ANCIENT ROME was built out of concrete, brick, and quarried rock. Today, you can find elegantly carved rock everywhere in the Forum, most obviously in the grand ruins of the site’s many temples and triumphal arches. But chunks of lovely fluted column and fragments of carved plinths, capitals, entablatures, and lintels are also strewn everywhere, often half buried in the dirt and heaped in rough piles.
Few visitors pause to think about the enormous labor it took to cut these rocks out of the ground, carve them, and carry them to Rome from distant quarries, or about the ingenuity needed to accomplish such a feat. Yet these scattered, broken stones often have fascinating stories to tell.
So it was with one piece of rock that caught my attention. It wasn’t in the Forum, but a five-minute walk away, to the southeast, in the Colosseum—the greatest of all Roman amphitheaters, built to gratify the Roman lust for gladiatorial combat and wild beast shows. I spied this piece of rock as I passed underneath an arch to climb from the Colosseum’s street-level entrance to the main promenade inside. I looked up briefly and noticed the arch’s keystone. For some reason—I don’t really know why; maybe it was something about the keystone’s sheer unremarkableness—I stopped to look more closely. I saw that the arch itself was huge—over seven meters high and four meters wide—but in the context of the amphitheater’s great structure it didn’t stand out. It was only one of eighty of the same size and construction that formed the bottom tier of the building’s immense first interior wall. But looking more closely I realized that while seemingly insignificant, the arch and its keystone were actually a brilliant feat of engineering.
The keystone in the Colosseum
The Roman arch is a type of structure that engineers designate a “voussoir arch” because its wedge-shaped stones are called voussoirs. It was an engineering breakthrough. Prior to its invention, someone building a bridge across a river, say, would usually span the horizontal distance using a long length of rigid material, like rock or timber. But this type of structure is vulnerable to collapse in the middle of its span. Worse, it’s bedeviled by a basic contradiction: if the span’s substance is thickened to make the bridge stronger, the span itself becomes heavier. Eventually, it becomes so heavy that the bridge is weakened, not strengthened. The voussoir arch circumvents this contradiction. It gains its strength not just from its constituent materials but also from its design: because it’s an arch, the structure’s load is distributed downward into the pillars on each side, and because its voussoirs are wedge shaped, up to a point their weight and that of any material they carry actually strengthens the structure by driving the voussoirs together more tightly.
The design of the voussoir arch gives it strength.
The Romans didn’t discover this idea; they probably adopted it from the Greeks and Etruscans.1 They did master the technology and worked out the surprisingly intricate engineering principles that govern the voussoir arch’s proper construction. Then they spread the technology far and wide, using it to build bridges, aqueducts, temples, domes, and amphitheaters throughout their empire. Today, a large proportion of the residue of ancient Rome—the most durable bits, in fact, like the occasional segments of aqueduct that can be spied in the French, Italian, or Spanish countryside—incorporate voussoir arches.
The keystone lodged above my head was massive: about two meters deep, one meter high, one and a half meters wide at the top of its wedge, and three-quarters of a meter wide at the bottom. It was made of travertine, a kind of limestone that has been quarried for millennia near the present-day town of Tivoli (known to the Romans as Tibur), about thirty kilometers to the east. Tivoli travertine is a cream color, streaked with darker indentations and small holes that make it look a bit like white cork. But here, over time, age, pollution, and years of inattention had allowed a thick layer of grime to coat the rock’s white surface, turning it a muddy brown.
Scruffy though this keystone was, I was struck by how precisely it had been cut: one couldn’t insert a razor blade between it and the adjacent stones. Romans didn’t have high-speed diamond-edged rock-cutting saws—they cut their stone by hand, with iron picks and chisels, one hammer blow at a time. The labor involved, and the sheer determination and patience, boggled my mind. I was struck most, though, by something more essential: stone is heavy stuff, and I estimated this particular keystone weighed about 5.7 metric tons.2 I started to think about how Roman engineers and laborers had put it there, imagining the giant crane they would have needed—constructed in an A-frame, festooned with wooden pulleys—with dozens of men straining on ropes at its base.
The powerful emperors Augustus, Caligula, and Nero had all considered erecting a huge amphitheater in the middle of Rome. But Nero’s death in 68 CE—and the brutal civil war that followed—spurred his successor, the emperor Vespasian, to build the Colosseum as a gift to the city’s people to show that his rule would be different from Nero’s vainglorious and cruel reign. Vespasian began construction between 72 and 75 CE, likely financing the scheme using spoils from his crushing victory over the Jews and the sacking of Jerusalem a few years before. His son Titus inaugurated the building in 80 CE with one hundred days of games in which some ten thousand beasts were killed.
