When Endymion, king of the Moon, devised war upon Phaethon, king of the Sun, he decreed that a race of spiders as big as the Cyclades should weave a web between Venus and his lunar dominion, to serve as the battlefield for their regal rumble. And in that region of the heavens he arrayed his army: the king himself led his elite Hippo-vultures in the clouds on the right wing, 80,000 strong; his other cavalry, mounted on giant birds with wings like lettuce leaves, held the left. The Moon's stalwart infantry held the centre, posted on the spider web: Millet-launchers and Garlic-fighters, and his light-armed Flea-archers and Wind-runners, whose long tunics carried them about like sailboats in the fierce winds of the celestial realm. To Endymion's Hippo-vultures, Phaethon opposed the Sun's Hippo-ants (and near two hundred feet long were the insects that bore these cavalry). On the opposite flank of the solar array came the Air-mosquitoes and the formidable radish-flinging Air-dancers. The spears of Phaethon's phalanx, in the centre, were stalks of asparagus, and their round shields were mushrooms. Phaethon's allies, the Cloud-centaurs, expected at any moment from the Milky Way, had not arrived in time for battle.
Signals were hoisted; donkeys (serving as trumpets) brayed; and the armies came together. On Endymion's right his terrifying Hippo-vultures turned their foes to flight before a blow was struck. But on the far flank Phaethon's Air-mosquitoes prevailed, and pursued the Moon's warriors towards the massed foot. There Endymion's infantry resisted stoutly, and there the Sun's rampaging right finally turned their backs, hastened in their panic by learning of the flight of their friends on the left. The victory was glorious, and many were taken alive and many slain. So much blood was shed that the clouds turned red – rather like at sunset, in fact, on earth.
But wait! Just as the triumphant army of the Moon was smugly erecting two trophies, one for the cavalry battle in the clouds, and one for the infantry struggle on the spider webs, scouts reported the coming of the Sun's belated ally Sagittarius, leading the gigantic Cloud-centaurs (their human torsos were the size of the Colossus of Rhodes). The Moon's soldiers had left their ranks as they scattered to plunder and pursue, and the formed Cloud-centaurs set upon them all disorderly, and put them to flight. They drove Endymion back behind his lunar bastions, slew his birds, threw down his trophies, and took prisoners – one of them the second-century ad satirist Lucian, from whom we have this True Story (Ver. hist. 1.12–18).
Lucian's humour in this satire depends on his ancient reader being familiar with – tired, indeed, to the point of irritation by – the conventions of battle description in the historians of ancient Greece and Rome: the pompous array of exotic troop types with curious attire and equipment, the blare of trumpets, the contrasting fates of the flanks, the reversals, the inevitable defeat of disorder by order.Footnote 1 And since the great preponderance of our information about both ancient combat in general and specific ancient battles comes from exactly those descriptions of battles in the historians of Greece and Rome whom the satirist was mocking, it is useful both for today's student of ancient warfare and today's student of the Greek and Roman historical authors to come to grips with the conventions of ancient battle description.
Accordingly, this article traces the evolution of the ‘battle piece’ in the Greek and Roman historical tradition.Footnote 2 It begins with a rude contrast between modern historical and ancient methods of describing battle; then investigates why ancient conventions developed as they did (a discussion split between Parts I and II); and finally considers how those conventions worked (or failed to) when applied to battles at sea. Like modern authors, ancient historians naturally differed in their treatment of battles, and proper justice is hardly done here to their idiosyncrasies (although these may be commented on in the notes) nor to the different ancient genres in which the authors worked (so proper ‘history’ as the ancients knew it must muck in with Caesar's Commentaries): but distinguishing branching streams and creeks is easier once the course of the main river – the practice of the majority – is understood.
The deeper question that this article addresses is the relationship between the desire of ancient historical authors to tell true stories about the past and the genre elements so prominent in their accounts of battles – genre elements some of which, as we shall see, became so powerful and entrenched as to encourage authors to write accounts more true of battles in general (broadly true) than true of the particular battle described, or sometimes to write accounts that were, in part or all, not true.Footnote 3
Some of the matter of this article will be familiar to or suspected by specialists on the classical authors discussed; but such specialists may nevertheless find use or pleasure in the collection and organization of their insights. Historians of ancient warfare, in contrast, are likely to be appalled – as was the author – by the degree to which ancient battle descriptions turn out to be mummified by convention. To historians this article is a cautionary tale, told by a resigned and disillusioned friend.
Battle description modern and ancient
Battle descriptions in today's histories are usually written backwards in a logical chain from the outcome of the battle. From the result of the battle, then, proceed in reverse order the fighting that created that result, the manoeuvres, the dispositions of the units of the armies that did that fighting, and, first of all, the plans of the commanders that disposed and set those units in motion (although the plans of one commander can, if a surprise lies in the future, be held back for dramatic effect). This strong logic disciplines the battle description: we hear of the climactic engagement, not what happens elsewhere; we hear of the units in at the kill, but rarely get a full account of the forces of either army; the terrain is described where it bears upon that decisive combat, but the rest of the battlefield is neglected. Similarly, differences in numbers or equipment between the contending sides, matters of supply or weather or chance, the quality of troops or weapons, or human foibles – stupidity, insubordination, over-boldness, cowardice – tend to appear in the account only where the main plot requires them, unless, of course, they offer comic anecdotes.
