TOK M27 essay prompt 3 asks: “In the production of knowledge, how can knowledge be reliable if it is built on assumptions? Discuss with reference to the human sciences and one other area of knowledge.”
The strongest response argues that assumptions do not automatically make knowledge unreliable. Reliability depends on whether assumptions are explicit, justified, tested where possible, limited in scope, and open to revision. Mathematics provides a useful second area of knowledge because its assumptions establish formal systems, whereas assumptions in the human sciences usually concern changing human behaviour.
Students should confirm the exact prescribed title using the official document supplied by their school and reproduce it exactly. The TOK essay is a 1,600-word, externally assessed task, so this guide should support your planning rather than replace your own argument.
What TOK M27 essay prompt 3 is asking
The title presents an apparent contradiction. If an assumption is accepted as a starting point, why should knowledge built upon it deserve trust?
Your task is not to claim that every assumption is dangerous or that assumption-free knowledge is possible. Instead, investigate when assumptions support reliable knowledge and when they distort it.
Four terms require careful interpretation:
-
Production of knowledge means the processes through which claims are formed, tested, criticized, and accepted by a knowledge community.
-
Assumptions include starting propositions, definitions, simplifications, axioms, and background methodological commitments.
-
Reliable knowledge is knowledge that can be depended upon within an appropriate scope because it is consistently produced and adequately justified.
-
Built on suggests that assumptions are foundational. Changing a central assumption may change the resulting conclusion.
Reliability is not identical to absolute certainty. A probabilistic psychological finding may be dependable within a defined population, while a mathematical theorem may be certain within an axiomatic system without accurately describing reality.
A focused underlying knowledge question is:
To what extent does reliability depend on the truth of assumptions rather than on the methods used to expose and manage them?
A defensible thesis
A strong thesis should be conditional:
Knowledge can remain reliable when built on assumptions because reliability comes from making important assumptions visible and subjecting their consequences to appropriate scrutiny. In the human sciences, this requires empirical testing, representative evidence, replication, and restricted generalization. In mathematics, conclusions can be deductively reliable relative to accepted axioms, although applying them to reality requires additional empirical assumptions.
This position distinguishes internal reliability from external reliability. Internal reliability asks whether a conclusion follows correctly from its starting points. External reliability asks whether those starting points represent the world to which the conclusion is applied.
Human sciences: assumptions organize inquiry
The human sciences include psychology, economics, sociology, and anthropology. The official IB TOK subject brief emphasizes that areas of knowledge use different methods and perspectives to produce and evaluate knowledge.
Human scientists cannot investigate behaviour without assumptions. Researchers assume that concepts such as intelligence, poverty, motivation, or well-being can be defined and measured. They also make assumptions about whether samples represent populations and whether findings can be generalized beyond their original context.
Such assumptions can be productive because they reduce complexity enough to support hypotheses and comparable evidence. Economics, for example, may model rational choice or stable preferences. These assumptions do not describe every decision literally, but they create a baseline against which behaviour can be predicted and deviations investigated.
Daniel Kahneman and Amos Tversky challenged simplified accounts of rational decision-making. As explained in Kahneman’s Nobel lecture, prospect theory treats decisions as reference-dependent and sensitive to perceived gains and losses. Knowledge progressed because existing assumptions produced predictions that could be tested and qualified.
However, hidden assumptions can undermine reliability. Henrich, Heine, and Norenzayan’s review of WEIRD research populations warned that participants from Western, educated, industrialized, rich, and democratic societies could not automatically represent humanity. The central problem was unsupported generalization, not the mere presence of assumptions.
The Open Science Collaboration replication project also found that replication effects across 100 psychology studies were generally smaller than the original effects. Although interpretations of replication rates remain debated, the project demonstrates how transparency and repeated testing can expose fragile assumptions. Human-science knowledge is often probabilistic, provisional, and context-sensitive, but those qualities need not make it unreliable when claims are appropriately limited.
Mathematics: assumptions establish systems of proof
Mathematical assumptions are commonly expressed through definitions and axioms. Once a formal system has been established, deductive proof determines what follows from it.