In as little as five years, the Colosseum’s architects, engineers, and laborers had erected a building with the dimensions and seating capacity of New York’s Yankee Stadium. It was, the architectural historian Rabun Taylor writes, “the most complex structure ever successfully completed in antiquity.” Not only could it support the combined weight and simultaneous swaying and stomping of over fifty thousand spectators, but its architecture also provided for the smooth circulation of these spectators (the stadium could be emptied in as little as fifteen minutes), while an elaborate arrangement of pipes brought perfumed water to fountains throughout the building and carried away waste from lavatories. “It is surpassed in daring, originality, and beauty by other buildings,” Taylor continues, “but as a monument to architectural process it stands alone.”3
Public buildings like the Colosseum were designed to astonish, awe, and even humble Rome’s citizens with the sheer audacity of their engineering. They were the physical expressions of an ideology of power. “They spoke of strength, control, and stability,” says Taylor. “The intent was to induce participatory pride and willing submission and allegiance to the emperor.” In fact, the Colosseum’s most spectacular views, available only from the uppermost seating farthest from the arena, were reserved for the common person, whose allegiance was most uncertain and most in need of reinforcement.4
The Romans began this monument to their power by draining and excavating a lake that had been on Nero’s palace grounds. They then laid an elliptical ring of cement and crushed stone—thirteen meters deep, fifty meters wide, and over five hundred meters in circumfer-ence—to serve as a foundation, reinforcing each side with brick walls three meters thick. On top of this foundation they placed a floor of travertine almost a meter thick, and then anchored the base blocks for the building’s main pillars to the floor with molten metal.5 On these blocks, the Romans erected the columns and arches of the amphitheater’s eighty radial walls that fanned outward from the central field like the spokes of the wheel—each intersecting the three immense outer walls that girdled the building.
The final building’s total mass was simply staggering. The builders extracted, moved, shaped, mixed (in the case of concrete), and assembled about a million metric tons of raw material, including 295,000 tons of travertine, 653,000 tons of concrete, 54,000 tons of tufa (a soft volcanic rock abundant along the west coast of Italy), 58,000 tons of clay brick, 6,000 tons of marble, and 300 tons of metal to connect the major stones.6
The Colosseum and its foundation originally contained about a million metric tons of rock, concrete, and brick.
That scruffy keystone in the Colosseum could tell a story, I thought, about Rome’s huge energy requirements and about how those requirements shaped the empire’s evolution. Empires run on energy. A central task of any empire is to produce, transport, and focus enough energy to maintain and extend its economic and political power. Acquiring and protecting the sources of this energy, the routes along which it’s carried, and the people and organizations responsible for generating and transporting it becomes a key job of an empire’s security and military forces.
Like any ancient society, Rome was powered in essence by the sun’s energy—absorbed by plants in Roman fields and converted into food.7 As I inspected the keystone, I asked myself what area of farmland was needed to grow the calories that powered the muscles that put just this stone in place? And how much was needed to power the construction of the whole Colosseum? This was not an idle question, I believed, but tied to Rome’s rise and fall.
The stone could also tell us something, I suspected, about the similarities and differences between Rome’s empire and our own of more recent times—including today’s American imperium.
Energy is the lifeblood of all societies. Just as we can understand fundamental things about the human body by tracing its flow of blood, we can understand much about a society’s activities by tracing its flows of energy.
Through the ages, the ways in which human beings have produced and used energy have sporadically advanced, and each advance has defined a new stage in technology and the civilization it sustains.8 Ancient Romans, for instance, not only lacked rock-cutting machines, they also had no dump trucks or hoists powered by electricity, gasoline, or diesel fuel. Almost all the work they applied to the job of cutting, moving, and lifting stone and other materials was done by human and animal muscles; the energy fueling those muscles came of course from food, and the food was mainly grain and hay, such as wheat, barley, millet, rye, spelt, and alfalfa grown by Roman farmers (most Romans ate only very small amounts of meat).9
We readily agree that energy is vital, but it’s harder to grasp its role in our affairs.10 When we think of energy, we usually think of the gasoline that fuels our cars, the electricity that lights our homes, and maybe the natural gas and coal that we burn in our power plants. In other words, we usually think of energy as fuel, and we tend to think that it’s useful because of the immediate services this fuel provides—services like transport, light, and heat.
Of course, these are critical services, but energy’s role in our lives is actually much more fundamental, essential, and subtle. We extract energy from our environment to create order out of disorder and complexity out of simplicity.11 We often use this order and complexity, in turn, to help us solve the problems we face—for instance, to shelter ourselves from our harsh environment and to protect ourselves from attack. Put simply, societies with access to lots of energy are generally more adaptive, resilient, and better at solving problems.
To understand the link between energy and complexity, we need to explore some of the scientific principles that govern energy’s behavior. Although people usually think of energy as a tangible thing, like gasoline, it’s actually not a material substance at all. It’s really a property of things. A handful of wheat kernels, a lead-acid battery, a stream of photons of light, or a rushing river all have the property of energy. This property can move from one place to another, and when it moves we can sometimes use the resulting flow of energy to do what physicists call “work”—that is, to change things in our physical world. For example, the energy in a rushing river can do work for us when it moves to another system like a water mill or turbine. The energy in a lead-acid battery does work only when it flows to a light bulb or an electric motor.
