How the book handles contested ideas

The evidence system

Much of what this book preserves is contested. Rather than leave it out or wave it away, every contested or frequently misstated claim gets the same treatment, in the same order.

Fair telling. The idea in its strongest honest form, from the primary sources where they exist.
Evidence status. A box that states the claim neutrally in one line and gives it one of five statuses.
Recent findings. What has changed in the last decade or so, dated and sourced.
Safety, wherever a system or experiment is described.

The five statuses

Proven: Established by replicated measurement or an unbroken documentary record. Stated plainly.
Plausible: Consistent with known physics and some evidence, but not yet nailed down.
Disputed: Credible evidence or argument on more than one side. Both sides get their strongest form.
Untested: Claimed, but no adequate test has been published either way.
Examined, not replicated: Tested at least once; independent replication has not reproduced the claimed effect.

There is no sixth status, and none of them is "debunked". A claim that was tested and not reproduced is Examined, not replicated: the record says what was tried and what was found, and leaves the door open to a better experiment.

What every box contains

The claim: Stated neutrally, in one line, so it can be agreed or disagreed with.
Status and type: One of the five statuses, plus whether the question is empirical (an experiment or record could decide it) or interpretive (no experiment yet distinguishes the readings).
What's established: The part nobody serious disputes, with sources.
What's contested or unknown: Where the disagreement actually lives.
What would settle it: The document, measurement or replication that would move the rating.

Rate claims, never people

Inventors, experimenters and theorists appear in this book because their ideas are worth keeping track of. A rating describes the evidence for one claim at one date. It is not a verdict on anyone's character or competence, and the same person can have one claim rated Proven and another Disputed.

An example

Who first conceived the rotating-field induction motor, Nikola Tesla or Galileo Ferraris? The book rates it Disputed (empirical): Tesla filed first and won in court; Ferraris's own dated footnote puts his bench experiments earlier; neither copied the other on any evidence found; a dated notebook or the Patent Office interference file would settle it. The full box is in the sample chapter.

On the same pages, the claim that Edison's laboratory, instruments and staff supported Harold Brown's 1888 animal tests is rated Proven, from consistent sworn testimony on both sides, while the claim that Edison directed those tests to discredit alternating current is Disputed (interpretive).

When a rating changes

Ratings are dated. When new research changes one, the change is recorded in the next printing and listed with corrections for that volume. If you have a source that should move a rating, send it through the contact form and choose "Correction".