The Demon-Haunted World

Sagan's case is that science is not a body of knowledge but a way of thinking — specifically, a set of procedures for catching one's own errors — and that a society dependent on technology while unable to evaluate claims is in a precarious position.

His warning, written in the mid-1990s, is that the combination of technological dependence and widespread inability to assess evidence produces a population that can be told anything.

A society dependent on technology and unable to evaluate claims can be told anything at all.

Devmark’s reading — not a quotation from the book

His central practical contribution and the most-extracted part of the book: a set of tools for evaluating claims.

Seek independent confirmation. Encourage debate from knowledgeable proponents of all views. Arguments from authority carry little weight. Spin more than one hypothesis. Do not become attached to your own. Quantify where possible. Every link in a chain of argument must work. Prefer the simpler explanation when both fit. Ask whether the claim can be falsified.

Accompanying the kit, a catalogue of common errors: ad hominem, argument from authority, appeal to consequences, special pleading, begging the question, observational selection, small-number statistics, the excluded middle, the straw man.

The value is that they are given with concrete examples from public debate rather than as abstract logical forms, which makes them recognisable in the wild.

His extended treatment of the abduction phenomenon, examined seriously rather than dismissed: the sincerity of the reporters, the role of sleep paralysis and hypnagogic states, the effects of hypnosis on memory, and the cultural patterning of the reports.

His method is the point. He takes the experiences as real experiences requiring explanation, and applies the kit to the explanations offered rather than to the people offering them.

His most reproduced parable: a claim that an invisible, incorporeal, heatless dragon occupies the garage, with every proposed test met by a further property that makes it undetectable.

His point is that a claim immunised against every possible test is not merely unproven but empty — the difference between an undetectable dragon and no dragon at all is nothing, and this is what falsifiability means in practice.

The difference between an undetectable dragon and no dragon is nothing. That is falsifiability in one image.

Devmark’s reading — not a quotation from the book

His chapters on the European witch trials, on the machinery of accusation and confession under torture, and on how a coherent system of belief produced tens of thousands of deaths.

His argument is that the beliefs were not stupid given the framework, and that the failure was the absence of any procedure for testing them. This is his strongest historical case for method as a moral matter.

His chapters on why school science fails: it delivers conclusions rather than the process of reaching them, discourages the questions that children naturally ask, and presents the subject as settled fact rather than as continuing inquiry.

His observation that children arrive curious and leave incurious, and that something in between is responsible, is the part of the book most frequently cited by teachers.

His recurring insistence that the two are not in tension: that the real universe is more interesting than the invented ones, and that scepticism without wonder is arid while wonder without scepticism is credulity.

This is his answer to the charge that scientific thinking drains meaning from experience, and it is made more convincingly here than anywhere else in his work.

The specific delusions are of their moment — abduction narratives, facilitated communication, recovered-memory therapy, channelling — and a reader will find the examples period-specific.

The diagnosis has not dated at all. His description of a public unable to evaluate claims, combined with media incentives that reward sensation, reads as more urgent now than when it was written.

It is the standard text of modern popular scepticism and is frequently described as Sagan's most important book, though it is less read than *Cosmos*.

Its lasting contribution is the kit, which is short enough to remember and general enough to apply to anything, and which has been reproduced independently of the book more than any other passage he wrote.