Some sources of energy are more useful for doing work than others, provided that we know what to do with the energy available in them. The high-quality energy available from gasoline, for instance, is better for doing work than the low-quality energy of the diffuse heat present in the ground under our feet, which isn’t much good for anything, except perhaps heating other things, like our buildings and homes.12 So sources of high-quality energy, such as oil, natural gas, and rushing rivers, are extraordinarily valuable, because we can use them to provide many different types of service.13
There are, fundamentally, only two forms of energy: kinetic energy, which is the energy of material in motion, and potential energy, which is energy that’s trapped in something. We can convert potential energy to kinetic energy and vice versa. When we burn gasoline to power a car engine, we’re converting the potential energy in the gasoline’s chemical bonds to the kinetic energy of the engine’s motion. On the other hand, when we use a river’s rushing water to turn an electric turbine and then use the turbine’s electricity to charge a battery, we’re transferring the river’s kinetic energy to the turbine to do work, and then we’re using this work to create potential energy in the battery.
Heat is a special type of kinetic energy. It’s stored in a material like the walls of our house or the air around us by the vibrations and other motion of the atoms and molecules making up that material. We often use heat as an intermediary form of energy to run our machines. For example, when gasoline is ignited in the cylinders of a car engine, high-temperature gases are created, and the heat of these gases provides the kinetic energy that drives the engine’s pistons. Heat also drives our jet and rocket engines. Understanding the nature of heat can tell us important things about the flow of energy in natural and human-made systems, such as how much work a given system can do.
This is the province of the sub-field of physics called thermodynamics. The nineteenth-century discovery of the laws of thermodynamics—among the most important in all science—was a stunning breakthrough. These laws tell us two essential things about our natural world. The first says that energy can’t be created or destroyed: the total energy of any system and its surroundings (generously defined) stays constant, whether the system is a mechanical device like a car engine, a biological system like a human body, or a social system like ancient Rome’s.14 The second law says that during the normal operation of most systems, energy degrades in quality: high-quality energy degrades to progressively lower-quality energy, with the end result being simply low-grade heat.15 It’s as if energy always flows down a slope from forms that we can use to do lots of work—like hoisting rocks in the Colosseum—to forms that aren’t very useful to us at all. And every time we use energy to do work, we further degrade its quality. As a system’s energy degrades, physicists say its “entropy”— often described as its disorder or randomness—increases.
To see all of this better, imagine you’re sealed inside a big box with a newspaper, a chair, and a battery wired to a light bulb. Let’s assume that the box is isolated from the rest of the world, so energy can’t flow into or out of it. Inside the box, high-quality chemical energy is concentrated in the battery; it’s also concentrated in you in the form of the sugars and proteins that you’ve obtained from food and that power your body. Physicists say that such energy is “coherent” and “ordered.” Sitting inside the box, you can read the newspaper because the battery converts its chemical energy into electrical energy that the bulb then transforms into light energy. That light—speeding away from the bulb—hits the newspaper, walls, and other things in the box where it’s absorbed and degraded into low-quality heat. The light’s energy is essentially spread around the box as heat, slightly increasing the incoherent and disordered vibrations of the molecules making up the things in the box. But the chemical energy in the battery is finite, so eventually the battery is exhausted, and the light goes out.
The light won’t last forever in a closed system.
You (the person in the box) are much like the battery: you convert your chemical energy into other forms of energy, such as kinetic energy as you move and lift things. If there’s a bicycle generator in the box, you can hop on board and use it to drive the light and finish reading your newspaper, warming up the box even more as the heat from your exertion is carried away from your body by your sweat. But just like the battery, you too will run down and eventually die. All the high-quality energy in the battery and your body will be then degraded, and the box itself will wind down—just the way a mechanical clock winds down—to a black three-dimensional space of uniform temperature.
We’ll see later that societies that don’t have access to enough high-quality energy are likely to disintegrate. The laws of thermodynamics tell us that despite the fact that energy can’t be created or destroyed, it does inevitably degrade, which makes it progressively less useful for work. And if it’s less useful for work, it’s less useful for maintaining a society’s complexity and resilience.
Our light-in-a-box illustration is entirely imaginary and artificial. It’s a “closed” system, which means it’s sealed off from the outside world. Scientists use such imaginary systems to tease out implications of their theories. But in the real world almost all physical, biological, technological, and social systems are “open.” They interact with their surroundings. Most important, these systems often extract high-quality energy from their environment to do work or to reduce disorder at their core, and they expel waste heat and material back into that environment. A city like ancient Rome, for example, imported from its hinterland timber, fresh water, and energy in the form of food and released back into its surroundings heat, sewage, and garbage (at its peak population, Rome probably produced around one million cubic meters of human waste each year).16 A modern car engine takes in fuel to do the work of moving the car, and it expels heat and exhaust gases into the atmosphere. A steel mill takes in raw materials like iron ore and high-quality energy (in the form of coking coal, for instance) to create coherent and ordered materials like steel rods; in the process, the mill discards heat, carbon dioxide, and pollution into nearby air and water.
