From the room, last lesson
Six sentences taken down during the policy rounds. Two claim too much, two claim too little, and two hedge the wrong part of the claim.
Six sentences. One problem each. The note says what is actually known.
1. Insulation always pays for itself within five years. — in the trials, on solid-wall properties
2. It may possibly perhaps be worth considering, potentially.
3. The evaluation possibly found a twenty per cent rise in ridership. — the evaluation certainly found it; whether it generalises is the open question
4. It could be argued that frequency matters more than coverage. — this is your own view
5. Everyone agrees heat pumps are the answer.
6. Cities of this size tend to possibly show a rebound.
The finding
Attached to the wrong part
Two of the errors in this area are visible from across a room. Over-hedging stacks instruments inside a single slot until the sentence claims nothing; under-hedging leaves every slot bare and claims everything. Both are easy to see and easy to fix, and between them they account for most of what gets marked in a B2 essay.
The third is not about quantity at all. The evaluation possibly found a twenty per cent rise. The evaluation definitely found it — that is what an evaluation is, a record of what was found. What is genuinely uncertain is whether the same rise would appear anywhere else. So the hedge has been attached to the measurement, which was never in doubt, and the part that is in doubt has been left completely bare.
This is the error that matters, and it is invisible for the usual reasons. The sentence is grammatical, it is hedged, and it sounds modest. It also misinforms twice over: a reader now believes the researchers are unsure of their own numbers, and has been given no reason to doubt the generalisation, which is the thing they should doubt.
And there is a fourth, which examiners meet more often than any of the others: it could be argued that… Argued by whom? By nobody in particular. That form does not limit a claim; it removes its author, which is a different move entirely, and it is why so much student writing reads as though nobody wrote it. If it is your view, say so with the right slot filled. If somebody else argued it, name them.
The repair for all four is one question, asked before you reach for a word: which part of this am I actually unsure about? Then use the instrument that reaches that part, once.
Three questions.
1. What is wrong with the evaluation possibly found a twenty per cent rise?
2. Why does that error misinform twice?
3. What does it could be argued that do?
Five errors, and the sound of a hedge
What goes wrong, and what a reader concludes
| Error | Example | Repair, and why |
|---|---|---|
| 1 · Stacking in one slot | Keep one instrument. Two in a slot cancel; three and four do damage. It may be the case that… | |
| 2 · The bare absolute | Fill the slot the absolute is standing in for — nearly always scope. On solid-wall properties, insulation appears to… | |
| 3 · The hedge on the wrong element | The evaluation found a 20% rise, which may not generalise. The measurement is not the uncertain part; the generalisation is. | |
| 4 · The vanished author | On this evidence, frequency appears to matter more. That form removes the writer rather than limiting the claim. | |
| 5 · Hedging what you know | Check, then state it. A hedge in front of something you could establish in ten seconds is a refusal to be useful. |
And one small double-limit that appears constantly. tend to possibly · may be able to potentially · roughly around approximately. Each of these limits the same slot twice. Tend to already contains its exceptions; nothing needs adding to it.
Out loud, and it is quick.
1 · A hedge is unstressed and fast. It may be an underestimate — may almost disappears. Stress it — it MAY be an underestimate — and you have produced a warning, or sarcasm.
2 · Keep the falling intonation. A rising tone on a statement turns every claim into a question, which is the spoken form of over-hedging and it cancels every careful written hedge you have chosen.
3 · The stress goes where the content is: it may be an UNDERestimate.
A hedge is part of the claim, not a comment on it — and that is exactly how it should sound.
Use the word in capitals to form a word that fits the gap. All eight belong to the vocabulary of evidence.
- The report says almost nothing about the scheme’s . EFFECT
- The data gathered so far are in both directions. CONCLUDE
- The estimate is not especially , as the authors say themselves. RELY
- The finding has not yet been by any other group. PRODUCE
- Air quality has improved at the three separated sites. MARK
- The claim, as stated, is simply . VERIFY
- The methodology was widely when the paper appeared. CRITIC
- There is considerable between the eleven sites. VARY
Four questions.
1. Which part of the evaluation found a 20% rise is genuinely uncertain?
2. What is wrong with tend to possibly?
3. How is a hedge said aloud?
4. You are about to write it could be argued that. What should you do instead?
Moving the hedge
Eight gaps from a note to a committee. Every option is grammatical; choose the one that puts the doubt where it belongs.
The 2022 evaluation a twenty per cent rise in ridership on the three routes where frequency was increased. That figure is not in doubt; what is in doubt is whether it to routes with a different passenger mix. Cities of comparable size to see smaller effects where the baseline service is already frequent, so a rise of that size here be optimistic.The evaluation is measured rather than modelled, which is the reason I am relying on it, and it been reproduced elsewhere, which is the reason I am not relying on it heavily. one further point, which is my own inference and not the report’s: the effect was concentrated in the first eighteen months, and the report say why.On the evidence available, then, an increase in frequency to be worth trialling on two routes rather than committing to across the network.
Eight sentences. Seven need work; one is correct as written.
1. The study possibly found a fall of four per cent. — the study found it; whether it holds elsewhere is the question
2. It may possibly perhaps be worth trialling.
3. Recycling always reduces landfill. — in schemes with kerbside separation
4. It could be argued that the frame is too short. — it is your own argument
5. Cities of this kind tend to possibly show a rebound.
6. The meeting was possibly on the fourteenth. — the date is in your calendar
7. On the evidence available, separation rather than provision appears to be doing the work. — correct as written
8. The figure may be roughly around approximately eight per cent.
Move the hedge
Task card · Speaking and writing · Pairs · 10 minutes
Five claims each, written fast about something you know — your city, your field, your course — and then audited by somebody else.
For every hedge, one question: is that the part you are actually unsure about?
- 3 minutes: write five claims. Hedge them however comes naturally; do not think about the unit while you write.
- Swap. For each hedge your partner finds, they ask one question: which part of this are you unsure about? Answer out loud before you touch the sentence.
- Then move it. Where the hedge is on the wrong element, rewrite the claim with the doubt in the right slot — and leave the certain part unhedged.
- Finish out loud: read your five repaired claims to your partner with the hedges unstressed and the intonation falling. Your partner says whether any of them sounded like a question.
Expect at least one of your five to have the doubt in the wrong place. It is the commonest error at this level and it is invisible from the inside, because the sentence sounds appropriately modest either way. The only way to catch it is the question — and the question has to be asked by somebody else, or answered out loud.
Where the doubt actually was
Two questions.
1. How many of your five had the hedge on the wrong element?
2. When you read the repaired claims aloud, what did your partner hear?
Which part, exactly
Two questions. Twenty seconds.
1. What is the question to ask before reaching for a hedge?
2. What is wrong with it could be argued that?
Next. Three hundred and sixty words on a sustainability question — written, where every slot you leave bare is a claim somebody can hold you to.