The main exception to this strong causal chain occurs when a contemporary author digresses to recount the experiences – the triumphs and tragedies – of single named soldiers or small groups of individuals, whose personal stories the author may know from letters or diaries or interviews, and whose adventures the author is likely to have been following from earlier in the narrative. The modern art of battle description consists particularly of the skilful mixing of shots from an imaginary camera hanging in the middle air that records the motions of the units whose fighting decides the battle – divisions, regiments, companies, platoons – and a lower-hanging camera that records the deeds of those individuals or small groups: enough shots from above to keep the reader oriented as to the general progress of the fighting, and to explain the result, but enough low camera following individual soldiers to provide human interest and prevent the account from becoming distant and abstract.Footnote 4 A camera hanging high in the heavens that can encompass the deeds of whole armies is available to the contemporary author, but is not frequently used: often just to record the strategic results of the battle, the casualties of both armies, and the retreat of the losers.
The battle descriptions of classical antiquity were very different.Footnote 5 And the most important contrast is that the modern logical chain is lacking, or attenuated; the elements that make up the account are present for their own purposes and do not drive, or discipline, each other.Footnote 6
The first structural element in an ancient battle description is not the general's plan but (as in Lucian's battle in the heavens) the arrays of the armies, which can be highly detailed. Sometimes the array is attributed to a general's plan, or presented as a reaction to the array of the enemy, but for the most part it is simply told for its own sake, and has no relationship to the enemy array.Footnote 7 Nor does the modern discipline operate that encourages the author to describe only the units he will need later in the battle narrative. Most contingents in ancient arrays (as again in Lucian's battle) are described there only, and play no role in the action that follows. There is a real sense that an array in an ancient battle description is an independent work of art, a passage of prose to be admired for itself, not to be put to work later.Footnote 8
Usually after the array (but sometimes before), the author can write for one or both opposing generals speeches encouraging their troops.Footnote 9 And that, and no more, is what those speeches usually do. Generals’ speeches (written in the author's own style with no, or only the slightest, nod to making them sound as if they were spoken by, for example, a Spartan or Carthaginian), adhering to loose, little-changing formulae, list the reasons why the soldiers should be confident of victory and urge them to fight hard. ‘Fight for your hearths and families!’ the defenders are told. ‘There is no place to retreat to!’ the aggressors are reminded. ‘Remember your ancestors!’ ‘Remember your past victories!’ ‘The gods are with you!’ Where both opposing generals are given speeches, they only rarely answer each other, as do the speeches historians write for political debates; nor does the author regularly foreshadow the outcome of the battle by giving the general who will win a more effective speech than the general who will lose. Sometimes the author mentions the effect that a speech had on the troops to whom it was addressed, but usually not.Footnote 10 A German scholar counted some seventy-five major battle speeches in the Greek and Latin historians from Thucydides to Procopius.Footnote 11 But after Thucydides there is a desperate sameness to the sentiments rehearsed in these speeches and many modern readers find them among the most ungratifying products of ancient literature. The tradition of ancient battle speeches is like Saehrimnir, the Swine of Asgard, from whom chops are cut each day to feed the heroes who feast in Valhalla, and who puts forth new chops the next day: Ever generous of his meat is Saehrimnir, but the meat is always pork.
Once array and speeches have been presented, it is time for the actual fighting. Ancient authors and readers preferred a low camera matched with a higher camera than today's authors use: the deeds of whole armies, mixed with the fates and deeds – ideally the heroic deeds – of named individuals.Footnote 12 Depiction of the motions of the contingents so lovingly drawn up in the array – with the middle height, unit-manoeuvre-recording camera modern authors depend on so heavily – is not unknown (Polyb. 5.82–5; Procop. 1.14.44–7), but it is not common,Footnote 13 and when it does happen, the motion of a contingent is frequently made the function of a named individual who heroically leads it into action: ‘at this point one may without question call Agesilaus brave’.Footnote 14 Sometimes, indeed, a battle description is no more than a series of episodes involving the fates and heroics of individuals,Footnote 15 but, even if not, there is at all times a strong tendency to make as many greater events as possible hang upon the acts of great personages.Footnote 16 The proper parallel is to the depiction of pre-gunpowder battles in film: the armies charge in masses but, once contact is made, we are shown only small-scale combats, usually involving the movie's main characters.
