Tracking Hannibal

Dum elephanti traiciuntur . . .
Livy, Ab Urbe Condita

I

There's a word missing. And, without it, without that single, indispensable qualifier, we'll never be entirely free of guesswork, idle speculation, academic conjecture. I'm not referring, here, to some ontological puzzle touching on, say, the nature of existence, but the location of a particular event at the outset of the Second Punic War: namely, the exact point at which Hannibal crossed the Rhone. Does it really matter, though? Who in fact cares about an event so relatively minor within the entire breadth—unraveling—of history itself? Why should this particular moment in the late summer of 218 B.C. be the focal point of such lavish attention? Such meticulous, painstaking examination?

There's a word missing. And history, a system that tends toward closure, irrefutable fact, finds itself in this instance confronted by an omission, a lacuna. Finds its tightly woven fabric momentarily rent. It may well be the underlying cause, indeed, of so much obsessive attention, why innumerable hypotheses have been emitted in regard to Hannibal's exact point of crossing.

The lacuna in itself draws the scholar's curiosity with all the involuting magnetism of some low pressure zone, some cyclonic eye. For this tiny textual oversight, this minor omission in the running narrative of Hannibal's exploits, has generated an entire literature based on interpolation alone. History, we come to realize, has a horror of omissions. It comes rushing, each time, into its gaps, silences, hiatuses. Comes rushing with a scarcely concealed voracity.

From Narbonne to the Rhone crossing, Polybius is explicit: Hannibal's armies had marched one thousand stadia (approximately 18 5 kilometers) to reach the banks of the river. But which banks, exactly? We can be certain that he'd taken, in crossing Languedoc, the Via Iberia (later known as the Via Domitia, a Roman Praetorian route). Passing through the territory of the Volcae, whose capital was Nîmes—a final stage-point on the Via Iberia before the Rhone itself-he would normally have reached the crossing at a level with Beaucaire. But did he, in fact?

It's far easier to imagine Hannibal and his armies than to fix their exact location at any given point. Easier to imagine the clouds of late summer dust that must have arisen from the trampling of fifty thousand foot soldiers, nine thousand Numidian horsemen on horseback, an inexhaustible quantity of pack animals, not to mention the soft, waddling gait of thirty-seven mounted elephants, striking terror as they went into the heart of incredulous Gaul. As for Hannibal, his name alone—"By the Grace of the God Baal"—speaks eloquently enough. Riding one of those immense, ungainly creatures, one can readily imagine that twenty-six-year-old warrior—already considered a genius in the dark art of strategy—as if enlightened from within. We have Livy's account (drawn, no doubt, from some Carthaginian source) of Hannibal's dream: how, only months earlier, a youth of divine aspect (iuuenem diuina specie ) had appeared to Hannibal in his sleep, sent by Jupiter himself to guide

the young warlord on his march into Italy.^[1] With Baal behind him and Jupiter—like a star—ahead, how could Hannibal help but feel enlightened, protected as he was by those invisible forces?

From the outset, he had three major material preoccupations. First he'd been obliged to bide his time, to camp with his immense armies in Cartagena (Carthago Novo), Spain, until the spring crops had been harvested and his troops adequately supplied with rations. This had meant delaying the entire campaign until late May at the earliest. Next came the Rhone crossing. At no point in his immensely perilous journey would his armies be more exposed, open to attack. In September, he knew, the Rhone's waterline would be at its lowest, and the river itself at its most amenable. No moment could be better chosen. Last came the greatest obstacle of all: the Alps, the imperious gateway onto the italic plains and, ultimately, Rome itself. These, as Hannibal well knew, had to be crossed before the Pleiades had set in the western sky (occidente iam sidere Vergiliarum )^[2] —before, that is, the first snows fell.

Considering the scale and daring of such an adventure, why should we—twenty-two hundred years later—confine ourselves to the petty details of one specific instance: those hypothetical crossing points in Hannibal's vast dé-marche? I find myself asking this question as I, too, enter the puzzle, following the banks of the Rhone this morning—geodetic survey maps in hand—traveling northward from the lowest conceivable crossing point, inspecting each alternative, each purported site, as I go. What, though, could anyone add to those endless pages of supposition already written, that endless catalog of postulates? I stare at the river, the perfectly flat, imperturbable slip of its pearl-grey waters, asking myself that very question. What's left to be determined?

