The World-State at 284 · Imperium Romanum

What Broke, and What Held

The audit of this volume must begin by separating two questions the crisis conflated. Did the Roman military system fail between 235 and 284? By every measure the sources permit: yes — the perimeter was penetrated on every front, two emperors died at enemy hands and a third in captivity, a province was abandoned, the empire fragmented into three states for over a decade, the currency effectively ceased to be silver, and the capital of the world built a wall. Did the Roman state fail? No — and the volume's most important finding is the asymmetry. Every seceding fragment governed in Roman forms, with Roman institutions, in Latin or Greek, minting Roman coin: Postumus and Zenobia alike claimed to be Rome's defenders rather than its heirs, and when Aurelian arrived the fragments rejoined rather than fought to the death. No barbarian confederation founded a kingdom on imperial soil in this century; no province seceded on ethnic or national lines; no religious movement raised a rebellion; and the reconquest, when force finally became available, took five years. The empire of Book VII was structurally far tougher than its narrative reads, and the reason lies in Book VI's chapters: the citizenship extended to everyone in 212, the provincial elites made Roman across two centuries of quiet, the army recruited from the frontier populations it defended. The crisis was a military and fiscal emergency inside a civilization that had no available alternative to itself.

What the emergency destroyed was the Principate — the Augustan fiction, audited in the last volume, that an emperor was a first citizen holding a restored Republic in trust. Between the auction of 193 and the acclamation of 284 that fiction was dismantled by the people who had always underwritten it: the army. Senators lost legionary command, then provincial command, then the throne itself; emperors ceased to come from Italy and the old order and came instead from Africa, Syria, Thrace, Arabia, Pannonia, Illyricum, Dalmatia — the frontier's own officer corps ruling the empire it garrisoned, exactly as the Bato of Book VI had predicted the wolves would. The century's twenty-odd violent deaths were the price of a succession system that had never existed, and Diocletian's answer — colleagues, sacral distance, a court in place of a household — was the admission that the Republic's clothes could no longer be worn.

What the emergency built is the army of Book VIII, and the volume closes on its shape because everything that follows uses it: the perimeter retained but no longer trusted alone; behind it the mobile field forces Gallienus improvised and Diocletian will formalize, cavalry-heavy, imperially attached; the cities walled — hundreds of them across Gaul, Spain, and the Balkans, their circuits often built from their own monuments and enclosing a fraction of their old area; the taxation converted toward requisition in kind, the soldier supplied rather than paid; and the frontier populations increasingly settled inside the line under service obligation, the practice that will make the fifth century's armies and, in time, its kingdoms. The empire that emerges in 284 is poorer, harder, more bureaucratic, more military, and demonstrably more durable in the East than anything the Antonines built: it will survive the next crisis, and the one after that, and in its eastern half it will outlast the volume that follows by a thousand years.

The last accounting is the sources', and it is this volume's peculiar legacy. The compendium has narrated the empire's maximum and its near-death from a Persian king's cliff, a garrison's abandoned paperwork, an emperor's private notebook, a rebel commander's requisition slips, a widow's unclaimed deeds, sixty thousand coin-hoards, and a marble column that tells the truth about engineering and lies about war. The great historians stopped: after Dio, no contemporary Roman narrative of Roman wars survives complete until Ammianus picks up the story in Book VIII's century — and the Editor has treated that silence as a fact about the age rather than a defect in it, because a governing class that has lost the command of armies stops writing their history. Book VIII will find both the narrative and the wars transformed: Ammianus Marcellinus, a serving officer, writing the last great Latin history; Julian's Persian disaster; Adrianople; the Rhine crossing of 406 and the sack of 410; the kingdoms crystallizing inside the wall, and the long negotiated dissolution that ends, on the fourth of September 476, with a German officer sending a boy emperor into retirement in Campania and the imperial regalia east to Constantinople with a note explaining that the West no longer required them. The perimeter this volume built, broke, and rebuilt will hold for a further century — and then, one province at a time, stop being Roman without ever quite being conquered. That is the last volume's subject. This one ends where it must: at Nicomedia, on the twentieth of November 284, with the army — as always, in this compendium, the army — raising a Dalmatian cavalry commander on a shield and asking him to make the world work again.

The evidence

The Epilogue, and the volume's apparatus in retrospect

Open the evidence file

The Epilogue's synthesis rests on the volume's cumulative evidence; its structural claims are the standard modern reconstructions, and the Editor names the debates rather than settling them: the "crisis" itself as a category is contested — Liebeschuetz, Witschel, and the regionalist school have argued that Africa, Egypt, and much of the East show continuity rather than collapse, and that the crisis was severest exactly where our sources' authors lived; the plagues' demography (Duncan-Jones and Harper against Bruun and Greenberg) is unresolved; Gallienus's cavalry reform's scale is disputed (de Blois against the older Homo picture); the Dacian evacuation's completeness remains an archaeological argument; and the fourth-century Kaisergeschichte hypothesis, on which the epitomators' agreement rests, is a reconstruction, not a text. Where this Epilogue asserts (the asymmetry between military failure and civilizational continuity; the army as the century's governing class; Diocletian's reorganization as an admission rather than an innovation), it asserts as the Editor.

The volume's cumulative source-map: narrative spine — Dio 68–80 (epitomated to 180, autoptic to 229), Herodian (180–238), then epitome and Byzantine excerpt only, with Dexippus and Eunapius visible beneath Zosimus and the new palimpsest fragments still revising the field; biography — the Historia Augusta, usable under audit to Caracalla and DGrade D thereafter, the single most damaging document in Roman historiography and the reason this volume's later chapters are built from stone; contemporary voices admitted — Marcus's Meditations, Fronto, Galen, Cyprian, Aristides, Sibylline Oracles 13; and the documentary-material stratum that has carried this volume more than any predecessor: Trajan's Column and the Adamclisi altar, the Maximus stele, Hadrian's Wall and the Lambaesis addresses, the Bar Kokhba letters and Babatha's archive, the Antonine distance slabs, the Column of Marcus Aurelius, the Severan jurists and the Constitutio Antoniniana, Dura-Europos entire, the Res Gestae Divi Saporis, the Aurelian Wall, and the coinage — whose silver content, mint-marks, and hoard-horizons are, for the years 235–284, the most trustworthy continuous record of Roman history in existence. Book VIII opens with the narrative restored and the empire changed: Ammianus, Zosimus's better sources, the Notitia Dignitatum, the church historians, and the codes — a documented world, dissolving.

Here ends Book VII of Imperium Romanum: the High Empire and the Long Crisis, from the accession of Nerva to the accession of Diocletian. Book VIII will treat the last age of the Western empire: the Tetrarchy and the wars of Constantine, Julian's march to Ctesiphon and the ruin at Adrianople, the Rhine crossing and the Gothic sack of Rome, Aetius and the Huns at the Catalaunian Plains, and the long dissolution that ends on the fourth of September 476 — the final volume of the compendium, and the one this series has been travelling toward since a Gallic army crossed the Allia.