Exactness is valuable, but human conversation is not a laboratory report.
We regularly speak when the number is not known, the category has fuzzy edges, the detail is irrelevant, the memory is incomplete, the relationship is socially delicate, or the listener already knows enough to reconstruct the rest.
- There were about thirty people.
- It happened around midnight.
- It was kind of strange.
- She works in finance and stuff.
- Bring a towel or something.
These forms do not represent failed precision. They are part of a systematic English toolkit for deliberate inexactness.
The relevant system includes approximators, vague category markers, general extenders, placeholder nouns and other expressions that loosen the boundary around what a speaker commits to.
For the wider architecture, begin with How English Works and the English Learning Hub. Vague language connects especially to Inferential Enrichment, Face Management and Politeness, Parenthetical and Comment Clauses, and Discourse Markers.
The shortest useful definition
Vague language is language whose boundaries are intentionally left less exact than they could be. Approximators are expressions that locate a quantity, time, degree or category near a target without committing to an exact value or membership.
Examples include:
- quantity: about twenty, roughly half, nearly ten, more or less;
- time: around six, sometime later, towards evening;
- category: kind of a workshop, sort of metallic, something like a receipt;
- general extension: and stuff, or something, and things like that.
The speaker is not necessarily ignorant. Sometimes the speaker knows the exact value but judges that exactness would add no useful value.
Approximation reduces commitment
Thirty people attended.
This sentence commits to thirty.
About thirty people attended.
Now the speaker commits only to a neighbourhood around thirty.
The exact width of that neighbourhood depends on context. About thirty seconds may tolerate a difference of a few seconds. About thirty million dollars may tolerate a much larger absolute difference. Approximation is scaled to the discourse domain.
About and around overlap but are not identical everywhere
- about fifty people
- around fifty people
- about six o’clock
- around six o’clock
Both commonly signal approximation.
Around retains stronger spatial associations in other contexts: walk around the building. About has separate topic uses: a book about science.
When used as approximators, however, both loosen the boundary around a target value.
Roughly, approximately and circa change register
English offers several approximation forms with different register profiles.
| Expression | Typical register | Example |
|---|---|---|
| about / around | neutral, everyday | about 20 minutes |
| roughly | neutral analytical | roughly one-third |
| approximately | formal/technical | approximately 2.5 km |
| circa | specialised, dates/history | circa 1900 |
| more or less | conversational/qualifying | more or less complete |
The underlying semantic job is similar, but the social and disciplinary signal changes.
Nearly and almost approach a boundary without crossing it
Nearly ten students arrived.
This normally suggests fewer than ten, but close to ten.
Almost finished means the completion boundary has not yet been reached.
This differs from about ten, which can include values slightly below or above ten.
almost/nearly 10 → approaches 10 from below about/around 10 → neighbourhood around 10
Approximation therefore has internal geometry.
Over and under can approximate from one side
- over a hundred people
- under ten minutes
- just over two metres
- just under fifty dollars
These expressions locate the value relative to a boundary while preserving direction.
Just narrows the distance from that boundary: just over a hundred suggests a number only modestly above one hundred.
This connects approximation to Focus Particles and scalar meaning.
Kind of and sort of loosen category membership
It’s kind of a museum.
The speaker does not fully commit to the category museum. Perhaps the place has museum-like exhibits but also functions as an archive, visitor centre or gallery.
Kind of and sort of can therefore mark fuzzy categorisation.
- It’s sort of a classroom.
- The surface is kind of metallic.
- He looked sort of worried.
The speaker places the target near a category without claiming perfect membership.
Category boundaries are often genuinely fuzzy
Not every concept has sharp mathematical edges.
Is a stool a chair? Is a tomato a vegetable in culinary language? Is a tiny printed booklet a book? Is a room with displays but no permanent collection a museum?
Natural categories often have central examples and borderline cases.
Vague category language allows speakers to respect that structure rather than forcing a false yes/no classification.
In that sense, vagueness can be more accurate than false precision.
Kind of can also soften degree
- I’m kind of tired.
- It’s sort of expensive.
- That was kind of unfair.
Here the expression does not merely classify. It reduces commitment to the degree of the adjective.
That was unfair is direct. That was kind of unfair leaves room for qualification, disagreement or politeness.
This makes approximators part of interpersonal grammar as well as semantic precision.
Vagueness can soften social risk
Compare:
- Your answer is wrong.
- Your answer is not quite right.
- There’s something a little off in the second step.
The later versions reduce the sharpness of the judgement.
This can protect face and keep correction collaborative. But vagueness can also become evasive if the speaker knows exactly what is wrong and refuses to say it.
Politeness and precision must be balanced against the reader’s need to act.
Stuff and things are general nouns with powerful discourse uses
We bought paper, pens and stuff.
The word stuff signals that the listed items belong to a broader, loosely defined category whose remaining members do not need to be named.
Things works similarly:
We talked about exams, university and things like that.
