Why Socratic Dialogue Is the beating Heart of Theory of Knowledge

Theory of Knowledge is the one IB subject that doesn't teach content. It teaches thinking about thinking. That's a strange thing to teach — and an even stranger thing to assess.

The IB's solution is the TOK Exhibition and the TOK Essay. Both ask students to do something deceptively hard: show how they know what they claim to know. That means identifying assumptions, weighing competing knowledge claims, and saying something true about the nature of knowledge itself — not just demonstrating knowledge of a subject.

Nothing builds that skill like Socratic dialogue. Not lecture. Not reading comprehension. Not a worksheet. Direct, structured dialogue where a student has to defend a claim, qualify it under pressure, and revise it in real time.

The challenge for TOK teachers: good Socratic questions are hard to write. They need to be genuinely open (no "correct" answer), specific enough to generate real discussion, and connected enough to the TOK framework that they serve the course. Most prompt books give you generic philosophical questions. What you need are questions your students can actually get somewhere with.

That's what these are.

10 Socratic Discussion Prompts for TOK

Each prompt below is organized around a TOK theme area. For each one, I've included:

1. Knowledge & the Knower

Prompt: "If you grew up in a completely different culture, would you still believe the things you believe now? How would you know?"

Why it works: This question surfaces the distinction between knowledge that is genuinely justified and knowledge that is locally conditioned. Students quickly distinguish between things that feel universal ("murder is wrong") and things that feel culturally relative. That distinction — between the necessary and the contingent — is exactly the kind of thinking the TOK essay rewards.

Follow-up: "You just said [X] is a universal truth. What would it look like if that belief were genuinely culture-specific? Can you imagine a context where it wouldn't hold?"

2. Knowledge & Technology

Prompt: "Wikipedia is more accurate than any individual encyclopaedia — but does that mean we know more than people did before Wikipedia? What's the difference between more information and more knowledge?"

Why it works: Students assume the internet has made us smarter. This question forces them to distinguish access to information from the development of understanding. The Socratic payoff: many students realize that "knowing more facts" and "knowing better" are different things — and that the conditions for knowledge include things technology doesn't provide (judgment, context, the willingness to sit with uncertainty).

Follow-up: "You said knowledge requires understanding, not just facts. Can an AI have understanding? How would you test for it?"

3. Knowledge & Language

Prompt: "Some concepts in physics — 'force,' 'energy,' 'mass' — don't translate cleanly into most languages. Does that mean speakers of those languages have less access to knowledge of physics? Or does it mean something different about what it means to 'know' physics?"

Why it works: This is a direct entry point into the relationship between language and knowledge, one of the core TOK knowledge frameworks. Students who initially say "physics works the same everywhere" are pushed to articulate what they mean by "works" — and what they find is that the answer involves more than operational success.

Follow-up: "You could calculate the answer correctly without having an intuition for what the concept means in your first language. Is that still knowledge? What's missing?"

4. Knowledge & Society

Prompt: "Whose knowledge counts as knowledge in a democracy? And is that the same question as 'whose knowledge is most useful'?"

Why it works: This connects the human sciences framework to questions about power and knowledge. Students see that democratic societies don't distribute epistemic authority equally — experts are granted status that non-experts don't have — and that this is both necessary and potentially problematic. It's a productive tension without a clean resolution, which is exactly what the TOK essay requires.

Follow-up: "Are there cases where a democracy should defer to expertise even when the majority disagrees? Who decides when that's appropriate?"

5. Knowledge & Ethics

Prompt: "If a neuroscientist could prove that free will doesn't exist, would that change anything about how we should treat criminals? Why or why not?"

Why it works: This question sits at the intersection of natural sciences (the neuroscience), ethics (how we should treat people), and the knower (whether our sense of moral responsibility depends on believing we have free will). Students who say "no, we should still punish crime" must articulate what their moral reasoning depends on. That often surfaces assumptions students didn't know they were making.

Follow-up: "You said [X] would still matter even if free will doesn't exist. Where does that sense of [X] come from — reason, intuition, social agreement? And if it's from one of those, does it survive the neuroscientist's proof?"

6. Knowledge & History

Prompt: "We can never know the past directly — only through traces and interpretations. Is history a knowledge discipline, a humanity, or something else? What does it need to be in order to count as knowledge at all?"

Why it works: The status of history within the TOK framework is genuinely contested, and that's the point. Students who uncritically treat historical knowledge as certain discover this question is harder than they expected. The challenge is that history is both evidence-based (satisfying) and interpretation-dependent (troubling).

Follow-up: "You said history can be knowledge if it meets [certain criteria]. Who gets to decide whether those criteria are met — historians, philosophers, the public? And does the answer change depending on who decides?"