Despite Polybius’ insistence (in denouncing an earlier author, 12.25e.1) on the necessity of the historian's possessing, and giving, a good sense of topography, discussion of battlefield terrain is rare, even in Polybius. Where it does happen (because a detail of terrain was highly important to the unfolding of the battle: Polyb. 3.83–4; Caes. B Gall. 2.18), the reader is usually told only the minimum needed to make sense of the account.Footnote 17 Here the methods of the ancients and the moderns are close, but not their motives. For, while an extensive general discussion of topography is driven out of modern battle descriptions by the discipline enforced by the logical chain of cause and effect, it is likely driven out of ancient battle description by a strong literary preference for single combats and depiction of small-scale episodes in the battle that allow for enargeia (‘vividness’, ‘immediacy’), excitement, fine writing, reversals, paradox, and, where possible, elephants.Footnote 18 And if the ancient author has no data amenable to such treatment, and is not inclined to make any up (as at App. Pun. 45–6; cf. Ach. Tat. 4.14) or borrow appealing ‘fighting filler’ from another author (both extremely common practices, alas), the actual description of combat can be extremely terse. ‘Putting shields against shields they pushed, they fought, they killed, and were killed’ (Xen. Hell. 4.3.19).Footnote 19 A striking disproportion in length between the pre-combat material (array, speeches) and the fighting itself is often, to the puzzlement of the modern reader, the result.
Where the depiction of the fighting is extended, patterns can sometimes be discerned. Local triumphs on one side can alternate with local triumphs on the other, allowing an exciting see-saw effect (as in Lucian's fantasy battle).Footnote 20 Or the army from whose point of view the story is told can progress through the action, overcoming a series of steadily more formidable obstacles, sometimes each with its own crisis – ‘here they would have been defeated, except that’ – and dramatic reversal.Footnote 21 This latter especially happens where (as in Livy, or the Alexander historians) the author chooses to tell the story of a battle strongly emphasizing the experience of one side (the Romans, the Macedonians), and so the account of the action itself naturally follows the deeds of that side.
The internal patterning of the action, as well as the mix of very high and very low camera that usually records the actual combat in ancient authors, can also, to our minds, have the result of isolating in the narrative the climax of the battle: the conventional moment when the ancient author ‘weighs the events, as it were, in the scales’, and states why one side won, why the opposing army was turned to flight – when he thumps his finger down on the crucial quality that allowed one side to prevail over the other.Footnote 22 This weighing is the point in the whole battle description where the author is most likely to enter his own text, appealing directly to his audience with his reasoning (Hdt. 8.86). But frequently there is no relationship between this climactic comparison and anything that has come before in the battle description; the author simply states that one side proved braver, or that the other fell into confusion.Footnote 23 Sometimes it develops in retrospect that a manoeuvre was decisive. But, if so, the author may simply incorporate a description of that manoeuvre (and even introduce the contingent making it) in his account of the climax (Tac. Agr. 37.2; Amm. Marc. 16.12.49–51). The accounts of the fighting and weighing can (to our sensibility at least) conflict with each other: for where an author has depicted the action in terms of one side overcoming a series of obstacles, the modern readers (with their expectations of the natural result of such a rising crescendo) can be quite perplexed when, in the end, the side that has overcome ever-greater obstacles wins the battle for reasons unrelated to those obstacles, or actually ends up losing (Livy 22.4–7).
Once one side has been turned to flight, the author describes the battle's results. The mass pursuit (as in Lucian's fantasy battle) is usually dealt with briefly and from a high-camera perspective, although the camera can swing down to record an adventure or tragedy of small groups or individuals (such as the capture of Lucian).Footnote 24 If the historical combatants had a formal way of indicating victory in battle – among the classical Greeks, the losers appealing for a truce to recover the bodies of their dead and the victors erecting their trophy – that process is usually explicitly recounted, even if those formalities have been told many times before in the author's descriptions of other battles, and even if, from a standpoint of strict narrative economy, the result of the battle is already obvious to the reader. It is far more common to give numbers of dead in a battle than to give the number of those who came to fight; both the Greeks and the Romans tallied the dead, and these numbers were thought worth reproducing (or sometimes inventing) as indicators of the magnitude of victory.
Although not all were used in every battle description, it is easy to enumerate the structural elements from which ancient battle descriptions were assembled: array, speeches, action, weighing, results. But what needs re-emphasis is the logical independence of each part from the others. The speeches rarely explain (or even discuss) the array, or drive or help to explain the subsequent action. The details of the array only rarely have significant consequences in the action, and even less frequently in the weighing. The action can, but need not, have a causal connection to the weighing; and although the weighing logically decides the battle, and so causes the results, the results are a structural element all their own, and the author usually stops and draws breath before recounting them as a sounding coda, rather than making his story flow directly from weighing to result. There is, in general, a great deal of standing up and taking deep breaths, and rather obvious and ceremonious transitioning from one stage of the battle description to the next. We moderns value a seamless description of battle, a jazz battle; even the transitions from the middle-camera tactical chain that leads back from the results of the battle to low-camera human interest we prefer to be accomplished as invisibly as possible. But Greek and Roman battle description was like a classical symphony, where the orchestra goes quiet between clearly defined movements, giving the audience time for a good cough.
