Developing a hypothesis to frame your investigation
Why the process starts here
- The Explanatory Notes list developing a hypothesis or theory first, before selecting critical texts.
- A hypothesis is a claim you make in advance and then test. It is not a topic, not a question, and not a summary of what you expect to find.
- It frames the investigation, deciding which critical texts are worth reading, which parts of them matter, and what would count as an answer. Without one, searching has no stopping rule.
- A theory can do the same job. The standard says hypothesis or theory, and a critical lens chosen in advance is a fixed set of questions.
What makes a hypothesis usable
- It makes a claim. How does the story present the sea wall? is a question, and a question cannot be wrong. The story presents the wall as a decision already made elsewhere can be wrong, which makes it testable.
- It is specific to your primary source. A claim true of most texts of that type frames nothing.
- It is contestable. If nobody informed could disagree, there is no debate to enter and no criticism worth using.
- It is answerable with evidence you can get. The author intended the wall as a symbol requires the author's mind; the text builds the wall out of language borrowed from official documents requires only the text.
- It is one claim, not three. A hypothesis joined by and is two investigations sharing a page.
Three shapes that work
- The mechanism claim — X works by doing Y. Names a method in the text and claims an effect. Best for language investigations.
- The against-the-obvious claim — X looks like A, but is doing B. The obvious reading gives you a position to argue against.
- The lens claim — read through F, X shows Y. The standard's own example is using feminist theory to interpret advertising language. Avoid a fourth shape, X is interesting because of Y, which cannot fail.
Turning a first noticing into a hypothesis
- Start from something you actually noticed — a word that recurs, a scene that sits oddly, a grammatical pattern — not from a theme.
- Say what it does. The word is repeated is an observation. The repetition transfers the decision from people to weather is the beginning of a claim.
- Push it until it could be wrong. Ask what someone would have to show you to make you withdraw it.
- Check that criticism could bear on it. If you cannot imagine what critical text would help, the claim is either too narrow or not about how the text works.
- Write it in one sentence and keep it visible. Every source you read gets judged against that sentence.
The three failures
- Too broad. This text explores identity. True of thousands of texts, so no critical text can confirm or damage it. The repair is to name a mechanism and a consequence.
- Too safe. The poem uses imagery to create atmosphere. Nobody would disagree. The repair is to make the claim narrower and riskier, not longer.
- The disguised summary. The story shows how a community responds to a decision. This describes the plot in abstract words. If it could not be false, it is a summary in a hypothesis costume.
When the evidence goes against you
- A hypothesis that survives untouched is usually one that was never at risk. Investigations are supposed to move, so revise rather than defend: narrow the claim, add the condition under which it holds, or reverse it.
- Say in your response that it changed, and why. That change is direct evidence the critical texts informed your understanding, so keep the original wording to set beside the final one.
What the grades look like here
- Achieved — a clear hypothesis, tested against the primary source and at least one critical text.
- Merit — discernment shows in the scope of the hypothesis: narrow enough to answer in the space, wide enough to be worth answering.
- Excellence — a perceptive investigation ends with a qualified hypothesis, the claim plus the condition under which it holds.
Worked Example
Worked Example
The Sea Wall is a short story written for these pages. A coastal settlement wins council funding for a sea wall; the narrator's grandmother refuses to sign the consent that would let it cross her land. The story ends:
The engineers came back in March with a new drawing. The wall went around her, out and in again, so that from the road it looked like a wave had already got through. In the minutes it was recorded as a variation. They left a gap where she was.
A student notices that nobody in the story ever decides anything on the page. Take that noticing to a usable hypothesis.
Answer:
Step 1 — state the noticing plainly. Nothing in the story is decided by a character we meet. An observation, not a claim about how the text works. Saying the absence of decision-makers makes the wall feel inevitable is better, but inevitable is vague and could be true of many stories.
Step 2 — name the mechanism. Funding is won. The drawing came back. The refusal is recorded as a variation. These are the constructions of minutes and reports: things happen, and no one does them.
The story borrows the grammar of official documents, so that decisions appear as events.
Step 3 — add the consequence. If decisions appear as events, the grandmother's refusal is the only action performed by a named person — and the text records even that in the passive.
Step 4 — test that it could be false. If the text used ordinary narrative grammar for the council's decisions, the claim collapses, so the hypothesis is testable. And criticism can bear on it: analysis of institutional language would test the mechanism, criticism about how texts represent authority the consequence.