Summary

Vaclav Smil, a Czech-Canadian energy scientist, retells world history as a history of energy: which "prime movers" (things that turn energy into motion — human and animal muscle, waterwheels, windmills, steam engines, internal combustion engines, turbines) and which fuels (wood, charcoal, crop waste, dung, then coal, oil and gas) each society used. For almost all of history, humans worked with a few hundred watts of muscle and biomass fuel; within about two centuries, fossil fuels and electricity multiplied the energy available per person many times over, creating modern farming, industry, cities and mobility. His two big lessons: energy sets limits and creates possibilities but does not determine history — religions, constitutions and art arose in low-energy societies, and many collapses had nothing to do with energy; and energy transitions are slow, each new fuel taking fifty to seventy-five years to win a large share of world supply, so the move away from fossil fuels will also take generations. He adds that beyond roughly 100–120 gigajoules of energy per person a year, quality of life stops improving, so rich countries could use far less. As a work of reference the book is outstanding — precise, wide-ranging, full of numbers and comparisons — but it reads as a catalogue more than an argument, says little about how slavery and colonialism fed Western energy use, and its scepticism about fast change has been both borne out (fossil fuels still supplied about 86% of world energy in 2025, the same share as in 2015) and challenged (in electricity, renewables overtook coal in 2025).

Chapter-by-Chapter Summary of Energy and Civilization: A History by Vaclav Smil

MIT Press, 2017; a much-expanded rewrite of Smil's Energy in World History (1994). Seven chapters, many boxed technical asides, a chronology and appendices of units and power ratings; about 190,000 words of text, with a reference list of several hundred works. The EPUB has no page numbers, so this note cites chapters. Chapters 1 and 7 were read closely; Chapters 2–6 by section openings, with searches of the full text for every figure used.