At first glance, open systems appear to violate the classical thermody-namic principle that disorder, randomness, and entropy always increase. After all, the mill produces low-entropy steel rods. And over time Rome’s internal arrangements became more ordered and complex, as its various social and technological parts became more diverse, specialized, and interdependent.
The principle that a system’s entropy must always increase, scientists eventually realized, applies only when its boundaries are defined to encompass virtually all its interactions with its surrounding environment. The system of a steel mill then includes the entire technological infrastructure that produces the iron ore and coke it uses as well as the atmosphere and waterways into which it expels its pollutants. And the system of ancient Rome included the solar energy provided by the sun and the city’s entire natural hinterland of land, forest, water, and air. Within this generous boundary, the average quality of the system’s energy always declines, and entropy always increases.
All the same, there can be parts of the broader system that have a very high degree of order: Rome was a zone of low entropy within its larger system. In fact, things like cities, ecosystems, and even our human bodies can create order and complexity spontaneously, decreasing their entropy even further in the process.17 Cities build elaborate transportation, water, and energy infrastructures; ecosystems become more biologically diverse as new species evolve; and human embryos develop into people, with all their complex organs and structure. How do such amazing things happen?
Scientists still aren’t sure. But they now know that systems like cities, ecosystems, or human bodies are, as they say, “far from thermodynamic equilibrium.”18 They can spontaneously create order inside themselves. But maintaining this order is a bit like holding a marble on the side of a bowl with your finger: the marble wants to sit at the bottom of the bowl—that’s its equilibrium point; so holding the marble on the side takes a constant input of energy. Similarly, cities, ecosystems, and human bodies must have a constant input of high-quality energy to maintain their complexity and order—their position far from thermodynamic equilibrium—in the face of nature’s relentless tendency toward degradation and disorder. And, as the system gets larger and more complex, more and more energy is needed to keep it operating.
All these ideas can help us grasp why the Roman empire fell and, ultimately, discern the fate of our own societies. The Romans employed astonishing technological prowess to construct buildings like the Colosseum. Less obviously but just as critically, they needed considerable social prowess to assemble themselves into work units, coordinate the efforts of these units, encourage specialization skills, and provide themselves with public services like governance, tax collection, and security. Codified laws regulated everything from money and debt to property rights, corporate organization, guilds, and the employment of laborers and slaves.19
Complex social organization doesn’t appear out of thin air. Courts must be staffed, functionaries paid, and armies fed and supplied with weapons. More fundamentally, to create and sustain organizations, rules, and laws, people must move around, communicate, discuss, argue, and negotiate with one another; educate and train one another; and record—in some stable medium like rock, parchment, or CD ROM— basic rules, contracts, and bargains. All these activities once again require high-quality energy.
So the Romans used farms to capture the sunlight falling on wide swathes of land around the Mediterranean basin. Some of the farms existed before the Romans arrived. Especially in the eastern Mediterranean—for instance, in modern-day Lebanon and Syria—the expanding empire often simply took over existing cities and their food-production and tax systems, while in northwestern Europe, in places like the Rhône valley, new land was sometimes converted into farms. But wherever the farms were located, they played a role in the Roman energy economy similar to that of solar battery chargers: they converted sunlight into a form of high-quality potential energy, especially fodder and grain, that was storable and transportable.
The Romans then focused this energy—they used their food batteries, so to speak—to create a productive, resilient, and phenomenally complex system of public buildings, manufacturing facilities, housing, roads, aqueducts, and social organization. And here’s the punch line: recent research (which we’ll come to later) shows that the Roman empire was eventually unable to generate enough high-quality energy to support its technical and social complexity. This shortfall—more than proximate events like incompetent emperors and invasion by Visigoths—was the fundamental cause of Rome’s fall. In other words, the empire’s loss of internal order, coherence, and complexity was, in significant part, a thermodynamic crisis. The empire tipped into irreversible decline precisely because it couldn’t feed its energy hunger.20
This was Rome’s fate. Will it be ours as well? A closer look at energy’s role in ancient Roman society will help us find out.
When I returned to Canada from Rome a few weeks after my visit there, I set out to see if I could follow my particular keystone’s journey from its beginnings to its final placement in the great arch. I felt that focusing on that single stone could help me understand energy’s role not only in the overall effort to build the Colosseum—and, more generally, in the life of Rome—but also in the empire’s ultimate decline.
I began by turning to my research assistant at the University of Toronto, Karen Frecker. A painter, pianist, and electric utility analyst with a truly remarkable mind, Karen is skilled at applying technical analysis to social questions. She set to work enthusiastically to find out how much farmland the Romans needed to build the Colosseum. She divided the project into two parts—demand and supply.
