Hadrian's wall in Northern England, as depicted in an illustration for the original edition of Rudyard Kipling's "On the Great Wall" (1906). In this image, the height of the wall is much exaggerated, but it is true that the Roman Empire had built a complex system of fortifications along its borders. However, as the empire's economy declined, it couldn't sustain any more the kind of complexity that was possible at its peak. Complexity has a cost and its returns are not always proportional to it. So it was for the Roman fortifications, that were abandoned in early 5th century A.D., so it may be for some of our infrastructures, such as nuclear energy or manned space exploration .
In the winter of 406 A.D. (or perhaps 405), the fortifications protecting the Western Roman Empire collapsed. Says Edward Gibbon in his "Decline and fall of the Roman Empire": "This memorable passage of the Suevi, the Vandals, the Alani, and the Burgundians, who never afterwards retreated, may be considered as the fall of the Roman empire in the countries beyond the Alps; and the barriers, which had so long separated the savage and the civilised nations of the earth, were from that fatal moment levelled with the ground" (Chapter XXXI book V)
Starting with Augustus, in early 1st century AD, the Romans had built and garrisoned a vast system of fortifications - that today we improperly call "limes" (*) - meant to protect the empire's borders. From the remnants we can still see today, the Roman defensive system was similar to the Great Wall of China, that existed during the same period. It must have been hugely complex and expensive as it included forts, walls, roads and all the infrastructure needed to sustain a military force.
The fortifications may have improved the effectiveness of the Roman Army in defending the empire. But the walls were also as a burden that locked huge amounts of money and resources. It must have become more and more difficult to sustain the cost of the border defenses as the empire declined. In this sense, Kipling's tale "On the Great Wall" rings true when it tells how difficult it was for the Romans of late 4th century to find resources to garrison Hadrian's wall, in Northern England.
We have here a good illustration of Tainter's principle of the diminishing returns of complexity. With time, not only the capability of the Roman Empire to afford complex structures diminished, it is also likely that the complexity of the structures increased, and with it the need for more and more resources. Eventually, it all came to an end with an especially cold winter in early 5th century, when the iced surfaces of the rivers in Germany let invaders cross the border, unopposed. Gibbon may be too dramatic in his statement of a sudden and complete abandonment of the fortifications but it is true that, with the 5th century, there are no more reports of the imperial borders being guarded by a stable military force.
But the end of the walls was not the end of the empire. Curiously, the demise of the fortifications in early 5th century may have freed resources and generated a minor renaissance for Rome. There followed a period of a few decades during which the Empire managed to regain control over most of its territory. The army was rebuilt and transformed it into a mobile force that turned out to be able to defeat invaders - including Attila the Hun - in every major battle fought in Europe in that period. The empire as a political structure was doomed anyway, but the disappearance of the border fortifications had unlocked the transition that was leading to the new world that we call today "Middle Ages."
The demise of complex structures is a characteristic of declining societies. Even in our world, we are witnessing similar events. With the last space shuttle flights planned for this year, we are seeing the end of the US manned space program, Apparently, we are abandoning a frontier that we can't defend any longer. Something similar may be happening with the nuclear energy program. Once hailed as a promise of "energy too cheap to meter", we have seen it stalling in the 1980s. Today, events such as the Fukushima disaster may be telling us that we don't have any more, perhaps we never had, the kind of resources needed to manage the enormous complexity of a nuclear energy system. Also here we may need to abandon an old frontier.
Abandoning a frontier, however, does not mean defeat. It means adaptation - it means thriving in different ways. In our case, it means that we need to use systems which are simpler and more effective than those which in earlier times had seemed to be the solutions to our problems. With space exploration, we have discovered that robot probes are both cheaper and more effective than human explorers. With energy, we may discover that renewables are a simpler and more resilient solution than nuclear. A world based on renewables cannot be the same thing as a world based on fossil fuels or nuclear energy, but it may well be the direction we are heading to, no matter whether we like the idea or not. Wherever we are going, in any case, the transition is unavoidable and we need to understand our limits; otherwise we will crash against them.