A literary history of ancient battle description
Fighting
Homer, of course, was at the root of both the Greek art of war and the Greek art of battle description.Footnote 25 His cultural authority was supreme, and a great part of the Iliad is given over to descriptions of fighting. The Iliad was thus the inevitable model when later Greeks wondered how to describe combat in writing; we could guess its influence even if we could not see it so clearly in early Greek prose, as we can, for example, in Herodotus’ account of the Battle of Thermopylae.Footnote 26 And we could guess the same even if we did not notice Homeric habits lovingly reproduced in historians’ battle descriptions for nearly a millennium: indifference to terrain, for example, and the emphasis on the din of battle – the trumpets, war cries, paeans, shouting, screaming, clash of weapons – taken over from Homer's noisy engagements.Footnote 27 As such, it is natural to begin an account of the evolution of the battle pieces in the ancient historians at the point of greatest Homeric influence – the depiction of combat – and proceed to lesser: first to Herodotus’ transformation of the Homeric catalogue into the array; then to Herodotus’ independent invention of the weighing passage; then (in Part II of this article) Thucydides’ creations of the battle speech and the results-of-battle passage. But this is not a history of authors imitating and adapting and innovating in a world purely of words, of birds twittering by a watering hole. The crocodile of reality often rises from beneath brown waters and sets the birds to flight.
The Iliad is the story of heroes, and much of the depiction of combat in the Iliad is that of a single hero against another – or a chain of others, or (more rarely) a messy mass of minor opponents. But the poem also values the context of these fights, so from time to time the camera rises up from the individual duelling to take in the wider battle, and, once the listener is properly oriented, swoops down again into the fighting of individuals.Footnote 28 Here lies the origin of the favoured ancient way of depicting the action of combat: as a sequence of close-ups of the deeds of individuals, interleaved with orienting shots from a lofty vantage – the mix of very low camera and very high.Footnote 29
In Homer the camera does not usually focus upon the events between the scale of the individual and the army because groups of warriors smaller than entire armies are usually brought on stage as a function of an individual hero who leads them into action; and upon that heroic leader, rather than his followers, the camera remains.Footnote 30 In likely consequence, the narrative dependence of the group on the individual was always, in a stronger or weaker form, characteristic of later ancient prose battle description, and later writers reproduce the tendency for contingents, when they are singled out during the action, to be led by a named individual. Thus the activities of the unit as unit, the group of ten, or the group of fifty, rarely feature: of the Spartan enomotia, or ‘sworn band’ of perhaps forty, we see nothing in battle descriptions, nor of the Roman contubernium (‘tentful’ of eight), nor of the Roman century, and we see very little of the Roman maniple of 120. The camera simply does not hang at that middle distance over the battle, following the deeds of units. And this ancient narrative habit has left us in profound ignorance of the mechanics of ancient combat, as military historians are quick to complain.Footnote 31
There are, of course, scenes of mass fighting in Homer, and they, too, left their mark on the battle descriptions in the historians. In Homer the presentation of mass action is usually generic, generalizing – ‘now as these advancing came to one place and encountered, they dashed their shields together and their spears, and the strength of armoured men in bronze, and the shields massive in the middle clashed against each other’ (Il. 4.446–9; trans. Lattimore). And so too in Tacitus: ‘sometimes, in the way of cavalry fights, there were alternate attacks and retreats; sometimes, as if they were fighting in a proper line of battle, with bodies and the press of weapons they pushed forward and were pushed back’ (Tac. Ann. 6.35.1).Footnote 32
Evident in that same passage of the Iliad is the way in which masses of warriors in Homer tend to behave like robot soldiers all performing the same martial deed at once (see also Il. 11.67–91). This Homeric narrative habit, mass action portrayed as the simultaneous deeds of the members of those masses – ‘now our men resisted, now they gave ground, and the most experienced of the barbarian warriors tried to force the enemy back by the pressure of their knees, but with supreme stubbornness they fought hand to hand, and shield-boss thudded on shield-boss’ (Amm. Marc. 16.12.37) – is exactly the habit we often find in the descriptions of combat in later authors. Later authors could, of course, have produced generalizing, robotic accounts of combat without Homeric influence, but Homer rode the historians like the parrot on the shoulder of Long John Silver: when money came to the pirate's mind it was apt to be ‘Pieces of Eight!’Footnote 33
