Did the Environment Kill Rome?

An article at Vox argues that environmental disasters destroyed the Roman Empire: “6 Ways Climate Change and Disease Helped Topple the Roman Empire,” by Kyle Harper, a professor of classics, whose article summarizes his book The Fate of Rome: Climate, Disease, and the End of an Empire. And in the process teaches us how not to do history. Sort of.

Harper’s sin is not being a bad writer. Or even a bad historian. His article and his book are well written, consistently engaging, deliciously fact-filled, and interesting. In fact I highly recommend his book, especially as an excellent piece of a complete puzzle. His sin, rather, lies in not actually testing the theories proposed. His constructed timelines are models of verification bias, not actually a test of his theory. You need to match social changes to the actual events correlating with them. Not just the ones you claim matter. That’s circular reasoning. Although this is just a special instance of a more general error: Not trying to prove your thesis false, before asserting it’s true. And all of the article’s other errors manifest a failure at that more general rule. Harper collects a fascinating body of facts—especially in his remarkable book, where he does show environmental changes mattered in some sense to every course of events. But his methodology, doesn’t always work.

I’m most concerned that by their brief and sensational nature, articles like Harper’s Vox piece run the danger of overselling the idea of environmental determinism. A dose of caution is warranted.

What Really Happened

Of course atheists do like to accuse the Christians of destroying Rome. “They took over, and then everything went to shit.” Except that’s not true. It commits the same errors as you could say this Vox article does. Christianity was a symptom, not a cause, of the Fall of Rome. As a phenomenon, Christianity can be blamed, in significant part, for leading us from the fall of Rome into the Dark Ages, and for sustaining those Dark Ages for centuries, holding us back from scientific and technological progress for almost a thousand years. As opposed to helping society rebuild and recover from what was befalling it (see my chapter on the Dark Ages in Christianity Is Not Great).

Though Christians would eventually figure out how to retool their ideology to recover and get us back on track, it took a thousand years, and that retooling amounted to repaganization (in its intellectual values, including the moral goodness of curiosity, the supremacy of empiricism, and a renewed belief in epistemic progress), a movement begun after the 11th century and pressed most concertedly in the 15th and 16th centuries. And that movement still struggled against intense opposition from other Christians even then (I document all this in my new book The Scientist in the Early Roman Empire).

The real causes of the Fall of Rome were ultimately human. In particular, by failing to establish a constitution that ensured the peaceful succession of power, the Roman Empire was perpetually thrown into civil war, and thus inherently unstable. Like a child continually playing with fire, it was statistically inevitable the whole house would burn down (a theory of structural failure well known to Harper: Fate, pp. 3 & 12). Indeed, if you removed that one causal factor, there is no evidence any of the other disasters that befell the Empire would have crashed it.

The second most proximate cause was the collapse of the Empire’s fiduciary economy, caused in large part by, you guessed it, their civil wars (and thus already inevitable, given the perpetual condition of civil war structurally inherent in the Empire). And the response was a catastrophically bad economic policy, including iron-fisted fascism and draconian price controls (and eventually, under Constantine, the de facto enslavement of much of the Empire’s farming class, with the invention of the serf), rather than state investment in its people’s survival and recovery. Human decisions, that doomed the Empire thereafter.

Everything else Rome could probably have handled. Because when times were good (civil war was at bay, and the economy flourishing) it always did. In such conditions, foreign invaders were always dealt with, famines and plagues easily recovered from, and environmental degradation always answered—as today—with technological innovations that actually left the Empire better off (a fact of which Harper is aware: Fate, pp. 19 & 35-38). The key was Pax Romana, Roman Peace. Break that, and everything else would eat them alive. As befell any other society of such scale.

This is not unparalleled in history. The Empires of Athens and China likewise underwent collapse and reconstruction when peace and prosperity crumbled as a result of inept human policy. But even they had foreign powers fully their equal or even far superior tearing at them. Rome never did. It’s collapse occurred even in the absence of any actually comparable military power, its foreign invaders only ever succeeding owing to Rome’s technical and economic decline. Hence an effect, not a cause, of that decline.

The Correct Timeline

Rome wavered between periods of peace and civil war all the way from the fall of the Republic to the fall of the Empire. But the most devastating, and clearly the mortal wound that eventually bled Rome out, was the greatest of all civil wars, a succession of them, which essentially lasted fifty years, from 235 to 284 A.D. (called the Crisis of the Third Century). During which time Rome endured over twenty emperors, six coups, and numerous failed rebellions, and was continually at war with itself, annihilating its population and disrupting its economic centers and trade networks. It actually split and reunited several times in the course of that fifty year span of near continual war.

(Also, a couple of attempts were made in the Third Century to create a new state religion to control the populace with…so using one like Christianity for that purpose was not an innovative idea, it’s just who won the game of musical chairs: Constantine was only the first to attempt that strategy who successfully avoided assassination; and his motives for choosing Christianity over its closest competitors were politically apt, given his enemies at the time and the lessons of recent history: see Chapter 18 of my book Not the Impossible Faith.)

As that continual stretch of civil wars approached an end, the fiduciary economy collapsed in the 270’s, an event analogous to our Great Depression. And contrary to what you often read, it was not currency debasement that caused that. In fact, fiat currencies are fully functional (we’ve already been living on them, worldwide, for a century). And the data show that they even worked fine during most of the civil warring of the 3rd century (producing mild inflation at best, which is infinitely sustainable, as we now experience).

