# Why "Society Did It" Explains Nothing *The Coleman Course · Part 1 · [[00-social-dynamics|all 24 laws]]* > **Law 1 — If no person appears in the explanation, it isn't an explanation yet.** Here's a sentence you've read a hundred times this year: > *Economic anxiety is driving the rise of populism.* Sounds like an explanation. Look closer: **nobody is in it.** Anxiety doesn't vote. Populism doesn't rise. People do things — and this sentence contains no people. It's two big facts with an arrow drawn between them, and the arrow is doing no work. James S. Coleman opened his 1990 masterwork *Foundations of Social Theory* with a rule for catching this trick, usually drawn as a diagram people call **Coleman's boat**: ```mermaid flowchart LR A["MACRO cause"] -.->|"✗ lazy arrow"| D["MACRO outcome"] A -->|"1 · DOWN"| B["a person's<br/>situation"] B -->|"2 · ACROSS"| C["a person's<br/>action"] C -->|"3 · UP"| D ``` Any real explanation of a group has to travel the long way round — three moves: 1. **Down** — what did one specific person see, fear, or believe that they didn't before? 2. **Across** — what did a *reasonable* person do about it? Not a stupid one. If your story needs people to be stupid, you've almost always misread their situation, not their intelligence. 3. **Up** — how did those individual actions combine? **And they almost never combine by addition**, because people react to each other on the way up. ## The most famous example is a rebuild job Coleman's own opening demonstration is worth knowing, because it's the canonical version of the trick. Max Weber's celebrated thesis — *the Protestant ethic caused modern capitalism* — is, as usually stated, a lazy arrow: one giant social fact (a religious doctrine) causing another (an economic system), with no humans in the sentence. Coleman rebuilds it as a boat: the doctrine changed what individual believers *valued* (down: work as a calling, thrift as virtue) → so individuals worked, saved and reinvested rather than consuming (across) → and the combination of many such behaviours, interacting through markets, produced sustained accumulation (up). Whether Weber was *right* is still argued. Coleman's point is different: only in the rebuilt form is the claim even *checkable*, because only there does it say what anyone actually did. That's the quiet function of Law 1. It doesn't tell you which explanations are true. It tells you which ones have reached the starting line. ## Watch it work on a live event Around 30–31 July 2026, somewhere between 60,000 and 72,000 people crossed from Morocco into Ceuta — a Spanish city of 83,000 — in about a day and a half. Between 67 and 75 died. Days later, nearly all had returned ([Euronews](https://www.euronews.com/my-europe/2026/08/01/ceuta-migrant-crisis-sparks-strong-reactions-from-all-over-europe), [Fortune](https://fortune.com/2026/08/10/ceuta-morocco-70000-migrants-eu-aid/)). The lazy explanation: *poverty causes migration surges.* But poverty in northern Morocco was the same in June, and June was quiet. **A constant cannot explain an event** — that single test kills half of all macro-to-macro arrows on contact. Now the three moves. **Down:** a rumour spreads — amplified by smuggling networks — that a Spanish court ruling means you won't be sent back. (The ruling was real; it didn't actually apply to land crossings. For the mechanism, that made no difference, which is itself a lesson: the boat runs on what people *believe*, not on what's true.) **Across:** if the door seems open today and shut tomorrow, going today is reasonable. **Up:** every person who crossed was living proof, to thousands watching, that crossing works — so each crossing raised the odds of the next. That's not addition; it's multiplication. Hundreds became tens of thousands within hours, then the whole thing reversed just as fast when the evidence flipped. The deaths carry the same lesson in its hardest form: nobody was killed by the *total*. People died in the few hours when arrivals-per-hour exceeded what one breakwater could pass. The lethal quantity was a **rate** — a fact that exists only at step three. Aggregate numbers ("60,000 crossed") literally cannot contain it. ## Where people misapply the law Two boundary markers, so the tool doesn't get overused. **It is not a demand for psychology.** "Down" asks what a person's *situation* became — what they could see, what their options were — not what childhood produced their temperament. Coleman's individuals are deliberately simple; the explanatory work is done by situations, which is why the method scales to millions of people you'll never meet. **And it is not a ban on talking about structures.** Structures matter enormously in this framework — as the thing that *shapes situations* (step 1) and the thing that *combines actions* (step 3). The ban is only on letting a structure act directly on another structure with no human in between. "The algorithm radicalises people" fails Law 1 as stated; "the algorithm shows each person an ever-narrower sample, and people reasonably update on what they're shown" passes — and immediately suggests what to measure. ## Use it this week Take any headline shaped like *"X is causing Y."* Try to fill in the three moves. One of four things happens: | You find | It means | |---|---| | All three fill in cleanly | A real explanation. Rare. Keep it. | | Step 1 is vague ("the mood") | Nobody checked what individuals actually saw | | Step 2 needs people to be stupid | Wrong situation, not wrong people | | Step 3 assumed to be simple addition | **The classic miss. Suspect a cascade.** | Do this for a week and most confident commentary stops sounding confident. That's the point — and the rest of this course is essentially a parts catalogue for the three arrows: Laws 2–5 for the down-and-across, Laws 6–12 for the ways actions combine, Laws 13–24 for what the combinations build. **Remember it as:** *no people in the sentence, no explanation in the sentence.* --- *Next: [[02-interest-and-control|Law 2 — power is what passes through your hands]] · [[00-social-dynamics|Index]] · Coleman, Foundations of Social Theory (1990), ch. 1.*