First, she studied the details of the Colosseum’s construction; calculated the total mass of the different types of materials used in the building; estimated how much work—or how much energy—was required to cut, move, and place all those materials; and estimated how many people and oxen were required to do that work. We called this the “demand side” of the problem. Second, she calculated how much agricultural land was needed to grow the food to feed these people and oxen. This was the “supply side” of the problem.
It sounded simple at first, but all these calculations turned out to be vastly more involved than we’d expected—and much more interesting.21 We made a series of assumptions about Rome’s technology, workforce, and agricultural practices and about the physics of Roman transportation and construction. To ensure these assumptions were sound we consulted ancient texts, literature on Rome, treatises on mechanics, and experts on the Colosseum’s architecture and engineering.
Early on, we learned about a crucial difference between Roman and modern buildings. Today’s buildings are constructed mostly of materials fabricated by men and women using huge quantities of energy—steel, aluminum, glass, plastics, brick, and composites. But Rome’s buildings were made mainly of rock that nature had provided. In those days, the materials that took the most energy to gather and shape for human use were metals—the lead used widely in water pipes, the silver in coins, and the iron and copper in armor, swords, and other weapons, as well as the iron used to hold stones together in many buildings. Tiles and bricks also required a lot of energy: after being fashioned from clay, straw, and small amounts of high-quality pozzolana sand, they were dried in the sun and baked in wood-fired kilns at temperatures reaching hundreds of degrees Celsius.22
As for the energy spent to cut, move, and place the Colosseum’s rock, my keystone told much of the story. Some 1,927 years before I spied it in its arch, workers at a Tibur quarry chiseled cracks into a natural bed of travertine. Then they drove wedges into the cracks, splitting off a rough six-ton block.23 The block was rolled on logs and hoisted with pulleys onto an oxcart pulled by two oxen. It then began its two-day, thirty-kilometer journey to Rome along a specially built road. According to ancient records, every day of the Colosseum’s construction, two hundred carts of stone destined for the building entered the city.24 Unlike my travertine keystone, though, the tufa used in the building was found at sites within a few kilometers of Rome, while the bricks and concrete were also likely produced from materials close at hand. Indeed, much of the aggregate used as the binding agent in the Colosseum’s concrete probably came from the rubble of Nero’s nearby demolished palace.25
Once the rough six-ton block arrived at the construction site, masons chiseled the rock into the precise wedge designed to fit between the voussoirs—called counter-keys—that would be its neighbors. Workers used a giant wooden crane, fitted with a block and tackle, to lift the counter-keys into the wood scaffold they had built to support the arch; then the keystone itself was hoisted and maneuvered into the space between them and joined with iron clamps. Now the arch had structural integrity, and the scaffolding could be removed.26
The A-frame cranes used for lifting the building’s rock were massive, sometimes twenty meters high. Some of these cranes were located around the inside of the Colosseum leaning outward, with stay cables made of hemp stretching behind them hundreds of meters to their anchors at the center of the arena. Others were positioned on the outside leaning inward, with their cables splaying from the building. All these cranes could tilt vertically, but none could move horizontally, so the builders had to release the stays and disassemble and assemble the cranes repeatedly.27 Most employed ropes with pulleys and hand-cranked winches, but treadmills powered the largest cranes of all: several workers, probably slaves, got inside a device that looked like a giant hamster wheel—up to eight meters in diameter and called, in Latin, majus tympanum. By clambering up the curved wall of the treadmill’s interior, they automatically turned a shaft that acted as the crane’s winch. Such devices could lift rocks weighing tens of tons. The Romans, as one scholar puts it, “considered it a point of honor … to work with blocks of enormous size simply out of the desire for technical achievement.”28
A-frame cranes lifted the rock for Level 1 of the Colosseum.
As the first floor was completed, the Romans rolled their cranes up the Colosseum’s sloped seating area so they could hoist the stone elements for the higher floors. Many of the walls of the upper levels were made of concrete faced with brick; this material was lighter than the first floor’s travertine and tufa, and it didn’t have to bear as much weight. Wet concrete was probably mixed at ground level and lifted by ropes tied to buckets; or workers carried it, along with loads of brick, on their backs up a winding route of stairways and scaffolding to the ongoing work on the upper levels. When construction finally reached the fourth and highest level, nearly fifty meters above the ground, the builders installed a row of columns—eighty in all—along the inside edge of the gallery. Each column shaft alone weighed nine tons, and together they held aloft a sturdy entablature and roof of travertine designed to shelter the spectators seated in the amphitheater’s upper reaches.29
Cranes were rolled up the sloped seating area to construct Level 4.