But as soon as banks, creditors, and merchants lose faith in the token currency the economy is built on—as tends to happen after decades of unending war establishes extreme uncertainty in the stability and credit of the government, and keeps disrupting trade centers and networks to the point of evoking panic—the system collapses. As ours did, from different causes, in 1929. Indeed, what happened to us shows that what currency you use, its intrinsic value, is actually irrelevant. As our Great Depression resulted not from a loss of faith in fiat currencies (which no nation had yet adopted), but from a loss of faith in the very gold and silver that currencies were then fully backed by. Indeed, we solved that problem by converting to fiat currencies, over which we have far more policy control; and we’ve been rising on a near continual tide of economic growth ever since.

The Romans answered their own fiduciary collapse, instead, with fascist price controls (and eventually, serfdom). Exactly the worst thing you can do. Imagine if the United States’ Civil War lasted fifty years instead of five, and then ended with the Great Depression, which was then answered by the embrace of fascism and a price-controlled economy. The U.S. would have been screwed. So, too, the Roman Empire. The doom of Rome was thereafter sealed. Especially as they still didn’t solve the underlying problem. There remained no system ensuring a peaceful succession of power, so Rome just continued its endless cycle of civil wars, until it was split internally into two separate empires, and finally torn to shreds by foreign invaders the Romans no longer had the economic, demographic, or technical ability to repel.

Fans of Late Antiquity (usually Christian historians) like to argue that the 4th and 5th centuries were actually economically glorious. But that’s another example of doing history wrong. In fact, Rome was in an aggregate economic decline from the 270s to its final ends. Centers of glory remained, but shrank in number. The net effect, was catastrophic loss. The end came in the West in the 6th century; while in the East, it surged in the 7th century by capitalizing militarily on the collapse of the West, but all for nought, as it then steadily shrank every century thereafter (economically, geographically, intellectually, and militarily), ending in a bare cinder that was eventually snuffed out in the 15th.

What those historians do wrong is look at the outrageous displays of wealth undertaken by the elite in Late Antiquity, or in certain rare centers of concentrated wealth, and the dominance of a small class of landholders below or around them, and then claim, “Look, see, everyone was rich and everything was prosperous!” When in fact, that “wealthy elite” and the landholding class were declining in size, and income disparity was increasing disastrously. For a useful analogy in fiction, think Panem. Those enormous displays of wealth, and the flourishing of landlords below that, do not reflect the economy as a whole, but simply the rise of a shrinking 1% hoarding and wasting so much wealth they were literally starving to death millions of desperate people, the doomed 99% who saw little but continual decline and ruin, their numbers dwindling precipitously by comparison. And not being replaced. While those that survived, were not being reliably fed or supplied in any way comparable to the real glory days centuries before.

While major capitols still saw impressive investment, most cities were shrinking in size or dying entirely. And production and trade fell continually in scale across the entire Empire. As all archaeological data confirm (e.g. the documented decline in urbanization, decline in trade, decline in industry, decline in population and consumption, etc.). Harper tries to deny this in his book at one point, but his evidence is simply incapable of redrawing the picture (Fate, pp. 175-88). The hard data are far more telling than a gaggle of non sequiturs and the selective evidence of a successful few.

The conclusion is: the Fall of Rome was caused by the Crisis of the Third Century; the Crisis of the Third Century was caused by the lack of a constitution ensuring a peaceful succession of power; the lack of a constitution ensuring a peaceful succession of power is a human failure, not at all related to any environmental disasters. Environmental disasters have effects on a society only in relation to how human decisions react to them. They are impotent by themselves. Devastating to people, but not to civilizations or empires. One may imagine exceptions that have yet to occur, like total thermonuclear war or a near-life-exterminating asteroid impact, in the face of which no human decision can effect recovery (except perhaps after some centuries of endured ruin). But no such event befell Rome. What did befall it, was of man’s making. And only better human decisions could have saved it.

The Harper Timeline: Was It Climate?

Harper argues that “Climate Change and Disease” were the culprits. He does allow in his title that they may only have “helped,” which would be a fair way to put it, but the text of his article (and his book) argues more for their actually being the cause. In reality, neither climate change nor disease matter, when people act to compensate for them, politically and technologically. So they “helped” only in the sense that they were the ordinary flu that killed the terminal AIDS victim. But they were going to die anyway. What finally did them in is not the underlying cause. Because the flu rarely kills anyone else. It’s definitely worthwhile to document these things (Harper’s book establishes they played a part). But claiming they were the cause is theoretical overreach.

Harper starts with contradictory causal explanations for the late 2nd century, at first saying a decline in food production must have resulted from a shift in climate, then saying a plague substantially reduced the population at exactly the same time. But those two disasters cancel each other out. How then can they have combined to cause any decline? “Maybe the loss of population caused such a crisis in the labor market that there wasn’t enough food to feed survivors?” All the evidence we have shows no such decline, but in fact ample labor, and more than adequate production. There were price adjustments. As all economies experience. But no food crisis is mentioned in any sources or indicated in any archaeology until after the Crisis of the Third Century was already hobbling the Empire.

The Romans usually could handle these things because they had a robust economy and infrastructure and were automating the agricultural industry, with water-powered flour factories, bread-kneading machines, and all manner of other labor-saving devices, freeing up and multiplying the productivity of existing labor well beyond that of any comparable society (as I also document in The Scientist in the Early Roman Empire). Only when those things failed in consequence of human policy did ever-present natural dangers dismantle the Empire.

It appears that Harper is starting with his hypothesis, and then looking for correlations that match his hypothesis. He is ignoring alternative correlations that may in fact have been the necessary and sufficient causes. In other words, he isn’t ever actually testing his hypothesis. But we can’t do that. We have to test our thesis against the best alternatives. Like, a fifty year civil war ending in total economic collapse. You know, things like that. And on top of that, Harper is ignoring that his own grab bag of causes contradict each other. So he’s not even trying to build a coherent causal model of what happened.

