How English Works · Social meaning · Part of the How English Works collection.
“Perhaps the room is already booked.”
“Perhaps we could use the other room.”
The two sentences share a word, but that word is not doing exactly the same work. In the first, the speaker presents a possible explanation rather than an established fact. In the second, the speaker offers a proposal without making it sound like an instruction everyone must accept immediately.
The distinction matters because English uses softening language for more than one purpose. A person can be certain about a fact and considerate about asking someone to act on it. They can be uncertain about a cause while giving a clear instruction to check it. They can also use a tentative-looking expression in a context where the listener has very little real choice.
Epistemic hedging limits commitment to a proposition; interpersonal hedging moderates how a contribution bears on another person. These are two important functions, not an exhaustive division of every hedge. An expression may serve both functions, and some hedges principally concern approximation, category fit or the precision of wording.
Cambridge’s grammar guidance includes modal expressions, tense and aspect, vague expressions and certain verbs among the resources used to soften language. Manchester’s Academic Phrasebank separately illustrates how cautious academic wording avoids overstatement and overgeneralisation. Reading those functions together is more useful than treating every perhaps or could as evidence of the same kind of uncertainty. Cambridge: hedges; Manchester: cautious academic language.
Ask what the hedge is changing
A useful first question is not “Which hedging word is present?” but “What would change if we removed it?” Compare “The room is already booked” with “Perhaps the room is already booked.” The second version does not assert the booking with the same commitment. A later discovery that the room is free would contradict the first assertion more directly than the second suggestion.
Now compare “Use the other room” with “Perhaps we could use the other room.” The change concerns how an action is proposed. The second wording makes space for discussion, at least on the surface. It does not simply announce a lower probability that the other room exists.
The difference becomes even clearer in “Could you send the timetable?” This is ordinarily heard as a request in an appropriate context, not merely as an inquiry into the recipient’s physical ability. The collection’s Communicative Force article explains that larger distinction between sentence form and social action.
Here, the narrower task is to identify which part of the message becomes tentative, indirect, approximate or less imposing. Without that diagnosis, removing hedges can make a sentence less accurate, less considerate or both.
Two questions are better than one scale of confidence
For a practical analysis, ask two independent questions. First, how strongly is the speaker committed to the claim? Second, how directly is the speaker asking the other person to respond? This is a teaching framework for comparing examples, not a validated numerical instrument.
A speaker can be confident and direct: “The total is incorrect. Correct it before submitting the file.” They can be confident and mitigated: “The total is incorrect. Could you correct it before submitting the file?” The factual commitment stays stable while the request changes presentation.
They can also be uncertain and direct: “The total may be incorrect. Check it against the source before submitting the file.” Or uncertain and mitigated: “The total may be incorrect. Could you check it against the source before submitting the file?” The uncertainty concerns the total, not the need for verification.
These four versions expose a common editing error. A writer tries to sound polite by changing “is incorrect” to “may be incorrect,” even when the error has already been confirmed. That needlessly weakens the factual account. Courtesy can be added to the request without making the established fact appear doubtful.
Epistemic caution should identify a real limit
Epistemic language concerns knowledge and commitment. In academic writing, a cautious expression can make room for exceptions, uncertainty about an explanation or limits in the evidence. Manchester’s guidance explicitly links hedging to avoiding claims of absolute certainty and excessive generalisation. Academic Phrasebank: being cautious.
In a fictional school survey, suppose twelve students report that they prefer a later starting time. “Students prefer a later starting time” may imply a broader population than the survey supports. “In this group of twelve respondents, a later starting time was preferred” makes the actual scope visible.
Adding perhaps to the broad statement does not necessarily solve the scope problem. “Perhaps students prefer a later starting time” is tentative, but it still fails to tell the reader who was asked. Sometimes the best calibration is not a hedging word at all. It is a more exact subject, timeframe or condition.
