Tag: secondary teaching

By Clay Shumate

Good questions for a Socratic seminar have no answer you could look up, no answer the teacher is holding, and at least two defensible readings in the text itself. That is the whole test. A question that a prepared student can settle in one sentence is a comprehension check — useful, but it will not carry thirty minutes of discussion.

Below are forty questions you can use tomorrow, sorted by what they are for, plus the way to write your own out of whatever text you are already teaching.

Key Takeaways

• Three kinds of question run a seminar: one opening question, several core questions you may never need, and closing questions that ask what changed.
• The research on “higher-order questions” is weaker than the training says. One meta-analysis of 20 studies found a significant positive effect; a second, of 14 studies, found the effect small. Both are true and both are old.
• What does hold up is the authentic question — one the teacher does not already have an answer to. In a study of more than 1,100 eighth and ninth graders, that predicted higher literature achievement.
• Wait longer than feels reasonable. Teachers average about one second. Three is the floor for a recall question, and for a genuinely open one there appears to be no ceiling.
• Students should be writing questions too. A seminar where only the teacher asks is a recitation with the chairs moved.

What Makes a Good Socratic Seminar Question?

Four things, and you can test all four in about a minute. Write your question down, then run it against this list before you take it anywhere near a class.

1. You cannot answer it in one sentence. Try. If you finish in a line, you have written a comprehension check. Keep it — just do not open with it, and file it with your other ways of checking what landed.
2. You do not already know the answer. This is the hard one, because teachers are trained to ask questions we can assess. A seminar question is one where you would genuinely take notes on the replies.
3. The text can support more than one answer. Not “what do you think” — opinions with no evidence anchor produce a conversation nobody can lose or learn from. The disagreement has to be arguable from the page.
4. It cannot be Googled. If a phone settles it in twenty seconds, it will get settled in twenty seconds.

The fourth one has quietly become the most important. A question like “what were the causes of the conflict” is a research task. “Which of the causes the author lists is doing the most work, and which one is there to make the argument look balanced?” is a seminar, because the answer is not written down anywhere.

The Three Kinds of Question a Seminar Needs

An opening question, a stack of core questions, and a closing question — and you should expect to use the opening one and maybe two others. The stack is insurance, not a script.

The opening question is the only one you are guaranteed to ask. It should be answerable from a first reading, hard enough to divide the room, and broad enough that four different students can enter it from four different places. It usually points at the whole text rather than a line.

Core questions go down into the text. These are the ones that name a paragraph, a word choice, a contradiction between two claims. Write six and plan to use two — and if you have five preps and six is not happening, write three. Three real ones beat six you assembled at 10pm. The worst thing you can do with a good stack is work through it, because then you are running the agenda and the students are answering it.

Closing questions leave the text. What does this mean for something the students actually deal with? What changed in your thinking, and whose comment did it? That second one belongs in the post-seminar writing, and it is the single best assessment item this format produces. One guard on it: ask students to name a classmate for credit, never for criticism. “What did this conversation get wrong?” is a question about the conversation, and it is worth saying so out loud before they write.

There is a fourth kind that never gets written down, and it is the one that matters most in the room: the follow-up. Building the next question out of what a student just said — researchers call it uptake — is what separates a discussion from a round of answers.

Does Asking Harder Questions Actually Improve Achievement?

Less clearly than the training slides claim, and it is worth knowing why before you rebuild your question bank around Bloom’s taxonomy. Two meta-analyses looked at the same question and did not agree.

Kathleen Cotton’s synthesis for the Northwest Regional Educational Laboratory lays both out. Redfield and Rousseau’s 1981 meta-analysis of 20 studies “concludes that asking higher cognitive questions has a significant and positive effect on student performance.” Samson and colleagues ran a second meta-analysis in 1987 across 14 studies and found that students exposed to higher cognitive questions did outperform others, “but that the effect size is small.”

A later critical review by Mikyeong Yang went further, concluding that this body of work “reported incoherent results” and that the evidence cannot support the expectation that higher-cognitive questions are more effective than fact questions. So: a real effect, a small effect, and a serious argument that the whole line of research is not measuring what it thinks it is.

