Constantinople Stopped the Arabs; Tours Got the Credit

Ask what stopped the Arab advance into Europe and the schoolbook answer arrives fast: Tours, 732, Charles Martel, the hammer that broke the tide. It is a good story and a small one. The better answer is a city, not a battle, and it had to hold twice. Constantinople took the main weight of the caliphate's expansion in the 670s and again, catastrophically for the attacker, in 717 and 718, and what happened at its walls did more to fix the shape of the medieval world than anything that happened in a field in Aquitaine. Tours stopped a raiding column at the ragged end of a very long supply line. Constantinople stopped an empire's main effort at the one place the empire actually meant to go through.

This is not a brief against the Franks, and it is not a claim that Charles Martel's victory was unimportant to the people who lived through it. It is a claim about proportion, and about how a memory gets built. The men who fought at Tours were defending real land against a real force, and Aquitaine's duke had already bled for it once before Charles arrived. But the size of a battle in the history books and the size of a battle in its own century are two different measurements, and for Tours the gap between them is unusually wide.

The city that had to hold, twice

Constantinople's first ordeal is the one modern scholarship is least sure how to describe. The traditional account, built mostly from later Byzantine chronicle material, has an Arab fleet blockading the city on and off across most of the 670s, from roughly 674 to 678, using the nearby Cyzicus peninsula as a winter base and probing the sea walls year after year until a Byzantine naval weapon — Greek fire, deployed here for what may be its first recorded use in battle — broke the Arab ships and ended the effort. That is the shape most surveys still give you. It has been challenged hard. The historian Marek Jankowiak, working back through the source chronology, has argued that the "siege" as usually described telescopes several separate and smaller episodes of Arab naval activity in the Aegean and Marmara into a single dramatic five-year blockade that may never have existed in that form; other scholars, including James Howard-Johnston, have likewise pressed on how much of the traditional narrative rests on later, tidied-up retelling rather than contemporary reporting. This uncertainty strengthens the essay's case rather than weakening it: even a messier, more intermittent set of Arab naval pushes against the imperial capital in the 670s, whatever its exact shape, absorbed effort and ended without the caliphate taking the city, and that failure sits at the front of everything that follows.

The second Arab attempt is not in doubt, and it is the one that matters most. In 717 the caliph Sulayman sent his brother Maslama ibn Abd al-Malik north with what ancient and modern historians alike treat as the largest single expedition the Umayyad caliphate ever mounted against a European target — an army that crossed into Thrace and a fleet, reported in the thousands of vessels, that sailed to blockade the city by sea. The timing looked disastrous for Byzantium: the empire had just been through a decade of rapid changes of emperor, and the man who met Maslama at the walls, Leo III the Isaurian, had seized the throne only months before the siege began. Leo turned out to be exactly the general the moment required. Constantinople's fleet used Greek fire against the Arab ships in the Bosporus, a weapon whose exact chemistry no ancient writer explains and modern chemists still argue over, but whose tactical effect the sources agree on: burning ships that other weapons could not touch. The winter of 717–718 was unusually severe, and the besieging army, strung along the land walls without adequate supply, began to starve and freeze in large numbers. Then a Bulgar army under Khan Tervel, whose kingdom lay just north of the city and had its own reasons to want the caliphate's forces gone from the region, attacked the besiegers and inflicted heavy losses. By the time Maslama withdrew in August 718 the retreating fleet was further battered by storms, and near-contemporary report of the losses — repeated, with real hesitation about exact numbers, across the Byzantine chronicle tradition — describes an army and navy effectively destroyed as a fighting force.

Our fullest narrative for this siege is the Chronicle of Theophanes the Confessor, a Byzantine monk writing in the early ninth century using earlier material now largely lost; his account is paraphrased throughout rather than quoted from any modern translation, since the copyright on those translations belongs to their translators, not to Theophanes. What Theophanes gives us, in outline, is siege, fire, famine, winter, a northern ally's intervention, and a retreat that ruined the caliphate's Mediterranean fleet for a generation. Whatever the precise casualty figures — and ancient chroniclers on every side of every siege in this period inflate numbers as a matter of course — the strategic fact is not contested: the caliphate's largest bid for the imperial capital failed completely, and it did not try again on that scale.

Tours stopped a raiding column at the ragged end of a very long supply line. Constantinople stopped an empire's main effort at the one place the empire actually meant to go through.

