Lectures

The twelve lectures, who gives them, and the evidence each one puts on the table.

Mondays, one hour, twelve of them. Every lecture has slides and they are linked from its page, before the lecture rather than after it — there is nothing in them worth withholding, and reading ahead makes the Thursday seminar better.

A lecture here is not a summary of the week. It is the place where the evidence gets put in front of you properly: the actual experiment, the actual numbers, and the actual disagreement where there is one. That is why several of these pages name a single paper and spend most of their length on it.

Three of us teach. Halvard takes weeks 1, 11 and 12: the week that sets the evidence standard, and the two that apply it to the course’s own sources and then to the people in the room. Rosalind takes the comparative weeks — 2, 3, 4, and then 8 through 10. Tuyet takes 5, 6 and 7, the three weeks where the machinery gets internal enough to model itself.

Two lectures argue against their own content in the last twenty minutes, and they are marked as such on the page. That is not a rhetorical device. Week 6 presents predictive processing and then gives you the dark room problem; week 9 presents Nagel and then gives you Akins. If a framework cannot survive the hour it was introduced in, you should find that out in the room.