Harper’s first thesis is stated thus:

It turns out the Romans were lucky. The centuries during which the empire was built and flourished are known even to climate scientists as the “Roman Climate Optimum.” From circa 200 BC to AD 150, it was warm, wet, and stable across much of the territory the Romans conquered. In an agricultural economy, these conditions were a major boost to GDP. The population swelled yet still there was enough food to feed everyone.

He is right the Romans were lucky. In fact most societies are lucky. In one way or another. It would be unlikely for any society to not be lucky on any dimension of advantage we can measure. Due to the laws of probability, if there are, say, ten dimensions in which a society may have an advantage, and “lucky” means enjoying an advantage only 1 in 10 societies experience on that dimension, the probability that any randomly selected society will be “lucky” on at least one of those ten dimensions equals one minus nine-tenths to the power of ten, or 1 – (9/10)^10, which happens to be about 65%. Roughly 2 in 3. Most societies will therefore be “lucky” in some sense.

So we can’t really use being lucky to explain unique events. Least of all because one can say the Romans were lucky in many ways. For example, unlike most other pre-modern societies, they had an advanced science-based technology that they were putting to use in subsistence and industry (from aqueducts that port water uphill, to mechanical water-powered sawmills, to crop rotation and fertilization, to advances in land reclamation). And Harper is not factoring into his causal model all the other ways they were lucky. Which other advantages, can compensate for losing their luck on some other dimension. Why would those other advantages fail in one century and not the other? There has to be a deciding factor. And by being the deciding factor, that’s the cause. Particularly as that cause would have had the same effect without the added environmental stressors.

Even if we don’t credit “luck,” for example, but instead simply a fact of declining agricultural production (never minding that, in his own imagined model, a plague already solved that problem), we still have a few methodological problems to overcome before we can conclude that that affected Rome’s decline.

(1) We first have to ask…was agricultural production even declining? Roman agriculture boomed considerably due to technological innovations. The effect of climate change may have been comparatively minimal.

When you look at graphs estimating temperature by century, there is really not so radical a variability until after Rome had fallen. And when you look at graphs estimating precipitation by century, there is a decline in the 3rd century, but it recovers in the 4th, falls again in the 5th, and recovers again in the 6th, and even in the 3rd century the loss relative to the 2nd century barely exceeded 2.5cm a year. So we shouldn’t expect that to have had much notable effect. None, certainly, the Romans couldn’t answer technologically.

What happened was something else. In the peak period the Romans were reclaiming and irrigating extensive lands previously unusable. They were developing more productive crop breeds. They were increasing yields per hectare and hectares cultivated. They were improving and extending distribution networks. Just as we have continued doing, even with a worse climate. So it does not automatically follow that a less favorable climate will actually cause a decline in production. We know too many examples of that not being the case.

So is there evidence of a real decline in Roman agroproduction? Only after 250. And only significantly after 300. Which suspiciously correlates with other, more obvious causes. Particularly as disease and war were actually reducing the population, thus abrogating any decline in production that climate may have been causing. By contrast, there is no notable correlation we can find in the evidence between agroproduction and temperature or precipitation change across the whole Empire by century.

(2) But we also have to ask…would declining agroproduction even be a problem? One would assume declining production would lead to reduced population. But that happened during the Black Death, too. Which wiped out 30-60% of Europe’s population. After which the economy did not decline, but stabilized, and soon boomed. So we can’t even establish that the model itself is plausible, were it even factual.

Declining populations do not automatically entail declining civilizations. They can actually have the opposite effect, by raising wages, driving innovation, and increasing opportunities for social mobility. The example of the Back Death is particularly apt, as the kind of declines Harper’s theory would entail, would be dwarfed by the scale of population reduction in the 14th century, and well into the 15th century as outbreaks continued. Yet it had the opposite effect that Harper’s model would predict. What happened to all the major European Empires after the Black Death? Oh right. So we can’t even establish the causal model is valid.

Harper tries anyway, arguing:

But from the middle of the second century, the climate became less reliable. The all-important annual Nile flood became erratic. Droughts and severe cold spells became more common. The Climate Optimum became much less optimal.

This is true, although maybe a tad hyperbolic. As you can see from actual graphs of reconstructed seasonal temperature (linked above and more below), the increase in these erraticisms is hard to prove large enough to matter. Until after Rome had fallen. And even then, the change is not all that great…compared to modern industrial runaway warming. During which we are flourishing with a population literally a hundred times larger than Rome’s. Harper’s model has a problem. How are we, in the actual worse conditions he describes, maintaining a population a hundred times larger? Technology. What did the Romans have when faced with comparable problems? Technology.

Nile flooding patterns were becoming less important when Romans improved and extended artificial irrigation networks—erratic Nile flooding only became a lasting problem when this infrastructure decayed, not when the environment did. Prior, Nile droughts were known in every century, not peculiar to later ones. And yet Rome prevailed and Egypt remained a massive food producer. Even after the 3rd century. Droughts everywhere else were likewise being abrogated by extensive aqueduct and reservoir development.

Quite simply, boom-bust agrocycles were always the plague of all ancient civilizations and had been for thousands of years; the problem had already long been solved (to the extent it ever could with pre-modern tech) even before the Romans, by state-subsidized storage and distribution networks for riding out bust seasons and evening-out water supply. Good or bad outcomes were therefore always a function of human competence or incompetence in building and managing those technologies (another fact of which Harper was aware; indeed he extensively describes and praises the Roman Empire’s infrastructure of resilience in Fate, pp. 54-62).

And yet even with occasional human failure causing trouble from time to time, the overall pattern remained sustainable. There is no evidence this pattern ever got worse, until the technology did. Likewise, as conditions became less favorable in North Africa, they got better in middle Europe. The Romans were already extending development north to take advantage of that, expanding production in Germany, Spain, and France. So they could have compensated for any losses in African production with increases in European production. And if things turned the other way, vice versa.

