I saw a video recently about function currying, and it triggered the realization that currying might solve a problem I’ve been pondering in the context of language parsing. The problem involves knowing how many arguments an operator expects, what’s called the arity of an operation or function. It can vary from zero to many.

But it occurred to me that, with currying, there could be a language where operations always take just one argument. And that would solve a challenge for a mathematical expression language I have in mind.