Here is what survives all three, and it is not about difficulty at all. Nystrand and Gamoran studied eighth- and ninth-grade English classes across sixteen middle schools feeding nine high schools, testing more than 1,100 students in fall and spring. Students whose teachers asked a higher proportion of authentic questions — ones the teacher did not already have an answer to — and who practiced uptake scored significantly higher on literature achievement.

Authentic is not the same as hard. “Analyze the author’s use of irony” is high on Bloom’s and completely inauthentic if you have a rubric in your hand describing the right answer. “Is the last line a joke?” is simple, open, and genuinely contested. Write for authentic, not for the taxonomy. It is the same trade as anywhere else on this site: real questions hand students something to do, which is what giving students a genuine decision rather than a cosmetic one actually looks like inside one class period.

One more number worth carrying, from the same Cotton synthesis: in ordinary classroom recitation, about 60 percent of teacher questions are lower cognitive, 20 percent higher cognitive, and 20 percent procedural. A seminar is one of the few structures that changes that ratio on purpose.

40 Questions for Socratic Seminar

Eight questions in each of five categories, written to be subject-general so you can drop your own text into them. Every one of these is a shell — swap in your author, your document, your data set.

Two things before you use them. They are subject-general on purpose, which means the question is only as safe as the text you put under it; choosing a text on a contested public issue is a decision that belongs to you and your school’s policy, not to a question bank on the internet. And the fourth group — the ones that surface disagreement — is asked to the room, never to a named student. “Who read that line differently?” is an invitation. “Marcus, you disagree, right?” is an ambush, and the student you do it to will remember it longer than the text.

Opening questions (use one)

1. What is the most important sentence in this text, and what does the rest of it do?
2. What is this author afraid of?
3. If you had to remove one paragraph without losing the argument, which one goes?
4. Who is this written for? How can you tell?
5. What does this text refuse to say?
6. Is the author describing the world or arguing about it?
7. What would have to be true for this to be right?
8. What question is this text an answer to?

Core questions: evidence and claims

1. Which claim here has the least support behind it?
2. Where does the author stop explaining and start assuming?
3. Are the two claims in the second and fourth paragraphs compatible?
4. What counts as evidence in this text, and who decided that?
5. Is the strongest objection to this argument in the text, or did the author leave it out?
6. Which detail would change your reading if it turned out to be wrong?
7. Does the author distinguish what they know from what they think?
8. If this were written by someone on the other side, which sentence would survive?

Core questions: language and choice

1. Why that word and not the obvious one?
2. What work is the title doing?
3. Where does the tone shift, and what happens right before it?
4. Who is given a name in this text, and who is not?
5. What is the effect of the thing the author repeats?
6. Is the ending earned by what came before it?
7. What would be lost if this were written plainly?
8. Which sentence do you think the author rewrote the most times?

Questions that surface disagreement

1. Who in this room read that line differently than you did?
2. What is the strongest version of the argument you disagree with?
3. Is this a disagreement about the text or about something outside it?
4. What would change your mind?
5. Can both readings be true at once?
6. What are we assuming that the text never says?
7. Which of us is using a word to mean two different things?
8. Is the disagreement about what happened or about whether it was right?

Closing and reflection questions

1. What did you come in believing that you are less sure of now?
2. Whose comment changed or sharpened your thinking, and what did they say?
3. What question do we still not have an answer to?
4. What would you need to read next to settle this?
5. Where does this argument show up outside of school?
6. If you had to explain this text to someone in one sentence, what would you leave out?
7. What did this conversation get wrong?
8. What is the question you wanted to ask and did not?

Question Stems Students Can Use

The fastest upgrade to a seminar is teaching students to ask each other questions instead of taking turns making statements. Most students arrive with one move — say your opinion and stop — and a short posted list fixes it faster than any amount of coaching.

Post six stems and refer to them by name. Do not post twenty; nobody reads twenty.

The other half of this is teaching students to write a question, since you are requiring one at the door. Give them one frame and it stops being mysterious: point at a specific place in the text, say what confused or bothered you about it, and end with a question you cannot answer yourself. “On page two the author says X, but earlier they said Y — which one are they actually arguing?” is a better seminar question than most of mine, and a fourteen-year-old can produce it on the second try.

For a student who found the text hard, say plainly that the question can be about the part they did understand. A question from paragraph one is a real contribution. Nobody has to have finished the text to have something worth asking about it.