In fact, there is no evidence of changes in climate causing any angst among Romans about supply until the middle of the third century. But by then they had already trashed their society with ceaseless internecine war. Which was the most obvious cause of their decline. Before that, localized variations due to their own deforestation and overproduction, and minor climate changes and routine seasonal variations (Wikipedia briefs some examples), including periodic droughts, could be offset by relocating centers of production or reclaiming unused land, and storing and redistributing the produce; and by improving infrastructure (again, technology).

There were other technological solutions already available should other problems arise; and the Romans frequently innovated new solutions whenever they came needed. So we can’t tell what effect those factors had on balancing out any problems one might assume obtained. Or later would have had, if Romans hadn’t stopped using them. Ultimately, beyond occasional famines that occurred in every century, before and during Rome’s rise, the sources never mention any long-term declines in food to feed the Empire. Until after the third century ruined everything. So we don’t really know that became a problem owing specifically to climate change.

Scientists love to over-speculate about history from ambiguous data that they don’t actually show any causal effect of. And Harper may be investing too much in the unproved speculations of such studies as this. Although he does a much better job than they do of trying to make a case. For a more cautious perspective, though, see the methodological sections of The Ancient Mediterranean Environment.

In the end, Harper’s attempt to conclude that all this “shows just how sensitive human societies can be to such change—now amplified in speed and scope by human activity” is not supported by any evidence from the Roman Empire. It could be true…of our current situation. But only because the scale and speed of climate change in modern times is vastly greater than any ever experienced in human history. To understand the effects of that, you simply won’t find past empires to compare with ours. And once again, it’s going to be human decisions that decide the outcome. Not climate change. Or disease…

The Harper Timeline: Was It Disease?

Harper is right that “in the AD 160s, at the apex of Roman dominance, the Empire fell victim to one of history’s first recorded pandemics—an event known as the ‘Antonine Plague‘.” Epidemics were known before then (most famously the plague that struck Athens in the 4th century B.C.). But the Roman Empire’s success in integrating a vast region with efficient transportation networks allowed a pandemic to strike for the first known time in the West.

Another such pandemic happened right amidst the Crisis of the Third Century (the Cyprian Plague from 250 to 270 A.D.). And another centuries later in the Eastern Empire, after the Western Empire had disintegrated (the Plague of Justinian in the 6th century). Each is estimated to have eliminated somewhere between a twelfth to a quarter of the population. So did plagues take down the Roman Empire?

Again, no evidence. For either the effect or the model.

The Roman Empire flourished for another whole lifetime after the Antonine plague. After it struck between 165-180 A.D., its subsequent economic effects were small or weakly attested (as even Harper’s own excellent research on the data occasionally admits), until no earlier than 235, when the Great “Civil War” Crisis began. That’s 55 years without collapse or even remarkable decline. The average human lifespan then was 48 years even for those who survived childhood. The plague was thus no more effectual than ordinary fluctuations century by century (indeed compare similar markers for the U.S. after the Great Depression, which are all substantially worse…and that didn’t end our civilization). So we can’t say the Antonine plague had that great an impact on the economy or politics. There were price adjustments. But no markers of decline followed. An entire generation was born and died between the plague’s end and the dawn of any truly alarming downturn.

Meanwhile, the Black Death, as I already pointed out, was far worse, and yet did not lead to collapse but in fact was immediately followed by a remarkable economic and political recovery in Europe. Influenza ravaged America and nations beyond, by some measures the worst pandemic in human history. Yet the American Empire was unphased. We went on to win two world wars immediately after and became the greatest terrestrial superpower. So even the model is not attested as a viable explanation of civilization collapse. In fact, I do not know of any example in the whole of human history of a plague downing a civilization. So why would we even consider it as a hypothesis? No examples of the effect despite repeated instances of the cause, supports no causal theory.

The only instances I know that come close, are the Athenian plague, and the American genocide. But both were compounded by far more obvious causes of their downfall: an equally powerful Sparta besieging the Athenians after the Athenian allies abandoned them in consequence of horrid governance from Athens (so, once again, human decisions were the factor; the plague would have been ineffectual but for that); and deliberate genocidal actions (from forced marches to mass murder) by more powerful empires (the U.K. and the U.S.), whose bringing of disease not only had atypical effects (due to a new population lacking immunity) but was also deliberate (the British Empire actively employing germ warfare to murder Native populations). Neither scenario is analogous to Rome.

Harper concedes the failure of his thesis by noting Rome was pretty much unaffected by the Antonine plague (indeed due to the resilience of the existing social system: Fate, pp. 115-18). And yet Harper then illogically claims “the resulting demographic crisis” from the next plague under Cyprian “triggered a total meltdown of the entire imperial system, known as the ‘crisis of the third century’.” Um. No. The Crisis of the Third Century began in 235, not 250. And it was a series of civil wars that lasted fifty years. Not a plague. And all the evidence shows the notable problem that followed the Crisis was the collapse of the fiduciary economy, a collapse that was caused by those extended back-to-back civil wars. There is no contemporary evidence that the plague was as notable a problem compared to those two factors: war and its effect on the currency. Had those other factors not been in place, how would the Cyprian plague have differed from the Antonine plague?

We can’t say the Crisis of the Third Century was “caused” by the Cyprian plague. That occurred deep within the Crisis, after civil war had already been crashing the Roman system for fifteen years. And there is no actual evidence of what effect it had, just speculation. It’s unclear how the Cyprian plague could have even had any negative effect in those circumstances. To the contrary, for all we know it helped end the Crisis by depleting the manpower needed to continue the civil wars that were the actual thing tearing Rome down.

