Vagueness
Published:
Some concepts have sharp boundaries.
An integer is even or odd.
Other concepts do not.
A person may be clearly tall.
Another clearly not tall.
Between them lies a region where the answer feels uncertain.
This is vagueness.
Vague Predicates
Examples include:
- tall,
- bald,
- heap,
- rich,
- old.
These words work perfectly well in ordinary life.
But they resist exact cutoff points.
Vagueness Is Not Ignorance
Suppose we do not know whether a sealed box contains gold.
That is uncertainty.
The predicate:
“contains gold”
is not vague.
By contrast, “is tall” may remain borderline even when height is known exactly.
Vagueness is semantic, not merely epistemic.
Vagueness Is Not Ambiguity
The word “bank” is ambiguous because it has distinct meanings:
- financial institution,
- riverbank.
“Tall” is not ambiguous in the same way.
Its problem is an unclear boundary.
Borderline Cases
A borderline case is an object for which the predicate does not clearly apply or clearly fail.
For example:
Is someone 179 cm tall?
The answer depends on context.
There may be no single natural threshold.
Context Dependence
A person may be tall:
- among children,
- not among professional basketball players.
Vague predicates often depend on comparison class.
Context partly fixes standards.
The Sorites Paradox
The sorites paradox exposes the problem dramatically.
Start with a heap of one million grains.
Remove one grain.
Surely it remains a heap.
Repeat.
Eventually only one grain remains.
Where did the heap disappear?
Sorites Structure
The reasoning seems to use a tolerance principle:
If n grains form a heap, then n−1 grains also form a heap.
Repeated application yields:
one grain forms a heap.
That is absurd.
One seemingly harmless step produces disaster when iterated.
Why the Sorites Matters
The paradox shows tension among three intuitions:
- Some cases are clearly heaps.
- Some cases are clearly not heaps.
- One grain cannot make the decisive difference.
All three feel plausible.
Together they conflict.
Sharp Cutoff Theory
One response says there really is a sharp boundary.
Perhaps:
42,731 grains = heap.
42,730 grains = not heap.
We simply do not know where the cutoff lies.
This view is associated with epistemicism.
Epistemicism
Epistemicists argue vague predicates have precise but unknowable boundaries.
The vagueness lies in our ignorance.
This preserves classical logic.
But many find the hidden exact boundary metaphysically strange.
Supervaluationism
Another approach considers multiple precise ways of sharpening a vague term.
A statement is:
- supertrue if true under every admissible sharpening,
- superfalse if false under every sharpening.
Borderline statements may be neither supertrue nor superfalse.
Fuzzy Logic
Fuzzy logic allows degrees of truth.
Instead of only:
0 or 1,
a predicate may take values between them.
Someone might be tall to degree:
0.7.
This models gradual membership.
But whether vagueness literally is degree-of-truth remains debated.
Degree vs Uncertainty
A fuzzy value is not the same as probability.
“0.7 tall” does not mean:
70% chance the person is tall.
It may mean:
the predicate applies to degree 0.7.
These are different concepts.
Contextualism
A contextualist view says standards shift during conversation.
As borderline cases are introduced, the threshold for “heap” or “tall” may move.
This explains some flexibility of ordinary language.
Psychological Categories
Human categories often have prototypes.
A robin feels like a more typical bird than a penguin.
Category membership can be clear while typicality is graded.
Vagueness and prototype structure interact but are not identical.
Scientific Terms
Science often tries to reduce vagueness.
Examples:
- temperature is measured numerically,
- legal adulthood has exact age thresholds.
Operational definitions create sharper boundaries.
But the underlying concepts may remain partly conventional.
Species
The concept of species can be vague in evolutionary transitions.
Populations diverge gradually.
There may be no single generation where one species suddenly becomes another.
Evolution produces continua.
Life
“Alive” also becomes difficult at boundaries.
Viruses challenge simple definitions.
Prions challenge them differently.
Categories shaped by central examples may become vague at the edges.
Consciousness
Later we will encounter similar questions:
Is every organism conscious?
Where exactly does consciousness begin?
Sharp binary labels may exceed available evidence.
Vagueness can be conceptual and epistemic at once.
Legal Vagueness
Law needs predictable boundaries.
Terms such as:
- reasonable,
- excessive,
- dangerous
are intentionally open-textured.
Not every concept can be replaced by a number.
Human judgment fills the gap.
Precision Has Costs
A perfectly sharp rule can create arbitrary discontinuities.
Suppose a benefit is available below an income threshold.
One dollar can create a large legal difference.
Precision solves vagueness by creating cliffs.
Measurement Does Not Eliminate Vagueness
Knowing exact height does not settle “tall.”
Measurement sharpens input.
The predicate’s standard remains a separate question.
Data precision and conceptual precision differ.
Vagueness in AI
Machine classifiers often output scores.
A system may assign:
0.51 probability to one label.
The binary decision comes later through a threshold.
This makes explicit something human categorization often hides.
Thresholds Are Decisions
A threshold can be:
- statistically optimized,
- legally mandated,
- socially chosen.
It is not always discovered in nature.
Classification frequently mixes facts and purposes.
Higher-Order Vagueness
Suppose we divide cases into:
- clearly tall,
- borderline,
- clearly not tall.
Now ask:
Where exactly does the borderline region begin?
The new boundary can itself be vague.
This is higher-order vagueness.
Why Classical Logic Struggles
Classical logic assumes propositions are true or false.
Vague predicates make this assumption difficult to apply cleanly.
Alternative semantic frameworks try to preserve useful reasoning without pretending every boundary is sharp.
Vagueness Is Useful
Vagueness is not merely defective language.
It allows flexible communication without specifying unnecessary precision.
“Meet me near noon.”
“Bring a large box.”
Exact numbers may be unnecessary.
Vagueness reduces cognitive cost.
The Philosophical Lesson
Vagueness appears when categories work well at clear cases but lack sharp boundaries at the edges.
The sorites paradox shows that ordinary tolerance cannot be extended indefinitely without trouble.
Language needs flexibility.
Logic demands precision.
Their tension is productive.
The Next Question
Logic also distinguishes another dimension.
Some statements are merely true.
Others seem necessarily true.
Some events actually happen.
Others are only possible.
To reason about these differences, we need the concepts of:
necessity and possibility.