Six titles, 1,600 words, two thirds of your TOK grade. You will spend longer choosing than you think, so here is each one unpacked: what it is really asking once you get past the phrasing, the words you have to pin down, the argument and the strongest objection to it, and examples that are specific enough to be worth writing about.
Two rules before you start. The areas of knowledge named in a title are a requirement, not a suggestion. And copy the title exactly, because paraphrasing it in your introduction is where most essays quietly become about something easier.
1. The artist or the historian on human experience
The word doing the work is appreciation, not accuracy. A historian can be right about a village in 1943 and leave you with no sense of being there; a novelist can invent the village and leave you unable to stop thinking about it. Decide what appreciation means to you and the essay writes itself.
- appreciation
- Emotional recognition, factual grasp, or a sense of scale? Pick one.
- human experience
- Individual inner life, or the collective patterns no individual can see?
- For: art gives you interiority. Guernica does not tell you the casualty figure; it shows what a bombing does to bodies.
- Against: art is made, so nothing can be checked. History is constrained by evidence, and constraint is what makes a claim worth something.
2. Changing values that still guide knowledge
The title concedes that values change and asks how guidance survives it. Do not spend 600 words proving they change. The move that works is separating a value from how it is applied: honesty has guided enquiry for centuries, while what counts as honest practice in a lab has moved a long way.
- values
- Split epistemic (accuracy, replicability) from ethical (consent, fairness). They change differently.
- guide
- Does a moving standard still constrain, or slide into relativism?
- For: the value holds still while the practice moves. Preregistering a hypothesis was not expected twenty years ago and is becoming standard now.
- Against: if values change, why is the current set better rather than just later? You need an answer that is a reason, not a fashion.
3. Reliability built on assumptions
The title grants that knowledge rests on assumptions, so explain how reliability survives that rather than denying it. Two angles: stated and tested assumptions versus buried ones, and what reliability even means.
- assumption
- Stated and testable, or invisible until someone names it?
- reliability
- Foundations certainly true, or results that hold within a known range?
- For: Newtonian mechanics rests on assumptions we know are false and still lands spacecraft, because its range of validity is understood.
- Against: buried assumptions cause real failures. WEIRD sampling meant psychology generalised from undergraduates for decades before anyone said so.
4. The Agatha Christie coincidence claim
Christie describes a two-stage method: notice everything, discard later. Your essay evaluates it. The key word is any, which puts the whole burden of quality control on the discarding stage. It maps onto how ideas are discovered versus how they are justified.
- coincidence
- A pattern with no reason attached to it yet.
- noticing
- Free, Christie assumes. Is it?
- For: mathematics has the strictest filter there is, so it can afford to be permissive about where conjectures come from. Ramanujan worked this way.
- Against: where the filter is statistical it leaks. Test enough hypotheses and something crosses the threshold by chance, which is what p-hacking is.
5. How much simplicity should count
Occam's razor on trial. Refuse to treat simplicity as one thing: parsimony, elegance and ease of use come apart constantly. Then ask whether simplicity tracks truth or just suits the knower.
- simplicity
- Fewer entities, a neater form, or easier to compute with?
- prioritized
- Ranked against what? Usually explanatory scope or fit with evidence.
- For: overfitting is a precise version of the argument: a more complex model predicts new data worse, so simplicity earns its place rather than being asserted.
- Against: the Standard Model is not simple by any measure and nobody counts that against it. Monocausal accounts of the First World War are simple and poor history.
6. The value of the paradoxical
The title assumes there is value and asks what kind, so argue about the kind. Be specific: a paradox is diagnostic. It marks a place the current framework cannot hold something, which is different from a puzzle the framework could solve given time.
- paradox
- Veridical (surprising but true), falsidical (hidden mistake), or antinomy (a real contradiction).
- counterintuitive
- Weaker. It only means our expectations, formed at human scale, are wrong.
- For: Einstein, Podolsky and Rosen offered a paradox as an objection, Bell made it testable, and the experiments won the 2022 Nobel Prize.
- Against: not every paradox is productive. Some dissolve into linguistic confusion, and the time spent is gone.
Choosing, and what examiners reward
The title you can write well is the one you already have examples for. Read all six, then write down two real examples for each without looking anything up. Whichever title gives you four, pick that one.
- Copy the title exactly onto your title page and do not paraphrase it anywhere.
- Define the load bearing word in your first 150 words, and hold that definition to the end.
- Use the areas of knowledge the title names. A brilliant essay on the wrong two scores badly.
- Give the opposing view its strongest form, then say why you still disagree.
- Cut the opening about how knowledge has always fascinated humanity. It costs 80 words and says nothing.
Two or three examples used properly beat six mentioned in passing. If you could swap an example for a different one without changing the argument, it is decoration.
In short
- Choose the title you have specific examples for, not the one that sounds interesting.
- Define the load bearing word early and hold it to the end.
- Treat the named areas of knowledge as a requirement.
- Give the counterargument its strongest form before you answer it.
- Take a position. A qualified answer beats an even handed survey.