Karen discovered ingenious ways to estimate the total mass of various materials in the Colosseum and the distances they traveled to the building.30 Then she plugged these figures into standard equations used to calculate the amount of energy needed to carry, push, pull, and hoist a particular mass a certain distance. This gave her an estimate—grounded in hard physics—of the minimum amount of human energy needed to move the building’s materials. But this was only a start, because it left out most of the work done to construct the Colosseum. Among other things, oxen were used for most of the transportation of materials from quarries to the site.31 Also, the Romans excavated and filled the foundation, drove oxcarts to and from the quarry, shaped the rocks at the construction site, built and removed the scaffolding, and mixed and laid concrete. All these activities were administered, and everyone involved was housed, clothed, and fed. In fact, over a dozen workers’ guilds participated in this astonishing endeavor, including construction workers, bronze workers, blacksmiths, carpenters, porters, brick makers, marble workers, pavement layers, and masons.32
So Karen calculated the work required for each one of these sub-tasks.33 As our project expanded (and as we increasingly questioned our sanity), she crunched through endless calculations and added hundreds of numbers to her lengthening spreadsheets.
Despite our attention to detail, we still had to leave many things out. For instance, our Colosseum was just a shell: it didn’t have any of the real building’s elaborate external and internal decorations, statuary, and water fountains—there were probably over 150 fountains in the building—or even internal piping for its lavatories.34 Also, because we weren’t able to find accurate data for the work involved in assembling, disassembling, and moving the cranes, we had to omit these important tasks from our estimate. Finally, the Colosseum project was, in thermodynamic terms, a classic open system. For example, the Romans had to harvest and transport timber for scaffolding and cranes, and they needed enormous amounts of firewood to slake lime, melt metal, and fire brick. Their wood-harvesting operation must have extended far into the city’s hinterland and involved not only woodcutters but also countless teams of oxen and drivers, road builders, and a host of people supplying tools and food to these workers and animals. To make our task tractable, however, we decided that all this work fell outside the boundary of the Colosseum system; we focused instead simply on the caloric needs of the people and animals directly involved in the building’s construction.
Even then, the final tally was staggering: erecting the Colosseum required more than 44 billion kilocalories of energy.35 Over 34 billion of these kilocalories went to feed the 1,806 oxen engaged mainly in transporting materials. More than 10 billion kilocalories powered the skilled and unskilled human laborers, which translates into 2,135 laborers working 220 days a year for five years.36
Some of our results on the demand side of the problem surprised us. We found that the scholarly literature’s standard estimate of the amount of concrete that Romans used—some six thousand metric tons—was far off the mark; the actual amount in the building alone is nearly thirty times larger. In fact, the Colosseum is a bit like an iceberg: most of its volume is underground, and almost 90 percent of this underground structure is concrete.37 Indeed, the visible travertine, tufa, and brick make up barely a third of its volume. We were surprised, too, to find that three-quarters of the total energy expenditure went to oxen. Supporting the workers when they weren’t working also took a lot of energy—after all, they had to be kept alive over holidays and rest days, and ancient Rome had many holidays.
Then there was the most surprising fact of all. When I first spied the keystone, all I could think of was the monumental effort needed to move rock and other heavy materials around the construction site and to hoist multi-ton stones and columns up to fifty meters in the air. But when we’d completed our demand-side analysis, we discovered that all these activities used only a small fraction—about 4 percent—of the total energy needed to build the Colosseum.
This was an eye-opener. It reminded me that human societies expend a great deal of energy in non-obvious ways and places—for example, to excavate and build the Colosseum’s foundation. Yet most of the time we don’t think about unseen but essential activities. And this would become a key discovery in my larger investigation: to understand energy’s role in our societies’ ability to adapt and survive, we need to develop the everyday habit of recognizing energy in all its uses and consequences.
Once we’d calculated the total number of calories needed to build the Colosseum, we could turn our attention to the supply question. How much cropland did the Romans need to grow the food that provided the calories for the animals and men who did all this work?
The Roman diet consisted of a mixture of grain (especially wheat), fruit (including olives and figs), legumes, small amounts of meat, vegetables, and wine.38 Oxen were fed hay, legumes, millet, clover and other grasses, tree foliage, wheat chaff, and bean husks.39 But to make our calculation manageable, Karen assumed Romans produced only two crops: wheat for humans and hay for oxen, with alfalfa the main type of hay.40 Although Romans did allow their oxen to graze on pastureland, it seems that working oxen were fed mainly cropped fodder. As Cato the Censor marvelously observed in his second-century BCE treatise De agricultura, oxen “should not be put out to graze except when they are not worked; for when they eat green stuff they expect it all the time, and it is then necessary to muzzle them while they plough.”41
A lot of Roman grain came from northern Africa, Egypt, Spain, Sardinia, and Sicily.42 The regions of Etruria and Campania—just north and south of Rome along Italy’s western coast—were also very fertile and probably provided some food for the surrounding area, including Rome. So grain came by ship across the Mediterranean and along the coast of the Italian peninsula.43 Arriving at the port of Ostia, located at the mouth of the Tiber, the grain was offloaded, weighed, sorted, and reloaded on barges that were pulled by men and oxen the twenty kilometers up the river to Rome.44 Records indicate that during the reign of Augustus, the city had to store annually half a million cubic meters of grain from Egypt and North Africa alone, enough to fill a cube almost as long as a football field on each side.45
Ancient Rome and its surrounding regions
In Roman times an average hectare of good cropland in Etruria produced 1,158 kilograms of wheat and 2,600 kilograms of dry alfalfa.46 But farmers always set aside some harvested seed for the next year’s planting, and a significant portion of any harvest was also lost to rot and vermin during storage and transport.47 Taking these factors into account, we estimated that a hectare of Roman agricultural land could produce wheat containing about 2.4 million kilocalories or dry alfalfa containing about 4 million kilocalories of usable energy.48 Yet we’d still neglected a key complication: many people were involved in growing and harvesting the wheat and alfalfa, and all of them had to be fed. In other words, Roman farmers had to burn calories to produce calories. So before arriving at a final figure for the total land needed to build the Colosseum, we had to determine how much labor was needed to farm a hectare of land.