Preface. Rewritten because energy studies grew so much after 1994.
1. Energy and Society. Energy is "the only universal currency": every action is a conversion of one form to another. Smil sets out the basic measures — energy density (energy per kilogram of a food or fuel), power density (energy per square metre of land, which limited how large pre-industrial cities could grow on wood and charcoal), conversion efficiency, energy return on investment (how much energy you get back for the energy spent obtaining it). He shows how efficiency gains often raise total use (the Jevons paradox) and warns that energy numbers miss crucial qualities: whether food is digestible, how accessible a fuel is, how smoky it burns. He rejects the claim, from Wilhelm Ostwald to Howard Odum, that cultural progress tracks energy use; quantifying is necessary but not sufficient.
2. Energy in Prehistory. Foragers lived at densities that, apart from some maritime peoples, never exceeded one person per square kilometre. Hunting large, fatty animals gave high energy returns; plant gathering was steadier. Farming did not start for one reason; it arose from many causes, not simply from energy need.
3. Traditional Farming. Agriculture as a long effort to get more food energy from each hectare: draft animals, irrigation, manuring, crop rotation and multicropping. Detailed cases from Egypt, China (more innovative than Egypt), Mesoamerica (no draft animals), Europe (behind China until the seventeenth century) and North America (the most labour-efficient farming in the world by 1900). An ordinary grain harvest could feed about ten times as many people as foraging on the same land, and the most intensive Asian farming more than five people per hectare — but famines persisted and yields rose very slowly.
4. Preindustrial Prime Movers and Fuels. Human labour, at a sustained 60–100 watts per person, was the main source of mechanical power for most of history. Draft animals, then waterwheels (common in some regions only after about 200 CE) and windmills (after about 900 CE), slowly added power; the largest waterwheels took about a thousand years to grow tenfold in capacity. Wood, charcoal, crop residues and dung supplied heat; charcoal-making lost much of the wood's energy. Chapters on household needs, transport (sail ships far outperformed land transport), building, metallurgy and warfare (gunpowder as the first major non-muscle energy in war).
5. Fossil Fuels, Primary Electricity, and Renewables. The "great transition": coal in England and China long before industrialisation; coke replacing charcoal in iron smelting; the steam engine, the first new prime mover in more than 800 years since the windmill; oil and the internal combustion engine; electricity, the most flexible form of energy. Oil supplied 3% of fossil energy in 1900, 21% by 1950, peaking at about 46% in 1972. The world still burns more wood and charcoal than ever, because rural populations in poor countries grew.
6. Fossil-Fueled Civilization. World energy use grew from about 20 exajoules in 1800 to 45 in 1900, 100 in 1950, 380 in 2000 and over 550 in 2015. Fossil energy transformed farming (fertiliser, machines), industry, transport, communication and economic growth. Smil's quality-of-life thresholds: below 5 gigajoules per person a year (commercial energy), basic needs cannot be met; around 42 GJ, industrialisation takes off; about 84 GJ marks incipient affluence. Decent diets, health care and schooling are possible at 40–50 GJ; low infant mortality and an HDI above 0.8 at 60–65 GJ; the world's best outcomes need at least 110 GJ, with no improvement above about 100–120 GJ — yet the leading EU economies and Japan used about 150 GJ, the US about 300 and Canada about 385 in 2015. The US, with twice Europe's energy use, had worse infant mortality and life expectancy. The chapter closes on urbanisation, political dependence on energy, weapons and environmental damage.
7. Energy in World History. The synthesis. Energy eras overlap: animate power dominated until the nineteenth century, and the split between muscle and machine is complete only for the richest fifth of humanity. Transitions are slow: coal took sixty years (1840–1900) to go from 5% to 50% of world supply; oil followed a nearly identical path from 5% in 1915 but never reached 50%; gas took 55 years to go from 5% to 25%. In 2015 solar and wind together were below 2% of primary energy, so "the odds are highly in favor of another protracted process". Prime movers became vastly more powerful, lighter and more efficient, and in Britain the cost of light fell to a ten-thousandth of what it was in 1500. Yet energy was not used rationally — much goes on frivolous consumption — and by 2015 the richest 10% of humanity used about 35% of the world's energy. In the section "Between Determinism and Choice", Smil argues energy is an indispensable lens but must not be made the principal explanation: ethics, religions, the US Constitution and great art arose in low-energy societies; freedoms and happiness do not track energy use; population growth in Qing China and the collapse of complex societies are poorly explained by energy. Controls — bits and harnesses, navigation, feedback devices, printing, institutions — matter as much as energy sources. He ends with uncertainty: a renewable transition must also replace fossil fuels as raw materials for fertiliser, plastics and steel, and success is not assured.

Critical Thinking Framework Analysis

Stage 1: Reconstruct the Argument

Standard Form:

Every human activity requires converting energy, and the amount and form of energy a society can control limits its population, cities, production, transport and warfare.
For most of history, societies were limited to muscle, biomass and modest water and wind power, which kept them small, poor and slow-changing.
Fossil fuels and electricity raised the energy available per person by an order of magnitude or more within about two centuries, making modern civilisation possible.
But culture, institutions, ideas and choices shape how energy is used, and many of history's great achievements and failures had no clear link to energy.
Past transitions to new energy sources took fifty to seventy-five years at the global scale, because they need vast infrastructure and old systems persist.
Beyond a moderate level of energy use, quality of life no longer improves.
Conclusion: energy is a necessary but not sufficient explanation of history; the transition away from fossil fuels will probably be slow, and rich societies could live well on much less energy.

Principle of Charity: Smil is not an energy determinist; he spends a whole section refuting that view and states that energy sets bounds without fixing what happens inside them. His forecast is probabilistic ("the odds are highly in favor"), not a claim that fast change is impossible; he says explicitly that the past pattern does not have to repeat. And his scepticism about fast transitions comes with an argument for using much less energy in rich countries, not with a defence of fossil fuels.

Clarification: Most of Chapters 2–6 is a reference survey (crop yields, waterwheel ratings, ship speeds) supporting points 2 and 3. The argument proper lives in Chapter 1's caveats, the transition data and quality-of-life thresholds (Chapters 6–7), and "The Limits of Energy Explanations".