The speaker trusts the listener to infer a category around the examples.
These are not empty words. They outsource part of category construction to shared knowledge.
General extenders keep categories open
- and stuff
- and things like that
- or something
- and so on
- or whatever
These expressions signal that the explicit list is not exhaustive.
Bring a pencil, ruler or something says the exact item is flexible within a practical category.
She works with budgets, forecasts and so on signals that more related tasks exist but do not need enumeration.
The listener completes the category from context.
Or something can mark uncertainty about the label
He’s an engineer or something.
The speaker is uncertain about the precise occupation but confident that it belongs in a nearby professional category.
The phrase therefore protects against overclaiming memory.
But it can also sound dismissive: He won some award or something may imply the detail is not important enough to remember.
Vagueness encodes attitude as well as uncertainty.
Or whatever can signal indifference
Take the bus or whatever.
The expression keeps the alternative set open, but it can also signal that the speaker does not care which option is chosen.
That can sound relaxed among friends or dismissive in a professional context.
The same semantic vagueness therefore carries different social meanings depending on relationship.
Something like can offer a prototype rather than an exact identity
We need something like a checklist.
The speaker is not necessarily asking for a literal checklist. The phrase offers a prototype: a tool with similar functions.
This is useful during design, brainstorming and early planning, when specifying the functional neighbourhood is more productive than prematurely fixing the final form.
Vague quantities can be precise enough for the task
If someone asks when dinner begins, around seven may be exactly the amount of precision needed.
If a surgeon asks for a drug dose, around seven milligrams may be dangerously inadequate depending on the medication and context.
The quality of vagueness therefore depends on the decision it supports.
Precision is not a virtue in the abstract. It is a requirement calibrated to consequence.
False precision can be worse than honest approximation
Suppose someone glances at a crowd and says There were 437 people without counting.
The number sounds precise but the evidence does not support that precision.
There were roughly four hundred people may be epistemically stronger because the language matches the quality of the observation.
Good English does not merely maximise precision. It matches commitment to evidence.
Vague language and memory
Human memory often preserves gist better than exact detail.
It happened sometime around 2018.
This sentence can honestly represent the speaker’s memory boundary.
A more precise date would be misleading unless verified.
Approximators therefore function as epistemic safety valves: they let language preserve useful information without pretending certainty the speaker does not possess.
Vagueness and shared knowledge
Bring all the cables and stuff.
Between strangers, this instruction may be poor. Between colleagues who packed the same equipment yesterday, it may be efficient.
Vague language depends heavily on common ground. The less shared knowledge speakers have, the more explicit they usually need to become.
This links directly to Common-Ground Management.
Vague language in conversation
Spoken English tolerates and often prefers more vagueness than formal technical prose because conversation is fast, collaborative and repairable.
Meet me around six near the station or somewhere there.
If that is not precise enough, the listener can ask a follow-up question. The communication system includes repair.
Writing often lacks immediate repair, so vague expressions must be judged more carefully.
Vague language in instructions
Instructions expose the limits of vagueness quickly.
- Add a little water. — acceptable in forgiving cooking contexts.
- Tighten it a bit. — potentially dangerous in engineering if torque matters.
- Wait around five minutes. — fine if the process tolerates variation.
- Install it somewhere near the sensor. — inadequate if exact placement affects function.
The required precision is determined by tolerance.
A strong technical writer asks not “is this phrase vague?” but “how much variation can the system safely tolerate?”
Vague language in science
Science often replaces everyday approximators with quantified uncertainty:
- confidence intervals;
- measurement uncertainty;
- ranges;
- probability estimates;
- orders of magnitude.
But approximation remains legitimate when measurement itself is approximate or when an order-of-magnitude statement is the correct analytical level.
Approximately 70% is scientifically useful if the underlying estimate warrants that resolution. Adding unnecessary decimal places does not make the claim more truthful.
Vague language in journalism
Journalism frequently uses controlled approximation before final figures are available:
- about 200 people
- roughly half the district
- around midnight
The wording should reflect source quality. If official records later establish an exact count, continuing to report a rough number may become unnecessarily vague. If counts remain contested, exact-looking figures can create false authority.
Approximation is part of source discipline.
Vague language in law and policy
Legal language sometimes needs open-textured standards such as reasonable, substantial, near or promptly because future situations cannot be enumerated exhaustively.
But vague drafting can also create uncertainty about rights and duties.
The problem is not vagueness itself. It is uncontrolled interpretive range where consequences are high.
Good policy language makes the necessary flexibility explicit and constrains it with definitions, examples, thresholds or review mechanisms where possible.
Vague language in narrative
Fiction uses vagueness to represent perception faithfully.
Something moved behind the curtain.
The narrator does not yet identify the entity because the character cannot.
It was somewhere near midnight. can create a hazy remembered atmosphere rather than an instrument-read timestamp.
Controlled vagueness therefore supports viewpoint, suspense and limited knowledge.