What actually happened outside Tours

Fourteen years later and a continent away, a very different kind of fighting produced the battle everyone remembers. Its context is Aquitaine, a duchy that Charles Martel, the Frankish mayor of the palace, did not control and had in fact invaded a few years before to bring it under Frankish authority. Duke Odo of Aquitaine had beaten an Arab army decisively at Toulouse in 721 — a victory now largely forgotten that was, if anything, a bigger check on Andalusi ambition in Gaul than Tours would be. But in 732 a new governor of al-Andalus, Abd al-Rahman al-Ghafiqi, led a force north from Spain, crushed Odo at the Garonne, and pushed on through Aquitaine toward the wealthy monastery of Saint-Martin at Tours. Odo, beaten and desperate, appealed to Charles, the same Frank he had been fighting a few years earlier. Charles marched south, met the Andalusi force somewhere between Tours and Poitiers — the exact site is not securely located, and the traditional October date, though widely repeated, has been argued by some scholars to belong instead to 733 or 734, a discrepancy that comes from the difficulty of reconciling Christian and Islamic chronicle dating for the year — and won. Abd al-Rahman was killed in the fighting; the invading force, its raiding season already loaded with captured goods and by most accounts more interested in getting home with the loot than in fighting a pitched battle it had not sought, withdrew south.

What was this force, in scale? Here modern military historians genuinely disagree, and the disagreement matters for how the battle should be weighed. Bernard Bachrach, reading the Frankish source tradition against what he takes to be realistic medieval logistics, has argued the numbers on both sides were much smaller than older popular accounts assumed, and that the engagement is best understood as a large raid, not an invasion force aiming to conquer Francia. David Nicolle and others have pushed back on parts of that minimalism, arguing the Andalusi force was a serious campaigning army rather than a raiding party, even if nothing like the "300,000 Saracens" of the old romantic tradition. What both camps agree on is the more important point for this essay: nobody now defends the nineteenth-century picture of a caliphate one battle away from planting its flag in Paris, still less across the Rhine. Abd al-Rahman was not marching to conquer Gaul. He was raiding a rich stretch of it, at the far, thin end of a supply and communications line that ran back over the Pyrenees to Cordoba, and when the raid's commander died the raid stopped, the way raids do.

None of this makes Charles Martel's victory trivial to the people it protected. It made his reputation, funded his family's rise, and Tours' abbey and its treasury were spared. But "spared a raid" and "saved Europe" are different sentences, and the historical record supports only the first one cleanly. The historian Paul Fouracre, working through the Frankish evidence in detail, treats 732 as one important episode in Charles's long consolidation of power within Francia — significant to Frankish political history, not obviously significant to the survival of Christian Europe as a civilization, which was never seriously at stake in Aquitaine the way it was at stake, twice, on the Bosporus.

Geography did more work than any general

The deeper problem with the "Tours saved Europe" story is that it imagines a single continuous highway of conquest running from Damascus to the English Channel, interrupted only by Frankish steel. The Mediterranean world the Umayyad caliphate actually operated in did not work that way. Andalusi forces raided across the Pyrenees repeatedly through the 720s and 730s — into Septimania, up the Rhone valley, as far as Autun in Burgundy in 725 — but raiding and holding territory are different projects with different logistics, and the caliphate's foothold north of the Pyrenees was always thin. Narbonne, the Arab-held city in Septimania that anchored what presence there was, held out for decades after Tours; it did not fall to the Franks until 759, under Charles's son Pippin, after another round of campaigning that included a Frankish check at the river Berre in 737. If Tours had truly broken the advance in 732, Narbonne would not still have needed taking twenty-seven years later.

What actually redirected the caliphate's energy was not a defeat in Gaul but a convulsion at its own center. The Umayyad caliphate collapsed in 750, overthrown by the Abbasid revolution, in an upheaval whose violence and scale dwarfed anything that happened at Tours; the one Umayyad prince who escaped the purge, Abd al-Rahman, fled all the way to Spain and in 756 set himself up there as an independent emir, cut off from Damascus and then from Baghdad for good. Al-Andalus became, from that point, a breakaway state fighting its own wars of survival, not the forward edge of a caliphate still reaching for the Rhine. A regime absorbed in a civil war on the far side of the Mediterranean was never going to press a conquest of Francia regardless of what happened outside Poitiers. The historian Hugh Kennedy, tracing the caliphate's political history across this period, treats the Abbasid revolution as the hinge event of the whole eighth century for exactly this reason: it is the moment the Umayyad project everywhere, not just in Gaul, stopped expanding and started fighting for its own life.

Constantinople sits on the other side of this geography problem, and it explains why the two theaters produced such different outcomes. The Bosporus is not a raiding frontier at the ragged edge of a supply line; it is a chokepoint, the hinge between Asia and Europe that any land force moving from the caliphate's core into the Balkans and beyond had to pass. That is precisely why the caliphate sent its largest expedition there rather than to Aquitaine, and precisely why its failure there mattered more. John Haldon's work on the seventh-century empire makes the material case for how close-run this was on the Byzantine side: an empire that had lost most of its wealthiest provinces to the same conquests, reduced to something closer to a fortified core than the empire of Justinian, and still managed, twice, to hold the one position that could not be allowed to fall.

The Umayyad project everywhere, not just in Gaul, stopped expanding and started fighting for its own life.

How the credit moved west

If Constantinople did the heavier work, why does Tours own the memory? The answer is a chain of retelling, each link written by someone with a reason to make Tours bigger than it was.