Two of these deserve a note. “What in the text makes you say that?” does most of the work of a facilitator and it is the one stem I would teach first — it is polite, it is not a challenge, and it forces the conversation back to the page every time it drifts. And “Can you say more?” is how a quiet student participates without having to produce a position of their own, which matters more than it sounds like.

How Long Should You Wait After Asking?

Three seconds minimum, and for a genuinely open question, longer than you think you can stand. This is the oldest finding in the questioning literature and it is still the one teachers violate most.

Mary Budd Rowe measured it in the early 1970s. Teachers allowed an average of about one second for a student to respond, and followed a student’s answer with a comment within roughly nine-tenths of a second. When those two waits were stretched to three to five seconds, student responses got longer, more students answered without being called on, speculative thinking went up, students began comparing their evidence to each other’s rather than to the teacher, and fewer questions went unanswered. Worth naming honestly: Rowe’s classrooms were elementary science. The effect has held up widely, but that is the population she measured.

One cheap adjustment that goes with the wait: put the opening question in writing as well as saying it. On the board, on the handout, anywhere it stays. A student who processes slowly, who missed the first three words, or who is rereading the text while you talk should not lose the question because it only existed as sound. It costs nothing and it changes who is able to answer.

Cotton’s synthesis adds the part that matters for seminars specifically: three seconds is the threshold for lower cognitive questions, and for higher cognitive questions there appears to be no threshold at all — students “seem to become more and more engaged and perform better and better the longer the teacher is willing to wait.”

I have tested the ceiling on this and I would not recommend the experiment. I once waited more than five minutes. It was one of the most awkward moments of my life, that section finished nine minutes behind my other classes, and it is the only time I have ever assigned homework. I am not telling that story as a technique. I am telling it because the price of waiting is real and it is worth knowing what it costs before you decide the silence is unbearable at eight seconds. Eight seconds costs you nothing.

How to Turn Your Own Text Into Questions in Ten Minutes

Read for the seams, not for the content. You are not looking for what the text says. You are looking for the places where it strains.

1. Mark every hedge. “Arguably,” “some have suggested,” “it is possible that.” Each one is a place the author was not sure, and each one is a question.
2. Find two sentences that do not quite fit together. Not a contradiction — a tension. “Are these both true?” is a seminar question that writes itself.
3. Note what is missing. Who is not quoted. Which objection is not answered. What the author had to leave out to make the piece work.
4. Write your opening question from the whole text, and three core questions from the marks you made.
5. Answer your own opening question in one sentence. If you can, go back to step two.

That is the whole process, and it gets faster. The fourth time you do it you will find yourself marking hedges while you read for other reasons, which is a better outcome than any question bank.

Questions That Wreck a Seminar

Four shapes, and all four are easy to write by accident.

• The question you already answered in class. Students will give you your own words back, correctly, and nothing will happen.
• The yes-or-no question with an obvious side. “Was the narrator wrong to lie?” gets twenty-eight yeses. Ask instead what the lie cost, and who paid.
• The question that requires a personal disclosure. “Have you ever been treated this way?” will get an honest answer from a fifteen-year-old in front of twenty-seven peers, and that is exactly the problem. Keep the question on the text and let students bring themselves to it if they choose. If a text is likely to pull the room in that direction, that is a conversation to have with an administrator before it is a seminar question.
• The question with a correct answer you will eventually supply. Students can smell this within four minutes, and once they have, the seminar becomes a guessing game about you.

The third one is the one worth being most careful with, because it usually comes from a good instinct. Relevance is not the same as exposure. A question can matter to a student’s life without requiring them to narrate it to the class.

What to Do Next

Take the text you are teaching this week and spend ten minutes on the five steps above. Write one opening question, three core questions, and one closing question. Then pull four questions from the bank here that fit the same text, so you have something in reserve you did not have to invent.

Ask your opening question once, and count to eight in your head before you do anything else. That is the whole practice. If you want the rest of the format — the norms, the seating, what to do about the students who will not talk — it is in the full guide to running a Socratic seminar. And if the questions you write keep coming out as comprehension checks, that is not a failure; it means you have a set of quick checks for understanding and a seminar question still to write.

