How to Run a Socratic Discussion That Doesn't Suck

Most Socratic discussions suck.

Not because the idea is bad. The idea is great. Students sit with an important question, listen carefully, make claims, test assumptions, deal with disagreement, and sometimes realize they were wrong. That is real education.

The problem is that what many schools call a "Socratic Discussion" is just a dressed-up participation exercise.

Students are told to "discuss," which usually means a few confident kids start talking, a few others try to sound smart, somebody repeats something from the text, and everyone else waits for it to be over. The teacher sits there pretending this is rigorous because the desks are in a circle.

It isn't.

A good Socratic Discussion is not just students talking. It is students thinking together, using questions to clarify ideas, uncover assumptions, test reasons, consider objections, and following an argument where it actually leads.

What a Socratic Discussion Actually Is

A real Socratic Discussion is a structured discussion built around an open, important question. Usually there is a common text, case, problem, or claim that gives the discussion something solid to work with. Students are supposed to do more than offer opinions. They are supposed to give reasons, respond to other people's ideas, and refine their thinking.

The goal is not for students to say what they already think. The goal is for them to examine what they think.

That distinction matters more than any protocol you run.

What It Actually Looks Like

I teach Philosophy and Critical Thinking to seventh graders. The first question I ask when we start the epistemology unit is this:

What does it mean for a claim to be true?

That is it. One question. And we can spend the entire class period on it. Forty-five minutes, sometimes more.

Here is what that looks like in practice. A student says, "Something is true if you believe it's true." A weaker discussion moves on. Someone agrees, someone disagrees, the teacher nods and calls on the next hand. A Socratic discussion does not move on. Instead, someone follows up: So everything you believe is true? Or more specifically: If you believe it's going to rain tomorrow, does that make it true that it will rain?

Now the students have to think. The definition they offered sounded fine until it was tested against a real case. Students work toward a definition, test it against examples, discover where it breaks down, revise it, and test it again. The other questions (what it means for a claim to be false, what it means for a claim to be a lie) come later in the unit, and each gets the same treatment.

I guide the process, but I do not hand them conclusions. We build the definitions together, and we come back to them every time something relevant comes up for the rest of the year.

The point is not efficiency. Forty-five minutes or more on the word "true" sounds like a terrible use of class time until the third week of the unit, when a student says, "Wait. Is that actually true, or do we just all believe it?" and you realize they are using the definition we built together. That does not happen by accident.

Why Most Discussions Go Bad

Usually for predictable reasons.

They start with a weak question

If the opening question has one obvious answer, the discussion is dead before it starts. A good Socratic question creates real intellectual pressure, something worth wrestling with.

Compare:

Bad: "What is confirmation bias?" Better: "Why is confirmation bias so hard to notice in ourselves?" Better still: "If we are biased in evaluating evidence, what should justify our confidence in any of our beliefs?"

The third question has teeth. It opens into epistemology, psychology, intellectual humility, and method. The first question is a vocabulary check.

The quality of the discussion depends heavily on the quality of the question. A shallow question produces a shallow discussion, regardless of what protocol surrounds it.

Students have not been taught how to discuss

Teachers often say students are bad at discussion. Usually what they mean is students are bad at doing something they were never actually taught.

We would never assign an analytical essay with no modeling, no examples, no criteria, and no feedback. But that is basically how most discussions run. Students are told to "have a discussion" as if good discussion is natural.

It isn't.

Students have to learn how to listen without just waiting to talk, disagree with a claim without attacking the person, ask a follow-up question that actually moves things forward, and use evidence instead of vibes. Those are learned behaviors. If you care about this kind of thinking, you have to teach them deliberately.

Give students language to work with:

* "Can you say more about what you mean?"

* "I want to challenge the assumption that..."

* "That seems true in this case, but what about this one?"

* "I agree with the conclusion but not the reasoning."

* "Can we distinguish those two ideas?"

These are not training wheels. They are tools. Students who have them run better discussions.

The teacher rescues the discussion too quickly

The teacher asks a real question. Students pause. Teacher panics.

Then comes the re-wording, the hinting, the gentle redirect toward the student who always answers, the subtle pressure to produce something so the silence goes away.

That kills the thinking.

If a student is trying to answer something like "What makes a belief justified?", you should want slowness. Fast answers to hard questions are usually just prepackaged opinions dressed up as insight. The pause is not the problem. The pause is where the thinking begins.

Participation gets confused with quality

Some students can talk a lot while doing almost no thinking. Others say one sentence that clarifies the whole issue.

The real questions to ask yourself after a discussion: Did students give reasons? Did they respond to each other rather than just take turns? Did anyone identify an assumption? Did anyone revise a claim? That is the work. Volume is not.

The Skill Nobody Teaches: Asking Better Questions

If you want a classroom built on Socratic thinking, students cannot remain passive recipients of your questions. They have to learn how to generate their own.

Not fake school questions. Not prompts they write because they were told to come up with three. Real questions: the kind that expose uncertainty, reveal confusion, and open inquiry instead of closing it.

Model what that looks like.

In philosophy, the difference between a weak question and a strong one usually comes down to whether it creates a genuine fork in the road:

"What is knowledge?" A definition question with a Wikipedia answer. "If someone arrives at the truth for bad reasons, do they know it?" Now you have a discussion.

In history: "Was this decision wrong?" is less useful than "Are we evaluating the decision based on what they knew, or what we know?"

In science: "What did the experiment show?" matters less than "What conclusion does this evidence actually support, and what conclusion would go beyond it?"

Same underlying move in every discipline: push past recall toward judgment.

A good classroom is not one where students can answer your questions. It is one where they have learned how to generate better ones.

Slow Down

This is the simplest and most ignored piece of advice in teaching.

Slow down.

Most teachers are afraid of silence because silence feels like failure. In discussions about complex ideas, silence usually means students are trying to think rather than just perform. That is a good thing, not a problem to solve.

Ask the question. Let it sit. Make students live in it for a few seconds.

Then, after a student responds, slow down again. This is just as important.

A lot of teachers hear something interesting and immediately interpret it, polish it, or redirect it. When you do that too fast, students learn that their job is to toss out a thought and wait to be told what it means. Instead, let the room deal with the response. Let someone build on it. Let someone push back. Let the idea become common property.

That is how discussion becomes collective thinking rather than serial answering.

I have spent multiple class periods on the question of what it means for a claim to be true, with middle school students, and not once has it felt wasted. The slowness is not a failure of pace. It is the point.

What You're Actually Teaching

If you care about Socratic thinking, you are teaching more than content.

You are teaching students how to ask better questions, give reasons, notice assumptions, tolerate uncertainty, evaluate evidence, distinguish stronger arguments from weaker ones, and revise their thinking without it feeling like defeat.

A discussion should not be a performance of "engagement." It should be a disciplined attempt to think more carefully with other people.

When it works, students do not leave thinking, "That was fun." They leave thinking, "That was harder than I expected," or "I had not thought about it that way," or "I am less sure than I was, but in a better way."

That is the goal. Not more discussion. Better thinking.

If students are learning to ask stronger questions, take reasons seriously, slow down before speaking, and treat discussion as a way of testing ideas rather than displaying themselves, then the discussion is doing something real.

And if that starts happening regularly, not just during formal discussions but across your class discussions in general, then you are not just running a Socratic Discussion.

You are teaching Socratically. That is the better thing to be doing.

*

Read more: 5 Critical Thinking Activities That Actually Work in Middle School

*

Try ThinkingEngine free — AI-powered Socratic dialogue for your classroom at thinkingengine.org