It turns out that we can evaluate any project intended to generate energy, including farming, in the same way that we’d evaluate a financial investment: we compare the size of the investment with the size of the return on that investment. In the case of energy projects, energy experts call this ratio the EROI, which stands for energy return on investment.49 We calculate the EROI by dividing the amount of energy a project produces by the amount it consumes. For a modern coal mine, for instance, we divide the useful energy in the coal that the mine produces by the total of all the energy needed to dig the coal from the ground and prepare it for burning, including the energy in the diesel fuel that powers the jackhammers, excavation shovels, and off-road dump trucks and the energy in the electricity that powers the machines that crush and sort the coal.
If you’re interested in the role of energy in human society—including our own societies, as we’ll see—keep an eye on the EROI for different sources of energy, because it’s one of the most useful statistics to compare the relative value of the sources. “An EROI of much greater than 1 to 1 is needed to run a society,” note the system ecologist Charles Hall and his colleagues, “because energy is also required to make the machines that use the energy, [and to] feed, house, train and provide health care for necessary workers and so on.”50 In agrarian societies like ancient Rome’s, where people did most of the farm work, food production also had to have an EROI of well over 1 to 1 for the society to support any kind of social and technological complexity. Indeed, in any complex society, those of us who aren’t farmers are essentially parasites on those of us who grow the sources of energy—the grain, vegetables, fruit, and meat—that keep all our bodies running.
To see the truth of these assertions, imagine an agrarian society in which human beings do all the work, including all farm work—where there are, in other words, no oxen or other draft animals—and which has an EROI of only 1 to 1. In this kind of society, the farmers themselves would consume all the energy generated by farming: everyone would have to work exclusively to feed himself or herself, so there wouldn’t be surplus energy to support people doing other things—even to build houses for the farmers, make their farm tools, or prepare their food. If this society’s EROI fell below 1 to 1, the farmers wouldn’t be able to generate enough energy even for themselves, and they’d slowly starve to death.
Two thousand years ago, Roman farms were often organized as large plantations—or latifundia—worked by slaves.51 Farming technologies were remarkably similar to those used today in many rural zones of Asia and Africa. Oxen pulled iron-tipped plows, and farmers sowed seeds by hand, fertilized their fields with manure, and separated grain from chaff using winnowing baskets. The agricultural historian M. S. Spurr has concluded that Romans needed about fifty-eight days of work each year to farm a hectare of wheat and about forty days to farm a hectare of hay.52 Since one day of labor burns about 3,000 kilocalories, farm workers therefore had to spend about 174,000 kilocalories to grow and harvest a hectare of wheat and about 120,000 kilocalories for a hectare of alfalfa. But this wasn’t the end of the story, of course, because farm laborers had to be fed during the portions of the year when they weren’t working in the field. Taking all these factors into account, we estimated that the Roman EROI for wheat was around 12 and for alfalfa around 27.53 In other words, for each kilogram of wheat that the Romans invested in farming, they got about twelve kilograms of wheat in return.
Karen had finally arrived at the last steps in the process of estimating the total land area needed to produce the energy to build the Colosseum. Now she could determine the crucial figures for the surplus energy generated per hectare: 2.2 million kilocalories for wheat and 3.8 million kilocalories for alfalfa. She then divided these figures into the number of calories needed by Colosseum laborers and oxen.
The results were impressive: to build the Colosseum the Romans had to dedicate, every year for five years, at least 19.8 square kilometers to grow wheat and 35.3 square kilometers to grow alfalfa. That’s a total of 55 square kilometers of land—or almost the area of the island of Manhattan.54 And to capture the solar energy needed to extract, move, carve, and hoist the single keystone that I spied on my visit to the Colosseum, they needed nearly 1,300 square meters of farmland, or about three times the area of a modern city lot.55
A crude but useful measure of a society’s complexity is its level of urbanization. By this measure, for an agrarian society the Roman empire was extraordinarily complex. The size of the city of Rome and the level of urbanization in key regions like the Italian peninsula and Egypt were probably unmatched until the industrial era.56 To sustain this urbanization, complexity, and order, the Romans had to invest immense effort to generate and transport energy.57 And this may have been the empire’s fatal weakness.