The Greek historian, Polybius, is both our first and our most indispensable source of information. His Historiae , covering the full fifty-three-year span

of the Second Punic War (beginning with its causes and concluding with Rome's total subjugation of the Carthaginian empire) was written only seventy years after Hannibal's astonishing blitzkrieg across Gaul. Drawing heavily on documented reports left by Hannibal's own historians, on eyewitness accounts, and on his own footwork (he would trace Hannibal's entire trajectory himself), Polybius was an utterly faithful, if somewhat verbose, chronicler of these events. Indeed, his own account often reads like an "engineer's report," as one modern historian put it.^[3] Why, then, hasn't Polybius himself furnished us with the missing word, the long-vanished locative? Given his respect for painstaking accuracy, why didn't he fix the exact crossing point forever in his written Historiae ? Polybius himself supplies us with an answer. In Book 3 he writes: " . . . in unknown regions, the enumeration of place names has, I feel, no more value than words devoid of all meaning, than a mere babble of sounds."^[4] Writing in Greek for readers totally unfamiliar with the geography of those scarcely explored regions, Polybius simply deleted most place names altogether. Instead, he employed what might be called spatial indices: cardinal points to indicate direction, and either stadia (a unit of length) or marching days to indicate distance. Both, of course, could be readily comprehended by his Greek readers; his descriptions were totally exempt from meaningless toponym.

What, then, about the documented reports left by Hannibal's own historians, those who'd accompanied the warrior throughout the various stages of his vast expedition? Polybius himself had drawn freely from these Carthaginian sources and had even berated them in his Historiae for cluttering their own accounts with toponyms, those place names devoid of any significance whatsoever. One thinks especially of Silenus—one of seven known Carthaginian sources—who was not only Hannibal's personal historian but also a companion-in-arms in

his march across Gaul. Certainly Silenus would have named the exact crossing point and the battle that immediately ensued. And, most certainly, he had. But along with the six others, Silenus's accounts have totally vanished. Nothing remains of their work except their names. History, here, has deprived us of everything but the vacuous signature of its historians.

Silence is a relative quality: there are degrees of silence just as there are degrees of sound. Here, for instance, with the loss of these seven contemporary eyewitness accounts, the silence of history itself has deepened, amplified. We've traced the event down to its most immediate, recorded sources and discovered that the sources themselves have disappeared. We have no choice but to move forward in time, to seek out classical historians who still happened to have access to those works, hoping, as we do, to detect some thin trickle of evidence.

Coelius Antipater, a Roman annalist working at about the same time as Polybius, compiled a monograph on the Second Punic War in seven consecutive volumes. Considered a highly reliable source by Valerius Maximus,^[5] it must certainly have contained our missing word. But Coelius's monograph has, alas, vanished as well. Like the works of those Carthaginian historians, it too has entered a level of silence that we can only consider as absolute. Searching for a scrap of historical information, we've encountered nothing but deliberate deletion on Polybius's part and the successive effacement of recorded testimony owing to the barbarity or brute indifference of the Dark Ages that followed.

There remains Livy. Considered along with Tacitus and Sallust as one of Rome's greatest historians, Livy wrote an account of Hannibal's expedition in Book 21 of his vast, sweeping, eight-hundred-year Ab urbe condita (History of Rome). The pages of that book contain some of the most celebrated passages in all of Latin literature. They belong, indeed, more to literature than

history, as Livy—with an aristocratic disdain for verifiable fact—worked his narrative to a high rhetorical polish. Drawing freely from not only Polybius and Coelius Antipater but also from the seven vanished Carthaginian histories, he conflated his materials into a single flowing narrative, a flumen orationis of his own. For all its beauty, however, it leaves us guessing. Rarely citing the names of his historical informants, Livy offers up a "cut-and-paste" version—as Sir Gavin de Beer put it—of Hannibal's great exploits.^[6] We'll never know, of course, which parts originated where. With Livy, we're continuously at the mercy of amalgamated material and, inevitably, this material comes to contradict itself. Furthermore, if Polybius chose to delete place names in favor of spatial indicators (distance and direction), Livy did exactly the same, but for the sake of tribal territories. He located events in history according to whose territory in which they happened to occur. Hannibal crossed the Rhone, he tells us, in the land of the Volcae. Alas, the "land of the Volcae" designates an area far too extensive, lying on either side of the Rhone, to lend itself to anything more than exacerbated guesswork, shallow speculation.