At best, we can speculate that the Crisis of the Third Century left Rome vulnerable to the Cyprian plague in a way it never otherwise would have been (likewise to a unique but brief drought in Egypt around the same time, that Harper documents in Fate). It could have exacerbated the conditions that caused the fiduciary collapse, for example. But we know there is no causal rule that ensures it would have. After all, there was no such collapse after the Antonine plague—or any of the other regional droughts and famines Rome had always experienced.

The Harper Timeline: Was It a Butterfly Flapping Its Wings in China?

Next Harper suggests climate change ended Rome by driving the Huns West, thus in turn driving the Goths south, and thus ravaging the Roman Empire into dust. But this plays a little loose with the timeline. The Goths were being driven south into Rome since the mid-2nd century. Which also makes this another blaming-of-the-flu for killing the AIDS victim. The Goth incursion of the 4th-to-5th centuries was not so different from that of any other century. They invaded in the 2nd century to no effect, and in the 3rd century as well, right in the middle of The Crisis, and yet failed to kill the Empire even then. So why would they have succeeded in the 5th century? Not because they invaded. The mere fact of their invading did not have that effect in any prior century. So the Huns can’t explain anything. And so neither can any climate change affecting Hun migration explain anything.

The Goths probably succeeded in the 5th century because the Empire was already jacked up and in ruin, from never having recovered from the Crisis of the Third Century. They were just killing with a feeble rock an already dying foe. But I should also point out that Harper’s entire model is overdeterministic. Was it climate change that drove the expansion of the Hun Empire West? We don’t actually know. The documented megadrought in the Hun homeland could just have simply withered or ended the Hun clans, or driven them East or South; that someone instead got the idea to mass-migrate militarily West is a human decision. You know, someone like this guy.

The Westward expansion was actually begun by the Hun kings two generations prior to that guy. But humans nevertheless. So we are again looking at the effect of human decisions, not climatic determinism (a fact even Harper has to admit: Fate, p. 191). Likewise, the Goths actually didn’t “invade” Rome this time. They mostly sought asylum, and were happy to assimilate and even serve in Roman armies. It was then their horrid treatment by the Romans that turned them against the Empire. In other words, a human failure, inept policy. Not climate. A different policy decision could have actually turned the Goth migration into Rome’s advantage. Just as immigration has consistently benefitted Western nations in modernity, this could have even contributed to a Roman recovery. All but for bad decisions made by Roman leaders (as again even Harper must admit: Fate, p. 193).

We must not attempt to erase the effect of human decisions on history, by imagining the deterministic causation of climate change. But for certain kings deciding to expand their empire West, there would have been no Westward expansion. But for Roman leaders enacting disastrous policies in reaction to that expansion, there would have been no disaster from it, but boon. This is how history usually proceeds. Had the Hun migration taken place centuries earlier, we can expect the Roman Empire would have effectively repelled or conquered and assimilated any peoples they displaced. As in fact the Roman Empire did. When it kept civil wars in check and its economy flourished. But when their system was destabilized, stressed, and in decline, everything changed. And that change in the means and attitude of the 4th century, was a product of the 3rd.

The Harper Timeline: Did Events After the Fall of Rome Cause the Fall of Rome?

Harper then goes on to explain events in the 6th century did Rome in. But Rome was already gone then. So this becomes a quibble over which Rome you mean. The original one, or the new one built out of desperation in the East, as it sloughed off its losses in the West? Although the mere fact that the Empire had shrunk in half (the West collapsing and the East retreating) is already what we mean by decline. Harper blames the continuation of that decline (after a brief surge of victories and wealth in the Eastern Empire under Justinian) on another plague and a “mini-ice-age.”

That resurgence, though, was really just the opportunistic feasting on the corpse of its recently deceased Western sibling, an unsustainable program without any plan for recovery, which Justinian never implemented, nor even imagined so far as we can tell. We should have expected decline to continue after that anyway, even without any plagues or quasi-ice-ages. And Harper concurs that, whatever the cause, by 650 A.D. the Eastern Empire was but a straggling shadow of anything before (Fate, p. 11). And it was all downhill from there.

It seems more obvious that the East was still dissolving from the impact of the Crisis of the Third Century that it never implemented any effective policy to fix, instead magnifying income disparity ever more, losing territory ever more, producing no new advances of any impact, and never implementing any of the reforms necessary to recover a civilization—which is a policy failure, not a deterministically inevitable result of nature. And in result, right after Justinian’s doomed program, as one could have predicted, the so-called “Eastern Empire” continually shrank for hundreds of years until its last city fell to foreign invaders even before Europe had discovered the New World.

Again, we already showed how plagues cannot explain the fall of civilizations. All recorded instances show plagues actually help civilizations or have no longstanding effect on them, except when other, more overwhelming causes of their destruction are in place. You have to remember disease was in every decade already a mass killer until modern times (a fact Harper himself extensively describes in Fate, pp. 7 & 68-91). Millions routinely died of cholera, tuberculosis, typhoid, malaria, and other diseases, in every generation, even at Rome’s peak. Half of all children born, never reached adulthood. Half of all adults, never reached the age of fifty. They were replaced almost immediately in Malthusian fashion. If you kill off a quarter of a family, you just make room for another son or daughter to be born and raised. The net impact on a population is nullified in less than a generation. As for all the ailments killing the Romans in every year, so for any other disease added in. That a plague is rapid and remarkable, doesn’t really change the demographic effect all that much. And as we have seen, its social effects are always temporary or aren’t even uniformly bad.

Which gets us to the “How do you test that?” part of historical reasoning. Did any of the many famines under the Roman Empire, including during its height (precisely when Harper claims it was “lucky” and enjoying a paradisical climate), do the Empire in? (Such as those under Augustus and Claudius and Domitian and Hadrian.) If not, why would plagues do so? You need to have a plausible causal theory. It needs to have evidence from other cases confirming the model. And the indicators of it happening, have to be present in the case you are trying to explain with that model; while if there are other, more obvious causes present, and you can’t rule them out, you can’t rule your novel theory in.