7. Knowledge & Mathematics

Prompt: "Mathematical objects — triangles, prime numbers, pi — existed before humans did. Does that mean mathematics is discovered, not invented? If so, where do mathematicians 'go' to discover them?"

Why it works: This question cuts to the heart of one of the deepest disputes in the philosophy of mathematics: platonism versus constructivism. Students who say "math is discovered" face the uncomfortable implication that mathematicians have some kind of access to a non-physical realm. Students who say "math is invented" must explain why it works so well in describing the physical world. There's no comfortable position here, which is exactly right.

Follow-up: "You said [math is discovered/invented]. Then if an alien species existed, would they have the same mathematics? What does your answer imply about the nature of mathematical truth?"

8. Knowledge & the Arts

Prompt: "Can art communicate knowledge that other areas of knowledge — natural sciences, mathematics, history — cannot? If so, what kind of knowledge is it, and how would you know if it were true?"

Why it works: This question asks students to articulate what they think art is for, which is harder than it sounds. Most students have strong intuitions about art's value without having thought carefully about whether those intuitions are epistemic (about knowledge) or purely aesthetic (about experience). Making that distinction explicit develops precisely the kind of reflective language the TOK essay demands.

Follow-up: "You said art can show us [X]. But can it prove X? What's the difference between showing and proving — and does that difference mean art is less valuable, or just different?"

9. Knowledge & Natural Sciences

Prompt: "A scientist runs the same experiment twice and gets different results. She blames measurement error. A humanist says: 'maybe there's no error — maybe the phenomenon itself is indeterminate.' Who is right, and how would you know?"

Why it works: This question introduces students to the concept of measurement uncertainty and the philosophy of science, while connecting to issues in statistics (what does it mean to replicate a result?) and even quantum physics (indeterminacy as a feature, not a bug). It's a productive anchor for discussing what scientists mean when they say they "know" something.

Follow-up: "You said [the scientist/the humanist] is more likely to be right. But what standard of evidence are you applying? Is that the same standard the natural sciences use? Is it the same standard that works in other areas of knowledge?"

10. Knowledge & Politics

Prompt: "During the pandemic, different countries knew different things about COVID — not because the science was different, but because their governments chose which knowledge to distribute. Is political knowledge a real category, or just a polite way of describing manipulated information?"

Why it works: This question connects to core TOK concerns about the social dimension of knowledge: who decides what counts as legitimate knowledge, and whether that decision can be separated from power. Students immediately recognize that governments, corporations, and institutions make knowledge claims — but whether those claims are epistemically legitimate requires careful analysis beyond "it's all propaganda."

Follow-up: "You said [some political knowledge claims are legitimate, some aren't]. How do you tell the difference? Is that distinction itself politically neutral, or are you just applying one ideology's standard?"

How to Facilitate These Prompts Effectively in a TOK Classroom

ThinkingEngine helps teachers run Socratic discussions at scale. See how it works →

Socratic dialogue is easy to do badly and hard to do well. Here's what to watch for.

Start with written reflection, not spoken discussion.

Before any dialogue, give students 3-4 minutes to write a one-sentence response to the prompt. This does two things: it commits them to a position (so they can't just agree with whoever speaks first) and it surfaces confusion they can then test against others' reasoning. The written record also gives you something to return to at the end of class for debrief.

Hold the follow-up. Then hold it again.

The biggest mistake in Socratic facilitation is letting students answer the follow-up and then moving on. A good Socratic dialogue is a chain of follow-ups, each one pressing slightly deeper than the last. When a student says "I disagree because..." — the next question is "but what would change your mind on that?" When they answer that — "if we found evidence that..." — the next question is "but how would that evidence be produced? Who would accept it?"

Five to seven rounds of this kind of persistent questioning is far more valuable than five to seven different students saying five to seven different things.

Name the reasoning moves out loud.

After the discussion, debrief on how people argued — not just what they said. Did anyone change their position? What made them do it? Did anyone raise a point that reframed the question entirely? These meta-level observations are what students carry forward to the TOK essay.

Using Socratic Dialogue at Scale with ThinkingEngine

The prompts above work beautifully in a seminar setting. The challenge: a 50-minute period gives you one round of dialogue. A full TOK course gives you 150+ class sessions.

ThinkingEngine is built to run Socratic dialogue one-on-one, at scale. You can assign a specific prompt as a pre-class activity — each student works through it individually, with an AI dialogue partner that follows their reasoning and pushes back. You get a session transcript for every student, which means formative assessment and targeted discussion prep in one step.

It's the difference between saying "let's discuss this" to 30 students and having 30 genuine, individualized dialogues before class ever starts.

Try a free Socratic session — no account required to start, works in any browser.

Related Articles

Ready to bring critical thinking into your classroom?

ThinkingEngine guides students through Socratic dialogue — questions that build reasoning, not recall. Free to start, no setup required.

Start Free →