Frequently Asked Questions

Sources

1. Cotton, Kathleen. Classroom Questioning. School Improvement Research Series, Northwest Regional Educational Laboratory, May 1988. Reports Redfield and Rousseau (1981), 20 studies, “significant and positive effect” for higher cognitive questions, against Samson et al. (1987), 14 studies, effect “small”; a three-second wait-time threshold for lower cognitive questions and no apparent threshold for higher cognitive ones; and a 60 / 20 / 20 split of lower cognitive, higher cognitive and procedural questions in ordinary recitation. https://educationnorthwest.org/sites/default/files/ClassroomQuestioning.pdf
2. Rowe, Mary Budd. Wait-Time and Rewards as Instructional Variables: Their Influence on Language, Logic, and Fate Control. 1972. ERIC ED061103. Teachers allowed an average of about one second for a response and commented within about nine-tenths of a second; extending both to three to five seconds produced longer responses, more unsolicited appropriate answers, more speculative thinking, more student-to-student evidence comparison and fewer unanswered questions. Population: elementary science classrooms. https://eric.ed.gov/?id=ED061103
3. Nystrand, Martin, and Adam Gamoran. Student Engagement: When Recitation Becomes Conversation. National Center on Effective Secondary Schools, 1990. ERIC ED323581. Eighth- and ninth-grade English classes in sixteen middle schools feeding nine high schools across eight Midwestern communities; over 1,100 students tested fall and spring. Higher proportions of authentic questions and uptake predicted significantly higher literature achievement. https://files.eric.ed.gov/fulltext/ED323581.pdf
4. Yang, Mikyeong. “A Critical Review of Research on Questioning in Education: Limitations of Its Positivistic Basis.” Asia Pacific Education Review, vol. 7, no. 2, 2006, pp. 195–204. Argues that the teacher-questioning meta-analyses “reported incoherent results” and that the evidence cannot support the expectation that higher-cognitive questions outperform fact questions. https://files.eric.ed.gov/fulltext/EJ752340.pdf
5. Applebee, Arthur N., Judith A. Langer, Martin Nystrand, and Adam Gamoran. “Discussion-Based Approaches to Developing Understanding: Classroom Instruction and Student Performance in Middle and High School English.” American Educational Research Journal, vol. 40, no. 3, 2003, pp. 685–730. 64 middle and high school English classrooms, grades 7, 8, 10, 11 and 12. https://eric.ed.gov/?id=EJ782328
6. National Paideia Center. “Paideia Seminar.” The seminar defined as a collaborative intellectual dialogue facilitated with open-ended questions about a text, run as a pre-seminar / seminar / post-seminar cycle. https://paideia.org/blogs/npc/paideia-socratic-seminar

By Clay Shumate

A Socratic seminar is a structured, whole-group discussion in which students question a shared text and each other while the teacher stays mostly silent. There is no winner and no correct answer to arrive at. The point is that students do the reasoning out loud, in front of each other, using evidence from something they all read.

That is the definition. What follows is the part most guides skip: what the research actually supports, what it does not, and how to run one in a real secondary class without the same five kids carrying it.

Key Takeaways

• A seminar is not a debate. Nobody is assigned a side, nobody is trying to win, and the question has to be one you cannot settle by looking something up.
• The strongest evidence is about discussion generally, not this format specifically. A study of 64 middle and high school English classrooms found discussion-based teaching predicted higher spring literacy performance — for low-achieving students as well as high-achieving ones.
• The most honest finding is a mixed one. A meta-analysis found discussion produced big jumps in student talk and real gains in text comprehension, but few approaches moved critical thinking and reasoning.
• Preparation is the whole game. A seminar with an unread text is twenty minutes of opinions.
• Silence is not failure. The most common way a teacher wrecks a seminar is by answering the question they just asked.

What Is a Socratic Seminar?

A Socratic seminar is a formal, text-based dialogue where students ask and answer open-ended questions with each other, and the teacher facilitates instead of teaching. The National Paideia Center — which has trained schools in this format for decades — defines it as “a collaborative intellectual dialogue facilitated with open-ended questions about a text.” That definition is doing more work than it looks like.

Every word in it rules something out. Collaborative rules out debate. Open-ended rules out questions with answers. About a text rules out a free-floating conversation on a topic. Take away any one and you have something else — possibly useful, but not a seminar. Here is the distinction that matters most in a secondary room, because teenagers assume it is a debate unless you tell them otherwise.

