arguments.callee

A reference to the currently executing function.

callee is a property of the arguments object. It can be used to refer to the currently executing function inside the function body of that function. This is useful when the name of the function is unknown, such as within a function expression with no name (also called "anonymous functions").

(The text below is largely adapted from a Stack Overflow answer by olliej)

Early versions of JavaScript did not allow named function expressions, and for this reason you could not make a recursive function expression.

For example, this syntax worked:

function factorial(n) {
  return n <= 1 ? 1 : factorial(n - 1) * n;
}

[1, 2, 3, 4, 5].map(factorial);

but:

[1, 2, 3, 4, 5].map(function (n) {
  return n <= 1 ? 1 : /* what goes here? */ (n - 1) * n;
});

did not. To get around this arguments.callee was added so you could do

[1, 2, 3, 4, 5].map(function (n) {
  return n <= 1 ? 1 : arguments.callee(n - 1) * n;
});

However, the design of arguments.callee has multiple issues. The first problem is that the recursive call will get a different this value. For example:

const global = this;

const sillyFunction = function (recursed) {
  if (this !== global) {
    console.log("This is:", this);
  } else {
    console.log("This is the global");
  }

  if (!recursed) {
    return arguments.callee(true);
  }
};

sillyFunction();
// This is the global
// This is: [object Arguments]

In addition, references to arguments.callee make inlining and tail recursion impossible in the general case. (You can achieve it in select cases through tracing, etc., but even the best code is suboptimal due to checks that would not otherwise be necessary.)

ECMAScript 3 resolved these issues by allowing named function expressions. For example:

[1, 2, 3, 4, 5].map(function factorial(n) {
  return n <= 1 ? 1 : factorial(n - 1) * n;
});

This has numerous benefits:

• the function can be called like any other from inside your code
• it does not create a variable in the outer scope (except for IE 8 and below)
• it has better performance than accessing the arguments object

Strict mode has banned other properties that leak stack information, like the caller property of functions. This is because looking at the call stack has one single major effect: it makes a large number of optimizations impossible, or much more difficult. For example, if you cannot guarantee that a function f will not call an unknown function, it is not possible to inline f.

function f(a, b, c, d, e) {
  return a ? b * c : d * e;
}

If the JavaScript interpreter cannot guarantee that all the provided arguments are numbers at the point that the call is made, it needs to either insert checks for all the arguments before the inlined code, or it cannot inline the function. This means any call site that may have been trivially inlinable accumulates a large number of guards. Now in this particular case a smart interpreter should be able to rearrange the checks to be more optimal and not check any values that would not be used. However in many cases that's just not possible and therefore it becomes impossible to inline.

A recursive function must be able to refer to itself. Typically, a function refers to itself by its name. However, an anonymous function (which can be created by a function expression or the Function constructor) does not have a name. Therefore if there is no accessible variable referring to it, the only way the function can refer to itself is by arguments.callee.

The following example defines a function, which, in turn, defines and returns a factorial function. This example isn't very practical, and there are nearly no cases where the same result cannot be achieved with named function expressions.

function create() {
  return function (n) {
    if (n <= 1) {
      return 1;
    }
    return n * arguments.callee(n - 1);
  };
}

const result = create()(5); // returns 120 (5 * 4 * 3 * 2 * 1)

Although function expressions can now be named, arrow functions always remain anonymous, which means they cannot reference themselves without being assigned to a variable first. Fortunately, in Lambda calculus there's a very good solution which allows a function to both be anonymous and self-referential. The technique is called a Y-combinator. Here we will not explain how it works, only that it works.

// The Y-combinator: a utility function!
const Y = (hof) => ((x) => x(x))((x) => hof((y) => x(x)(y)));

console.log(
  [1, 2, 3, 4, 5].map(
    // Wrap the higher-order function in the Y-combinator
    // "factorial" is not a function's name: it's introduced as a parameter
    Y((factorial) => (n) => (n <= 1 ? 1 : factorial(n - 1) * n)),
  ),
);
// [ 1, 2, 6, 24, 120 ]

• Functions guide
• Functions
• arguments
• Function.prototype.caller