SOURCE: Barry Mazur, When Is One Thing Equal to Some Other Thing?
Anyone who writes a number in the form 1729 implies a method of calculation: one thousand, plus seven hundreds, plus two tens, plus nine ones.
Different than writing tick-marks |||||||||||||||||||||||…¹⁷²⁹ which would imply
Different than Roman numerals MDCCXXIX,
hexadecimal 6C1,
or the most agnostic way to write a number, via its prime factorization ”the fourth prime ⨯ the sixth prime ⨯ the eighth prime”.
They’re all ways of calculating the number, but they’re not the number itself.
Daphne
We could agree to call this number some agreed-upon name, like ”Daphne” and use a symbol ₯ for shorthand. 1729 is no more her name than is 6C1.
Or we could refer to Daphne by property without implying a particular calculation: “the smallest sum of two cubes, which can be written two different ways”.
Or we could denote Daphne by equation:
- ₯ = 9^3 + 10^3, or
- Daphne ₯ is the number that solves the equation 12^3 - ₯ = 1^3.
It’s the same way with the square root of two. Its name is no more √2 than ⨿2 or ¶2.
Just like 1729, √2 is merely a notation. What makes √2 be √2 is the property it has.
Numbers aren’t numerals, they’re … uh … things.
All of the above is meant to drive a wedge between numbers as written on paper and numbers as they “exist” abstractly.
![]()
Numbers don’t need numerals. And you can talk about numbers without knowing how to write them. Just agree on some symbol like π and use π whenever you want to talk about the number you don’t know how to write.
It sounds trivial talking about an integer, but the difference between
- properties of numbers,
- ways to calculate numbers, and
- the numbers themselves
is good to keep in mind when you’re thinking deeply about
- the real line (measure theory),
- algebraic numbers (Galois theory),
- transcendental numbers,
- p-adic numbers,
- complex numbers or other bodies of numbers.