Finally, Harper is really stretching with his bit about a “mini-ice-age.” Once again the causal model fails. The more famous “Little Ice Age” began in the late Renaissance, in Western Europe, not in the Middle East six hundred years earlier. It also lasted longer, and produced colder temperatures, and yet had precisely the opposite effect. But not only does the model fail, the evidence doesn’t even support it in the case Harper is proposing. Harper shows no causal effect of the prior, smaller cooling period, on agrocproduction or anything else. As we all know, mere correlation is not causation. Human failure remains the more obvious explanation of continued decline in the Eastern Empire.

Harper’s facts aren’t quite right, either:

We have long known that in the year 536 there was no summer; for about 15 months, the sun seemed to shine only dimly, unnerving people worldwide. In recent years, careful work on tree rings and polar ice cores has clarified what happened. First, in AD 536, there was a massive [volcanic] eruption in the Northern Hemisphere. Second, in AD 539/40, a tropical volcano erupted. The result was not just a year of darkness but truly staggering global cooling: The decade 536 to 545 was the coldest decade of the last 2000 years, with average summer temperatures in Europe falling by up to 2.5 degrees Celsius. And this was no passing phenomenon. For a century and a half, colder temperatures prevailed across large parts of the Northern Hemisphere.

Here, Harper’s science is wrong.

The latest studies show temperature declines were as bad or greater, and longer, under the Little Ice Age, which peaked in the 17th century. See this graph for worldwide data; red indicates the best and most recent scientific results, published in 2005; corroborated by the Klimenko study for Northeastern Europe; and for Europe as a whole, this study found the mini-ice age cooling Harper is talking about actually started in the 4th century and not the 6th, and for Summer temperatures was never any worse than the Little Ice Age. Yet the effect was exactly the opposite. Similarly, in Fate, p. 254, Harper tries to blame a decline in solar radiation in the 7th century, right in the middle of the Mini Ice Age, for furthering the collapse of the Eastern Empire; with a graph that shows the exact same decline in 1500, also in the middle of the Little Ice Age, yet which had exacty the opposite effect!

Once again, Harper’s model fails when tested against other cases. It has no support in any other cases. And it isn’t backed by enough evidence for the particular case in question to be sure it’s what caused any systemic collapse. Justinian’s building programs to mitigate the rising problem of excess precipitation (Fate, p. 255), for example, were simply too little, because his Empire was too weak. Odds are, Hadrian would have fared better. Likewise, the drying of Africa (Fate, p. 256) was as much affected by dying infrastructure as with shifting climate; and the rising rainfalls in Europe could have been cultivated to compensate for declining production in Africa. The failure to do that, is a human failure. The Empire had already disintegrated in the West. That the East was already suffering that loss was not some new climatic event.


You have to test your theories by trying to prove them false. You can’t just cherry pick any correlation that matches your theory and declare victory. That’s illogical. You also have to get your timelines correct, lest you stumble on false premises like “the Crisis of the Third Century was caused by the Plague of Cyprian” or “the Goths only started invading the Roman Empire after Huns fleeing drought pushed them to.” And you have to count misses as misses. If you are admitting the Antonine Plague had no effect, why are you still arguing plagues were causing decline? The causal assumption is refuted by countless contrary cases (the Black Death; the Modern Influenza Pandemic).

Harper occasionally admits these cautions in Fate. But if anyone takes Harper’s argument too far, trying to erase the causal role of human decisions from the course of historical events, then we would be entering the realm of pseudoscience. Climate change and disease simply don’t have the deterministic causal effects on civilizations and empires that Harper perhaps too strongly implies. There is no evidence in the whole of human history, West and East, of that ever being the case. They can be contributing factors, like many other things are. But ultimately it’s human decision making that will decide what those factors actually cause to happen. Leadership, innovation, technology, culture, policy, ideology. These decide the fate of nations. Everything else is just another decision waiting for humans to make.


  1. There was also the factor of lead poisoning from all the lead smelting going on around Rome. That and eating and drinking with lead utensils.

    1. There’s actually no evidence that had any effect. Because it was constantly present. Yet Rome thrived for hundreds of years. It was no different a factor than constant malaria, cholera, malnutrition, and everything else that impacts ancient (and sadly many third world) civilizations. The net effect was always overcome. Rome progressed and flourished in spite of it all.

      In truth, lead exposure was probably greater in early 20th century America due to the atmosphere being thoroughly polluted with it thanks to the auto revolution and leaded gasoline. Already from mid-19th to late 20th century lead use in canning food was probably comparable to exposure in antiquity. Ancient employment of lead was not much more extensive than that, e.g. lead pipes were calcified and thus did not leach as much, and lead utensils had more impact on the ancient middle and lower class than the elite, but the effect was likely comparable to middle and lower class Americans eating leaded canned food, and giving their kids lead toys, painting their houses with lead paints, etc. Add to that leaded air thanks to gasoline, and we really can’t claim Rome was any worse than us.

    2. Quite an interesting read. Thank you. Chalmers Johnson wrote a bit about Rome versus the United States, and what parallels could be drawn between the two as regards their respective Republics and Rome’s descent from a Republic to Imperialism. Chalmers Johnson thought that the U.S. will go one of two ways, maybe: either the U.S. go the way of Rome, and sacrifice its Republic to maintain empire; or, it may go the way of Britain after WWII, and sacrifice its empire (albeit begrudgingly and not smoothly) to save its Republic. Interesting thoughts. I happen to like Chalmers Johnson, and his unofficial trilogy of books criticizing the U.S. imperial juggernaut.