The existing Claim Calibration article owns the wider evidence-to-claim relationship. This article concentrates on distinguishing that evidential work from the social softening that may look similar on the page.
Interpersonal caution should not falsify the task
Interpersonal hedging can make a request, suggestion or disagreement less abrupt. Cambridge’s politeness guidance describes indirect forms and also recognises contexts in which direct imperatives are appropriate, including instructions and warnings. The point is not that every direct sentence is rude, but that grammatical choices work within a social situation. Cambridge: politeness.
Suppose a teacher needs a corrected file before the class begins. “Could you send the corrected file before 9 a.m., please?” can be considerate while keeping the deadline clear. “Send it whenever you have a moment” changes the apparent timing. If 9 a.m. is necessary, the second version misrepresents the practical task.
The relevant distinction is between softening the manner of a request and changing its conditions. A courtesy phrase can acknowledge another person’s effort. It should not invent flexibility that will later be withdrawn.
This is also why an apparently polite question from someone with authority is not necessarily an invitation to decline. The wording and the actual options need to be read together. A good writer makes essential requirements explicit rather than relying on hierarchy to repair ambiguous language.
“I think” does not have one fixed interpretation
“I think the library closes at six” can express an uncertain recollection. The speaker may be inviting verification. In “I think we should hear the quieter members of the group,” the expression can frame a proposal as the speaker’s contribution rather than an unqualified command.
Context may make the second statement quite firm. A chairperson who has repeatedly asked for balanced participation may be issuing a clear intervention through a personally framed expression. The words I think do not establish that the speaker lacks conviction.
Nor should those words automatically be removed from every piece of writing. If the phrase merely repeats what the sentence already makes obvious, deletion may improve the line. If it marks a shift from reporting a source to offering the writer’s own judgment, it may have a useful boundary function.
The test is functional: does the phrase identify uncertainty, ownership of a judgment, interpersonal caution or some combination? The answer determines whether to retain it, replace it or express the distinction more directly.
“Could” may concern possibility, ability or a request
“The delay could be caused by a timetable error” offers a possible explanation. “She could read the sign from the back of the room” reports an ability in that situation. “Could you read the sign aloud?” asks for an action. Treating all three as low-confidence assertions would misread the grammar and the exchange.
When a learner says that a modal makes a sentence less certain, ask what is less certain. In a request, it may not be useful to describe the effect as factual probability at all. The modal form may be part of a conventional way of asking without using an imperative.
Cambridge’s discussion of hedges includes modal expressions among several softening resources. That guidance supports examining function rather than assigning one meaning to every occurrence of the form. Grammar reference.
For editing, rewrite the sentence in plain terms before deciding what to change. Is the speaker saying “This explanation is possible,” “This person had the ability,” or “Please do this”? The paraphrase reveals the job that the modal construction is doing.
Approximation is a third question, not a weaker personality
“The meeting lasted about an hour” is not necessarily a timid version of “The meeting lasted an hour.” The speaker may know the duration only approximately, or exact minutes may not matter. The expression changes the precision of the quantity.
Cambridge’s treatment of vague expressions includes approximate quantities and times, such as expressions using about or around. Its discussion of kind of and sort of also distinguishes softening from exact categorisation. Vague expressions; Kind of and sort of.
In an original example, “It is a sort of reading journal” may indicate that the object resembles a journal without meeting every expectation of the category. Replacing the phrase with “It is a reading journal” could make the classification sharper than the writer intends.
But approximation has limits. “Arrive around nine” and “The examination begins at nine” do not make the same practical commitment. Choose the level of precision that fits the task rather than treating approximate wording as universally sophisticated or universally weak.
The word “just” can minimise more than a sentence
“Could you just check this paragraph?” often presents the request as small. That may be appropriate when the task really is a quick check. It becomes a poor fit when “check” actually means rewriting a long report, verifying several sources and resolving contradictory data.
In that case, the problem is not simply the presence of a hedge. The wording understates the work. A more considerate request would specify the task and let the recipient judge whether they can do it: “Could you review the argument and source references? Please tell me whether the proposed deadline is workable.”