Finally, as he does with the passage above, Homer repeats his descriptions of mass fighting (see also Il. 8.60–3), and as an oral poet he endlessly repeats the same words and phrases in his depictions of single combat.Footnote 34 Licensed by this, we guess, and by the generic, generalizing quality of Homeric scenes of mass combat, battle scenes in the ancient historians were subject to being stuffed with generic, generalizing ‘fighting filler’ – sometimes borrowed like old clothes from other authors, or even (as in Homer) from other passages in the same author, or from half-remembered rhetorical training in boyhood, where ‘battle description’ was a topic (briefly) taught in school – to make the action longer, fuller, and more exciting.Footnote 35 The modern reader is best advised to accept such passages in the spirit in which they were written, as offering at best broad accuracy, rather than a linked-to-time-and-place account of actual events. So when Tacitus describes Rome's Batavian cohorts at the Battle of Mons Graupius striking the Britons with the bosses of their shields (Agr. 36.2), a deed we encounter frequently in other Latin authors, today's reader is far safer accepting such buffeting with shields as something Roman soldiers in general did in combat, rather than something that necessarily happened at that moment during the real Battle of Mons Graupius.Footnote 36
If Homeric camera work had powerful consequences for later battle descriptions, so did other aspects of Homeric narrative. Just as in Homer, later battle description betrays a certain intolerance of going back in narrative time to pick up another strand in a story – ‘they had prevailed upon the right, but before that, on the left’. This quality of narrative in Homer – known as Zielinski's Law – was certainly subject to exceptions, as scholars have pointed out. But its equivocal status as an iron law need not detain us: it was at least a Homeric habit.Footnote 37 Did the same habit in the ancient historians result from imitating Homer, or did Homer and the historians find the same solution to a similar literary challenge? No matter. But certainly events in ancient battle descriptions, and especially in the action itself, do tend to be presented as happening in a sequence one after another (the action in Lucian's battle is typical), even when, if we think about it, they must have been happening simultaneously (Xen. Hell. 4.2.20–3, 4.3.17–19). Even when simultaneity is asserted (‘meanwhile’), the author still behaves as if events came in chronological sequence – subsequent, non-simultaneous events nearly always following from the last of the ‘simultaneous’ events (Polyb. 1.27–28; Caes. B Gall. 2.23–7).Footnote 38
Other features of the structure of the action in the historians have Homeric resonances as well. An alternation of successes in historians’ accounts of the combat recalls the alternation of heroic killings in many Homeric battle scenes, and the competing narrative habit of depicting one side overcoming successive obstacles evokes epic's way of depicting the aristeia of a Homeric hero.Footnote 39 But here again it is hard to say whether later authors were imitating Homer, or just drawing, in common with Homer, on story-telling modes. Very likely both.
As fundamental as Homer was to setting the conventions for the literary depiction of combat, he had a co-conspirator, and that co-conspirator was the reality of ancient fighting. Later ancient authors’ tendency to describe the action of armies as unities also reflected the fact that, more often than not, ancient armies really did operate as unities, and the individual contingents of which they were made up did not manoeuvre independently. Any student of military history can think of ancient land battles where a general's deployment or manoeuvring of his army proved important or decisive – Epaminondas at Leuctra, Hannibal at Cannae. But our ability to count up such commanders’ battles on our fingers reveals how rare they actually appear to have been.Footnote 40 For the most part, from the fifth century bc to the sixth century ad, ancient armies, deployed in a line before the action commenced, simply walked or ran straight ahead at the signal, and each soldier fought the enemy soldier he met. Manoeuvres ordered in mid-battle – such as Alexander's turning around his cavalry at Gaugamela (Arr. Anab. 3.15.1) or the Roman flank attack at Cynoscephalae (Polyb. 18.25)Footnote 41 – are even rarer, and were usually a desperate reaction to looming disaster. No ancient general wanted to manoeuvre his army dynamically, commanding new motions during the course of the fighting. And for the most part, too, the various contingents of which an ancient army was made up shared each other's fate. For, although it might be possible to point out where in the great line a rout began, when a rout did begin, it tended to infect the whole army and carry all contingents in flight from the field. Again, stories survive of exceptions – contingents hanging on grimly as their comrades fled around them (Thuc. 4.96.3; Xen. Hell. 4.2.20) – and sometimes the right flank of each army would rout the other army's left, producing a pin-wheel effect.Footnote 42 But such outcomes are marked as unusual.
There were good practical reasons for the fact that an ancient army usually acted as a single vast organism, bent upon the simplest possible frontal attack on the enemy. Communications were poor, and chains of command weak or nearly non-existent in some armies. Not all ancient armies were formally trained, or trained well, and even in those that took systematic training – the Spartans, the Romans of the empire – the degree of mass drill (rather than individual fitness or weapons training) and the standard of drill achieved may have been substantially lower than that of European armies after 1500.Footnote 43 But it is easy to make too much of the practical obstacles to dynamic command in antiquity. European generals faced many of these same difficulties in pre-twentieth-century Europe but, despite them, Europe evolved the culture of aggressive manoeuvre to which we are heir.