Does the Research Actually Support Discussion-Based Teaching?

Yes for discussion in general — and more cautiously than most professional development admits when it comes to thinking. Two studies point in slightly different directions, and both are worth knowing before you build a unit around this.

The strong one is squarely in our grade band. Applebee, Langer, Nystrand and Gamoran studied 64 middle and high school English classrooms across grades 7, 8, 10, 11 and 12. Controlling for fall performance and background, discussion-based approaches were significantly related to spring literacy performance. Their own summary of why: students in discussion-heavy rooms “internalize the knowledge and skills necessary to engage in challenging literacy tasks on their own.” And the finding held “for low-achieving as well as high-achieving students” — which is the opposite of what people assume about discussion, and the single best argument for not reserving it for honors sections.

An earlier study from the same research group found the mechanism. Across sixteen middle schools feeding nine high schools in eight Midwestern communities, more than 1,100 eighth and ninth graders were tested in fall and spring. Students whose teachers asked a higher proportion of authentic questions — questions the teacher did not already have an answer to — and who practiced uptake, building the next question out of what a student just said, scored significantly higher on literature achievement. Not more questions. Different questions.

There is an equity finding buried in the first study that deserves more attention than it gets. Applebee and colleagues note that their interpretation is “complicated because instruction is unequally distributed across tracks.” Translated: the discussion-based teaching that helps low-achieving students the most was not what low-achieving students were mostly getting. If a school is going to act on this research at all, the action is not “add seminars” — it is checking which sections currently get them.

Now the cautious one, and it is the reason this section exists. Murphy and colleagues ran a meta-analysis of classroom discussion approaches in the Journal of Educational Psychology. They found “strong increases in the amount of student talk and concomitant reductions in teacher talk, as well as substantial improvements in text comprehension.” Then they found this: “Few approaches to discussion were effective at increasing students’ literal or inferential comprehension and critical thinking and reasoning.”

Read that twice. The talk goes up reliably. Comprehension of the text goes up. Whether students get better at thinking is not established by the pooled evidence, and the majority of the studies in that analysis were conducted in fourth through sixth grade, not with teenagers. If someone tells you seminars teach critical thinking, that is a hope, not a result.

The same gap shows up in a study built specifically around Socratic dialogue. Mahoney and colleagues ran an eight-week Socratic dialogue series in Dutch vocational secondary education with 85 students and five teachers. Teachers found it deliverable but cognitively demanding, and all five asked for more facilitation training. The study’s own limitations section says there was “no measurement of actual critical thinking skill development.” A paper with “learning to think critically” in the title did not measure whether anyone did. The researchers said so themselves; it is a warning about how this format gets sold, not about them.

One more, because the number gets quoted at teachers without its age attached. England’s Education Endowment Foundation ran a randomised controlled trial of Dialogic Teaching across 76 schools and roughly 5,000 pupils and found two additional months’ progress in English and science, with larger gains for pupils on free school meals. Those pupils were nine and ten years old. The practices probably transfer to a secondary room. The effect size does not automatically come with them.

The defensible version: structured, text-anchored discussion with authentic questions is one of the better-evidenced things you can do with a secondary class, and the evidence for it is about comprehension and engagement rather than about producing better reasoners in eight weeks.

One implication for anyone above the classroom level: this is not a practice you mandate in August and audit in October. Every teacher in the Dutch study asked for more training in facilitation specifically, and facilitation is the whole skill. A department that runs four seminars and talks about them afterward will get further than a district that requires one per unit.

What Do You Need Before the First Seminar?

Three things: a text worth arguing about, a question with no answer, and students who have actually read it. Skip any one and the seminar fails in a predictable way.

The text. Short beats long. One or two pages that every student can hold and mark up beats a chapter half of them skimmed. It needs genuine ambiguity — a primary source with a point of view, a court opinion, a poem, a paragraph from a science paper where the authors hedge. If the text has one obvious reading, there is nothing to discuss and students will know it within four minutes.