There's a word missing: a single, precise, irrefutable place name. And, in its absence, the question itself remains posed—history itself suspended—over the geographical particulars of a few dusty, tumultuous days, twenty-two hundred years earlier.

Here I am, then, reading Polybius along the banks of the Rhone in my own rudimentary Greek, murmuring phrases to the waters before me, as if the waters themselves could be awakened from their somnolence, stirred—on syllables alone—from centuries of flat, placid self-absorption. I'm at Arles, this morning. Or, rather, at Fourques, facing Arles across the waters of the Rhone which are moving, here, at no less than two meters a second. A traditional cross-

ing place since protohistoric times, Fourques-Arles represents the lowest—southernmost—point yet proposed as that of Hannibal's. In fact, in the absence of a single certifiable crossing, more than thirty carefully researched, thoroughly plausible locations have been proposed. From Seneca to Napoleon Bonaparte, the hypotheses abound. One of them, of course, is correct, for certainly Hannibal must have crossed somewhere. But where, exactly? Grosso modo , there are two schools of thought: that his crossing point was fairly close to the sea (thus, possibly, at Fourques-Arles) or that it was as far removed from the sea as Hannibal could reach within the allotted marching time. For the Carthaginian had to contend with not only a vast army of hostile Gauls, already massed on the opposite bank of the Rhone, but also a Roman consular force, led by Publius Cornelius Scipio, which was hard on his heels. If Polybius suggests, on three separate occasions, Hannibal's proximity to the sea throughout these events, he also tells us that Hannibal crossed the Rhone at a distance from the sea equivalent to "four marching days" (3.42.1). Here he enters into complete contradiction with himself and introduces one of the errors that will set historians astray for two millennia. Which of these two versions should we believe? Furthermore, when Polybius indicates "four marching days" from the sea, we have to ask ourselves, once again, from where? From Fos, the Roman Fossa Maritima, or some other coastal station, further to the west? As to the marching days themselves, how should we calculate that distance? Scipio's full-sized consular force, we know, moved at a rate of fifteen kilometers a day, whereas Xenophon indicated a rate of twenty-two kilometers a day for Cyrus's legions, and Caesar himself boasted of nothing less than twenty-seven kilometers a day in his own conquest of Gaul. Which of these distances, finally, constitutes a marching day in Polybian terms?

If the hypothesis of a crossing point somewhere near the sea (for example,

Fourques-Arles, or Beaucaire-Tarascon) has the advantage of a certain in-eluctability—these were traditional points of crossing, fully equipped with ferries, rafts, coracles—a point further to the north satisfies more fully the descriptions we have of Hannibal's immense strategic cunning. Livy clearly stipulates (21.31.2) that Hannibal, in removing himself as far as possible from the coastline, felt he might avert the Roman forces altogether. Even better, we have a single lapidary phrase from Zonaras: "Hannibal himself always avoided the obvious way."^[7]

Whichever hypothesis we tend to support, we find ourselves confronted by inherent contradictions. The closer we read—the more exacting our examination—the more enigmatic and ambivalent our findings. For each grounded argument, there's a counterargument; for each proof, a refutation. Have we entered, here, some kind of parable, some esoteric object lesson, say, on the nature of knowledge itself?

What, in material terms, am I looking for this morning? First, an area along the west bank of the Rhone broad enough to allow an army of sixty thousand men to mass in martial order and prepare their attack upon the facing bank. In short, a vast staging area.^[8] Part of that preparation would include the felling of trees (poplars mostly) for the sake of constructing, in haste, the rafts, barks, and pontoons necessary for that crossing. Next, there'd have to be a relatively flat plain en face . Hannibal wouldn't have given his enemies the advantage of a natural shelter behind rocks, ledges, hillsides; he would have forced them down onto the same merciless level as himself: that of the riverbank. He would have chosen, quite deliberately, his own beachhead.

