Hedging and Precise Language: Saying Exactly How Sure You Are
B2 · Writing Lab · about 30 minutes
1 · Learn
Hedging is not weakness; it is measurement. “Smoking causes cancer” is proven — say it plainly. “This diet causes weight loss” is a hope wearing a lab coat — hedge it: “may contribute to”, “tends to”, “in some studies”. Readers, examiners, and editors trust writers whose certainty matches their evidence, and they punish the two failures symmetrically: overclaiming and mush.
The toolkit: modal verbs (may, might, could), softening adverbs (relatively, broadly, largely), frequency phrases (tend to, in most cases, on average), and evidence frames (the data suggest, it appears that). Equally important is the booster used honestly: clearly, demonstrably, without exception — reserved for what you can prove.
2 · See an Example
One Claim, Three Calibrations
OVERCLAIMED: “Coffee destroys sleep and makes workers unproductive.”
MUSH: “It could perhaps be argued that coffee might possibly have some kind of effect on sleep in some people sometimes.”
CALIBRATED: “Coffee drunk after midday tends to shorten sleep that night, in most studies by roughly half an hour to an hour. The effect is relatively small for morning coffee and considerably larger for people who metabolize caffeine slowly. None of this makes coffee an enemy; it makes timing a variable worth managing.”
Notice what the calibrated version does: it limits the claim by TIME (after midday), by POPULATION (slow metabolizers), by SIZE (half an hour to an hour), and by SOURCE (in most studies). Four limits, and the claim becomes simultaneously humbler and far harder to attack.
3 · Analyze the Example
- The three failures on one line. Overclaim (destroys), mush (could perhaps possibly), calibrated (tends to, in most studies). Hedging is a dial, not a switch.
- Limits that do the work. Time, population, size, source — the four fences. A claim fenced on four sides can be defended from any of them.
- Boosters kept in reserve. “None of this makes coffee an enemy” is a firm line INSIDE a hedged paragraph — contrast makes both registers believable.
- Precision about size. “Roughly half an hour to an hour” — ranges are how honest people report variable effects. Fake exactness (47 minutes) is overclaiming in a costume.
Useful Vocabulary
Use hedging where evidence is partial.
Prices tend to rise in summer.
It is relatively cheap.
Advertisers overclaim daily.
People metabolize caffeine at different speeds.
Useful Expressions
- … tends to …
- In most studies / cases, …
- The data suggest that …
- … is relatively / considerably …
- None of this makes X an enemy; it makes …
Common Mistakes
✗ “This method always works for everyone and guarantees perfect results.”
✓ “This method works for most learners who practice daily; it works best for reading and vocabulary, less well for pronunciation.”
Always/everyone/guarantees are the three horsemen of overclaiming. One exception — and readers supply it gleefully — kills the whole text.
✗ “It might perhaps be suggested that possibly the results could be somewhat better in certain situations.”
✓ “The results were better in the morning groups, possibly because the rooms were quieter.”
Mush is hedging stacked until meaning drowns. ONE hedge per claim, placed exactly where the uncertainty lives.
✗ “The treatment improved patients' sleep by 43.7 minutes.” (from a study of eleven people with self-reported sleep)
✓ “In a small study, patients reported sleeping roughly three-quarters of an hour longer.”
False precision is dishonesty with decimals. Match your numbers' clothing to their evidence.
4 · Guided Practice
Calibrate: “Exercise cures depression.”
See a model answer
Example: “Regular exercise reduces the symptoms of mild to moderate depression in many patients, in several trials by an amount comparable to medication — though severe depression requires professional treatment.” (Limits: severity, population, comparison, boundary.)
De-mush: “It could perhaps be that the new system might possibly save some money in some departments.”
See a model answer
“The new system could save money in two departments — the pilot suggests around 10% — though the savings are not yet proven at scale.”
Choose where the hedge belongs: “Our survey of 30 students proves that all students hate the cafeteria.”
See a model answer
Scope and size: “In our survey of 30 students, most disliked the cafeteria” — 'proves' and 'all students' are the overclaims; the hedge belongs on scope, not on the students' feelings.
5 · Write — Your Task
✍️ Take three claims from your world (about learning, health, technology, your job) and write each one calibrated: limit it by time, population, size, or source as needed. Then take one claim you can fully prove and write it with an honest booster. Add a line under each explaining your calibration choices.
Goal: 4 claims + explanations · Length: 200–260 words
- Hunt the absolutes first: always, never, everyone, proves, guarantees.
- Place ONE hedge exactly where the uncertainty lives.
- Keep one claim at full strength — calibration includes knowing when NOT to hedge.
6 · Check — Self-Check
Work through your checklist honestly, then run the Self-Check. It counts words and sentences and spots mechanical points — it does not grade your English. You are the editor.
Read a model answer for the task
Claim 1 (calibrated): “Learning apps appear to help beginners most: in the first six months, daily app users in several studies learned vocabulary measurably faster than classroom-only peers. The advantage largely disappears at intermediate levels, where conversation does the heavy lifting.” Claim 2 (calibrated): “Our school's new timetable seems to have reduced late arrivals — by roughly a fifth this term — though winter weather makes the comparison imperfect.” Claim 3 (full strength): “The library's Saturday closure demonstrably hurt the students who work weekdays: weekend visits by that group fell to zero, for the simplest of reasons. Some claims need no hedge; they need a witness.”
7 · Improve & Rewrite
Weak version — rewrite it in your notebook first, then compare:
It has been proven that social media always makes all teenagers depressed and unhappy. Studies show it is 100% bad for everyone. Also maybe it is a little bit not good for adults possibly.
See an improved version
The evidence on social media and teenage mood is real but narrower than the headlines. Large studies tend to find small average effects — heavier use is associated with somewhat worse sleep and mood for a minority of teenagers, particularly girls who use image-based platforms late at night. For most teenagers, the measurable effects hover near zero, and for some — the isolated, the ill, the far from friends — the same platforms are a genuine lifeline. Precision is not a smaller claim; it is a claim that survives contact with the data.
Quick Quiz
Keep going
📘 Grammar deep-dive: Modals of Deduction: must / can't / might have🌱 Vocabulary Builder: the B1 Health Lifestyle set🎓 TOEFL Writing: the exam version of this skill← Previous lesson · B2 — all lessons · Next: B2 level page →
Ad space