The last lecture. A list of everything this course got wrong, left open or could not support, assembled on purpose rather than apologised for, and handed over.
24 May 2027
Where. Outdoors, venue posted by the Friday before. Opens with the ten minutes, for the last time.
Provisional, like the rest of the back half, though if the vote replaces the final lecture with something better that will be a good outcome rather than an awkward one.
Nothing is due. The log is not collected until after the whole cohort’s exams, and it is a log, so there is nothing to write now beyond tidying it.
The accounting#
A course called Unfinished Thinking that ended with a summary would be getting its own joke wrong. So here is the ledger instead.
The central claim is untested. That a loop repeats because the thinking never gets the conditions under which it would finish. It is assembled from real findings, incubation during undemanding tasks, attention residue, the phone break that measures like no break. Nobody has tested the claim directly, including us. It is the best account we could build, and it is still a hunch with good parts.
Three things we refused to claim. They are listed on the deck and they have not moved all semester. The temptation to quietly upgrade them at the end of a course is exactly what this lecture exists to resist.
Two traditions poked holes we never filled. Week 8 gave you Morita, saying the fix is not mechanical, and the Friendship Bench, saying it is not private. Both have at least as much support as the course’s own claim. Neither is in it.
Most of it is small. Week 9 put the numbers in one table. Worry postponement, d = 0.31 and 0.19. Expressive writing, about d = 0.15. Exercise on state anxiety, a quarter of a standard deviation. The honest summary of a semester is that several of these help a bit, one of them measurably backfires, and the biggest number on the syllabus is also the oldest and least re-examined.
Some of it we got wrong on the way and fixed in public. The sequencing of this course was built on concrete-to-abstract evidence, then corrected against Luiten, Ames and Ackerson’s meta-analysis, which says learners do better with a brief organising idea up front. That is why week 1 states the argument in a sentence and then abandons it until week 5. The Lectures page carries the full correction, including the contested paper that says the whole approach might be backwards.
One thing we could not find at all. The assessment calendar in this course is built around avoiding everyone else’s deadlines. We could not find a study isolating clustered against spread deadlines and measuring wellbeing. Assessment labels it design reasoning, not a finding, because that is what it is.
What actually transfers#
Strip the content out and three things survive.
Noticing the seam. Week 1 asked at which layer the group stopped learning about the message and started projecting. That question works on anything, and it does not need any of the research to be true.
Checking what you are crediting. You had a break and credited the sugar. That error is everywhere once you can see it, and it is most of what separates a technique that works from one you believe works.
Knowing what would change your mind. The last crit made you state a prediction specific enough to be wrong. The same move applies to every claim in this course, including ours.
Where to keep going#
Not here. The log asks what you found elsewhere that worked better than anything we taught you, and those answers change the syllabus. That is not a nice gesture at the end of semester. Three of the crits you sat this year came out of a previous cohort answering that question.
Nobody teaching this has stopped overthinking. That was on the Policies page from week 1 and it is still true in week 12.
Discussion#
- Which claim from this semester are you least convinced by, and what would it take?
- What did you find outside this course that beat anything in it?
- If you had to delete one week to make room for something better, which, and what goes in?
- The course argues a thought needs conditions in order to finish. Has anything you were carrying in February actually finished?
Teaching team
- Marisol Quaye, convenor
- Idris Fenn, tutor