TOEIC Link Writing Hedging and Epistemic Modality for Calibrated Claims: Saying Exactly as Much as the Evidence Allows
Every claim a TOEIC Link Writing candidate makes carries an implicit strength. "The new policy will increase productivity" claims certainty; "the new policy is likely to increase productivity" claims a strong probability; "the new policy may increase productivity" claims a possibility. The scorer reading the response is checking whether the strength of the claim matches the strength of the support offered for it, and the language that sets that strength is hedging and epistemic modality — the modal verbs, adverbs, and reporting verbs that tell the reader how confident the writer is entitled to be. A candidate who writes every claim at full certainty will sooner or later assert something the evidence does not support, and a candidate who hedges everything will read as unable to commit to a position. Both failures are calibration failures, and both cost marks on tasks scored for the quality of argument.
The difficulty is that the modal system feels like a matter of politeness or style when it is actually a matter of accuracy. "Suggests" and "proves" are not two ways of saying the same thing; they are two different claims about how strongly the evidence points. Choosing between them is not decoration — it is deciding what the response actually asserts. A candidate who treats modality as interchangeable will attach a strong verb to weak evidence and produce an overclaim, or attach a weak verb to strong evidence and throw away a point that was there to be made.
This article is the hedging-and-modality discipline for TOEIC Link Writing. The guide covers the epistemic scale from possibility to certainty, the reporting verbs that calibrate claims about sources, the adverbs and phrases that fine-tune strength, and the proofreading pass that catches the mismatch between claim and evidence before it costs a mark.
The epistemic scale from possibility to certainty
Epistemic modals sit on a scale of confidence, and matching the modal to the strength of the evidence is the core of a calibrated claim.
"May," "might," and "could" claim possibility — the evidence is compatible with the claim but does not compel it. These are the right tools when the writer is raising a plausible outcome without asserting it will happen. "Remote work may reduce commuting costs" claims only that the reduction is possible, which is defensible even on thin evidence. The trap is using possibility modals where the evidence actually supports more, which understates a claim the writer was entitled to make strongly.
"Is likely to," "probably," and "should" claim probability — the evidence points toward the claim without guaranteeing it. This middle band is where most well-supported argument claims belong, because most argument evidence establishes a tendency rather than a certainty. "The policy is likely to improve retention" claims the evidence favors the outcome, which is the honest strength of most real support. Candidates who skip this band and jump to certainty are the ones who overclaim.
"Will," "is," and unmodalized statements claim certainty — the evidence establishes the claim beyond reasonable doubt. Reserve the bare assertion for claims the writer can fully defend: definitions, arithmetic, and outcomes the evidence directly establishes. "The report shows sales rose" is certain because the report is the evidence; "the report shows sales will keep rising" is an overclaim because a past figure does not guarantee a future trend. The shift from describing evidence to predicting from it is exactly where certainty must downgrade to probability.
Reporting verbs that calibrate claims about sources
When a response refers to what a source says, the reporting verb sets how strongly the source is being read, and the wrong verb misrepresents the source.
"States," "reports," and "notes" claim neutral attribution — the source said this, no more. These verbs are the safe default for summarizing what a prompt document contains. "The memo states that the deadline is Friday" attributes the claim to the memo without endorsing or strengthening it. Use these whenever the task is to represent a source accurately rather than to evaluate it.
"Suggests," "indicates," and "implies" claim inference — the source points toward a conclusion it does not state outright. These verbs are the right tool when the writer is drawing a defensible inference from a source. "The falling attendance figures suggest declining interest" is honest because the figures indicate but do not state the interest claim. The quantifier and determiner precision discipline that governs data claims applies here too: the inference verb must match how far the data actually reaches.
"Proves," "demonstrates," and "establishes" claim conclusive support — reserve them for evidence that genuinely settles the point. These are the strongest reporting verbs, and attaching them to suggestive evidence is the reporting-verb version of an overclaim. Most argument evidence suggests; very little proves. A candidate who writes "the survey proves customers prefer the new design" when the survey shows a 55 percent preference has overstated what the source establishes, and the scorer reading against the source will mark it.
Adverbs and phrases that fine-tune strength
Between the modal bands, adverbs and hedging phrases let the writer place a claim precisely on the confidence scale.
Degree adverbs shift a claim up or down the scale. "Clearly," "certainly," and "undoubtedly" raise confidence; "arguably," "possibly," and "to some extent" lower it. These adverbs are precision instruments, and stacking them is the error — "it is possibly arguably true" hedges twice and reads as evasive. Choose one adverb that sets the intended strength and let it stand, in line with the redundancy and wordiness elimination discipline that keeps a sentence from carrying two words for one job.
Scope-limiting phrases confine a claim to where the evidence holds. "In most cases," "under these conditions," and "for this group" restrict a claim to the domain the evidence covers, which is often the difference between a defensible claim and an overgeneralization. "The training improves performance for new hires" is defensible where "the training improves performance" overreaches. Adding the scope is not hedging for its own sake — it is stating the true extent of the claim.
Attribution phrases separate the writer's view from a reported one. "According to the report," "the author argues," and "it has been claimed that" mark a claim as belonging to a source rather than to the writer, which lets the response present a position it is about to challenge without appearing to endorse it. This separation is essential in argument tasks that require acknowledging a counterposition before rebutting it.
The calibration proofreading pass
Before submitting, the candidate runs a pass that checks only whether each claim's strength matches its support.
Check one: does each strong claim have strong support? Find every certainty claim — every "will," every "proves," every unhedged assertion — and ask whether the evidence in the response actually establishes it. Any strong claim resting on suggestive evidence is downgraded to a probability modal. This single check catches most overclaims.
Check two: is any well-supported claim needlessly hedged? Find every "may" and "might" and ask whether the evidence actually supports more. A claim the writer could defend at "is likely to" but stated at "may" gives away rhetorical force for no reason. Calibration cuts both ways: underclaiming is as much a mismatch as overclaiming.
Check three: is the hedging single, not stacked? Find any sentence with more than one hedge — "it may possibly be the case that" — and cut it to one. Stacked hedges read as evasion and dilute the claim past the point of usefulness. One well-chosen modal or adverb sets the strength; the rest is noise, and the tone and register control discipline that keeps prose from sounding uncertain depends on removing it.
A response that passes all three checks reads as the work of someone who said exactly as much as the evidence allowed — no more, no less. That calibration is precisely what the TOEIC Link Writing rubric rewards when it scores the quality of argument, and the modal system, invisible when it works, is where the calibration is set.