Why Rome actually fell: plagues, slavery, & ice age — Kyle Harper

Dwarkesh Podcast 1h24 3 min #88
Why Rome actually fell: plagues, slavery, & ice age — Kyle Harper
Watch on YouTube

Summary

  • Kyle Harper, a historian and provost emeritus at the University of Oklahoma, argues that the fall of the Roman Empire was driven not just by politics or military decline but by environmental shocks—especially pandemics and abrupt climate change—that together caused catastrophic population collapse and economic disruption.
    • The Roman Empire at its height (1st–2nd centuries CE) was a vast, urbanized, trade-rich Mediterranean superpower with Rome as the world’s largest city.
    • By the 6th–7th centuries, Rome had shrunk to 50,000–100,000 people—just 5–10% of its peak—and the western empire had fragmented.
    • Harper emphasizes that while internal weaknesses existed, the empire might have endured in a reduced but coherent form (like China’s dynastic cycles) were it not for two massive exogenous shocks in the 6th century: the Plague of Justinian and a severe volcanic-induced cold period.

The double shock: plague and climate collapse

  • The Plague of Justinian (541 CE onward) was caused by Yersinia pestis, the same bacterium behind the Black Death.
    • It killed an estimated 50–60% of affected populations in cities and regions—far exceeding normal annual death rates of 3–4%.
    • Unlike later waves, this first pandemic likely traveled from Central Asia through Indian Ocean trade routes into the Red Sea and then the Mediterranean.
    • Ancient DNA evidence shows the pathogen originated in the Tian Shan Mountains (Central Asia), but its spread to Rome may have gone via Gujarat and maritime commerce.
  • A series of major volcanic eruptions in the 6th century triggered abrupt global cooling (1–2°C), disrupting agriculture for decades.
    • Sulfur aerosols in the stratosphere reflected sunlight, shortening growing seasons and causing crop failures.
    • This “Roman Little Ice Age” led to widespread famine just as the plague hit, creating a devastating synergy.
  • Together, these shocks overwhelmed the empire’s resilience at a critical moment—Emperor Justinian had nearly reconquered Italy and North Africa when the plague struck.

Why Rome didn’t industrialize

  • Despite impressive economic complexity—long-distance trade, financial markets, urbanization, and proto-industrial production—Rome never sparked sustained technological takeoff.
    • It achieved “Smithian growth” via specialization and market exchange but lacked the scientific foundation for continuous innovation.
    • Key missing ingredients:
      • Basic science: No systematic investigation of natural laws.
      • Empiricism: Knowledge relied on authority (e.g., Aristotle), not experiment.
      • Useful knowledge: Engineers and theorists didn’t collaborate; no feedback loop between abstract science and practical application.
  • Even with abundant slave labor (10–20% of the population), Harper argues slavery wasn’t the main barrier to mechanization—many advanced sectors used slaves, but without scientific progress, productivity gains stalled regardless.

Slavery in the Roman world

  • Slavery was central to Rome’s economy from the 2nd century BCE onward, fueled by conquest and sustained by markets that turned forced labor into profit.
    • Slaves worked in plantations (olive oil, wine), households, mines, and even skilled roles (tutors, nurses).
    • Unlike modern racialized slavery, Roman slavery was justified through legal status and conquest ideology—not biology or race.
  • Despite its brutality, there was no abolition movement.
    • Justifications included the idea that slaves were “spared” in war and now owed service.
    • Control relied on a mix of violence, repression, and incentives like manumission (earning freedom).
  • Large-scale revolts were rare (e.g., Spartacus) due to pervasive psychological and institutional control, though resistance likely occurred in unrecorded forms.

Disease, cognition, and human development

  • Infectious disease has profoundly shaped human biology and cognition over millennia.
    • Pre-industrial populations were shorter and likely had lower average cognitive capacity due to chronic infection and malnutrition during childhood.
      • The immune system is metabolically expensive; fighting disease diverts energy from brain development.
      • Even geniuses like Newton emerged from a population with a lower mean but wide distribution of ability.
  • The transition to agriculture (~10,000 years ago) increased disease burden:
    • Sedentism, proximity to waste, and dense settlements enabled pathogens to spread.
    • Farmers outcompeted hunter-gatherers not just through higher fertility and energy capture, but possibly via disease exposure—similar to European contact with Native Americans.
  • Some innovations (e.g., scurvy prevention, vaccination) took centuries to discover because they required understanding invisible mechanisms—highlighting how hard certain problems are without germ theory.

Future risks and de-extinction

  • Synthetic biology raises pandemic concerns: While evolution constrains pathogens (e.g., high virulence can limit transmission), outliers like Yersinia pestis show nature can produce highly lethal, vector-borne diseases.
    • Plague is unusual because it’s zoonotic (we’re collateral damage) and flea-borne—so it faces no evolutionary pressure to reduce human virulence.
    • Future threats may come from unexpected sources (e.g., prions, fungi) where we lack medical countermeasures.
  • De-extinction (e.g., woolly mammoths) is scientifically intriguing but ecologically limited:
    • A species isn’t just a genome—it needs an ecosystem. The mammoth steppe is gone.
    • Even if we recreate extinct animals, without habitats, they’d exist only in captivity.
    • Our current choices about biodiversity will shape Earth’s evolutionary trajectory for millennia—making conservation urgent regardless of future tech.

Harper’s use of AI in research

  • Harper now uses AI tools like Deep Research and LLMs daily as conversational research partners.
    • They help synthesize cross-disciplinary literature, extract data from PDFs, and explore new questions.
    • However, he sees AI as augmenting—not replacing—human historians, especially in framing original questions and creative synthesis.
Back to Dwarkesh Podcast