CHAPTER 02
The Tree of Knowledge
Read It
The puzzle isn't why Sapiens eventually dominated—it's why it took so long. For tens of thousands of years, our ancestors were just one human species among several, with no obvious edge. Then, around 70,000 years ago, something shifted. Not a bigger brain or a better throwing arm, but a rewiring that produced a new kind of language. Most discussions of language focus on vocabulary or vocal range, but the real breakthrough was stranger: the ability to talk about things that don't exist. That single capacity unlocked gossip, which let groups grow to about 150 people, and then fiction, which let them grow without limit. The chapter's real subject is the hierarchy of communication—from precise description to social tracking to shared myth—and why only the third level could produce cooperation among strangers at scale.
Open full image ↗Draw It
The progression from factual language to gossip to fiction is easy to flatten into a list of three features. But the power lies in the levels: each one solves a problem the previous level can't. Precise description helps coordinate hunting, but doesn't build trust. Gossip builds trust, but only among people who know each other. Fiction builds trust among strangers by giving them a common belief that no one has ever met. Mapping this as a hierarchy makes the ceiling explicit: chimpanzees and archaic humans were capped at small groups because trust required personal familiarity. Sapiens broke through because shared fictions made personal familiarity unnecessary. The visual choice is to foreground the thresholds, not the features.
Rethink It
This distinction matters whenever a team grows beyond the size where everyone knows everyone. A startup of ten people runs on gossip—who's reliable, who's struggling, who's aligned. At a hundred and fifty, that breaks down. What replaces it is fiction: mission statements, values, org charts, roadmaps. None of these exist in the physical world, but they let strangers coordinate. The mistake is treating these fictions as mere decoration. They are the load-bearing structure. When a company says "we need to get back to our culture," what they often mean is that the shared fictions have stopped being believed, and the cooperation they enabled is starting to fray.
Take It With You
The boundary of this insight is that not all fictions are equal. Some enable cooperation; others enable exploitation. The chapter doesn't dwell on that distinction, but it's the natural next question. For now, the takeaway is simpler: if you want to scale cooperation, don't start by optimizing communication tools. Start by asking what shared belief the people involved actually hold. If the answer is nothing, no amount of Slack or documentation will save you.