When Empires Stop Believing Their Own Stories

Ammianus Marcellinus served as an officer in the Roman army through the campaigns of the mid-fourth century, and when he sat down in the 380s to write the history of his own time, the empire around him was still functioning. The roads still carried trade. The legal codes still applied. The grain still arrived in the cities from the provinces, and the imperial bureaucracy still recorded its arrival in the formal Latin of an institutional order that had governed the Mediterranean world for four hundred years. And yet what comes through in his writing, beneath the conventional praise of emperors and the formal cadences of imperial historiography, is something quieter and harder to place. He writes about the institutions he serves in the past tense, even as they continue to operate around him. He chronicles their workings with the precision of a man documenting something whose meaning has already shifted, though no one has yet announced the shift, and the apparatus continues turning over as it has always done.

The question this essay wants to hold is not when Rome fell, because in any structural sense that Ammianus would have recognised, Rome had not yet fallen and would not fall in the West for nearly another century. The question is when the people inside the empire stopped believing the story it told about itself, and what happened in the long interval between that loss of belief and the eventual institutional reckoning. The Roman case is not unique in this respect. It is one of several historical instances in which the narrative that holds an imperial order together loses its credibility well before the material structures it sustains begin to give way. The pattern is worth examining now, in a register more careful than the usual rhetoric of decline, because some version of it appears to be unfolding in our own moment, and the historical record suggests that what people do during the interval matters in ways that are not always visible at the time.

What a narrative does for a political order

Imperial narratives are not ornamental, and treating them as ornament is one of the more common errors in how we read political history. The stories that empires tell about themselves perform a structural function, which is to translate the daily fact of hierarchy into something that can be lived with by the people inside it. They convert coercion into consent, extraction into legitimate exchange, the asymmetries of power into the appearance of shared destiny. The Roman language of civilitas and lawful order, the British language of commerce and civilising mission, the American language of opportunity and freedom: each of these performs the same essential work, which is to give the people inside the system a reason to participate that does not require constant force to maintain.

When the narrative is intact, the system absorbs its own contradictions with surprising efficiency. The wage that does not quite cover the rent, the promotion that goes to the better-connected candidate, the legal outcome that favours capital over labour: each of these gets metabolised, by the person experiencing it, as a piece of personal or local misfortune rather than as evidence of a structural pattern. The narrative does not need to deny that bad things happen. It only needs to provide a frame in which the bad things appear as exceptions to a fundamentally just arrangement, and to keep that frame intact enough that the exceptions never aggregate into a different kind of conclusion.

What collapses, when an imperial narrative reaches the end of its useful life, is not the events themselves but the interpretive frame. The same wage, the same denied promotion, the same political dysfunction, all of which were once read as individual setbacks to be solved through individual effort, begin to be read as structural betrayals. The losses do not change. The story that explains them does. Historians tend to identify this threshold more clearly in retrospect than the people living through it did at the time, partly because the change is gradual and partly because the structures continue to operate during the shift, which creates the impression that nothing fundamental has yet happened. The work of this essay is to make the threshold more visible by looking at how other societies have crossed it.

Rome and the slow exhaustion of civilitas

The Roman case is instructive because the gap between narrative and reality opened slowly and was bridged for a long time by the sheer momentum of institutional inertia. The story Rome told about itself, in its mature imperial form, rested on the idea that the empire was a vehicle for civilisation, that Roman citizenship conferred genuine dignity and meaningful participation in the res publica, and that the burdens of belonging were balanced by real protections and shared advantages. By the late fourth century this story had been straining for generations. The Constitutio Antoniniana of 212, which extended citizenship to virtually all free inhabitants of the empire, had over time emptied citizenship of much of its earlier content. The legal category remained, but the substantive participation it had once implied had been progressively narrowed by the consolidation of authority in the imperial household and its administrative apparatus. Citizenship had become, for most of the population, a status without a corresponding share in power.

What is most interesting in the historical record is not that this gap existed, since gaps of this kind exist in most political orders, but what the educated Roman class did with their growing awareness of it. Some doubled down on the official story, producing increasingly elaborate defenses of imperial virtue in panegyrics that were already, in their own time, recognised as overstatements. Some withdrew from public life into philosophical retreat, into the older Stoic and Platonic schools or, increasingly, into the early Christian communities that offered an entirely different account of meaning and belonging, one that did not require Rome’s narrative to make sense of a person’s life. Some occupied positions inside the imperial bureaucracy while privately refusing to invest in it emotionally, a posture that produced effective administration and quietly hollow institutions in roughly equal measure.