Euclidean geometry illustrates this structure. Altering assumptions about parallel lines produces non-Euclidean geometries, showing that Euclidean foundations are not the only logically possible ones. The Stanford Encyclopedia of Philosophy’s account of geometry explains the philosophical importance of these developments.
The existence of multiple geometries does not make geometry unreliable. A theorem can be reliably established relative to stated axioms because other mathematicians can inspect each inference and reproduce the proof. Its underlying form is conditional: if the axioms hold, the theorem follows.
Nevertheless, mathematical reliability has boundaries. Gödel’s incompleteness theorems show, under specific technical conditions, that sufficiently expressive consistent formal systems cannot prove every arithmetical truth within themselves. Such a system also cannot establish its own consistency entirely through its internal resources. The Encyclopedia of Mathematics overview summarizes these limits.
Gödel does not demonstrate that ordinary mathematics is untrustworthy. Instead, the example shows that proof operates within formal frameworks whose adequacy cannot always be established completely from inside those frameworks.
Pure and applied mathematics require different judgments
Pure mathematics can be internally reliable even when its axioms are not empirical descriptions. Applied mathematics introduces another requirement: the formal structure must correspond sufficiently well to reality.
An economic model may use flawless algebra while assuming complete information, stable preferences, or independent variables. If those assumptions fail, the calculations remain mathematically valid while the prediction becomes empirically unreliable.
| Area of knowledge | Typical assumptions | Main reliability test | Central limitation |
|---|---|---|---|
| Human sciences | Representative samples, measurable concepts, behavioural regularities | Empirical testing, replication, triangulation | People differ across cultures and contexts |
| Pure mathematics | Axioms, definitions, logical rules | Deductive proof and consistency | Conclusions remain conditional on the system |
| Applied mathematics | Formal structures adequately represent reality | Predictive success and observation | Valid calculations may use unrealistic models |
The comparison reveals that assumptions serve different functions. In pure mathematics, an axiom defines a system rather than reporting an observed fact. In the human sciences, assumptions about people must confront evidence when claims are intended to explain actual behaviour. RevisionDojo’s guides to the five TOK areas of knowledge and mathematics as an area of knowledge explain this distinction further.
A practical essay structure
A balanced structure should develop comparison throughout rather than presenting two disconnected mini-essays.
| Section | Purpose | Possible example |
|---|---|---|
| Introduction | Define assumptions and reliability; establish a conditional thesis | Contrast behavioural models with axioms |
| Human sciences claim | Show how simplification makes inquiry possible | Rational-choice models and prospect theory |
| Human sciences qualification | Examine hidden assumptions and overgeneralization | WEIRD samples or replication |
| Mathematics claim | Explain how axioms enable deductive reliability | Euclidean and non-Euclidean geometry |
| Mathematics qualification | Separate proof from real-world application | A mathematically valid behavioural model |
| Conclusion | Compare the AOKs and answer the title directly | Introduce no new example |
After every example, identify the assumption, explain how it affects knowledge production, and evaluate the mechanism used to test it. RevisionDojo’s TOK essay structure guide and guide to examples and evidence can help turn this plan into a coherent argument.
Common mistakes and final checks
Avoid treating every assumption as an unsupported guess. An assumption may be an axiom, definition, idealization, or provisional empirical claim, and each requires a different form of evaluation.
Do not equate reliability with certainty or claim that mathematics proves its axioms from nothing. Proof establishes what follows from axioms; application requires separate evidence that the assumptions fit reality.
Before submitting, check that:
-
Each example addresses the production of knowledge.
-
Both areas of knowledge receive substantial analysis.
-
Claims and counterclaims lead to reasoned judgments.
-
You compare internal validity with external applicability.
-
Your conclusion identifies when assumptions deserve trust.
RevisionDojo’s guide to improving a TOK essay provides a useful final revision process.
Conclusion
Knowledge can be reliable despite being built on assumptions because assumptions are unavoidable starting points, not automatic defects. In the human sciences, reliability requires empirical, methodological, and cross-cultural scrutiny. In mathematics, explicit axioms support powerful deductive conclusions, although real-world applications need additional justification. RevisionDojo’s TOK resource hub, Jojo AI, and Coursework Review can help you test the coherence of your argument while preserving your own reasoning and wording.