What Mathematicians Actually Do: Research and Proof

Ask most people what a mathematician does and they imagine someone performing enormous calculations, perhaps faster than a calculator. In reality, research mathematicians spend very little time calculating. Their real work is discovering new patterns, asking the right questions and, above all, proving that their answers are true. This article looks behind the scenes at how mathematical research actually happens, and walks through the kind of reasoning that lies at its heart.

Mathematics Is Not Finished

A common belief is that all mathematics was worked out centuries ago. In fact, tens of thousands of new research papers in mathematics are published every year, and the subject is growing faster than ever. Some branches, such as algebraic topology or cryptography, barely existed a hundred years ago. Every one of those papers contains something new: a theorem nobody had proved before.

The Research Cycle

Although every mathematician works differently, research usually follows a recognisable pattern.

1. Exploration. A mathematician studies examples, draws diagrams, runs computer experiments and reads earlier work, looking for patterns.
2. Conjecture. A pattern is turned into a precise statement that the mathematician believes is true but cannot yet prove.
3. Attempted proof. This is the long part. Most attempts fail. Each failure, however, reveals something about why the problem is hard.
4. Refinement. Often the original conjecture turns out to be false or too ambitious, and is modified. Counterexamples are valuable discoveries in their own right.
5. Proof and write-up. Once a complete argument is found, it is written carefully, checked, and shared with colleagues.
6. Peer review. The paper is submitted to a journal, where expert referees scrutinise every step. This can take months or even years.

Many mathematicians describe spending weeks or months stuck, followed by a sudden moment of insight, often while walking, showering or doing something else entirely. The French mathematician Henri Poincaré famously described realising a key idea in 1880s research as he stepped onto a bus.

What Counts as a Proof?

A proof is a logical argument that establishes a statement beyond any doubt, starting from accepted definitions and axioms. Evidence, no matter how strong, is not enough. The statement "\( n^2 + n + 41 \) is prime for every natural number \( n \)" is true for \( n = 0, 1, 2, \ldots, 39 \), but fails at \( n = 40 \), where it equals \( 40^2 + 40 + 41 = 1681 = 41^2 \). Forty successful cases proved nothing.

Mathematicians use several standard proof techniques: direct proof, proof by contradiction, proof by contrapositive, proof by induction and proof by construction. Here are two classic examples that every mathematician knows.

Worked Example 1: There Are Infinitely Many Primes

This proof appears in Euclid's Elements, written around 300 BC, and is still admired for its elegance.

Claim: there is no largest prime number.

Proof (by contradiction): suppose there are only finitely many primes, and list all of them: \( p_1, p_2, \ldots, p_k \). Form the number

$$ N = p_1 p_2 \cdots p_k + 1 $$

Now reason step by step:

1. Dividing \( N \) by any prime \( p_i \) on the list leaves remainder 1, so none of them divides \( N \).
2. But every whole number greater than 1 has at least one prime factor. So \( N \) has a prime factor that is not on our list.
3. This contradicts the assumption that the list contained every prime. Therefore there are infinitely many primes.

Illustration: take the primes 2, 3, 5, 7, 11, 13. Then \( N = 30030 + 1 = 30031 \). Interestingly, 30031 is not itself prime: \( 30031 = 59 \times 509 \). But its prime factors 59 and 509 are both missing from the original list, exactly as the proof guarantees. This is a common misunderstanding: the proof does not say \( N \) is prime, only that it has a new prime factor.

Worked Example 2: Proof by Induction

Claim: for every natural number \( n \),

$$ 1 + 2 + 3 + \cdots + n = \frac{n(n + 1)}{2} $$

Induction works like a line of dominoes: show the first one falls, and show each falling domino knocks over the next.

Base case (\( n = 1 \)): the left side is 1, and the right side is \( \tfrac{1 \times 2}{2} = 1 \). True.

Inductive step: assume the formula holds for some \( n = k \):

$$ 1 + 2 + \cdots + k = \frac{k(k + 1)}{2} $$

Add the next term, \( k + 1 \), to both sides:

$$ 1 + 2 + \cdots + k + (k + 1) = \frac{k(k + 1)}{2} + (k + 1) $$

Take out the common factor \( (k + 1) \):

$$ = (k + 1)\left(\frac{k}{2} + 1\right) = \frac{(k + 1)(k + 2)}{2} $$

This is exactly the formula with \( n = k + 1 \). So if it holds for \( k \), it holds for \( k + 1 \). Since it holds for 1, it holds for 2, then 3, and so on for every natural number.

Check: for \( n = 100 \), the formula gives \( \tfrac{100 \times 101}{2} = 5050 \), the answer the young Carl Friedrich Gauss is traditionally said to have found in seconds at school.

Collaboration and Community

The image of the lonely genius is mostly a myth. Today most research papers have two or more authors, and mathematicians constantly discuss ideas at seminars, conferences and over coffee. The Hungarian mathematician Paul Erdős wrote papers with more than 500 co-authors and travelled the world with a suitcase, staying with collaborators. Mathematicians still jokingly compute their Erdős number: the number of co-authorship links connecting them to him.

Most research mathematicians work at universities, combining research with teaching. Others work in industry, finance, cryptography, data science, government agencies or technology companies, applying the same skills of precise reasoning to practical problems.

Landmark Proofs of Recent Times

• Fermat's Last Theorem: stated by Pierre de Fermat around 1637, it was finally proved by Andrew Wiles, who announced a proof in 1993. A gap was found, and Wiles fixed it with Richard Taylor in 1994. Wiles had worked on it largely in secret for about seven years.
• The Four Colour Theorem: proved in 1976 by Kenneth Appel and Wolfgang Haken, the first major theorem to rely essentially on a computer check.
• The Poincaré Conjecture: proved by Grigori Perelman in papers posted in 2002 and 2003. He declined both the Fields Medal and a million-dollar prize.

Today, computer proof assistants such as Lean and Coq allow every logical step of a proof to be verified by machine. Thomas Hales's proof of the Kepler conjecture on sphere packing was formally verified this way in 2014.

Did You Know?

• The Fields Medal, often called the Nobel Prize of mathematics, is awarded every four years to mathematicians under 40. In 2014 Maryam Mirzakhani became the first woman to receive it.
• Some proofs are enormous: the classification of finite simple groups spans tens of thousands of pages across hundreds of papers.
• Mathematicians often value a proof for its beauty. Erdős spoke of "The Book", an imaginary volume in which God keeps the most elegant proof of every theorem.

Key Takeaways

• Research mathematicians create new mathematics: they explore, conjecture, prove and publish.
• A proof establishes truth with certainty; examples alone are never enough.
• Proof by contradiction and proof by induction are two of the most powerful techniques.
• Modern mathematics is highly collaborative and increasingly supported by computers.
• Famous problems like Fermat's Last Theorem can take centuries, and remarkable persistence, to solve.

Finding proofs difficult in your course or assignment? Amathify's experts and PhD scholars offer step-by-step solutions and assignment help via WhatsApp or the Ask a Question form.