Hidden Elements:

That global primary energy shares are the right measure of a transition. Measured in useful energy or in electricity, new sources look larger, because fossil fuels waste much of their energy as heat.
That the future will resemble the past enough for past transition speeds to be a guide, although solar panels and batteries are factory products with steep cost declines, unlike coal mines or oil fields.
That quality of life can be captured by infant mortality, life expectancy, schooling and the Human Development Index, not by comfort, mobility or choice.
That the main energy flows worth counting are fuels and machines, leaving out coerced human labour in the colonial economy.

Stage 2: Sceptical & Rhetorical Analysis

Identify Fallacies:

Induction from a small sample. Three transitions (wood to coal, coal to oil, oil to gas) are a thin basis for a general law of transition speed, and they were transitions to fuels that were themselves extracted and burned. Smil partly guards against this, but the forecast leans on the pattern.
Choice of measure tilts the result. Using primary energy makes wind and solar look small, because their electricity is counted at its own energy content while fossil fuels are counted including the heat lost in power stations. Ketan Joshi (RenewEconomy, 2021) made this criticism of Smil's later work.
Selective comparison. The US versus Europe and Japan is a strong case that more energy does not buy better health, but health also depends on inequality, health-care systems and diet; the comparison does not isolate energy.
Omission rather than fallacy. Slavery appears only in passing, as one of the coercive ways elites have controlled energy flows; the Atlantic slave trade as a source of energy for European colonies is not analysed.

Expose Rhetoric: Smil's style is the opposite of emotional: dense, numerical, with lists of references. That itself persuades — readers trust the numbers — and makes his occasional sharp judgments stand out ("tens of millions of people annually take intercontinental flights to generic beaches in order to acquire skin cancer faster"; "more than 500 varieties of breakfast cereals"). His dismissals of "energy myths" and failed forecasts (Lewis Strauss's "too cheap to meter") work by ridicule of past optimists, which can carry over unfairly to present ones.

Comparable Example: "Telephones took seventy years to reach half of American homes; therefore mobile phones will take seventy years to spread." Mobile phones reached most of the world in about two decades, because they were cheap mass-produced devices that needed little fixed infrastructure. The parallel shows Smil's transition argument is strongest for systems that need mines, pipelines and power plants, and weaker for modular technologies made in factories — which is what solar panels and batteries are.

Stage 3: Overcome Bias & Seek Refutation

Puncture the Echo Chamber:

Transitions can be fast. Sovacool (UNU-WIDER working paper, 2016) found national transitions that took from 1 to 16 years, and argued that transition speed depends on what is measured and where. Grubler, Wilson and Nemet (Energy Research & Social Science, 2016) replied that such cases compare unlike things — but agreed that scale and context, not a fixed law, set the pace.
S-curves, not straight lines. Kingsmill Bond and colleagues at RMI argue that solar and wind follow learning curves and exponential growth, so their share accelerates once they become the cheapest option; on this view fossil fuel use for electricity plateaued around 2018 and is set to decline (Inside Climate News, January 2023).
The colonial blind spot. Bart Hawkins Kreps (Resilience, June 2017) asks what of "the 20 million or more slaves taken from Africa to work in the European colonies" — a large energy subsidy to European economies that the book barely mentions. Chris Smaje (Resilience, March 2018) adds that Smil's discussion of farming "involution" neglects colonial economic structures.

Falsification:

If solar and wind together reached a large share of world primary energy (say 20%) within about twenty-five years of reaching 2%, the "protracted process" forecast would fail.
If countries using well above 120 GJ per person showed consistently better health and education than those at 100 GJ, the saturation claim would fail.
If careful historical work showed that energy availability, not institutions or culture, explained most differences in when and where societies industrialised, Smil's anti-determinist position would weaken.

Look Outside the Frame (WYSIATI):

Climate change is discussed, but mainly as one environmental consequence among many; the book does not treat it as a reason the next transition may be forced faster than past ones.
Storage and batteries, which change how intermittent renewables fit into a system, get little attention.
Global South agency: the history is told largely through Europe, China and the US; Africa, South Asia and the Middle East appear mostly as places of low energy use.
Energy and inequality within countries, beyond international averages.

