Cassandra has moved. Ugo Bardi publishes now on a new site called "The Seneca Effect."

Monday, July 15, 2013

The punctuated collapse of the Roman Empire

I defined as the "Seneca Cliff" the tendency of some systems to collapse after having peaked. Here I start from some considerations about whether the collapse could be smooth or an uneven process that we could define as "punctuated." I am taking the Roman Empire as an example and showing that it did decline much faster than it grew. But the decline was surely far from smooth.

The idea of an impending collapse of our civilization is already bad enough in itself, but it has this little extra-twist that collapse may be given more speed by what I called the "Seneca Cliff," from the words of the Roman Philosopher who had noted first that, "Fortune is slow, but ruin is rapid". The concept of the Seneca Cliff seems to have gained some traction over the Web and many people have been discussing it. Recently, I found an interesting comment on this point by Jason Heppenstall on his blog "22 billion energy slaves". He summarizes the debate as:

"In the fast-collapse camp are the likes of Dmitry Orlov (who bases his assessment on his experience of seeing the USSR implode) and Ugo Bardi, who expects a ‘Seneca’s Cliff’ dropoff. James Kunstler, Michael Ruppert and any number of others can probably also be added to the fast-collapse camp.

By comparison, the likes of John Michael Greer reckon we are in for a drawn-out era of terminal decline punctuated by serious crises which, at the time, will seem rather severe to all involved but which will give way to plateaux of relative stability, albeit at a lower level of energy throughput."

Actually, the two camps may not be in such a radical disagreement with each other as they are described. The idea of the fast (or Seneca-like) collapse does not necessarily mean that collapse will be continuous or smooth. The model that describes the Seneca effect does give that kind of output, but models are - as usual - just approximations. The real world may follow the curve in a series of "bumps" that will give an impression of recovery to the people who will experience the painful descent period.

So, collapse may very well be "punctuated: a series of periods of temporary stability, separated by severe crashes. But it may still be much faster than the previous growth had been. I discussed this point already in my first post on the Seneca Effect, but let me return on this subject and let me consider one of the best known cases of societal collapse: that of the Roman Empire.

First of all: some qualitative considerations. Rome's foundation goes back to 753 BC; the end of the Western Empire is usually taken as 476 AD, with the dethroning of the last Western Emperor, Romulus Augustus. Now, in between these two dates, a time span of more than 1200 years, the Empire peaked. When was that?

The answer depends on which parameter we are considering but it seems clear that, whatever choice we make, the peak was not midway - it was much later. The Empire was still strong and powerful during the 2nd century AD and we might take the age of Emperor Trajan as the peak (he died in 117 AD) as "peak empire." We may also note that up to the time of Emperor Marcus Aurelius (who died in 180 AD), the empire didn't show evident signs of weakness, so we could take the peak as occurring in mid or late 2nd century AD. In the end, the exact date doesn't matter: the Empire took around 900 years to go from the foundation of Rome to the 2nd century peak. Then, it took just 400 years - probably less than that - for the Empire to wither and disappear. An asymmetric, Seneca-like collapse, indeed.

We also have some quantitative data on the Empire's cycle. For instance, look at this image from Wikipedia.

It shows the size of the Roman military over the Empire's span of existence. WIth all the uncertainties involved, also this image shows a typical "Seneca" shape for both the Western and the Eastern parts of the Empire. Decline is faster than growth, indeed.

There are other indicators that we can consider about the collapse of the Roman Empire. In many cases, we don't have sufficient data to say much, but in some, we can say that collapse was, indeed, abrupt. For instance, you can give a look to a well known image taken from Joseph Tainter's book "The Collapse of Complex Societies"

The figure shows the content of silver in the Roman "denarius" which by the 3rd century AD, had become pure copper. Note how the decline starts slow, but then goes on faster and faster. Seneca himself would have understood this phenomenon very well.

So, the Roman Empire seems to have been hit by a "Seneca collapse" and that tells us that the occurrence of this kind of rapid decline may be commonplace for the entities we call "civilizations" or "empires".

It is also true, however, that the Roman collapse was far from being smooth. It went through periods of apparent stability, interrupted by periods of extremely fast descent. The chroniclers of the time described these periods of crisis, but none of them seem to have connected the dots: they never saw that each crisis was linked to the preceding one and leading to the next one. Punctuated collapse seemed to be invisible to the ancient Romans, just as it is for us, today.


  1. Hi Ugo, I think we're talking about two different things here. One is symmetry and the other is speed, and I completely agree with you that things come apart faster than they prosper. In fact, as far as I remember, Hubbert's Peak symmetry was based on the assumption that nuclear power would be ready on time to fully substitute oil's decline.

    This being said, I think the Druid is totally right when he points out that full collapse to a new equilibrium will take place over generations. That's how previous collapses have occurred so far and I don't see any major reason to think that this time will be different.

  2. Thank you for another interesting post Ugo. But for us to draw any conclusions on the prospect of our own future we must first establish two things: first, when did our empire/culture/civilization start? 1400? 1800? 1850? second, when did we peak? 1970-ish (when energy per capita and world GDP growth was the highest)? Or 2005-2008?

    If we use the Roman example it seems that the decline is 3 times faster than the expansion (3/4 pre peak, 1/4 post peak). IF we put the start of our civilization at 1800 and the peak at 1970, then that would put our "end-of-decline" at 2027. That seems a bit too soon. But still not soon enough from a climate perspective.