Polybius tells us that several days before the attack, Hannibal (as the inventor, most likely, of commando tactics) sent Hannon, one of his most trusted generals, two hundred stadia (thirty-seven kilometers) northward with a con-

tingent of troops. There, where an island lay in the midst of the Rhone, the waters would be at their widest, and, consequently, their most shallow. Hannon's troops could swim across at this very point, floating on their wooden shields, dragging behind them bloated goatskins laden with their impedimenta.

I needed to locate an island, then, situated a certain number of kilometers north of two facing plains: one on which Hannibal's troops might have been deployed; the other on which he would have engaged the Gauls in combat. Altogether, these three locations constituted the geographical determinants by which I might postulate Hannibal's crossing point. Furthermore, there was a fourth factor: Hannon, having crossed the Rhone and landed on the opposite bank with his swift, lightly equipped, elite corps, sent Hannibal a smoke signal (ex loco edito fumo , Livy 21.27.7) from a hill just behind or beside the battle site itself. He did so, according to Polybius (3.43.1), only a moment before daybreak. With that signal, the battle commenced. In a conjoined maneuver, Hannon took the Gauls by surprise from the rear as Hannibal, in a frontal attack, led a first wave of light cavalry across the river. By nightfall, we read, Hannibal's victory would be assured.^[9] Firmly established on the east bank of the Rhone, nothing but the Alps would stand between the Carthaginian and his ultimate ambition: Rome itself.

Along with my three other determinants, I had to locate the hill "just behind or beside the battle site." All four of these locations had to coincide with one another, mesh topologically. So, starting at Fourques, I set out. Carrying battered Loeb Classic editions of both Polybius and Livy, a sheath of notes I'd taken from the works of the specialists, and a complete set of geodetic survey maps covering the lower reaches of the Rhone valley, I made my way gradually northward. I was delayed at the outset by a ground fog that kept me from taking clear cross-river sightings, but a mistral arose by nine that morning

which virtually delivered the landscape, rendered it crystalline. Along with the mistral came the sun. A clear, mid-winter day in Provence now lay before me in which to visit a whole host of proposed sites, to verify or dismiss each of them as I went.

From the survey maps alone, I was well aware that any number of islands lay in the middle of the Rhone; several of them could easily qualify as the one that Hannon crossed with his small troop of commandos. I could readily discern, for instance, the Ile de la Barthelasse off Avignon, the Ile de la Piboulette between Ardoise and Caderousse, and—further to the north—the Ile du Malatras in the vicinity of Pont-Saint-Esprit. Indeed, each of these islands has been proposed as Hannon's by a number of specialists over the years. But did any of them lie thirty-seven kilometers north of a pair of facing plains? And, if so, did the plain on the opposite (eastern) bank possess an overhanging ledge or hillock from which a smoke signal—undetected by an enemy, just beneath—might have been lit?

After traveling over ninety kilometers northward to Pont-Saint-Esprit, inspecting each postulated site, triangulating, as I went, between the various topological determinants involved, I came, at last, to an irrefutable conclusion. None of the sites proposed or suggested corresponded with the existent physical realities. If nearly all of them satisfied three of the given characteristics, not one of them coincided with all four. If, for instance, I found a number of matching plains, several of which lay thirty-seven kilometers south of an island, I never found the rock overhang or hillock befitting. Or, if I did, I wouldn't find the island or the plains in their appropriate contextual positions. There was always a piece missing in this four-piece puzzle I was playing with history. Or was history, in fact, playing with me? Laying out false directives for the sake of concealing some cherished moment all the more consummately?

II

I would only belabor the issue by extending my area of research into the Alps themselves. For there, the same textual problems exist, but multiplied a thousandfold. Perhaps no single event in Roman antiquity has given rise to so much controversy as the exact passage Hannibal took in crossing the Alps. Be it the Rhone or the Alps, however, it's curious, I find, that each of these unspecified points involves passage. Each involves a single, perilous, all-determinate moment of territorial penetration in which the Carthaginian entered—forced his entry—into an entirely new region. Why, then, have these points, so crucial in themselves, gone unnamed? Even Livy, writing only a century and a half after Hannibal's passage, remarked that there existed amongst his contemporaries serious doubt as to which Alpine pass Hannibal actually took (ambigi quanam Alpes [Hannibal] transierit , 21.38.6). Dennis Proctor, in his brilliant study, "Hannibal's March in History," refers to the silence of those contemporaries. "Varro, Pompey, Strabo, Cornelius Nepos, Appian, and Ammianus Marcellinus," he writes, "apart from Polybius and Livy themselves, all mentioned Hannibal's pass in one context or another, but not one of them gave it a name or referred to it in any other way than as Hannibal's pass . . ."^[10]

