Skip to content

7.4 Constructive Proofs

Proving existence by giving explicit witnesses, algorithms, or methods of construction.

A constructive proof proves existence by producing an object. It does not merely show that failure is impossible. It gives a witness, an algorithm, or a method that builds the required object.

The basic form is:

xX, P(x). \exists x \in X,\ P(x).

A constructive proof supplies a specific xx and verifies that P(x)P(x) holds.

For example, to prove that there exists an even prime number, we can exhibit the number 22. It is prime, and it is even. The proof is complete because the witness is explicit.

Constructive proofs are especially important when existence must lead to computation. If a theorem says that a solution exists, a constructive proof may also explain how to find it. This makes the proof useful for algorithms, formal verification, and applied mathematics.

Consider the statement:

There exist integers xx and yy such that

48x+18y=6. 48x + 18y = 6.

A constructive proof can use Euclid’s algorithm:

48=218+12 48 = 2 \cdot 18 + 12 18=112+6. 18 = 1 \cdot 12 + 6.

Now solve backward:

6=1812 6 = 18 - 12

and

12=48218. 12 = 48 - 2 \cdot 18.

Substitute:

6=18(48218) 6 = 18 - (48 - 2 \cdot 18)

so

6=31848. 6 = 3 \cdot 18 - 48.

Thus one witness is

x=1,y=3. x = -1,\quad y = 3.

This proves existence by constructing the integers.

A nonconstructive proof may show that an object must exist without identifying it. Such a proof can be valid in classical mathematics, but it may not provide computational content.

The difference matters. A proof that a solution exists may not tell us how to compute the solution. A constructive proof gives more information: it turns existence into data.

Constructive reasoning is closely connected to algorithms. An algorithmic proof often has this shape:

Given input satisfying certain assumptions, perform these steps. The steps terminate. The output satisfies the required property.

For example, to prove that every finite nonempty set of integers has a minimum, one can give a procedure: inspect the first element, keep the smallest value seen so far, and update it while scanning the set. When the scan ends, the stored value is the minimum.

This proof constructs the minimum rather than relying on an abstract existence principle.

Constructive proofs also require care with disjunctions. To prove

PQ, P \lor Q,

one should prove one side and identify which side holds. It is not enough to show that both cannot be false unless one is working classically.

Similarly, to prove

x, P(x), \exists x,\ P(x),

one should provide the witness xx.

This gives constructive logic a stronger evidential standard. Truth is tied to evidence.

Constructive proof is not limited to simple examples. Many major theorems have constructive versions. In numerical mathematics, one may prove existence of an approximation and give an algorithm to compute it. In algebra, one may prove that a basis exists by giving a procedure under finite conditions. In computer science, one may prove that a program terminates and returns a correct output.

However, constructive proofs may be harder to find. Some classical proofs use contradiction, excluded middle, compactness, or choice principles in ways that do not produce witnesses. Replacing them with constructive arguments often requires more detailed structure.

Constructive proof also affects definitions. A statement such as “there exists a real number with property PP” may be interpreted differently depending on whether real numbers are treated as completed objects or as computable approximations.

This is why constructive mathematics often tracks more information: how objects are represented, how choices are made, and how witnesses are obtained.

A useful test is simple: after reading an existence proof, ask what object was produced. If no object or method was produced, the proof is nonconstructive. If a witness or algorithm was produced, the proof is constructive.

Constructive proofs make mathematics operational. They do not only certify that something exists. They explain how to obtain it.