In fact, the major reasons why the ancients usually fought simple battles were probably less practical than cultural and intellectual. First, most ancient generals participated in the expectation of most ancient warriors at all periods that battle was, at root, not a contest of intellect but a contest of courage, and so yearned, as Livy has a Roman general express it, ‘to fight on a clear and broad field, where, without any fear of ambush, the affair could be conducted on the basis of true courage’.Footnote 44 A battle fought otherwise – with manoeuvres and cunning stratagems – deprived soldier and general alike of the chance to display their courage and that of their city or nation, and victory by such means might leave the losers without a psychological sense of their own defeat, and so deprive the victory of its wider significance.Footnote 45 ‘Who wins is not a victor unless the defeated admits it’, as the poet Ennius put it.Footnote 46
Ancient commanders also dreaded lest their armies or fleets fall into disorder (ataxia in Greek). They feared this on the basis of experience, but also on the basis of their reading. ‘The whole Punic army fell into disorder’, wrote Polybius, ‘and there was an utter rout’ (1.19.10–11). Ataxia is the most common cause of defeat in Thucydides, and remains an overwhelmingly common peril in subsequent Greek and Roman accounts of battles, so common as to be mocked by Lucian in his fantasy battle.Footnote 47 Disorder was a physical getting-of-things-out-of-place in an army or fleet, but it led by an almost ineluctable process to dire psychological consequences, to flight and defeat.Footnote 48 The ancient conception of tactics – of military ‘ordering’ in Greek – was primarily defensive: not a matter of getting the better of the enemy, but of preventing your own army from falling into disorder. Tactics are like music, a Greek military writer might muse: their purpose is to produce harmony (Onasander 10.3). In the Roman Vegetius, too, tactics are about preventing disorder, preventing panic (1.9, 1.26). The surviving Greek tactical manuals spend very little time on teaching their reader how to manoeuvre against the enemy, and mostly discuss how an army ought to be subdivided (and the terminology for those subdivisions and their commanders), and give instructions for putting troops into formations and the proper commands for changing those formations in place: instructions, in other words, for avoiding disorder.Footnote 49 Generals’ fear of disorder covered the ancient battlefield with mental molasses, and made ancient commanders extremely reluctant either to dispose their troops on the battlefield in an unusual way or to attempt to manoeuvre their troops in action – to inspire, as they would think, intentional disorder in their forces.Footnote 50
Practicality in a world of – to our minds – often undertrained and under-officered armies; cultural preferences for ‘straight-up’ battles of courage against courage; and a dread of disorder and its consequences: all these things combined to make the reality of ancient land battle surprisingly simple by modern standards. Ancient armies did for the most part act as unities, and ancient habits of battle description, which usually depicted them that way, only reflected that greater truth.
Emphasis on battle conceived as an arena to display courage reminds us that the literary habit of representing the heroics of individuals on the battlefield was undergirded in every ancient generation by the fact of real individual heroics performed on real battlefields, whether by supreme commanders, officers, or common soldiers. All but the first can be proved independently of the battle descriptions themselves by mentions elsewhere in Greek writings of warriors receiving awards for being the ‘best’ in battle and being honoured for courage, mentions eventually confirmed by such honours appearing on the heroes’ tombstones, and of Romans receiving decorations for valour – of which the Romans evolved an elaborate hierarchy – or promotion in the army ‘for courage’, awards mentioned here and there in surviving writings but overwhelmingly confirmed by hundreds of Roman gravestones on which decorations are boasted of or depicted.Footnote 51 Scattered details – criticism of the practice, a list of the eight wounds suffered by Alexander the Great (Plut. Mor. 327A–B), Roman politicians’ habit of disrobing to show their war wounds, and the legal right of Romans to hang on their houses spoils they had personally taken in war, there to remain even if the houses were sold – imply, again without reference to battle descriptions, that high commanders often also fought in a heroic mode.Footnote 52 Nor should we be surprised: why should not Greek warriors drink from the same spring of inspiration – the Iliad – as Greek authors? And if the Romans were less likely to be influenced directly by the Iliad, they had their own tradition of heroic single combat, also attested in sources other than battle descriptions, and had a powerful culture of creating, and then imitating, heroic exempla. Roman generals, if they did not imitate Achilles directly, in some periods imitated him at second hand by imitating Alexander the Great.Footnote 53
Array
The Iliad, too, is the origin of the battle array depicted in the ancient historians, but in an odd and unexpected way. For Homer was never very interested in offering systematic depictions of the relative positions of his heroes during battles. Even when he divides up the Trojans into five contingents to attack the wall around the Achaean ships (Il. 12.80–107), he does not place them in space relative to one another, and it takes a very acute reading of the attack itself to realize that the first-listed contingent will attack the centre, the second, third, and fourth will be on the left, and the fifth on the right.Footnote 54 Epic exercised a severe discipline over descriptions of the dispositions of troops (rather like that in our own written battle descriptions): the listener was given what he needed to know to follow the subsequent fighting, and no more.
