Arguably, this is because the common definition of modular arithmetic is itself zero-based: “modulo N” maps all integers into the numbers [0,N-1]. It is fully possible to define a “%” operator that instead mapped the integers into the range [1,N], which might be more natural in 1-based languages?