The first link is Frankish. The near-contemporary Latin chronicle tradition that first describes the 732 battle was written within the Carolingian world, for audiences and patrons whose interest was Charles Martel's dynasty — the family that would produce Pippin the Short and then Charlemagne. A victory that saved Francia's most sacred shrine from plunder was, understandably, worth dramatizing by writers whose patrons descended from the victor. That is not an accusation of dishonesty; it is how court-adjacent chronicle writing worked across the early medieval world, Byzantine included. But it means the earliest surviving accounts of Tours were never neutral measurements of the battle's size against, say, the siege of Constantinople fifteen years earlier — an event those same Frankish writers had no particular reason to know or care about in detail.

The second link is Enlightenment historiography, and specifically Edward Gibbon. In The Decline and Fall of the Roman Empire, published across 1776 to 1789, Gibbon built one of history's most famous counterfactuals around Tours: had Abd al-Rahman won, he suggested, the Arab advance might have swept on unchecked to the Rhine and even the English Channel, and — in the version of the passage most often repeated — a victorious army might have gone on to teach the Koran in the schools of Oxford. That image is paraphrased rather than quoted from Gibbon's actual sentence: its substance is secure, its exact eighteenth-century wording less so, and Gibbon is worth representing accurately rather than approximately. The counterfactual is a magnificent piece of prose and a poor piece of strategic reasoning: it treats an eighth-century raiding force, already turning for home with its plunder when it met Charles, as though it were an army with both the intention and the supply capacity to occupy the length of Francia and cross the Channel. Gibbon was writing about a caliphate whose actual, real-world main effort against Christian Europe had already been broken, twice, at Constantinople — a fact his own book records in its account of the sieges of 674–678 and 717–718 — without drawing the comparison that would have deflated his Tours passage.

The third link is the nineteenth century's taste for lists. Edward Creasy's The Fifteen Decisive Battles of the World, published in 1851, took Gibbon's counterfactual and turned it into canon: Tours became, by Creasy's title alone, one of the handful of battles that supposedly determined the entire subsequent course of world history, ranked alongside Marathon and Waterloo. Creasy's book was enormously popular and enormously influential on how battles got taught in English-language schools for a century afterward; its selection criteria were frankly about which battles made the best narrative turning points for a Western readership, not a comparative audit of which sieges actually cost an expanding empire the most. Constantinople's defenses never made his list. A city that held, twice, does not offer the same clean counterfactual as a raiding column that got killed with its own general.

None of this is a conspiracy. It is closer to a filing error, repeated by each generation of retellers because the earlier filing had already set the shelf. Byzantium's long defense of its own capital was recorded by Byzantine writers, for a Byzantine and later Orthodox readership, in a chronicle tradition that Western European historiography engaged with far less than it engaged with its own Frankish and then Enlightenment sources. The event that did more to stop the caliphate's European ambitions ended up in someone else's history.

Weighing the two defenses fairly

None of this is a case against the men who fought at Tours, and it should not be read as one. Charles Martel's army met a real threat to a real place and beat it in the field, and Odo before him had already shown, at Toulouse, that Andalusi forces in Gaul could be defeated outright. The soldiers on both sides at Tours fought a real battle with real stakes for the people of Aquitaine. Equally, the defenders of Constantinople were not saints and the attackers were not villains: Maslama's expedition in 717–718 was a serious, well-planned effort by a competent commander against a defended capital, and its failure owed as much to a hard winter and a Bulgar alliance beyond Byzantine control as to anything Leo III did by design. Leo deserves the credit history gives Charles Martel and rarely gives him: a new emperor, ruling for barely a year, who held a siege that several of his more established predecessors might well have lost.

What the comparison should change is not admiration but proportion. The Mediterranean world of the seventh and eighth centuries was not a single front waiting on a single battle. It was a patchwork of chokepoints, supply lines, and raiding seasons, and the caliphate's actual attempt to break into Christian Europe by main force met its two hardest tests at Constantinople, not at Tours. The conquests elsewhere in this period moved at the speed of administration and generations, not of any single battle, which is its own argument against treating any one field as decisive. And what happened to Anatolia several centuries later, when a different and much later Turkish expansion did succeed where the Arab sieges had failed, is a useful reminder that Constantinople's walls bought time, not permanence — the city that held against Maslama in 718 fell, in the end, to someone else entirely, seven centuries on. That does not diminish 718. It only shows what the phrase "decisive battle" always overpromises: no single victory, Tours included, ever settles anything for good.

The schoolbook answer survives because it is simpler to teach and because the writers who first told it had every reason to make it the whole story. The fuller answer runs through two long, grinding sieges on the Bosporus, a civil war in Damascus that had nothing to do with any Frankish sword, and a city, Constantinople, that in a fair accounting deserves the credit that a chapel of Frankish tradition, an English gentleman's counterfactual, and a Victorian list-maker gave instead to a field outside Tours. The Christian communities that endured under Arab rule in the centuries that followed lived with the consequences of where the caliphate's expansion actually stopped, and it did not stop at Poitiers.