Every argument in the week's reading, written out for someone who has not opened the source. Plus a cheat sheet for the week before an exam.
Science cannot be made value-free, so somebody chooses which values it runs on. Two models for choosing, one chapter that refuses the question, and the woman barred from the French academy who got hypotheses right in 1740.
Research cut into nine steps, with a value entering at almost every one. Then the study of ignorance. One class only, since Monday is Thanksgiving.
Women entered primatology and saw things nobody else had seen, which should have been impossible. The six questions the rest of the book answers.
Two chapters of Sismondo. Whether science rewards the right people, and why the phrase everybody argues about is seven different claims.
Two short readings answering the same question from opposite ends. Why a scientific community changes its mind, and what the machinery for knowing people is for.
Nine texts across 2,400 years, all disagreeing about what counts as knowledge. Plato, Schadewaldt, Bacon, Comte, Rousseau, the Vienna Circle, Lynn White, Dennett and Foucault.
Twelve findings on why explanations fail, each with the bad version and the good version written out. Collected because the same faults kept turning up in the lecture theatre and in these guides.
Your notes stay in your own browser. Highlights, comments and notes are saved on the device you are using and they survive a reload, a closed tab and a restart. They do not travel to another computer or another browser, they are never sent anywhere, and nobody else can see them. Clearing your browsing data for this site removes them, and a private window keeps them only until you close it.
The guides get rewritten often, whenever a better way to answer a question turns up. A highlight is anchored to the words around it, so it survives edits elsewhere on the page. If the exact paragraph you marked gets rewritten, that highlight goes. Sorry. It is the price of the guide getting better.
These guides were written with Claude, from the readings. Treat them as good AI writing: mostly right, and capable of being wrong. They are dependable on the shape of an argument, on who said what, and on how the texts collide, which is what they are for.
They get weaker where a long passage has been compressed into a short one. Compression is where detail goes missing, and where a claim can come out stronger or flatter than the author put it. Anything you plan to quote, or anything that decides an exam answer, is worth checking against the reading itself. The page numbers are there so that check takes a minute.