Salvian of Marseille, writing in the fifth century, offers one of the more striking documents of this period. He observed that the Roman poor in some provinces were beginning to prefer life under the so-called barbarian peoples then settling within and beyond the imperial frontiers, because the tax burden and the corruption inside the empire had become more punishing to ordinary life than the disorder outside it. This is not a claim that needed to be true in every particular to be historically significant. What matters is that it could be written and read by literate Romans without seeming absurd. The narrative had thinned to the point where the alternative was no longer unthinkable.

The lesson the Roman case offers is not that the empire collapsed because the story failed. The story failed first, and the institutional collapse followed at its own pace, which in the West took roughly a century and in the East was deferred by an active and quite deliberate reworking of the narrative under Justinian and his successors, who kept a version of the Roman story going for another thousand years by modifying it substantially. What the historical record suggests is that the relationship between narrative exhaustion and material collapse is real but loose, and that the interval between them is where most of the consequential historical work tends to get done.

Britain and the embarrassment of the civilising story

The British case is useful for a different reason. The gap between narrative and reality was sharper, the timeline more compressed, and the moment of recognition more easily dated. The British Empire told itself, and the world, that it was spreading civilisation, free trade, the rule of law, and humane governance. The actual mechanism, at the level of operational reality, was extraction at a scale that depopulated parts of the subcontinent during recurring famines, reorganised global agriculture around the dietary and industrial needs of the metropole, and produced what economic historians have come to estimate as one of the largest involuntary transfers of wealth in human history. The story and the reality could be held in the same mind for a long time, with the help of some considerable rhetorical labour, but the labour itself became increasingly difficult as the twentieth century progressed.

What is worth attending to in the British case is the specific moment when the administrative class itself began to find the story uncomfortable to maintain. The Bengal famine of 1943, in which the wartime requisition policies of the British government contributed materially to the deaths of perhaps three million people in eastern India, produced a quiet but unmistakable shift in the internal correspondence of the Indian Civil Service. The partition of India and Pakistan in 1947 was managed with a haste that suggested an administrative class that had ceased to believe its own legitimating account of its presence on the subcontinent. The Suez crisis of 1956 is perhaps the clearest single inflection point. The military failure was technically recoverable, in the sense that the actual operational losses were modest. The narrative failure was not. After Suez, the educated British class could no longer tell itself the imperial story with a straight face, and the political will to bear the rising costs of empire dissolved with a speed that surprised the people who had assumed those costs would continue to be borne indefinitely.

The point to draw from the British case is that the empire was not lost because Britain ran out of soldiers or money in any absolute sense. The capacity to believe that holding the empire was a moral act, or even a defensible one, ran out first, and the political and economic withdrawal followed. The decolonisation that unfolded across the late 1940s, the 1950s, and the 1960s was, from inside the British administrative apparatus, less a series of defeats than a series of negotiated retreats from a story that had become awkward to maintain in front of an increasingly literate and globally connected population.

It is also worth acknowledging, in the same breath, that the loss of the imperial narrative did not produce a corresponding moral renewal in British public life. It produced a new set of stories that performed similar legitimating functions in less explicit registers: the special relationship with the United States, the Commonwealth as a softer afterlife of empire, eventually the European project, and more recently a nostalgic and historically loose account of British identity that has reshaped the country’s politics in ways that are still being worked out. The exhaustion of one narrative is rarely the end of narrative as such. It is the beginning of a search for what comes next, and the historical record indicates that the search can produce results of widely varying quality.

What disbelief looks like from the inside

The American case is unfinished and ought to be treated as such. What can be said with reasonable confidence is that the narrative which has held the postwar American order together is showing the kind of stress that previous examples in this essay have suggested as historically significant. The American story, in its mature twentieth-century form, has rested on a set of mutually reinforcing claims: that the system rewards effort, that mobility between economic positions is real and broadly available, that the country possesses an unusual capacity for self-correction through its political institutions, and that the future will, in some meaningful sense, be better than the past for the people who participate in good faith. The empirical record on each of these claims has been deteriorating for at least two decades, in ways that have been thoroughly documented in the economic and sociological literature.

The development that is harder to capture in the data, and that the historical pattern in this essay suggests is more significant than the data itself, is the change in how the deterioration is now being talked about. Until quite recently, the dominant interpretive frame for disappointing outcomes within the American system was personal or temporary. The lost job was an occasion for retraining. The stagnant wage was a problem to be addressed through harder work, geographic relocation, or smarter financial decisions. The political dysfunction was a phase that the system would eventually correct, because the system had always corrected itself before, and there was no obvious reason to believe that this would change. What has shifted, and the shift is now visible across a range of registers from the public-opinion data to the texture of ordinary conversation, is that this interpretive frame has become harder to maintain even for the people whose continued belief in it would most plausibly serve their interests. The same outcomes that were once read as personal are increasingly being read as structural. The vocabulary people use to describe their working lives, their economic prospects, and their political institutions has moved away from the language of opportunity and self-correction toward the language of endurance, exposure, and structural fatigue.

