I have run Socratic seminars for two decades, and I have bombed plenty of them. The failures taught me more than the successes: a seminar doesn't fail in the circle — it fails in the design decisions made before anyone sits down.
1. The question has to be genuinely contestable
"What is the theme of the text?" produces recitation. "Does this author believe progress is possible?" produces argument. If you already know the answer you're hoping to hear, it isn't a seminar question.
2. Evidence on the table, literally
Students speak with annotated text in front of them and a norm: claims point to lines. The sentence "where do you see that?" should be the most common in the room — and it should come from students, not you.
3. Prepare the quiet students in writing first
A ten-minute written response before the seminar means every student enters holding a position. Cold-calling becomes an invitation ("You wrote something about this — say more") instead of an ambush.
4. Track the talk, visibly
Map who has spoken. Students self-correct when participation becomes visible — the dominators throttle back, the silent notice their absence.
5. End with revision, not applause
The seminar isn't the product. The revised thinking that follows it is.
Close with five minutes of writing: What did someone say that complicated your view? What would you now change in your argument? That written turn is where the seminar becomes learning — and it's the artifact worth assessing.