This curious lacuna generated doubt, confusion, or a certain embarrassed silence on the part of those early historians. Even more it created, with time, a plethora of speculation. A first monograph on the subject appeared in 1535, followed, in each century, by hundreds of refutations, correctives, counterproposals. Scaliger, Casaubon, Gibbon, Napoleon, and Mommsen, among many others, would make their contributions. None, however, would entirely concur. Even today, come late October and the proverbial "setting of the Pleiades"—the moment, that is, of Hannibal's crossing—the Alpine passes are visited by

a seemingly inexhaustible number of prospecting Hannibal scholars. What do they expect to find at those altitudes? Having already compared the irreconcilable textual differences between Polybius and Livy, they'll attempt to determine the material conditions prevalent, at that moment, at some specific col. Measuring, say, the exact angle of each feasible slope or the "adhesion potential" of a hypothetical four-ton elephant on the freshly fallen snows of that slope, they'll arrive, quite often, at some entirely new hypothesis of their own. Questioning forest rangers, chamois hunters, road maintenance workers, priests in their mountain parishes, they'll come to ever new conclusions. Given that there exists more than forty negotiable passes between the Great Saint Bernard and the Col de Larche (the passes themselves ranging from mule trails to tarred roads), and that most of those passes possess a variety of approach routes of their own, the possibilities are virtually limitless. So, too, is the number of hypotheses advanced.

There is, quite clearly, a word missing. Or, more exactly, words missing from the pages of those first, founding, historical documents. Here, as elsewhere, history is riddled. Instead of a locative, we discover, in its place, a lacuna. Where an all-determinate event in history occurred, we're given, within the running fabric of those earliest narratives, a deliberate deletion in one case, a careless oversight in another, a succession of eloquent silences in yet others. Why, then, haven't historians simply accepted the good advice, say, of Marc Bloch? "Having explored every given possibility," he suggested, "there comes a time when a scholar's greatest duty is simply to confess ignorance and admit to it openly."^[11] Why, to the contrary, this insistence? Why have an endless number of trekkers tested one col after another, written—on speculation alone—an incalculable quantity of monographs touching on Hannibal's "true Alpine pass," or spent long days, as I have done, attempting to determine the exact point at which the Carthaginian crossed the Rhone?

"The problem, here," as Ulrich Kahrstedt, a German classicist, astutely observed, "isn't topographical so much as historico-literary."^[12] And, as I've already suggested, the nature of history is one that tends toward closure, conclusiveness, a seamless weave of sequential events. Most of the time, we feel ourselves an integral part of that continuum. That unbroken pattern. When, however, that pattern undergoes rupture, suffers hiatus for even the briefest instant, a certain intellectual anxiety results. Call it a zone of mnemonic instability. The void, we feel, must be filled; the lacuna, rectified. We come rushing into those areas of omission with a fervor bordering on the obsessional. Like bees, their metabolism stimulated at a certain given temperature, we swarm to the indeterminate: attempt to enter the very vocables that history, inadvertently or not, has left open.

Hannibal was fortunate, I find myself thinking. It's already late afternoon, now, as I gaze across the Rhone at one of the last hypothetical crossing points and muse on that doomed twenty-six-year-old warrior. On the opposite side, light is striking the ruins of a medieval watchtower—this very instant—a rich, carmine red. Yes, Hannibal was fortunate, for he'd managed to escape—at certain critical moments—his own historians. Those moments, and most notably the two perilous episodes involving passage, penetration, and the conquest of an entirely new territory, remain as if suspended, abeyant. They belong to a blank language. An empty script. Historians will go on postulating, speculating, demolishing one thesis for the sake of advancing another. It won't matter. As I watch the channel buoys bob in the draft of that deep, unremitting current, I'm fully convinced that that language will remain blank; that script, empty. The book, complete with the inaudible trumpeting of its thirty-seven volatilized elephants, will lie forever open.