Rome fell for real
The transformation school described a gentle handoff. Archaeology — femur length, coinage, shipwrecks — breaks in the fifth century. Majorian almost reversed it. Then the fleet was sabotaged.
For a generation the polite view was that Rome did not fall. It transformed. Barbarians came, in Henri Pirenne’s phrase, not to destroy the empire but to take part in its benefits. Peter Brown taught students to say Late Antiquity and mean a coherent age, not a wreck. Walter Goffart’s Toronto school put the settlements on a tax-share ledger: accommodation, not conquest. The bones disagree.
Bret Devereaux’s synthesis, sitting on the wiki next to Bryan Ward-Perkins and Peter Heather, is a pile of independent indicators. Femur length. Animal-bone consumption. Pottery. Coinage. Shipwreck frequency. They break in the fifth-century West. The drop from 450 to 550 is, in that read, the sharpest change in the data at any point. The strong form of continuity — that nothing of value was lost — is not sustainable. Gibbon’s disaster was real. His causes, Christianity and moral decay, are not the ones these people will sign.
No slack
Heather’s cascade starts in the third century, not the fifth. Sassanid Persia rises. From then on a much higher share of Roman men and money sits permanently on the eastern frontier — on the order of a third of the military. The slack the empire had used to absorb a shock in Gaul or on the Danube is gone.
On the European rim, Germanic societies had been living next to Roman demand for generations. Trade, subsidies, and frontier service fused smaller groups into larger, more durable coalitions. New words for authority, Heather notes, evolve from words that meant war-leader. When the Huns detonate in 376, two migration waves — 376 to 380, then 405 to 408 — hit Western field armies that could have beaten the same groups arriving separately. Amalgamations of three immigrant blocs, each 10,000-plus warriors, are a different problem. The central empire, in his line, did not pass away quietly. It was fought to extinction over seventy years.
Africa, then the loop
Every lost province is lost tax. The catastrophic loss is North Africa. The Vandals take it between 429 and 439. Grain that had fed as many as 300,000 mouths in Italy stops. Famine follows. The army is paid from the same soil. A smaller army invites the next seizure. That is the fiscal-military feedback, and it is why Africa sits at the center of every serious modern account.
Goffart still has cases. Some Burgundian and Visigothic settlements look like the tax-share story he wrote. Vandal Africa does not. Neither do Hunnic raiding and the post-405 Rhine crossings. The chain on the wiki splits the difference: the fiscally fatal losses were violent; some other settlements were genuine accommodations. Pirenne’s seventh-century Mediterranean rupture, the Arab cut of the sea, looks better against newer climate and archaeology than his implication that the fifth-century West was mostly continuous.
The last recovery
It almost went the other way. Aetius’s coalition stopped Attila in 451. Aetius himself was murdered in October 454. Majorian, emperor from 457 to 461, retook Gaul and parts of Hispania, reformed the administration, and built a fleet of 300 ships for a crossing against the Vandals. The fleet was destroyed from the Carthaginiensian shore “through traitors.” Ricimer deposed him. His successors, until 476, were puppets of barbarian generals or of the Eastern court. Romulus Augustulus’s deposition by Odoacer formalized a fact already on the ground.
By the fifth century the Western army is put at roughly 100,000 to 150,000 effectives — too few against coordinated coalitions. That range is the wiki’s, not a new count.
I might be writing this sentence, not in England, but in a still-extant province of Britannia.
Bryan Ward-Perkins, on the contingency of the Western fall
The plague that was not the whole story
Kyle Harper’s environmental catastrophism leaned on the Antonine Plague — 165 to 180 CE, likely arriving with Roman troops at Seleucia, reaching Rome by 166 and most provinces by 172. Traditional mortality estimates clustered around ten percent of an empire of roughly seventy-five million. Co-emperor Lucius Verus died in 169; Marcus Aurelius recruited gladiators and slaves to fill the army. Indian Ocean trade from Egypt suffered a major setback. The army-reads-as-nearly-extinct in contemporary sources sit next to a 2024 PLOS One epidemiological model that cuts maximum excess mortality to about seven percent for bubonic-plus-rats scenarios, about five for measles, about one for standard smallpox — with the authors concluding secondary literature had overestimated impact.
The plague was real; the structural-blow case has weakened. Harper’s framing absorbed dramatic environmental change; a BMCR review of his book flagged methodological circularity — minimalists in the footnotes. Reconciling Cassius Dio’s two thousand deaths per day in Rome in 189 with the models requires exaggeration, a separate epidemic, or parameters the model missed. For the Western fifth-century collapse, Antonine mortality is a third-century stressor in the cascade, not the knockout punch — part of the no-slack story, not a monocause.
The ice that came too late for the West
The Late Antique Little Ice Age — volcanic eruptions in 536, 539/540, and 547, cooling sustained into the 660s — is well attested in ice cores and tree rings. Summer temperatures in Europe fell as much as 2.5 degrees Celsius below normal in 536; aerosol loading exceeded Tambora’s Year Without a Summer. Scandinavia may have lost up to half its population in the volcanic window; Arabia’s fertility rose on cooler, wetter conditions, feeding into later expansion narratives; Elusa in the Negev shrank from the mid-sixth century.
Romulus Augustulus was gone in 476. LALIA begins in 536 — sixty years after the Western political collapse. Lombard and Slavic migrations, the Justinianic Plague from 541, and Pirenne’s seventh-century Mediterranean rupture are successor-state stories, not explanations for why the Western army could not retake Africa in Majorian’s day. Harper’s long arc treats 200 to 650 CE as one deteriorating curve; the strongest LALIA evidence concentrates after the West is already gone. Climate and plague belong inside the fall narrative as stressors that shaped the post-Roman world and weakened the East — not as triggers that caused the Western deposition itself.
The half that lived
The East faced Persia, the same plagues, the same debased currency, the same Germanic pressure, and the same Christian state. It lasted a thousand years longer. A.H.M. Jones’s observation, still the one people reach for: most of the weaknesses scholars discuss were common to both halves. The East’s apparent advantage was geography. The European frontier from the Rhine mouth to the Danube is some 2,000 kilometres and can be crossed. Constantinople had the Theodosian Walls, the Marmara, the Bosphorus, and a short, defensible line. Germanic armies lacked the siegecraft and the ships. Syria, Egypt, and Asia Minor — the rich provinces — stayed in the Eastern fisc. Ward-Perkins’s version is blunter: a thin band of sea, not structural superiority.
That counterfactual kills the monocausal-internal list. If decadence or Christianity or lead or institutional rot caused the fall, the East should have fallen on the same timetable. It did not. Geography is the precondition. Later Eastern adaptations — surviving civilian administration, the post-Justinianic recovery, the thema system — needed that precondition to be possible. The plague and ice sections above restate why environmental catastrophism no longer carries the Western fall by itself — timing and revised mortality, not denial that either happened.