When Herodotus, then, faced the prospect of describing the array of the many Greek contingents at Plataea, he could not simply copy arrays in Homer. He possessed a long list of the Greek contingents, their respective numbers, and their places in the line, although he had no information about the role, if any, that most of them played in the subsequent battle. How much of that list to include? Herodotus put in all of it. For there was another Homeric model available, one that Herodotus had already borrowed for describing the army of the Persians that invaded Greece (Hdt. 7.61–99), and for his list of states contributing to the Greek fleet at Salamis (Hdt. 8.43–8). This was the Homeric catalogue, and especially the long Catalogue of Ships (Il. 2.494–759), with its list of the number of ships each Achaean chieftain led to Troy.Footnote 55
Now the Iliadic Catalogue of Ships – closely followed by a shorter Catalogue of Trojans (Il. 2.815–77) – is a thing in itself, not subordinate to what comes before and after it in the poem, its independence marked by the poet giving it its own invocation to the Muses (Il. 2.484–93). Although not entirely useless for orienting the reader for events afterwards, that was hardly a catalogue's primary purpose; nor was there any expectation on the part of the listener that all of the persons or contingents listed in a Homeric catalogue would reappear later in the poem; nor did a catalogue react to another catalogue; nor did a catalogue move the story along. Indeed, when Thucydides offered his own tribute to Homer by including a catalogue of Athenian and Syracusan allies (7.57–8) as a prelude to the climactic battle in the Great Harbour at Syracuse, he seems to have valued the catalogue in part for creating a tension-heightening delay in the action – narrative retardation, as our literary colleagues call it.
The decision of Herodotus to model his Plataea array upon a Homeric catalogue had profound consequences for his narrative, and for authors after him, either emulating him or drawing independently upon Homer. For it established the array of contingents as a separate and elaborate element of battle description, an element whose details (the individual contingents) had no right to be mentioned subsequently, which was rarely related to the enemy array, and which was quite independent of the fighting itself and not logically subordinate to it: it removed the array from the Homeric discipline of telling the audience only what they needed to know.
So it is that at Herodotus’ Plataea, the arrays of both sides (the Greeks at length, the Persians more briefly, 9.28–32) are described at a point a full ten days before the battle proper is joined – ten days full of events, which take several pages to recount (9.33–50; cf. Procop. 1.13.19–1.14.34). Last of those events is the decision of the Greeks (denied access to the river for water, and their supply lines threatened by the Persians) to withdraw into a protected position between two branches of a river – and so to dissolve the very array Herodotus had described (9.51). This they do at night, in grave confusion, except that one Spartan contingent refuses to retreat, is set upon by the Persians in the morning, and must be rescued by the rest of the Spartans. That fight between the Persians and the isolated Spartans decides the battle (9.52–68). Of the twenty Greek contingents lovingly catalogued by Herodotus during his array, only three – the Spartans, their friends the Tegeans, and the Athenians (who are allowed their own little victory over the Thebans, but the reader is not told where it happens) – even appear in the description of the fighting (although the Megarians and Phliasians are mentioned later because they come to grief during the pursuit, and the Corinthians are alluded to as cowering in the hills [9.69]).
Herodotus’ tale of the Greek array at Plataea is, then, a literary artefact radically separate from the battle itself that follows.Footnote 56 Even though the Greeks do not, in the end, seem to have fought in that array at all, Herodotus felt it worth recording for its own sake, just as Homer provided the Catalogue of Ships for its own sake.Footnote 57 And what Herodotus created, his successors received from him – or themselves borrowed from Homer.Footnote 58 A strict dependence of the array of contingents on the subsequent action – the latter imposing a criterion of relevance on the former – never developed. The phalanx of the Boeotians at Thucydides’ Battle of Delium (4.93) is made up of eight contingents, he tells us, but only two of them – the victorious Thebans and the doomed men of Thespiae – are accorded any description of their contribution in the fighting that follows (4.96). Xenophon draws up the Athenian fleet at the Battle of Arginusae in twelve contingents, but only one of them is named in his description of the action (1.6.29–33; cf. 4.2.16–23, and Dion. Hal. Ant. Rom. 20.1–3). A battle description like that of Thucydides’ Mantinea (5.67), where, of the sixteen contingents arrayed on both sides, ten are picked up again in the action that follows, and the author gives the impression that he knows, and cares, about their locations as the battle develops, stands out for its rarity. Lucian's fantasy battle, where, of eleven contingents described in the array, only two are named in the subsequent action, is far more typical.
