Rome's Crisis of the Third Century Explained with Adrian Goldsworthy

ELI5/TLDR

For about two hundred years Rome had been calm and rich. Then, starting in the 230s AD, it spent roughly fifty years tearing itself apart: somewhere north of thirty men claimed the throne, and almost all of them were murdered within months. While the Romans were busy killing each other, new enemies on the borders — a revived Persia in the east, German tribes across the rivers in the north — started winning fights the Romans used to win easily. Adrian Goldsworthy’s main point is that the crisis wasn’t caused by Rome suddenly forgetting how to fight. The army still knew exactly what to do. The problem was that you can’t run a stable empire when every successful general becomes a death threat to the emperor, and you can’t fight three frontiers at once with two armies.

The Full Story

What “the crisis” actually means

The first thing Goldsworthy does is puncture the tidy label. Nobody living through it knew they were in “the third-century crisis” — that’s a name historians glued on afterward. What it points to is a period, roughly the 230s through the 280s, where emperors came and went at a dizzying rate.

You’re talking about, you know, at least 30 or so emperors in about 40 years.

The exact count is fuzzy because it depends on a judgment call: when a man declared himself emperor and was dead within a few months, do we count him as a real emperor or just a failed bid? Either way, this is a level of chaos — civil war, political murder, usurpation — that has no equal anywhere else in Roman history. Only a matter of degree, he stresses, because Rome was never as stable as the postcards suggest.

To see how far things fell, you need the “before.” The first and second centuries AD were remarkably calm. This was the Rome of the great surviving monuments — the Colosseum, the temples, much of what tourists think of as “ancient Athens” was actually built under the emperor Hadrian. The population peaked, the economy peaked, goods and people and ideas (Christianity among them) traveled freely from one end of the empire to the other. Then it all seems to go wrong.

A quick map of the emperors

Here the chronology matters, so take it slowly. The imperial system was invented by Augustus after the bloody civil wars of the late Republic. The official name is the “principate” — the rule of one man, the princeps (literally “first citizen”), who pretended to be just a senior magistrate serving the state. The pretense fooled nobody for long. As Goldsworthy puts it, quoting an old joke about a philosopher who let the emperor Hadrian win an argument:

Well, you don’t argue with someone who’s got 30 legions.

That’s the real basis of imperial power. Roman emperors were military dictators — sometimes benevolent and competent ones, like Augustus, Trajan, Hadrian, Marcus Aurelius.

The rot Goldsworthy traces back earlier than most people place it, to the Severan dynasty:

Commodus (the mad emperor from Gladiator) is murdered in 192. The system breaks.
Pertinax, remarkably the son of a slave, lasts three months — murdered by the imperial bodyguard when he won’t pay the bonus he promised them.
Septimius Severus wins the four-year civil war that follows, beating two rival governors. His deathbed advice to his sons: “love one another, indulge the soldiers, despise everyone else.”
Caracalla follows only part of it — he murders his brother Geta, indulges the soldiers, despises everyone else. He’s later stabbed by a bodyguard while relieving himself behind a bush on a journey east.
His killer’s commander, Macrinus, makes himself emperor. This is the hinge moment, because Macrinus is not a blood relative and not even a senator — the first time someone from outside the traditional elite is accepted.
Elagabalus, a bizarre teenage priest-emperor pushed forward by Severus’s female relatives on the fiction that he’s Caracalla’s bastard, is murdered by his own family for being an embarrassment.
Severus Alexander lasts a while, then is denounced and killed by his troops in 235 — and that, formally, is where the worst of the crisis begins.

The deeper point in that sequence: once an outsider like Macrinus can hold the throne, the old brake is gone. Previously you needed to be a senator of real standing to be taken seriously. Now anyone who could get troops to shout for him had a shot.

The two threats abroad

While Rome chewed on itself, two external problems sharpened at exactly the wrong moment.

First, in 224, a man named Ardashir overthrew the Parthian king and founded the Sasanian dynasty in Persia. Goldsworthy is careful here: this wasn’t some brand-new super-power. It was the same empire, the same aristocracy, mostly just changing sides after a civil war. But — and this is a pattern he keeps returning to — the winner of a civil war loves a foreign war. It’s “good, clean glory,” and the easiest way to reunite the people you just fought.

The Sasanians under Ardashir and his son Shapur were genuinely dangerous. They captured the emperor Valerian alive — a humiliation without precedent. They forced Philip the Arab to buy peace. Gordian III died on campaign against them; Shapur claimed the kill, though the likelier story is that Gordian was murdered by his own men. (“There’s definitely an asterisk on that one.”)

