feel himself strong enough to try conclusions with that indispensable figure Parmenio;⁸² but Memnon's mercenaries, who had been instrumental in achieving his humiliation, were quite another matter. On them he took prompt and savage vengeance, camouflaging his personal motives by the pretence that he was executing justice on behalf of the Hellenic League. His initial débâcle may also provide a possible explanation for the minuscule size of the Macedonian casualty-lists in our sources. As an overall estimate they are ludicrous, a fact which every scholar has acknowledged. If the final battle took place in the way Ptolemy claims it did, by direct frontal assault, the one thing we can say with absolute assurance is that Alexander's losses would have been murderously heavy, almost on the scale of those suffered (in not dissimilar circumstances) by the Light Brigade during the Crimean War. But if we take them as the casualties suffered by the thirteen squadrons which charged across the river with Alexander, and by them alone, they at once fall into place — even down to the nine foot-soldiers, who will have belonged to that ‘one file of the infantry’, καὶ τῶν πεζῶν μίαν τάξιν, included in the spearhead.⁸³ Alexander had statues erected at Dium to the twenty-five Companions who fell at the Granicus — another unique gesture, never to be repeated: it is significant that all of them are said to have been killed ‘in the first assault’, ἐν τῇ πρώτῃ προσβολῇ.⁸⁴ To commemorate the faithful few, and them alone, would have been a superbly contemptuous gesture, very much in line with all we know of Alexander's character.

Now it only remained to put the record straight for propaganda purposes. There was no need to tamper with the final battle; only to transfer its setting. What had to be eliminated, at all costs, was that disastrous, ill-conceived, and humiliating initial charge. So the two separate engagements were run into one, and the scene of the final conflict changed from dawn to evening, from the Adrasteian plain to the river-bank of the Granicus. Callisthenes (or whoever was responsible) had to do the job in a hurry; small wonder that some loose ends and tell-tale inconsistencies remained, that the stitching of the join could be seen by those who cared to look for it. Memnon's role in the defence was carefully obliterated, though (as we have seen) not quite carefully enough; the Persian battle-plan was put, unchanged, into a new context which made it appear perverse to the point of insanity (itself an excellent piece of propaganda); and the king's deed of personal ἀρετή was increased beyond measure as a result.

No one would dare to publish the truth during Alexander's lifetime: too many high officials had connived at its falsification. Nor, indeed, was the real story one that reflected overmuch credit on anyone concerned — except, perhaps, on Parmenio. The battle had, after all, been won; and human memory is mercifully short. But discrepancies — mostly caused by unthinking adherence to the truth except at specifically sensitive points — were bound to find their way into the official version. Lastly, one of Diodorus' sources utilized a tradition which put on record the true facts of Alexander's dawn manoeuvres. The genesis of this tradition can be no more than a matter for speculation; but it appears, severely truncated, in Diodorus' own narrative, and is hinted at by Justin andPolyaenus.⁸⁵ If this hypothesis should be correct, it shows us the one occasion in his whole career when Alexander suffered a personal defeat — and by so doing renders him one degree more credible as a human being.

I do not for one moment suppose that the theory here put forward solves the enigma of the Granicus beyond any reasonable doubt, and I am well aware of the arguments that can be brought against it. Diodorus is a notoriously uncritical and unreliable source (or transmitter of sources); his contaminated account of the battle of Issus would hardly encourage one to accept him on the Granicus were it not for the (to me) unavoidable considerations advanced above. Nor, let me freely confess, do I find it intrinsically plausible that — as one of my more cogent critics represents the case I propose — of our two accounts one (Arrian) is a deliberate falsification, combining (roughly) the first half of the first battle with the second half of the second; while the other (Diodorus), coincidentally but by pure accident, omits the first battle and gives us only the second one'. I simply find this less unlikely than the alternative possibilities. Again while the motives of Ptolemy and Aristobulus in this matter are clear enough, why should any source hostile to Alexander not have instantly jumped on the first, abortive, attack on the Granicus, and given it maximum detrimental publicity — as indeed happened with so many other incidents, known to far fewer people, which Alexander's propagandists afterwards suppressed or distorted? To this question I can see no answer — any more than I understand how, supposing Arrian to be telling the truth, the Diodorus version (so much saner and more commonsensical by comparison) ever got launched. The one postulate raises just as many problems as the other. It may be that there was no botched afternoon attack, and that Alexander crossed at dawn without further demur. It may even be true (a point hard to determine without on-the-spot topographical investigation) that he forced the crossing ab initio, though I find this improbable, to say the least. But in either case the very real difficulties I have outlined still need to be explained. (It will not do, for instance, to dismiss Diodorus' account of the battle's preliminaries as a piece of rhetorical fiction straight from the Issus stock-pot. Alexander repeated his basic dispositions in almost every major battle he fought: the cliché, if cliché there be, is tactical rather than rhetorical.) I would claim no more than that my hypothesis answers more questions than it raises. Perhaps in the last resort Davis was right, and the enigma must be pronounced insoluble.