This is not collapse, and the essay is not interested in pretending that it is. What it appears to be, if the historical pattern means anything, is the texture of disbelief beginning to spread through a population that still depends, both materially and psychologically, on the structures whose legitimating story it can no longer hold with much conviction. What distinguishes the present moment from the historical cases this essay has examined is the speed at which the narrative has thinned, and the absence, so far, of a coherent replacement that could perform the same legitimating work in a different register. The interval has opened. What gets done inside it is the part that is not yet settled.

What the interval is actually for

The historical examples are not offered as evidence that decline is inevitable, and the essay would be doing the reader a disservice if it left that impression. The Roman and British cases are offered for a different reason, which is that they suggest the period between narrative exhaustion and material reorganisation is a period of unusual openness, in which the work that ordinary people do at smaller scales tends to carry more historical weight than it does during periods of institutional stability. This is, I think, the part of the pattern that is most worth attending to, and the part that the more familiar narratives of decline tend to obscure.

Rome’s interval produced, among other things, the monastic networks that preserved Greek and Latin literacy through the period of institutional collapse that followed in the West. The work was done by communities that had no realistic expectation of restoring the empire and were not trying to. They were trying to maintain something that seemed to them worth maintaining: a body of texts, a practice of attention, a form of communal life. The historical consequence of that work, which became visible only across centuries, was the survival of intellectual continuity into the medieval and eventually the early modern periods. The Carolingian renaissance, the cathedral schools of the twelfth century, the universities that emerged from them, all rested on a base of preserved material that the monastic communities had quietly carried through the interval without any sense that they were doing what posterity would eventually understand them to have done.

Britain’s interval, briefer and less catastrophic, produced the welfare state, the National Health Service, and the postwar social settlement that reorganised the country around premises substantially different from those of imperial Britain. The work was done by political coalitions, trade unions, friendly societies, and reform movements that had been building their organisational and intellectual capacity through the long period of imperial decline, and were positioned to act when the political space opened in 1945. They did not produce a perfect settlement, and the settlement they produced has been substantially eroded in the half-century since. What they produced was nonetheless a real and consequential alternative, built during the interval by people who had been quietly preparing for it.

The historical pattern suggests that the dissolution of a dominant story creates space, and what fills that space depends substantially on what is already being built when the space opens. The institutions, communities, and practices that turn out to matter in such intervals are rarely the ones that were dominant inside the failing structure. They are the ones that had been growing at smaller scales, often in less visible places, often without any clear sense that what they were doing would eventually be needed. This is a description of the historical record rather than a piece of advice, but it carries an implication that seems worth naming. The work of building meaning, capacity, and continuity at one’s own scale, in one’s own community, in one’s own field of practice, tends to feel disproportionate to the scale of what is failing at the level of the larger system. The historical pattern indicates that this disproportion is the norm rather than the exception. The cathedrals were built by people who did not see them finished. The institutions that outlasted Rome were begun by individuals who had no reasonable expectation that their work would survive them, and who did the work anyway.

A closing reflection

The narrative that has held our political and economic order together is showing its age, and the signs of strain are real enough that pretending otherwise has itself become a kind of effort. That is a difficult thing to sit with, and the essay does not want to minimise the difficulty. It is also a condition with historical precedent, and the precedent suggests, with the qualifications appropriate to any reading of history, that the question worth holding is not whether the story can be rescued. The historical record indicates that stories at this stage rarely can be. The more useful question is what gets built in the long interval between the exhaustion of one legitimating account and the eventual emergence of whatever comes after it.

The writers, builders, teachers, organisers, and quiet institutional makers of every previous interval did not know, while they were inside it, that the interval was what they were inside of. They kept making the things that seemed to them worth making, on whatever scale was available to them, often without any clear sense of how the work would matter or whether it would last. The historical pattern, taken as a whole, suggests that this is more often the right response to such a moment than the more dramatic alternatives that tend to attract more attention at the time. The work that turned out to carry forward was the work that was being done quietly while the larger story was still in the process of losing its hold.

* * *

For a longer historical treatment of this pattern across four imperial forms, you can read Echoes of Empire: A Short History of Repeating Mistakes, available as a free guide.

Time’s Mirror publishes weekly essays at the intersection of history, economics, and psychology. You can subscribe at timesmirror.substack.com.