      “A nation can be one or the other, a democracy or an imperialist, but it can’t be both. If it sticks to imperialism, it will, like the old Roman Republic, on which so much of our system was modeled, lose its democracy to a domestic dictatorship.”


  2. Thanks Dr Carrier for that book’s review. Funny because I started to watch back HBO Rome Tv show this week… Actually the way they were treating slaves and other ethnics like Goths as inferior beings makes me feel like we are not missing any of that…

    I always thought that Roman arrogance and treachous ways were the main reason for the Empire fall.

    It’s interesting to have another view, but I never thought that climate changes could have been a major factor.

    And you do explain it well.

    1. Well, the arrogance and treacherous ways of Americans hasn’t really held us back much. Think Native Americans, Jim Crowe, our treatment of the Chinese, the Irish, the Japanese internment. It’s made the ride rough. But economy and empire keep advancing. Things get better on the other measure only but slowly.

  3. “Leadership, innovation, technology, culture, policy, ideology. These decide the fate of nations.”

    What is the major determinant of these factors?

    1. There is no primary determinant of them. Other than people and the choices they make.

      The ontological foundation of history as a field of knowledge is a single universal conclusion: the past causes the future.

      Beyond that, everything varies.

  4. We can see the problems, was anyone even dimly aware at the time of any of the problems; let alone aware and in a position to consciously do something to ameliorate them? You have pointed out the succession problem and the passing up of several opportunities to do something about it by the Senate for instance but that is just one substantial challenge; without sufficient knowledge or awareness comprehensive solutions are hardly possible: sooner or later the Res Publica would probably be knocked off it’s bike.

    1. Only in the sense that all systems have a nonzero probability of failure, and on an infinite timeline all nonzero probabilities approach 1.

      But that can mean thousands of years (like Egypt and China). The thing we are seeking to explain is why the Roman Empire, in spite of its economic, technological, and other successes, only lasted five hundred years.

      As to did they see it coming, that’s two different questions. The inevitability of collapse was a component of every major philosophical system of the time, including Stoicism and Aristotelianism, though there were plenty of elites who didn’t embrace the cycles theory or expected the next cycle to be thousands of years out. They were too close to events to think maybe these constant civil wars would be the culprit (indeed by the time the pattern might have become obvious, it was too late). Nor did they have yet a worked out theory of fiduciary currency and thus they had no currency policy.

      On the other hand, plenty of elites were well aware that the principate was volatile and a threat to their own freedom and wellbeing, and thus plenty of attempts were made to restore the Republic that all just happened to fail. If there were a more conscious understanding (as we now have in hindsight) of just how urgent and sweepingly important that project really was, more may have backed it, thus ensuring its success (which is how we can describe the American Revolution, IMO).

      In a sense, the first failure and thus the first such opportunity for success that was missed, was “the” Civil War itself: had Caesar been defeated by the Republicans, or had Caesar realized a scheme of restoring rather than usurping the Republic, and effecting systems to ameliorate rather than exacerbate internal state violence, there may still have been an Empire, but a democratic one, and thus a more sustainable one. (Other failed attempts occurred after the deaths of Nero and Aurelius; and yet another at some point in the third century; there may have been others.)

  5. So how do you test the hypothesis that it was the constitutional issues that were the determining factor in the 3rd century? (… it’s annoying enough that we can’t look in on the alternate universe with the control-group Rome that magically doesn’t have that particular problem and see what changes…)

    i.e., it seems like one could refute this by pointing to the 50+ years of civil war in the Late Republic from the time of Sulla to the final victory of Octavian, the same plague of opportunistic generals/billionaires and lack of clear succession,

    … and yet Rome emerges from THAT period stronger than ever.

    1. Technically it did. The Republic was destroyed.

      But it recovered well into the Empire, so your question is apt: one would want to ask, why didn’t that recovery happen after the third century? Partly because the first century B.C. was actually a period of comparatively few internal conflicts matched by continuing successful military expansion (the empire nearly doubled in size in the first century BC). Nothing at all like the twenty emperors and rampant rebellions and coups of the third century. And consequently there was no loss of confidence in the credit of the state, so no currency collapse.

      Indeed, the first civil war in the 1st century B.C. (the Social War) ended in expanding citizenship and integrating the Italians, precisely the right way to proceed to effect recovery. In fact, that all but ensured the Empire’s success (contrast with the completely opposite way the Goths were treated in the 4th century A.D.). Had Rome continued that province by province, it may have developed a stronger constitution for its future survival; whereas had Rome tried to not resolve that early civil war in Italy in such a manner, it may have been digging its grave.

      (Whereas when Rome finally did expand citizenship to the provinces, under Caracalla in the early 3rd A.D., the Empire was so deep into a class-divided monarchy that citizenship did not empower anyone or mollify anyone keen to rebel, and thus didn’t really help; in fact, citizenship had lost a lot of its social value by the previous division of citizens into honestiores and humiliores with different legal systems for both, replicating a system all but identical to the citizen-vs.-noncitizen system before it, and thus expanding citizenship did not really change much of anything.)

      The second civil war in the first century B.C., was Sulla’s coup. The third, Caesar’s. The fourth, the consolidation of Octavian. Each enacted reforms that made some things worse and some things better. But the net effect was doubling the Empire’s size and power and unleashing the economic powerhouse of newly acquired and now strongly controlled territories.

      These later civil wars were brief with effective periods of recovery and expansion, and even reform, in between (Sulla’s war, only 83-81; Caesar’s, almost a lifetime later, 48-45; Octavian’s over a decade later, only 32-30, and peace for a lifetime after). The Crisis of the Third Century was this times ten: six complete regime changes, almost no periods of peace, no effective reforms, regional rebellions, and most reigns lasting only a few years or less; all persistent without rest. Resulting in loss of confidence any regime could recover at all, resulting in the fiduciary collapse. Hence, very different scenarios.