Two constraints on that choice, both easy to miss when you are excited about a text. Every student has to be able to read it. A twelfth-grade primary source in a mixed ninth-grade class produces a conversation among the strongest readers while everyone else decodes. If the text is hard, read it aloud together or gloss the four words that will stop people — but do not let reading level decide who participates. And check the text against your own building’s ground. Ambiguity is the point; a text likely to pull a student into disclosing something personal in front of twenty-eight peers is a different situation, and it is worth a conversation with an administrator before it is worth a seminar.

The opening question. Write it down before the seminar and test it against one standard: could a well-prepared student answer this in one sentence and be finished? If yes, it is a comprehension check, not a seminar question. “What does the author claim?” is a warm-up. “Is the author being honest about what they do not know?” is a seminar. If writing them is the part you get stuck on, there are forty subject-general seminar questions and a ten-minute way to write your own.

The reading. This is what quietly kills most first attempts. The Paideia model puts multiple close readings in a pre-seminar phase before any discussion and writing in a post-seminar phase afterward. That three-part shape is the format, not an optional extra. Read the text once ten minutes before the circle forms and you get opinions instead of evidence — then conclude your class cannot handle seminars. Your class is fine. The prep was missing.

A note on when they read it. The obvious move is to send the text home the night before, and I would push back on that — not on principle, but because it decides in advance that the students who read at home get to participate and the ones who do not, do not. Read it in class. Ten minutes of everyone reading the same page in the same room is not lost time; it is the thing that makes the next twenty minutes possible.

One practical note about the room. The tables in my room are trapezoids pushed together in groups rather than desks in rows, which is closer to a discussion shape than a lecture hall — but it still is not a circle, and the furniture has to move. Build that into the clock. A seminar that starts with four minutes of scraping chairs starts four minutes late, every time, and you will feel it at the end.

How Do You Actually Run One?

Six moves, in order, and the hardest is the fifth. You can run this sequence in a single period and repeat it until it stops feeling like an event.

The sequence is deliberately boring in the middle. Step five is the interesting one, and the one I get wrong most often.

In my own room I have landed on roughly a fifteen-count before I respond to something that is off track — and I mean roughly. It is an average, not a rule, because real classes change day to day and a number you enforce mechanically stops being judgment. A seminar asks for that same discipline in a different situation. You ask the opening question, nobody talks, and everything in you wants to rephrase it. Rephrasing is how you teach a room that if they wait long enough, you will do the work. Ask it once. Then sit there.

The related move is refusing to evaluate. The instant you say “good point,” you have told twenty-eight teenagers that the goal is to produce points you approve of, and the conversation reorients toward you. Write the comment down instead. Nod. Look at somebody else. The Applebee study’s mechanism — authentic questions and uptake — only works if students believe you do not already have the answer, and praise is the fastest way to convince them you do.

What Are the Rules, and Who Enforces Them?

Keep the list short enough that students can hold it, and hand enforcement to them by the third seminar. Long norm lists get read aloud once and ignored. Five or six behaviors, posted, referred to by name, is enough.

The norms that matter protect the conversation from its two failure modes: nobody talking, and three people talking. Everything else is manners.

There is one behavior worth planning for: the student who is performing for the room rather than talking to it. Do not handle that inside the circle. Naming it publicly makes it a bigger performance, and the seminar stops. Let it pass, keep the conversation moving to someone else, and take it up with the student afterward at normal volume. A seminar is a bad place to win an argument with a teenager.

Worth saying clearly, because it is where teenagers get this wrong: disagreeing with an idea is the job. A seminar where everyone agrees has not happened yet. What is off-limits is going after the person instead of the claim, and the fix is a sentence frame, not a lecture — “I read that line differently, because…” does the whole job. The broader argument for why this works in a secondary room is the same one behind holding students to an adult standard of disagreement rather than a polite one.

What About the Fishbowl Format?

A fishbowl Socratic seminar puts one group in an inner circle discussing while an outer circle observes, then swaps them. It exists to solve two real problems: a class of thirty is too big for one conversation, and students who never get observed never get feedback on how they discuss.

It works, with one condition. The outer circle needs a job. An outer ring with nothing to do is a ring of spectators with phones. Give each outer student one inner-circle partner to track, and one thing to record — how many times their partner cited the text, or the one question their partner asked that moved the conversation. Keep the record behavioral: what the partner did, not how well they did it. “You went back to the text three times” is feedback. “You seemed nervous” is not, and a fifteen-year-old should not be handed that by a classmate. Hand the note to the partner rather than reading it to the room. That feedback is usually more useful than anything I would have said, because it comes from a peer who was watching only them — but it belongs to the person it is about.