But Goldsworthy resists the decline narrative even here. The Sasanian attacks, however violent, were never more than large-scale raids — Shapur sacked Antioch twice but never tried to hold Roman land. Rome still had more people and more money. The Persians were big and aggressive, not an equal. And it cut both ways: Palmyra’s ruler, husband of the famous Zenobia, twice raided right up to the walls of the Persian capital, Ctesiphon. Each side was strongest when the other was distracted.

Second, the German-speaking tribes across the Danube and Rhine started breaking through more often. Goldsworthy throws cold water on the idea that these tribes had transformed. Compare Caesar’s description of Germanic warfare with Ammianus Marcellinus’s, written over four centuries later, and there’s barely any difference — same scale, same tactics, maybe even fewer warriors later. No lasting tribal super-state ever emerged; charismatic leaders like Arminius rose and their power died with them. The tribes weren’t stronger. The Romans were just weaker and more distracted.

Why the army didn’t actually break

This is the heart of Goldsworthy’s argument, and it’s the part worth slowing down for. Yes, the Romans lost far more battles in this period than before. But that doesn’t mean their military system failed. The tactics, the equipment, the doctrine — all still sound.

They know the theory. They know what should work… But to achieve that requires training, preparation, good organization, good supply, and critically good leadership.

When you lose, it’s usually not because your system is bad. It’s because you sent a poorly prepared, poorly supplied, outnumbered army — because half of it was off fighting a civil war somewhere else. Civil wars wreck recruitment, training, and promotion. A legion of 5,000 might get split, half on each side; how do you even handle promotions when your two halves are at war? Every internal fight bleeds the system, because you’re spending your strength on yourself.

He extends this skepticism to the standard story that the empire “recovered” in the fourth century under strongmen like Diocletian and Constantine. It only looks like a recovery next to the catastrophic middle of the third. Over the long run, Rome was spending more and getting less, and those famous fourth-century emperors won their biggest victories against other Romans, not foreign enemies — which is precisely the disease, not the cure.

There’s a nice methodological aside woven through all this: we simply lack the evidence to be sure of much. No detailed battle accounts survive for the third century the way Caesar’s do for Gaul. The Antonine plagues that ravaged the empire under Marcus Aurelius almost certainly gutted the population and the army, but we have only dramatic descriptions, no numbers. We can watch the silver content of the coinage collapse — from over 70% to under 1% — a government quietly stretching its resources and hoping nobody notices. But the actual economic damage can’t be measured. Goldsworthy treats these gaps honestly, as guesses clearly labeled as guesses.

The fix, and what it cost

How did Rome eventually stabilize? Through structural change, much of it begun under the Severans. Two moves matter:

Splitting power so no one man could rebel. Britain, once a single province with 40,000–50,000 troops under one senatorial governor — enough resources to challenge for the throne — gets split into two provinces, then four, then five. Military and civilian command get separated. The result: much harder for any individual to gather everything needed to revolt. The cost: also much harder for a loyal commander to scramble together what he needs when 40,000 barbarians actually show up, because he might control the troops but not the food or the pay.

The tetrarchy — more than one emperor at a time. Diocletian’s solution to the “three fronts, two armies” problem was to have four emperors (two senior Augusti, two junior Caesars), each with mobile troops, each able to rush to a crisis like a firefighter. It genuinely increased Rome’s capacity to handle simultaneous threats.

But it had a fatal flaw: it relied on one terrifying strongman, Diocletian, to hold the others in line. The neat four-man system didn’t survive his retirement by more than a few months. You quickly get five claimants instead, Constantine proclaimed by the army in York, and another round of civil wars in which Constantine spent half his reign fighting other Romans.

The logic that drove all of it: your Roman rival will take your job and kill you; the foreigner usually won’t. So the priority was always the internal enemy. Valerian was captured by Persians, Decius killed by Goths, but most emperors died at the hands of other Romans, often in a palace conspiracy.

A quieter, stranger world

The era also changed what an emperor was. Augustus had pretended to be first among equals — a senator who walked the streets, dined with peers, greeted petitioners with a handshake. The later emperor became Dominus — “lord, master,” the word a slave used for its owner. You no longer shook his hand; you prostrated yourself and, if lucky, kissed the hem of his robe. Ammianus describes an emperor riding into Rome so rigidly that he didn’t even scratch his nose at the cheering crowd.