      One might also argue the Republican system, being more democratic (especially after the expansion of voting rights after the Social War), was more capable of recovery than a monarchy, which was at the mercy of the decisions of twenty dictators felled by dozens of assassinations and never freed from dealing with perpetual internecine war.

      But one could also argue the Republic was weak precisely because it collapsed into monarchy. It’s a different question what one could have changed to make the Republic less prone to those structural failures and thus less susceptible to collapsing into monarchy. I think ideology, culture, human choices, chance victories, had parts to play there, too.

  6. I would expand on your theory somewhat. I think that the basic contract of the Roman republic, and of the Roman empire (and of most empires), was that Rome would give people peace and technology. In return, the people would pay taxes to Rome, and suffer the sometimes unjust and brutal repression of the central power. But as the Republic failed, the elites started to forget about this contract. It became more about their personal opulence and grandeur.

    For a couple of centuries, this worked out sort of fine, because the empire was still run by professionals who knew how to handle various crises.

    But eventually, the rot got bad enough that it was not obvious to the subjects of Rome that the empire benefited them. The peace was disturbed because the elites were fighting each other, mainly to attain personal status. Often supported by lower classes (not particularly the lowest classes, but those below the very elites) who were promised that the system would be fixed, if they only supported the right man for emperor. But of course it was not fixed.

    So the economy was damaged from all these wars, but at the same time the elites still wanted their shiny villas and opulent lifestyles. So the population was taxed more, even to the point of serfdom. Which fuelled more rebellion, which cost more to repress and damaged the economy further, getting into a vicious spiral where less and less of the taxes paid were of any benefit to the populace. Perhaps breaking free from Rome was still not truly rational, but it became rational enough that the amount of rebellion became impossible to curb.

    I think that blaming the ‘constitution’ is a bit too specific. For instance, China had a concept known as the ‘mandate of heaven’ that justified the rule of an emperor, but that at the same time justified rebellion against an emperor who did not serve the interest of the people. So even though China suffered terrible rulers, civil wars and partial declines, the empire still managed to right itself eventually, for thousands of years.

    1. I think I would dispute that the Roman Republic ever had much of a basic contract with respect to The People in general.

      As I understand it, the situation in 500 BC was that they’d just booted their last Etruscan king and, faced with the question of which leading family would take over the monarchy, given no obvious choice and the prospect of a bloody and protracted war of succession, they decided to try Something New.

      Greek ideas being in the air, there being at least some level of trade (they got their copycat pantheon from somewhere, after all), it’s a fair bet the more educated folks at least vaguely knew about them. But their reaction to direct democracy would almost certainly have been something along the lines of “F–K NO”.

      The Republic was essentially the patricians’ power-sharing agreement. At some point they created assemblies for the military and the plebes as a safety valve so that they’d at least have a chance of hearing in advance about problems they might be needing to address, but there was never any point at which the Senate and the patrician class didn’t hold all of the cards that mattered.

      And, this not being Athens, and the French Revolution being 2200 years off, there’d be no reason for anyone to have expected otherwise.

      As for taxation, in the provinces, taxes were basically “bring enough back to make this worth our while” and, beyond that, whatever else the proconsul could get away with collecting to line his own pockets. The smarter ones, like Julius Caesar, eventually figured out they needed something more sustainable; Caesar made that work in Gaul and that became the model for later, but by that point the Republic was pretty much toast anyway.

      1. I’m not talking about some modern day concept of universal human rights or democracy. By Rome giving ‘people’ peace and technology, I do not mean ‘The People’ in this modern sense. I mean the people with some degree of power. Local elites who could decide to side with some usurper, or a neighbouring empire. Average folks would only matter in the sense that if they were truly starving, they might desperately rise and topple the local elites (but clearly not the empire). So local elites want a system that prevents that.

        I’m not talking about democracy, but about having a certain sense that rulers cannot purely be the strongest robbers. An empire does not survive for long if peace is only upheld through fear.

        Again, compare with the Chinese ‘Mandate of Heaven’. It is not that China was a democracy, or that it was egalitarian. Even the best emperors were shockingly brutal by modern standards, and lived in seemingly unlimited grandiosity while the masses were at the brink of starvation even in good times. But there was this basic idea or understanding among the elites, that an emperor could not afford to be seen as a menace to the rest of society. Because if he did not understand that, he would be toppled by someone who had that understanding, and that would be seen as just by the educated elites. Other elites would support such a rebellion not simply because they had some personal ties with the usurper, but because they believed that restoring this order was vital for the long term prosperity of themselves, their families, and the culture they saw themselves as part of.

        I think that there was such an understanding in the Roman Republic, perhaps not from the very beginning, but clearly visible by the way the fall of the Republic was seen as a disaster by many in the educated class, including many who held real power.

        But I think that after the fall of the Republic, any such understanding was gradually lost, and that became the doom of the Empire.

  7. One factor often overlooked is the Fall of Jerusalem and the subsequent loss of income from the crossroads between Africa, Asia, and Europe. Most importantly, the military blow given to the Romans in the Bar Kochba revolt in which ( I believe) 3 Roman legions were involved to put down the revolt.

    1. There isn’t any actual evidence of such an economic effect. Of either war. Indeed, 3 legions indicates a rather insignificant war by Roman standards. There were 28 legions in the Empire. Half the number that served under Augustus in the Roman Civil War. Literally. Augustus and Antony fought that war with 50 legions. Not 3. But 50. And yet, the economy boomed in result. So your thinking doesn’t align with facts or precedent.


Add a Comment (For Patrons & Select Persons Only)

This site uses Akismet to reduce spam. Learn how your comment data is processed.