Do not make it your first seminar of the year, though. Students end up learning two unfamiliar procedures at once and the period goes to logistics.

How Do You Handle the Students Who Won’t Talk?

Stop treating silence as a single problem, because it is at least three problems with different fixes. Some students have nothing prepared. Some have something prepared and cannot find an opening. Some are genuinely afraid of the sound of their own voice in a quiet room, and no amount of encouragement touches that.

A study of quiet students by Medaille and Usinger is instructive here, with a caveat stated up front: the participants were ten undergraduates, not teenagers, so treat it as a description of an experience rather than a finding about your seventh period. Nine of the ten struggled with instructor expectations for verbal participation. Six reported physical reactions — trembling, blushing, stuttering — when speaking aloud. And one of them described the thing I now believe is the actual mechanism: “When I speak [in class] I plan out what I want to say.”

That is not shyness. That is a student who needs a draft, and the fix is not calling on them warmly — it is giving the whole room ninety seconds to write before the circle opens, so the students who need a draft have one and nobody is singled out for needing it.

Three more moves that cost nothing:

• Give the quiet student a job that is not an opinion. Tracking which page the group keeps returning to, or reporting the one question nobody answered, is a real contribution that does not require volunteering a view.
• Use written entry. A sticky note with one question on it, collected at the door and read aloud anonymously, gets a reticent student’s thinking into the room without their voice attached to it. Read them yourself before you read any of them out. Anonymous is not the same as safe — a question can identify the student who wrote it, or disclose something that should not be discussed in a circle, and the two seconds it takes to scan the stack is the whole safeguard.
• Do not grade volume. More on that next, because it is the single change that most reliably changes who talks.

I would not promise that a seminar fixes this. Some students will speak twice all year and write the best post-seminar reflection in the class. That is a real outcome, and the format should be built to catch it.

Should You Grade a Socratic Seminar?

Grade the preparation and the reflection. Do not grade how many times a student spoke. The moment participation counts, you have created an incentive to talk rather than to think, and the students who most need the practice are the ones the incentive punishes.

Count what you can defend. Did they annotate the text? Did they bring a written question? Does the post-seminar writing show they changed or sharpened a position — and can they name whose comment did it? That last one is the best assessment item this format produces, because only a student who was listening can answer it. Three comments in a circle are easy to fake. “I came in thinking X, and the point about the second paragraph is why I do not think that anymore” is not.

This is also the answer when a parent asks why their quiet child is not being penalized, or why their talkative one is not being rewarded: the grade is on the thinking you can see on paper, and it is the same standard for everyone in the room. If your school requires a participation grade, take it from the written pre- and post-work and say out loud, to the students, that the talking is not scored. There is a four-criterion rubric that scores it this way, free and printed on the page. Then hold to it. The first seminar after they believe you is a different conversation. This is the same logic behind treating ungraded checks as information rather than points — the second you attach a score, you stop learning what students actually think.

Four Ways a Seminar Goes Wrong

Every one of these is a design problem, not a student problem. If a seminar dies, the autopsy is almost always in the prep.

1. The question had an answer. Students find the answer in six minutes and then sit there. Fix: write the question the night before and try to answer it yourself in one sentence. If you can, it is not the question.
2. The teacher kept talking. You clarified, you rephrased, you summarized, and the conversation routed through you every time. Fix: count your own turns. More than four in a thirty-minute seminar and you are running a recitation with the chairs moved.
3. Three students ran it. Not because they are hogging it — because they are the only ones who prepared. Fix: require a written question from everyone, collected as they come in. Make it an expectation, not a bar — a student who arrives without one writes it in the first two minutes while the furniture moves. Nobody gets locked out of the conversation for being unprepared; they get two minutes and then they are in it.
4. It happened once. A seminar in October and a seminar in March are two events. Students never get past the awkwardness of the format itself. Fix: shorter and more often. Fifteen minutes on a half-page every other week will outperform two forty-minute showpieces.

The fourth is the one I would push hardest. Frequency turns this from a special activity into the way the class talks, and the same logic runs through any structure that hands students the thinking: the first run is about the structure, and only the fourth or fifth is about the content.