At the height of the empire in the first and second centuries CE, the population of Rome itself probably approached 1 million and may even have reached 1.5 million.58 The entire empire’s population is far harder to estimate because solid data are almost nonexistent.59 On balance, evidence suggests that it likely peaked in the neighborhood of 60 million toward the end of the second century. Of this population, between 15 and 20 percent lived in a network of several thousand cities.60 Most of these cities were small, with 10,000 to 15,000 people, while a handful like Alexandria, Antioch, and Carthage had a few hundred thousand. Urbanization was likely over 30 percent in Egypt, while it might have been over 20 percent in the Italian peninsula, given the massive size of Rome itself.61
No European city again approached 1 million people until early nineteenth-century London, and as recently as the late eighteenth century the populations of many Western societies were still around 10 percent urban.62 The historical demographer E. A. Wrigley writes,
The ultimate reason for the comparatively low levels of urbanization in pre-industrial societies is not far to seek. To live, one must eat. Only if levels of output per head in agriculture rise to the point where one man on the land can feed ten, twenty or even fifty off the land can a very high degree of urbanization be reached. [In] many pre-industrial societies levels of output per head were so modest, and so variable from year to year with the vagaries of the harvest, that there was no margin available to support any considerable proportion of the population outside agriculture.63
The city of Rome, where at one point hundreds of thousands of people survived on the emperor’s grain handouts, required at least 8,800 square kilometers of agricultural land to grow enough wheat to feed itself, an area not much smaller than the entire country of Lebanon today. The population of the Roman empire in total required more than 530,000 square kilometers, or an area equivalent to modern-day France.
Around the Mediterranean basin and deep inland along the fertile valleys of great rivers like the Rhône and the Nile, Roman engineers built irrigation systems, networks of waterways for transportation, and grain, oil, and wine storage facilities.64 Boatbuilders constructed an armada of grain ships.65 Meanwhile, officials surveyed the countryside, classified its land by quality and use, and established and enforced property rights—all practices that allowed them to identify, and bring under the authority of Roman law, taxable cropland and individuals.66 Indeed, the late empire’s most important tax, which probably generated at least 90 percent of its revenue, was levied on agriculture.67
What does this initial analysis of Roman energy tell us about humanity’s predicament at the beginning of the twenty-first century? It highlights, I believe, two critical lessons.
First, when it comes to the exigencies of energy, our rich, high-tech Western societies aren’t any different from poor developing societies or, for that matter, from ancient Rome. All our societies require enormous flows of high-quality energy just to sustain, let alone raise, their complexity and order (to keep themselves, in the clumsy terminology of physics, far from thermodynamic equilibrium). Without constant inputs of high-quality energy, complex societies aren’t resilient to external shock. In fact, they almost certainly can’t endure. These ever-present dangers drive societies to relentlessly search for energy sources with the highest possible return on investment (EROI). They also drive societies to aggressively control and organize the territories that supply their energy and to extend their interests, engagements, and often their political and economic domination far beyond their current borders—as we see today with American involvement in Iraq and the Persian Gulf.
The second lesson is less obvious but more important: after a certain point in time, without dramatic new technologies for finding and using energy, a society’s return on its investments to produce energy—its EROI—starts to decline. The Roman empire was locked into a food-based energy system. As the empire expanded and matured; as it exploited, and in some cases exhausted, the Mediterranean region’s best cropland and then moved on to cultivate poorer lands; and as its grain supply lines snaked farther and farther from its major cities, it had to work harder and harder to produce each additional ton of grain.
Today humankind is facing the same trend with many of its vital energy sources, like conventional oil, natural gas, and hydropower. We’ve already found and tapped the biggest and most accessible oil and gas fields, and we’ve already exploited the best hydropower sites, as we’ll see in chapter 4. Now, as we’re drilling deeper and going farther abroad for our oil and gas, and as we’re turning to alternatives like tar sands and nuclear power, we’re finding that we are steadily spending increasing amounts of energy to get energy.
Even though today’s societies confront the same energy exigencies as ancient Rome, they’re different in one key respect: they’re vastly more complex and ordered, and they’re much further from thermodynamic equilibrium. In other words, our societies are like the marble that wants to roll back to the bottom of the bowl, and compared with ancient Rome we’re holding that marble much farther up the bowl’s side. Colossal flows of high-quality energy make this possible. If we can’t sustain these flows, our societies will fall back toward equilibrium—which means, essentially, that their complexity will unravel. And that unraveling, should it occur, would make Rome’s decline pale by comparison.