This example illustrates a distinction between reducing verbal pressure and acknowledging actual cost. A small word can make a request sound light without making the work lighter. Good interpersonal language should not depend on the receiver accepting an inaccurate description of the burden.
Also distinguish this use from other meanings of just, such as recency or exclusivity. “She has just arrived” is not automatically a polite hedge. A word’s dictionary form is not a substitute for analysing its local function.
Locate the hedge’s scope
“The timetable may explain the absence” hedges a possible explanation. “Some students were absent” limits the group. “The timetable explains part of the absence pattern” limits explanatory coverage. The three sentences do not express interchangeable degrees of caution.
A rewrite can preserve the topic while changing the scope of uncertainty. “All students may need support” is different from “Some students need support.” In the first, the possibility extends to everyone. In the second, the writer asserts a need for a subset. A cautious-looking word does not automatically make the same claim as a narrower quantifier.
For students, underline the exact phrase affected by the hedge. Is it the cause, the number of people, the duration, the category or the action being proposed? Then paraphrase the limitation in ordinary language.
The wider grammar is developed in Quantification and Scope. In social communication, the same attention to scope prevents a polite expression from quietly changing a factual commitment.
Do not convert word choice into an unsupported probability
A reader should not assume that might always means one numerical probability while probably means another. The cited grammar resources do not supply such a universal conversion table, and ordinary messages do not necessarily define one.
If numerical uncertainty matters, the writer should provide the relevant estimate and explain its basis rather than expect an everyday modal to carry exact quantities. If no defensible estimate exists, a number invented to make the message look precise would be worse than a clearly bounded verbal account.
In a classroom discussion, “This explanation is possible, but we have not checked the timetable” is more informative than assigning a percentage without evidence. It states a candidate and the missing verification step.
Likewise, “I think” in a request should not be converted into a confidence score. Its principal function may be conversational positioning. The reader needs to identify the activity before interpreting the degree of commitment.
Separate observation from explanation
Consider a fictional classroom record: six students arrived after the scheduled start. That observation is established for the purpose of the example. A timetable change is one possible explanation, but it has not been checked.
A weak report says, “There may have been some lateness, perhaps because the timetable was confusing.” The uncertainty spreads over the observation as well as the explanation. The reader can no longer tell which part is known.
A clearer version says, “Six students arrived after the scheduled start. A timetable change may have contributed; we have not yet checked that explanation.” The observation is stated directly, the explanation remains tentative and the missing check is visible.
This is why good caution is selective. It attaches to the unsupported step rather than fogging the entire paragraph. The result is often more confident in tone because the writer has made the boundary between knowledge and interpretation explicit.
Separate a confirmed correction from the invitation to discuss it
Suppose a document states that a meeting is on Thursday, while the agreed calendar states Friday. The writer has checked the current calendar. “I might be wrong, but perhaps Thursday is not quite right” introduces uncertainty that does not reflect the verified discrepancy.
Try: “The agreed calendar lists Friday, not Thursday. Could you update the document? Please let me know if there is a newer calendar I have not seen.” The correction is explicit. The final sentence remains open to a genuinely relevant possibility: a newer source.
This is different from performative self-doubt. The writer is not pretending to know less than they know. They identify the basis of the statement and the condition under which it would need revision.
The same structure works in academic discussion: state the evidence, identify the inference and make the revision condition visible. It allows disagreement without requiring either excessive certainty or a ritual of apologising for having a view.
A polite disagreement still needs a disagreement
“That is an interesting point” may be a genuine response, but it does not tell the listener whether the speaker agrees. “There are perhaps a few aspects we could consider” may sound careful while leaving the actual objection undisclosed.
A clearer contribution could say, “I agree that the plan is simpler. My concern is that it removes the only opportunity for students to ask questions.” The first sentence acknowledges a real strength. The second identifies the precise disagreement. Neither requires an attack on the proposer.
