Refining Your Thesis
The sentence you write last, and put first
- The thesis appears at the start of an essay and should be written near the end of your planning.
- You cannot know what your evidence supports until you have looked at it.
- Writing the thesis first is how students end up defending a position their evidence does not reach.
The thesis should be written...
It appears first but should be written near the end of planning.
Narrow it until it excludes
- A thesis should be specific enough to constrain what follows.
- Test: name a paragraph your thesis rules out. If you cannot, it is not directing anything.
- "Education matters" rules nothing out; "this measure improves the figure it reports and nothing else" rules out most of the essay you might have written.
A thesis is specific enough when you can name a paragraph it ____ out.
If it rules nothing out, it is not directing the essay.
Qualify it to the ground you can hold
- An unqualified thesis is defeated by one counter-example.
- Narrow it to the ground your evidence actually covers: in these cases, where X holds, for this group.
- A qualified thesis is not a weaker thesis — it is a defensible one.
What does this thesis need?
Diagnose each thesis by what is missing.
A qualified thesis is a weaker thesis.
It is a defensible one — an unqualified claim is defeated by a single counter-example.
A qualified thesis is not a weak one. An unqualified claim is defeated by a single counter-example, while a claim narrowed to the ground your evidence covers survives it — which is why qualification is also the most reliable route to the sophistication point.
Match each fault to its repair.
Three faults, three different repairs — and the third is the one students resist.
Change it when the evidence says so
- If, while drafting, the evidence does not support the thesis you began with, change the thesis.
- Examiners reward an argument that fits its evidence.
- There are no marks for having predicted your own conclusion.
Select all the checks this lesson gives before committing to a thesis.
Sentence count is never a test; reach, constraint and arguability are.
Four checks before you commit
- Could a reasonable person disagree? Can I name a paragraph it rules out?
- Does it respond to the prompt as asked? Does my evidence actually reach this far?
- A thesis that passes all four is worth the sentence it takes.
First draft: "Standardised testing harms students." One counter-example sinks it. Revised: "Where a school's funding depends on its test scores, standardised testing changes what is taught before it changes what is learned." Narrower, arguable, and the evidence reaches it.
Write the thesis after you know your evidence. Narrow it until it rules something out, qualify it to the ground your evidence covers, and change it if the evidence does not fit. Check: arguable, constraining, responsive, and reachable by your evidence.