What to Do Next

Pick a one-page text you already teach — something with a hedge or a contradiction in it — and write one question you cannot answer in a sentence. Give students ten minutes in class to read it, mark two places they disagree, and write one question. Move the furniture. Ask your question once, and then do not talk for as long as you can stand.

Then do it again in two weeks. The first one will be rough, and that is not information about your students. Run four before you decide whether this works in your room.

If you want the underlying argument for why handing teenagers this much of the conversation is worth the mess, it is the same one behind giving students real decisions rather than cosmetic ones. A seminar is just the version of that argument you can run on a Tuesday.

Frequently Asked Questions

Sources

1. Applebee, Arthur N., Judith A. Langer, Martin Nystrand, and Adam Gamoran. “Discussion-Based Approaches to Developing Understanding: Classroom Instruction and Student Performance in Middle and High School English.” American Educational Research Journal, vol. 40, no. 3, 2003, pp. 685–730. 64 middle and high school English classrooms, grades 7, 8, 10, 11 and 12; discussion-based approaches significantly related to spring performance controlling for fall performance, and effective for low-achieving as well as high-achieving students. https://eric.ed.gov/?id=EJ782328
2. Murphy, P. Karen, Ian A. G. Wilkinson, Anna O. Soter, Maeghan N. Hennessey, and John F. Alexander. “Examining the Effects of Classroom Discussion on Students’ Comprehension of Text: A Meta-Analysis.” Journal of Educational Psychology, vol. 101, no. 3, Aug. 2009, pp. 740–764. Strong increases in student talk and substantial improvements in text comprehension; “few approaches to discussion were effective at increasing students’ literal or inferential comprehension and critical thinking and reasoning.” Majority of included studies were conducted in grades 4–6. https://eric.ed.gov/?id=EJ861185
3. Nystrand, Martin, and Adam Gamoran. Student Engagement: When Recitation Becomes Conversation. National Center on Effective Secondary Schools, 1990. ERIC ED323581. Eighth- and ninth-grade English classes in sixteen middle schools feeding nine high schools across eight Midwestern communities; over 1,100 students tested fall and spring, 1987–88 and 1988–89. Higher proportions of authentic questions and uptake predicted significantly higher literature achievement. https://files.eric.ed.gov/fulltext/ED323581.pdf
4. Jay, Tim, et al. Dialogic Teaching: Evaluation Report and Executive Summary. Education Endowment Foundation, July 2017. ERIC ED581114. Three-level clustered randomised controlled trial, 38 intervention schools (2,492 pupils) and 38 control schools (2,466 pupils); +2 months English, +2 months science, +1 month maths; three-padlock security rating. Pupils were in Year 5, aged nine and ten — this is not a secondary-school result. Evaluators note 21% of pupils excluded from analysis. https://files.eric.ed.gov/fulltext/ED581114.pdf
5. Mahoney, Bethany, Ron Oostdam, Hessel Nieuwelink, and Jaap Schuitema. “Learning to Think Critically Through Socratic Dialogue: Evaluating a Series of Lessons Designed for Secondary Vocational Education.” Thinking Skills and Creativity, vol. 50, 2023, article 101422. 85 students and 5 teachers, Netherlands, eight weekly lessons; teachers found the format deliverable but demanding and all five requested more facilitation training. The study states as a limitation that there was no measurement of actual critical thinking skill development. https://www.sciencedirect.com/science/article/pii/S1871187123001906
6. Medaille, Ann, and Janet Usinger. “Quiet Students’ Experiences with the Physical, Pedagogical, and Psychosocial Aspects of the Classroom Environment.” Educational Research: Theory and Practice, vol. 31, no. 2, 2020, pp. 41–55. Qualitative study of ten upper-division undergraduates — not secondary students; nine of ten struggled with verbal participation expectations and six reported physical symptoms when speaking aloud. https://files.eric.ed.gov/fulltext/EJ1274336.pdf
7. National Paideia Center. “Paideia Seminar.” Definition of the seminar as a collaborative intellectual dialogue facilitated with open-ended questions about a text, and the pre-seminar / seminar / post-seminar cycle. https://paideia.org/blogs/npc/paideia-socratic-seminar