Vagueness can reveal character
A speaker who constantly says kind of, sort of and maybe may sound cautious, uncertain, diplomatic or simply conversational.
Another speaker may use precise numbers and categorical labels to project authority—even when that precision is unsupported.
Writers can use these patterns for characterisation, but should avoid simplistic conclusions. Speech habits vary by community, context and personality.
Vagueness and politeness
Could you move it a little?
The speaker leaves the exact distance open, giving the listener some autonomy.
Could we maybe meet around three? combines several softeners: modal question form, maybe, and approximate time.
This reduces imposition, but too much softening can make the actual request difficult to identify.
Social delicacy should not erase operational clarity.
Vagueness and deception are not the same
A speaker can be vague because exactness is impossible, irrelevant or socially inappropriate.
A speaker can also use vagueness strategically to avoid accountability:
Some mistakes were made.
Who made them? How many? How serious?
The sentence may be appropriately cautious if facts are still emerging, or evasive if the facts are known and responsibility is being hidden.
Critical reading evaluates the relationship between linguistic precision and available evidence.
Vagueness can be quantified
Instead of eliminating all vague language, a writer can replace uncontrolled vagueness with bounded ranges.
- soon → within two working days
- near the entrance → within five metres of the entrance
- most students → 78% of students, if measured
- around six → between 5:45 and 6:15, if that tolerance matters
Precision is a design choice. The writer can widen or narrow the permitted interpretation according to the reader’s task.
Most, many, several and a few encode fuzzy quantities
English quantifiers often have flexible contextual boundaries.
- many students
- several problems
- a few minutes
- most cases
Most has a clearer logical core—more than half—while many and several are highly context-sensitive. Five errors may be “many” in a one-page proof but very few in a million-line dataset.
Quantitative vagueness is relative to expectations and scale.
Approximately does not excuse bad measurement
Adding an approximator does not make an unsupported number responsible.
Approximately 80% still requires a source or measurement process capable of supporting an estimate near 80%.
Vague language calibrates commitment; it cannot manufacture evidence.
Vagueness in AI-generated writing
Generated text can produce two opposite failures.
One is unsupported precision: exact-looking claims, percentages or dates without adequate evidence.
The other is empty vagueness: various factors, many aspects, somewhat significant, things like that where the reader actually needs names, mechanisms and boundaries.
A strong edit asks two questions: how much does the evidence justify, and how much does the reader need?
The correct level of precision lies where those two constraints meet.
A compact vagueness map
| Expression | Main job | Typical interpretation |
|---|---|---|
| about / around | two-sided approximation | near a value |
| almost / nearly | one-sided approximation | close to but below boundary |
| roughly / approximately | formalised approximation | estimated value |
| kind of / sort of | fuzzy category or degree | partial/approximate membership |
| stuff / things | general category placeholder | related unnamed items |
| or something / and so on | open extension | non-exhaustive alternatives |
Common learner errors
- Assuming vague language is always poor English.
- Using kind of and sort of repeatedly until every claim becomes weak.
- Using stuff or things where a formal reader needs exact categories.
- Using false precision when the evidence supports only an estimate.
- Using approximation in safety-critical instructions where tolerance is too narrow.
- Using vague group claims to avoid naming who acted or how many were affected.
- Failing to distinguish almost ten from about ten.
A practical method for students
- Find the vague expression. Number, time, degree, category or open-ended list?
- Ask why precision is missing. Unknown, unnecessary, socially softened, remembered approximately or strategically avoided?
- Estimate the interpretation range. What values or category members remain possible?
- Check consequence. How much precision does the reader need to act safely and correctly?
- Check evidence. Does the wording match what is actually known?
- Replace or retain deliberately. Use a range, exact value or explicit category if the task demands it; keep the vagueness if it is the more truthful or useful representation.
Why vague language belongs inside How English Works
Vague language shows that communication is not a contest to produce the maximum number of decimal places.
English lets speakers widen a numerical interval, soften a category boundary, leave a list open, represent imperfect memory, reduce social force and match commitment to evidence.
Sometimes vagueness is laziness. Sometimes it is evasion. But sometimes it is the most precise representation of what a person genuinely knows and what a listener genuinely needs.
The skill is not eliminating inexactness. It is controlling the size, reason and consequence of the uncertainty you leave in the sentence.
Teaching Guide
Teach vague language by varying consequences. Ask students whether around seven is precise enough for dinner, a train departure, a chemical reaction and a surgical dose. The form stays similar while the acceptable tolerance changes.
Next, separate numerical approximation from category approximation and general extenders. Students should classify about thirty, kind of metallic, or something and and stuff by the boundary each expression loosens.
Finish with editing for evidence. Give students exact-looking claims supported only by rough observation, and vague claims supported by precise measurements. Ask them to recalibrate each sentence. Mastery means matching language resolution to evidence, consequence and reader need.
Continue through the English system: How English Works · English Learning Hub · Inferential Enrichment · Face Management and Politeness