Recent Developments Concurring or Debunking the Claims

(Concurring) The fossil share has hardly moved. Smil records fossil fuels at 86% of world primary energy in 2015. The Energy Institute's Statistical Review of World Energy 2026 reports 86% of total energy supply in 2025, with supply above 600 exajoules (up 1.7%) and renewables adding 3.3 EJ (reported by DieselNet, July 2026; the Energy Institute page itself was not opened). A decade on, the global energy mix has barely shifted — strong support for the "protracted process" forecast, at the level of total energy.
(Refining) Electricity is moving faster. Ember's Global Electricity Review 2026 (21 April 2026): in 2025 renewables supplied 33.8% of world electricity and overtook coal (33.0%) for the first time in a century; solar grew by a record 636 TWh (30%), and fossil generation fell slightly (38 TWh, 0.2%). In the electricity sector, the transition looks more like an S-curve than Smil's sixty-year pattern; in total energy (point 1) it does not yet.
(Concurring, with a lower threshold) Quality of life saturates. Jackson and colleagues (Ecosphere, 2022) found eight of nine well-being measures rise steeply and then level off between 10 and 75 GJ per person per year, five of them (including life expectancy and infant mortality) at 10–30 GJ; only air quality kept improving up to about 125 GJ. Today's world average of 79 GJ could in principle give everyone 95% or more of the best performance. This confirms Smil's saturation idea but puts the plateau lower than his 100–120 GJ. Kikstra and colleagues (Environmental Research Letters, 2021) estimate that decent living for all would need less than half the energy most 1.5°C scenarios project.
(Refining) Land use of renewables. Hannah Ritchie (Our World in Data) estimates nuclear needs about 50 times less land than coal and 18–27 times less than ground-mounted solar per unit of electricity, but argues the extra land, much of it unproductive, is an acceptable price — accepting Smil's point about low power density while rejecting it as a barrier.
(Refining) Disputes over Smil's method. Ketan Joshi (RenewEconomy, 11 May 2021) argued that a later Smil graphic of "failed" renewable forecasts mislabelled targets and scenarios as forecasts, and that his preference for primary energy understates renewables. This criticism targets later work, but the same method underlies this book's transition figures.

Additional Recommended Reading

Vaclav Smil, How the World Really Works (2022). The short, popular version of the same arguments, focused on the present. See How the World Really Works, The Science Behind How We Got Here and Where We're Going by Vaclav Smil.
Richard Rhodes, Energy: A Human History (2018). A narrative history of the same transitions, through the people who made them.
Kenneth Pomeranz, The Great Divergence (2000). Why Europe, not China, industrialised first, giving coal and colonies central roles — the counterweight to Smil's treatment of colonialism.
Arnulf Grubler, "Energy transitions research: Insights and cautionary tales" (Energy Policy, 2012). The academic case on transition speed.

Web Research: Recent Studies & External Reviews

Reviews. Bill Gates (gatesnotes, December 2017): Smil "goes deep and broad" on how energy conversions drove progress; Gates's own page could not be read, so the wording comes via CNBC. Bart Hawkins Kreps, Resilience, 21 June 2017: praises the coverage and illustrations; faults the neglect of slavery and colonial extraction. Chris Smaje, Resilience, 8 March 2018: the most critical, on Smil's views about fossil fuel limits, nuclear power and colonial context. A peer-reviewed review by C. Warren appeared in Local Environment 23(5), 2018 (not read).

Research. Energy Institute Statistical Review 2026, via DieselNet; Ember, Global Electricity Review 2026; Sovacool 2016; Grubler, Wilson and Nemet 2016; Inside Climate News on Kingsmill Bond and RMI, January 2023; Jackson et al., Ecosphere 2022; Kikstra et al. 2021, via ScienceDaily; Our World in Data on land use per energy source; Ketan Joshi, RenewEconomy, May 2021.

Where the evidence is thin. No reviews in Nature, Science or major newspapers were found; the book was reviewed mostly by bloggers and in one academic journal. The 2025 fossil share comes from a secondary report, and the Energy Institute's methods have changed since BP's 2016 edition that Smil used, so the two 86% figures are close but not strictly comparable.