The most elaborated depiction of arrays tended to be driven not by what the reader needed to know to follow the subsequent action but (where a historian had the necessary information or imagination), at least in part, by the presence of troops differentiated and exotic enough to give his readers pleasure from hearing about them. Naturally, then, the golden age of the elaborate array in the historians is the era after Alexander, when the battles of his successors offered a crowd of strange troop types from distant, wonderful places, with scythe-bearing chariots and elephants, and at sea stately quadriremes and quinqueremes, sixes and sevens.Footnote 59 And this is, of course, the sort of long, loving array of curiosities that Lucian was parodying in his fantasy battle. On the other hand, when an author chose strongly to emphasize the experience of one of the armies at the expense of the other, that side might be given a full array and the opposition a more cursory one, as sometimes happens to Alexander's opponents, despite their appealing exoticism (Arr. Anab. 2.8.3–2.9.4, 3.11.3–3.12.5). A fuller array might be given, too, when the historian felt that the dignity of the battle deserved extended, solemn treatment – a principle well illustrated by the desperate attempt of Diodorus to gin up a Macedonian array for the climactic battle where Greek freedom was lost, Chaeronea (Diod. Sic. 16.86), but where his sources provided no information, reducing Diodorus to guessing that (er…) King Philip commanded ‘picked men’ on one wing (but he does not say which one), and his son, Alexander, ruled the other, assisted by (um…) ‘the most famous of his generals’, and the (also unnamed) units of the rest of the army were stationed (ah…) ‘where the occasion demanded’.Footnote 60
Homer's catalogues were one fist of the one–two punch that gave his literary successors the pre-battle array. The other is an unexpected Homeric roundhouse that passed through reality on its way to thump the historical tradition. Arraying soldiers is, in the Iliad, an excellence in which the heroes compete, an areté. Long did the Athenians gloat that the poet had sung of Menestheus, the otherwise forgettable leader of their obscure contingent against Troy, that ‘never on earth before had there been a man born like him for the arrangement in order of horses and shielded fighters. Nestor alone could challenge him, since he was far older’ (Il. 2.553–5, trans. Lattimore). And in generations after, when tradition or a system of precedence among contingents did not tie their hands, commanders continued to compete in the heroic quality of arraying, leaders in the wars of the successors to Alexander ‘vying with each other in this skill’.Footnote 61 It was thus by mentioning the act of arraying or depicting the array that an ancient author displayed the general's skill in ‘tactics’ – ‘ordering’ – one of the two main branches of ancient military thinking.Footnote 62 For an author to depict an array in high detail, therefore, was not merely decorative, but, where a general had planned it, the account of the array displayed to the reader the excellence of the general.Footnote 63 A great captain such as Philip II simply had to be seen to array his troops for his greatest battle, Chaeronea. So Diodorus invented an array for him.
Weighing
After array and action comes the climax of the battle, the weighing, the moment where one side is said to defeat the other, and where very frequently the author intervenes in his text to give a summarizing reason for the victory. We see such interventions first in Herodotus, at Salamis and Plataea (Hdt. 8.86, 9.62–3).Footnote 64 In this Herodotus did not, probably, borrow from Homer. Since in Homeric narrative the actions of masses are usually dependent on the actions of heroes leading them or fighting against them – for the most part the masses are stage machinery for the heroes – in Homer it is the heroes who cause the great changes in momentum in the battle.Footnote 65 Thus Homer usually has no need to stand outside his own account and argue that, despite the heroism of the warriors on both sides, the Greeks prevailed because of their better order (at Salamis) or superior equipment and experience (at Plataea), or to note that Spartan bravery prevailed over the superior tactics of Sparta's enemies (at Mantinea; Thuc. 5.72.2).
Herodotus needs to weigh the two sides because what has passed during the action of the battle as he has described it – for the most part a series of small-scale adventures – is viewed with too low a camera to account for the outcome of the battle on a larger scale. The weighing can thus be elided in an account where the battle result flows more easily from the action (as at Marathon, Hdt. 6.111–13, where the successes of the Athenian flanks prove decisive; compare Dion. Hal. Ant. Rom. 6.33; Polyb. 1.26–8), or from a military trick.Footnote 66 Nor is the summary comparison needed where an author pursues the Homeric course of granting to small-scale adventures the power to decide the battle, when the heroic deed of an individual, for example, produces the snowballing rout of an entire enemy army.Footnote 67 Nor yet is a weighing needed where the author wishes to imply a cosmic level of causation – fate, the will of the gods, the destiny of the Romans – operating at a level above the details of the battle, as sometimes appears to be the case in Livy.Footnote 68
But the place and continuing importance of the weighing in the canon of elements of the ancient battle description are emphasized by its prominence in the ‘short-form’ battle descriptions of the historians – those countless passages where an author chooses not to present a full account of a battle. Here the author usually does no more than state the fact of the battle, and that one side defeated the other – but he frequently adds a few words about the reason for that defeat: that one side was braver, or more experienced, or greater in number.Footnote 69 Even when the battle description is compressed nearly to nothing, the notice of why one side prevailed over the other is one of the last things to be squeezed out.
The weighing did not, however, need to be brief. In the hands of Polybius (also the author of a lost book on tactics), it can grow into a long passage of technical military analysis, and (as at Drepanum [Polyb. 1.51] or Clastidium [Polyb. 2.33; cf. 1.33–4, 1.61, 2.30]) can swell to dwarf the rest of the battle description, and metastasize through it and into the tissue around it. Polybius’ famous contrast of the Macedonian phalanx and Roman legion (18.28–32) is just such a weighing passage, appended to his description of Cynoscephalae, a weighing passage that has grown so out of proportion to the battle description itself that Polybius (including a much briefer weighing passage in the battle description proper, 18.25) has to put it in afterwards, which in turn requires him fastidiously to re-end the battle when he is finished with his long comparison (18.33.1). And other authors, too, found uses for weighing passages outside the climax of the battle as a whole: in an action sequence where a number of incidents are related, some can feature weighing passages of their own.Footnote 70
‘Battle Description in the Ancient Historians, Part II: Speeches, Results, and Sea Battles’ will appear in Greece & Rome 64.2.