If the consequence is uncertain, say so at the relevant point: “My concern is that the plan could reduce opportunities for questions, especially in larger groups.” The hedge belongs to the predicted effect, not to the speaker’s right to raise the concern.
This develops a particular distinction inside Face Management. Social consideration is most useful when the substantive contribution remains recoverable.
When several hedges accumulate, inspect each one
“It might possibly perhaps be somewhat useful to consider a review” leaves the reader with several layers of tentativeness. Some may be redundant. Others may concern different things, but the sentence does not make those differences easy to recover.
Do not solve the problem by banning all hedges. First identify the intended message. If the writer recommends a review because a specific part is uncertain, say: “I recommend reviewing the attendance figures because the latest file may be incomplete.” The recommendation is clear; the uncertainty has a location.
If the writer is offering an optional idea rather than recommending action, “One option is to review the attendance figures” may fit better. If a check is required, “Please verify the attendance figures against the latest file” should not be softened into an apparently optional suggestion.
The editing question is therefore not how many cautious words can be removed. It is whether each surviving expression has a distinct and defensible job.
A full revision case
Original draft: “I was just wondering whether the totals might possibly be a little inaccurate, so perhaps you could have another look whenever convenient.” Assume that two incorrect totals have already been verified and that the file must be corrected before it is distributed at noon.
The draft has three problems. It presents confirmed errors as speculative. It understates the required work by leaving “another look” undefined. It replaces a real deadline with open-ended convenience. These are not merely matters of style.
Revised version: “Two totals in the file do not match the source table. Could you correct rows 4 and 7 before the noon distribution, please? I have marked the relevant entries. Let me know now if the deadline is not workable so that we can arrange another reviewer.”
The revised message is firm about the established discrepancy, considerate in making the request and explicit about the next action. It provides a genuine way to raise a constraint rather than suggesting that the task is optional and later treating it as compulsory.
Change the facts and the wording should change. If the totals have not been verified, “The totals may not match the source table; please check rows 4 and 7” is appropriate. The same professional tone can carry a different evidence state.
Teaching the distinction without teaching timidity
Begin with paired sentences in which a hedge has different functions. Ask learners whether the speaker is limiting a factual claim, softening an action or approximating a quantity. Let them explain in ordinary language before introducing terms such as epistemic.
Next, keep the facts fixed and vary only the request. This helps learners discover that they can be courteous without pretending uncertainty. Then keep the request fixed and vary the evidence: confirmed, suspected or unknown. The factual wording should respond to that change.
For more advanced writers, include a short paragraph containing both observations and explanations. Ask them to mark which sentences need qualification and why. The goal is selective precision, not a paragraph in which every claim carries the same weak modal.
These are suggested exercises, not claims of measured learning gains. The companion Stance and Hedging vocabulary article remains the route for building lexical range. The present article adds the pragmatic distinction between what the speaker knows and how the speaker approaches another person.
The final test: truth, action and relationship
Before sending a cautious sentence, identify its factual commitment, its requested action and its interpersonal tone. Check each separately. Does the factual wording match the evidence? Can the recipient tell what to do? Does the message acknowledge the other person without obscuring the task?
A hedge is useful when it marks a real limit or makes an interaction more considerate while preserving what matters. It is unhelpful when it spreads uncertainty over known facts, disguises an obligation as a free choice or makes the required action impossible to identify.
Careful English does not choose between confidence and courtesy. It puts confidence where the evidence supports it, caution where uncertainty belongs and respect into the way the message reaches another person.
Sources and connected reading
The grammar references linked above support the distinctions among cautious claims, indirect requests and approximation. The two-question comparison, classroom records and revision cases are original illustrations. They do not assign universal probability values to hedging words or claim that one style is appropriate in every community.
Continue with Professional Force Control for institutional messages, Euphemism and the Meaning-Preservation Test for factual integrity during softening, and How English Works for the full collection.