The Senate, meanwhile, faded out of military life. Power and legitimacy now flowed through the army and the equestrian order — the social class just below the Senate, but tens of thousands strong rather than five or six hundred. Goldsworthy questions the flattering old story that these equestrian officers were better “professionals.” There’s no evidence of formal training or merit selection, and they don’t seem to have performed any better than the senators before them. The likelier truth is partly that plague and turnover had simply thinned the senatorial ranks.

The twist at the end

Goldsworthy closes with an idea that surprised even him. We assume the “normal” emperor sat in Rome (the Tiberius model). But Augustus spent more of his reign away from Italy than in it, visited nearly every province, and repeatedly shared near-equal powers with trusted lieutenants like Agrippa — so that someone could always be wherever the trouble was. In other words, the multi-emperor, mobile-command idea that looks like Diocletian’s radical fix may actually echo what Augustus quietly did at the very start. Maybe the rigid one-man-in-Rome model was never the real plan — just an accident of Tiberius being old and tired.

Key Takeaways

The “third-century crisis” runs roughly 235–284/285 AD: 30+ claimed emperors in about 40 years, almost all murdered within months. The name is a hindsight label, not how anyone experienced it.
The instability started earlier than the textbook date, with the Severan dynasty: Commodus (d. 192) → Pertinax (3 months) → Septimius Severus → Caracalla (murders brother Geta) → Macrinus → Elagabalus → Severus Alexander (killed 235).
Macrinus is the hinge: the first emperor who was neither a blood relative nor even a senator. Once the throne is open to outsiders, the old restraint collapses.
In 224, Ardashir overthrew the Parthians and founded the Sasanian dynasty in Persia — same empire, new aggressive leadership. His son Shapur captured the emperor Valerian alive and sacked Antioch twice, but the Sasanians only raided; they never tried to annex Roman territory. Rome was still richer and more populous.
Sasanian warfare differed from Parthian: heavy armored cavalry (cataphracts) and war elephants charging in, versus the Parthian horse-archer who wears you down and shoots while fleeing (the “Parthian shot”).
The Germanic tribes across the Rhine/Danube were not meaningfully stronger than in Caesar’s day; no lasting tribal state ever formed. Rome was weaker and distracted, not out-evolved.
Goldsworthy’s core thesis: the Roman military system never broke. Defeats came from civil wars wrecking supply, training, promotion, and leadership — and from sending under-resourced armies because troops were tied up fighting other Romans.
The coinage was debased through the century — silver content falling from over 70% to under 1% — visible evidence of a government stretching thin, though the true economic impact is unmeasurable.
The Antonine plagues (under Marcus Aurelius) likely devastated population and the army, hitting crowded cities and barracks hardest, but no statistics survive.
The structural fixes: provinces split smaller (Britain went from 1 to eventually 5), military and civilian command separated, and Diocletian’s tetrarchy of four emperors with mobile armies — all to limit usurpation and cover multiple fronts.
The tetrarchy depended on one strongman and collapsed within months of Diocletian’s retirement, producing Constantine and fresh civil wars.
The emperor’s role shifted from Augustus’s “first among equals” (princeps) to the remote Dominus you prostrate before; the Senate lost military relevance to the far larger equestrian order.
Goldsworthy’s closing provocation: Augustus actually ruled mobile, mostly outside Rome, sharing power with deputies like Agrippa — so the multi-emperor model looks less like Diocletian’s invention and more like a return to the founder’s original design.

Claude’s Take

This is a strong episode and Goldsworthy is exactly who you want narrating it — a working scholar of the Roman army who refuses to let a clean story override messy evidence. The whole conversation is an extended exercise in not over-reading the sources, and he’s transparent every time he’s guessing.

The single most useful idea here is his inversion of the usual decline narrative. The lazy version says Rome’s army decayed and the barbarians got strong. Goldsworthy says: the army was fine, the enemies were the same, and the actual machine that broke was political — a system where every competent general was a mortal threat to the emperor, so the emperors spent their strength fighting each other. That reframing is genuinely clarifying, and it’s the kind of structural point that travels well beyond Rome.

The format costs it a point. This is two enthusiasts riffing, and the host, while clearly well-read, talks a lot — some answers are buried under long preambles, and the audio transcription mangles names badly (Sasanian as “Cisanian,” Ardashir as “Adashia,” Ctesiphon as “Tessifon”). Nothing factually wrong on Goldsworthy’s end that I caught; the Augustus-as-proto-tetrarch idea at the close is speculative and he flags it as such. An 8: serious, honest, and the central thesis earns its keep, docked only for the rambling conversational packaging.