    1. Good point, Daniel. It is difficult to make a quantitative estimation, but it is sure that things in our time are moving much faster than in Roman times. So, the collapse will be much faster!

    2. What other civilization faced such a multitude of problems?: overpopulation, climate change, global corporatist rule, chemical pollution, resource depletion, 6th mass extinction, ocean overfishing and acidification, global financial instability, mounting social disparities and injustices, etc...

  3. Daniel, very interesting point! Here is my guess...

    Our civilization is defined by the extensive use of fossil fuels; therefore, I would suggest as starting date 1750, beginning of industrial revolution. As the peak point, it should be the peak of all fossil fuels, which I'm not sure if it's already passed (peak oil was around 2005, natural gas will peak some-when before 2020, and peak coal was probably already passed in terms of net energy). Let's say that fossil fuels net energy peaked in 2000. If we assume decline 3 times faster than expansion,as you did, end of decline would be at around 2085.

    1. 1750 was preceeded by 250 years of the renaissance the scientific revolution and the age of discovery for the western european countries. All of these formed the foundation of the later industrial revolution. Just like Rome was just another city-state for a few hundred years. I'd put it as 1500.

  4. Good to show the Seneca cliff to raise general awareness of:

    You know there is thing called collapse, and perhaps you should pay attention to it.

    Beyond that comparing the steepness of the slopes on each of the curve's sides makes little sense to me.

    Collapse represents system failure, a breakage, not the slowly constrained exponential growth of society as it builds itself up to a peak going into reverse. I imagine the curve as a sum of interdependent mathematical functions. The functions all crash hard at the end because dependencies between the different functions fail leading to rapid decline feedback.

    Your explanation of the economics of copper mining I read a few weeks ago here provides an example.

    As copper ore depletes and becomes less concentrated a point is reached where increasing fuel costs suddenly makes copper mining uneconomical and all copper mining stops. Mining stops; the copper supply chain fails and its failure causes other supply chains dependent on copper to fail. Multiple chains fail from the failure of the first. Failure of total supply chain interdependencies grows at an exponential rate and rapid collapse is the result.


  5. If you didn't stack the Eastern Roman Empire on top of the Western Empire, and extended its dating to 1453 when Constantinople fell, you could get a very different, very drawn out, story.

    Rome faced a number of bottlenecks when facing off with very tough opponents. Only survivor bias makes their progress look smooth. If they had not recovered from losses to the Samites, the Poe Valley (Northern Italy) Gauls who sacked Rome, or had lost to Carthage, you would have seen the more common cause of a quick termination of Empires: military defeat.

  6. There also may be different kinds of empires now operating to consider here?

    Are we talking only about "the empire" or "the age" or epoch of fossil fuels? That started around 1750 and could well end by the end of this century in either a single or a series of collapses. (as mentioned above in an earlier comment)

    But there are perhaps a couple of other ages or empires also to consider. One is the age of the overall assault by Western nations on the rest of the world and its subsequent domination which started around 500 years ago. We are now rapidly moving back to a world where the East and the South (and not only the North and the so called West) play key roles.

    So a second "age" that we have witnessed recently is that of the British Empire (and other lesser European colonial powers) which then turned over the reins of power to the American Empire or the so called Pax Americana after the end of World War II. This so called Pax Americana seems to be in the process of unraveling right now and is giving rise to a so called "multipolar" world. (the transition may end up being relatively peaceful or it may not be; so far it has been punctuated by various relatively contained resource wars)

    The United Nations system as currently configured and with its present institutional components and their respective make ups, seems to be itself a kind of transitional order and a manifestation and tool of both the Pax Americana as well as of the emerging multipolar world order. And it is clearly showing signs of serious stress and strain and of the need for deep reforms. (this could be seen as a kind of harbinger of changes to come or at least of the need for changes in world governance)

    The two transitions or the two possible collapses described above are happening roughly at the same time and are related and inter-meshed. Since one of the Pax Americana's prime objectives has been to guarantee the securing, shipping and provision of the main fossil fuel -petroleum - around the world. (and in particular to the "imperial" and favored nations)

    And so is the "Roman Empire" that we are talking about above which is now collapsing the "Empire or the Age of fossil fuels"? Or is it the collapse of the American Empire (or Pax Americana for those who may prefer that term to more imperial terminology) . Or is it an interaction of the two?

    Many earlier empires (including the Roman empire) relied in good measure on slave labor. But they didn't necessarily collapse because the world ran out of slave labour. They typically collapsed because of imperial over-reach and a combination of economic and financial and other mismanagement factors. Which factors are going to cause collapse now? The energy sources, the financial system, the economic system, the political system, or various kinds of mis-management and of imperial over-reaching? (and "over-reaching" does not necessarily have to be only military over-reach it can also mean everyone in the whole world now driving an automobile which consumes the remaining fossil fuel that is available. (a kind of "collective over-reach vis a vis remaining resources)

    So I think a more fine grained and parsed analysis is probably needed about which empires we are discussing and how the various "empires" and their various features and dimensions may be interacting.

  7. You should divide your time period into two distinct part namely the roman republic period and the roman empire period. Officially, the empire start in 30BC( we can take 44BC also as it is then when Julius Caesar's appointment as perpetual dictator), reach a maximum in 180AD and end in 476. Hence, we can see that the time of rise and fall is actually similar in time.



Ugo Bardi is a member of the Club of Rome, faculty member of the University of Florence, and the author of "Extracted" (Chelsea Green 2014), "The Seneca Effect" (Springer 2017), and Before the Collapse (Springer 2019)