Lorraine Daston, director emerita of the Max Planck Institute for the History of Science, reviews Roopika Risam’s Data Empire — an 11,000-year history of information as an instrument of organization, control, and domination. Risam’s subtitle is her thesis: “The Power of Information to Organize, Control, and Dominate.” “We are the data species,” she writes.

The earliest written texts were not epics but ledgers:

  • Sumerian clay tablets listing sheep, barley, and wine to be taxed and shipped
  • Imperial accounting that tallied labor and materials for irrigation, walls, roads, and pyramids
  • Suriname’s colonial land surveys earmarking profit in sugar and timber
  • The Atlantic slave trade’s ledgers and insurance policies on human cargo

Risam’s ambivalence runs through every era: data curses the governed but blesses historians. A Theban land dispute between a deed-holding Khay and a memory-rich Mose; Qin Shi Huang standardizing weights and burning records; the Spanish casta system’s racial taxonomy; Frances Perkins assigning Social Security numbers from a former Coca-Cola factory. Even the Dewey Decimal system, Daston notes, smuggles in ethnocentric prejudice — more numbers for Christianity than for other world religions.

The modern catalogue is bleak: the IDF’s AI-driven “Lavender” database identifying bombing targets in Gaza, regime-critical posts blocking Chinese citizens from trains, keystroke-level workplace monitoring. Hopeful experiments keep getting crushed — socialist Chile’s real-time transport data scuttling a CIA-backed truckers’ strike before the coup, the Homebrew Computer Club’s free-sharing ethos swallowed by Bill Gates’s monopoly pricing.

Daston’s critique is that Risam sometimes grants data a life of its own, letting human agency disappear — local knowledge repeatedly trumped the most accurate maps, from the Portuguese and Ottomans in the Indian Ocean onward.

The book closes with two futures a century out: one comfortable-but-controlled, with ubiquitous surveillance; one of data for the public, by the public — worker-approved algorithms, unweaponized health data. The difference, Risam concludes, is not data but us. After 11,000 years of being data-driven, can we finally drive data?